the evolution of protein kinase inhibitors from ...shokatlab.ucsf.edu/pdfs/21548788.pdfshown is the...

BI80CH31-Shokat ARI 16 May 2011 13:48

The Evolution of ProteinKinase Inhibitors fromAntagonists to Agonists ofCellular SignalingArvin C. Dar1 and Kevan M. Shokat1,2

1Howard Hughes Medical Institute and Department of Cellular and MolecularPharmacology, University of California, San Francisco, California 94158;email: [email protected], [email protected] of Chemistry, University of California, Berkeley, California 94720

Annu. Rev. Biochem. 2011. 80:769–95

First published online as a Review in Advance onMay 6, 2011

The Annual Review of Biochemistry is online atbiochem.annualreviews.org

This article’s doi:10.1146/annurev-biochem-090308-173656

Copyright c© 2011 by Annual Reviews.All rights reserved

0066-4154/11/0707-0769$20.00

Keywords

protein kinases, kinase inhibitor, drug discovery, signal transduction

Abstract

Kinases are highly regulated enzymes with diverse mechanismscontrolling their catalytic output. Over time, chemical discovery effortsfor kinases have produced ATP-competitive compounds, allostericregulators, irreversible binders, and highly specific inhibitors. Thesedistinct classes of small molecules have revealed many novel aspectsabout kinase-mediated signaling, and some have progressed fromsimple tool compounds into clinically validated therapeutics. Thisreview explores several small-molecule inhibitors for kinases high-lighting elaborate mechanisms by which kinase function is modulated.A complete surprise of targeted kinase drug discovery has been thefinding of ATP-competitive inhibitors that behave as agonists, ratherthan antagonists, of their direct kinase target. These studies hint ata connection between ATP-binding site occupancy and networks ofcommunication that are independent of kinase catalysis. Indeed, kinaseinhibitors that induce changes in protein localization, protein-proteininteractions, and even enhancement of catalytic activity of the targetkinase have been found. The relevance of these findings to thetherapeutic efficacy of kinase inhibitors and to the future identificationof new classes of drug targets is discussed.

769

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Contents

INTRODUCTION. . . . . . . . . . . . . . . . . 770PROTEIN KINASE INHIBITORS

THAT BIND IN THE ATPSITE . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773Promiscuous Inhibitors . . . . . . . . . . . 773A Selective Inhibitor of the Map

Kinase: P38 . . . . . . . . . . . . . . . . . . . 773An Unusual MEK Inhibitor that

Binds Adjacent to ATP . . . . . . . . 775Conformation-Specific Inhibitors:

Gleevec . . . . . . . . . . . . . . . . . . . . . . . 775The Interplay Between Inhibitor

Specificity and KinaseConformation . . . . . . . . . . . . . . . . . 776

Selective Inhibition by Li+ Ions . . . 777Chemical Genetic Kinase

Inhibitors . . . . . . . . . . . . . . . . . . . . . 778Structural Bioinformatics–Based

Design . . . . . . . . . . . . . . . . . . . . . . . . 779Structural Bioinformatics–Based

Targeting of ParasiteKinases . . . . . . . . . . . . . . . . . . . . . . . 779

KINASE INHIBITORS THATBIND OUTSIDE OF THEATP POCKET . . . . . . . . . . . . . . . . . . 780

Protein-Based NaturallyOccurring Selective KinaseInhibitors . . . . . . . . . . . . . . . . . . . . . 780

An Allosteric Inhibitor Competesfor Myristate . . . . . . . . . . . . . . . . . . 782

An Interdomain Inhibitorof AKT . . . . . . . . . . . . . . . . . . . . . . . 783

Kinase Activation via Bindingof the PIF Pocket. . . . . . . . . . . . . . 783

IN VITRO INHIBITORSBECOME ACTIVATORS INCELLS . . . . . . . . . . . . . . . . . . . . . . . . . . 784Bypassing the Kinase Activity to

Activate an RNase Domain . . . . 784Ligand Occupancy of the

ATP Site Templates PKCMaturation . . . . . . . . . . . . . . . . . . . . 785

Inhibitor Hijacking of AKTActivation . . . . . . . . . . . . . . . . . . . . . 787

A RAF Inhibitor That StimulatesRAF Activity . . . . . . . . . . . . . . . . . . 788

Kinase-Inactivating Mutationsand Cancer. . . . . . . . . . . . . . . . . . . . 789

SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . 791

INTRODUCTIONEukaryotic protein kinases have emergedas key components in most, if not all, signaltransduction cascades. The conserved nature ofthe ATP active sites of protein kinases and thelarge number of kinases in eukaryotic genomeshave made the interrogation of each kinasea challenge to traditional pharmacologicalmethods used to study other enzymes. In spiteof these challenges, the number of selectiveprotein kinase inhibitors has exploded in thepast ten years largely as a consequence of in-tense efforts from the pharmaceutical industryto discover inhibitors of kinases dysregulatedin cancer and inflammatory disease. Proteinkinases have been the topic of a large body ofreview literature. We point the reader to the

following articles pertaining to kinases anddisease, inhibitor design, and the analysis of ki-nases with chemical tools: References 1, 2, and3, respectively. The aim of this report is (a) toexplore the various classes of kinase inhibitors,(b) provide insight into the mechanistic detailsabout kinase signaling gained through theapplication of these inhibitors, (c) to highlightrecent examples where kinase inhibitors havebeen found to paradoxically stimulate signalingwithin cells, and (d ) to discuss the implicationsof these findings to the future development ofkinase-directed therapeutics.

This review begins with the biochemical andstructural lessons learned from protein kinaseinhibitors that bind in the highly conservedATP pocket, including the molecular basis

770 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Mg2+α

β

γ

Protein kinase active site

Hin

ge

a b

C-lobe

N-lobe

Adenosinepocket

Tyrphospho-acceptor

Figure 1A representative protein kinase ATP substrate complex. Shown is the insulin receptor kinase in complex withATP and a small peptide substrate (PDB ID 1IR3). (a) The ATP-binding pocket within the protein kinasedomain lies deep within a cleft bound by the N-terminal lobe, hinge regions, and C-terminal lobe: proteinkinase, gray surface; ATP, sticks; Mg2+ ions, green spheres. (b) Active-site magnification: critical H-bondsbetween ATP and the main-chain atoms of the hinge are highlighted (dashed lines). The γ-phosphate of ATPis poised for transfer to the hydroxyl group of the peptide substrate (in this case, the acceptor residue istyrosine).

Active site: in kinases,the ATP-bindingpocket and regionwhere phosphorylationoccurs

Gleevec: (trade name,Imatinib) inhibitor ofBcr-Abl and the firstapproved kinaseinhibitor drug

for inhibitor selectivity against closely relatedkinases. In Figure 1 we show the structureof a representative protein kinase active site.Although the conserved ATP pocket is wherethe majority of kinase inhibitors bind, it haslong been recognized that targeting pocketsoutside of the nucleotide region offer a majordiscrimination filter because the phospho-acceptor pocket and domains outside of thekinase catalytic domain are highly divergentacross the family. Examples of such non-ATPcompetitive kinase inhibitors from naturalsources as well as synthetic molecules providea functional picture of how regions outside ofthe catalytic site of kinases can regulate kinasefunction and how next-generation inhibitorsmay exploit divergent binding pockets.

We next consider how kinase inhibitorsregulate cellular pathways inside cells. Thegoal of most kinase inhibitor discovery is toblock the pathway mediated by the kinase

of interest in cells, animals, and, potentially,patients. Currently nine kinase inhibitors areFDA approved (Table 1), providing a wealth ofinformation about how tumors respond to suchtargeted therapy. The two drugs with the mostdramatic success target kinases mutated in par-ticular cancer types: Gleevec, approved nearly10 years ago to treat BCR-Abl-driven chronicmyelogenous leukemia cancers, and PLX4032(currently in phase II trials), a recent investiga-tional drug, that targets BRAF(V600E) mutanttumors in metastatic melanoma patients. Bothdrugs show dramatic effects in tumor cells con-taining somatically mutated activated kinases.

Several recent examples of kinase inhibitorsdesigned initially to antagonize signaling para-doxically induce pathway activation. PLX4032exhibits completely unexpected activation ofBRAF signaling in cells lacking mutated BRAF,which has generated new insights into BRAFsignaling while also complicating the clinical

www.annualreviews.org • The Evolution of Protein Kinase Inhibitors 771

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Table 1

Drug Structure FDA approval date Indication

Imatinib 2001 Chronic myelogenous leukemia Gastrointestinal stromal tumors

Gefitinib 2003 Nonsmall cell lung cancer

Sorafenib 2005 Renal cell carcinomaHepatocellular carcinoma

Erlotinib 2005

Chronic myelogenous leukemiaPhiladelphia chromosomepositive acute lymphoblastic leukemia

Sunitinib 2006Renal cell carcinomaImatinib-resistant gastrointestinalstromal tumor

Dasatinib 2006

Pancreatic cancer

Nilotinib 2007 Imatinib-resistant chronic myelogenous leukemia

Lapatinib 2007 HER2+ breast cancer

Pazopanib 2009 Advanced renal cell carcinoma

N

N

N OHN

NN

O

N

N

F

O

O

O N

O

NO

O

HN

HN

HN

OS

N

NN N N

OF

NH

NH

NH

NH

NO

O

O

OO

N

N

NO

N

N

N

N

HN SO

O

O

O

F

N N

N NN

SO O

O

N

F

HN S

O

OF

PLX4032Under

regulatoryreview

BRAF mutantmetastatic melanoma

Cl

Cl

NH

N

N

NH

NH

NH

NH

NH

CF3

Cl

Cl

Cl

OHHN

HN

CF3

NH2

772 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


development of this investigational drug.Realization that small-molecule kinase in-hibitors can actually activate kinase signalingcascades suggests new insights and opportu-nities for pharmacological manipulation ofprotein kinase function. Several recent exam-ples of kinase activators have been reported,which significantly expands our understandingof the complexity of kinase regulation and howsmall molecules can rewire signaling cascades.

The horizon for selective small-moleculekinase inhibitors remains the discovery ofcompounds that bind outside of the ATPsite and can attain high specificity and highpotency. An unexplored opportunity for acompletely new category of kinase regulatorlies in the recent results from cancer genomesequencing, which reveal that a number ofkinase mutations that drive oncogenesis inac-tivate kinase catalytic activity, yet drive cancer.Answering the question of how to inhibit acatalytically dead kinase that causes diseasemay lead to new therapeutic opportunities.

PROTEIN KINASE INHIBITORSTHAT BIND IN THE ATP SITE

Promiscuous Inhibitors

The development of protein kinase inhibitorsdates back 25 years beginning with the foun-dational studies of natural products. Thesemolecules inhibited the activity of the firstprotein kinases to be assayed biochemicallyafter purification from natural sources. In theseearly studies the bis-indoylmaleimide stau-rosporine stood out as a highly potent inhibitor(IC50 = 2.7 nM) of protein kinase C (PKC)that was not competitive with the PKC acti-vator diacylglycerol (4). Recent kinome-widescreening efforts of staurosporine have revealedthat this natural product is a potent inhibitorof >90% of kinases (5). Structural studiesconfirm that this molecule binds in the ATPpocket despite its much larger molecular size(Figure 2a). The staurosporine inhibitor repre-sents the first highly potent inhibitor of proteinkinases and demonstrates that, despite the

ATP-competitive nature of such ligands, theycan out-compete the high cellular concentra-tion of ATP and effectively block kinase func-tion. Structurally, staurosporine revealed theimportance of the inhibitor binding in subpock-ets of the kinase active site that are not occupiedby ATP, generating greater affinity than thatachieved by the substrate ATP (6). The largesize of staurosporine and its broad kinase inhi-bition suggested that such subpockets were notrestricted to particular kinases but that most,if not all, protein kinase family members con-tained additional pockets for inhibitor binding.

Although staurosporine ushered in the eraof kinase inhibitors, its uniform cellular induc-tion of apoptosis, likely due to its promiscuousinhibition of many kinases, made it a poor drugcandidate and caused concern that selectivekinase inhibition would be unachievable. Othernatural products, such as the flavonol quercetin(7), revealed new chemotypes that fit in theATP pocket, and these led to developmentof synthetic analogs such as the tyrphostinsas inhibitors of receptor tyrosine kinases (8).These early examples progressively narroweddown the spectrum of kinases inhibited bysmall molecules, although the first truly se-lective chemotypes eluded rational design andresearchers have relied on unbiased cellular-and pathway-based screens.

A Selective Inhibitor of the MapKinase: P38

In an effort to identify new anti-inflammatoryagents, SmithKline Beecham scientistsscreened for molecules that could block proin-flamatory cytokine (IL-1 and TNF) productionby monocytes in response to inflammatorystimuli (9). They identified a simple chemotype,the triarylimidazole exemplified by SB202190,that blocked cytokine production. Thestructure of the molecule did not suggestany known target in the pathway, and thus asearch using a radiolabelled analog and a pho-toaffinity probe was used to identify the novelkinases p38α and p38β as the targets of the hitmolecule. Remarkably, the inhibitors are highlyselective for these two MAP kinases relative to


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


HN

NN O

H

O

N

HN

N

S

O

F

a Staurosporine b SB203580 c Overlay:staurosporine + SB203580

H3CO

H3C

NHCH3

Figure 2Examples of ATP competitive and reversible inhibitors. (a) The natural product and general kinase inhibitor, staurosporine, in complexwith CDK2. The gatekeeper phenylalanine residue in CDK2 is shown as a gray surface (PDB ID 1AQ1). (b) The pyridinyl imidazolep38 inhibitor SB203580, an analog of SB202190, in complex with the ERK2 mutant Q105T. The gatekeeper mutant threonine residueis shown as a blue surface (PDB ID 1PME). (c) The Phe gatekeeper in CDK2 obstructs SB203580 binding. Shown is a structuraloverlay of both structures from panels a and b.

Gatekeeper: residueadjacent to N-6 ofadenine in the ATPpocket of kinases(corresponds toThr-338 in c-Src)

the many other MAP kinases. Structural studiesand sequence alignments of the map kinasefamily revealed that a single nonconservedresidue in the ATP pocket, now termed thegatekeeper residue, is largely responsible forthe high p38α selectivity of the triarylimidazole(Figure 2b,c) (10). The gatekeeper residuein p38α is a Thr, whereas in other MAPkinases, such as JNK, the gatekeeper residueis a Met. This provides a selectivity filter fortriaryl-imidazole binding to p38 and not JNK,despite their close homology. The gatekeeperresidue is a critical determinant of inhibitorselectivity as it is positioned near the N6 amineof ATP in all kinases and can be either a verylarge amino acid such as Phe or a much smallerresidue such as Thr.

The discovery of SB202190 opened the fieldof kinase inhibitor development to a wealthof drug discovery efforts. Not only did thiscompound lead to additional inhibitors for p38such as SB203580 (11), but it also revealedthe possibility of finding completely new het-erocyclic “drug-like” structures that could dis-tinguish between closely related kinases. Theimportance of small differences in amino acididentity in kinases could be read out by smallmolecules allowing for highly selective kinaseinhibition. A great wealth of chemical synthesisand screening was pursued, leading to the dis-covery of potent inhibitors of other kinases suchas PP1 for Src family kinases (12) and other Thrgatekeeper kinases including PD 153035, whichis directed at EGFR receptor kinases (13). In

774 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Type I

1QCFHCK-PP1

Helix aC

DFG-IN

Type III

1S9JMEK1-PD

Helix aC

N

N NN

N

N

NHN

HN N

PP1 STI-571 PD318088

O

OH

HN O

O

Br

F

F

HN

F

I

HO

a b c

NH2

ATP non-competitive

Hin

ge

Hin

ge

Hin

ge

Hin

ge

Adenosine binding pocket

Benzamideportion of

STI-571

Type II

MEK1inhibitor

PD318088binding

site

1FPUAbl-STI571

Helix aC

DFG-OUT

Figure 3Type I, II, and III kinase inhibitors. (a) Type I inhibitors form H-bonds with the kinase hinge and occupy the adenosine binding pocket(blue). Type I inhibitors do not require a specific conformation of key structural elements, including helix αC (red ) or the DFGMg2+-binding site (white sticks), for binding. Shown is the inhibitor PP1 in complex with HCK (PDB ID 1QCF). (b) Type II inhibitorsoccupy the adenosine pocket, but unlike type I binders, they induce a configuration of DFG residues termed DFG-OUT (the D of theDFG flipped 180◦ relative to the active state conformation). Shown is STI-571 in complex with Abl (PDB ID 1FPU). The benzamideportion of STI-571 (highlighted purple) is a common pharmacaphore found in type II binders. (c) Type III inhibitors block kinase activitywithout displacing ATP. Shown is the MEK1 inhibitor PD318088 binding site ( green) (PDB ID 1S9J).

Type I inhibitor:a small molecule thatbinds to the activeconformation of akinase in the ATPpocket

Type II inhibitor:a small molecule thatbinds to the inactive(usually DFG-OUT)conformation of akinase

Type III inhibitor: anon-ATP competitivekinase inhibitor

Figure 3, we show representative type I, II,and III inhibitors, which are described in thefollowing section.

An Unusual MEK Inhibitor that BindsAdjacent to ATP

The high frequency of Ras mutations leadingto activation of the Raf-MEK-Erk kinaseshas directed a great deal of kinase inhibitordiscovery efforts for Raf (a MAPKKK), MEK(a MAPKK), and Erk (a MAPK). These threekinases constitute a sequential cascade ofkinase substrate relationships. On the basis ofthis organization, Parke-Davis carried out apathway-based screen reading out the activityof Erk (14). The inhibitor identified in thisscreen, PD 098059, was based on a chromonescaffold and was shown to prevent MEK acti-vation of MAPK (15). That PD 098059 could

inhibit MEK and not other closely relatedkinases suggested it was another example of akinase inhibitor with unique selectivity. Thecocrystal structure of MEK1 with PD 318088(a variant of the core inhibitor) showed sur-prising simultaneous binding of ATP and theinhibitor (Figure 3c) (16). The inhibitor wasbound adjacent to ATP in a pocket partiallyexisting in the MEK1-ATP structure. This wasthe first inhibitor determined to bind in thekinase active site without making any contactsin common with ATP, now known generallyas type III binders.

Conformation-SpecificInhibitors: Gleevec

MEK inhibitors provided some of the first in-sights into conformational control of kinasesby inhibitors and dramatically highlighted the


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


DFG-OUT:a conformation ofkinases in which theDFG triplet occupiesan “out” conformationincapable ofcomplexing Mg in thecatalytically essentialstep of phosphatetransfer

potential of pockets adjacent to the nucleotidesubstrate. The concept that closely related ki-nases may exhibit similar active conformationsyet distinct inactive conformations, providingopportunities for inhibitor selectivity, becamea foundation for inhibitor design following thereport of the STI-571 (a Gleevec analog) struc-ture in Abl tyrosine kinase. The STI-571 in-hibitor exhibits substantial selectivity for Ablover Src. Despite near-identical drug-bindingpockets within these two tyrosine kinases, STI-571 inhibits Abl approximately 1000-fold morepotently than Src (17), which approaches theselectivity seen in purely biological systems.For example, potassium channels are approx-imately 105 more permeable for K+ over Na+

(18). The cocrystal structure of STI-571 boundto Abl revealed a previously unknown confor-mational change from the active conformationwherein the triplet motif, DFG, shifted 180◦

to an inactive conformation in a crankshaft-like motion (Figure 3b) (19). That STI-571required a large conformational change of thetarget Abl suggested an energetic explanationof its selectivity against Src. If Src were unableto adopt the “DFG-OUT” conformation re-quired by STI-571 binding, then this would ex-plain how kinases with almost identical aminoacids in contact with the drug could be differ-entially inhibited. Presumably, the amino acidsin the second sphere surrounding the ATP sitecontrolled such conformational flexibility.

This concept of differential conformationalbias between kinases took hold because it of-fered yet another mechanistic approach for se-lective inhibitor design. The challenge with thismechanistic approach, however, lies in the lackof structural models that would allow predic-tion of those kinases capable of adopting partic-ular noncatalytically competent conformations.In practice, many of the unusual conformationsof kinases identified structurally have requireddiscovery of selective inhibitors to stabilize theconformation in crystal structure studies (20).In this area, molecular modeling and dockingalgorithms could prove particularly instrumen-tal for the development of new classes of kinaseinhibitors.

The Interplay Between InhibitorSpecificity and Kinase Conformation

Crystallographic analysis of Src in complexwith Imatinib revealed Src in the identicalDFG-OUT conformation that was observedin the Abl-Imatinib complex (19, 21). Thestructure suggested that Src could in factadopt the inactive conformation required forImatinib binding but could not be used toexplain the relatively weak binding affinitybetween Src and Imatinib.

The paradox could be summarized asfollows: Two homologous kinases in identicalconformations are bound to the same drugusing near-identical drug-binding residues,but one binds with much higher potency to thedrug than does the other. To explain the weakbinding between Src and Imatinib, researcherssuggested the presence of a large energetic bar-rier inherent to Src that precludes binding toImatinib with affinity similar to Abl. Efforts tomake Src more sensitive to STI-571/Imatinibthrough point mutations that made Src moreAbl-like were largely unsuccessful, furthersupporting the idea that a conformational bias,distributed through the respective protein ki-nase domains, must explain the large differencein binding to Imatinib between Abl and Src(21).

A model suggesting that the DFG-OUTconformation was disfavored in Src made asimple testable prediction: No potent DFG-OUT inhibitors of Src should exist. If suchDFG-OUT Src binders exist, then Src doesnot pay a large energetic penalty for adoptingthis conformation and there must be someother explanation for how Src eludes STI-571inhibition. In three different studies exploringthree distinct chemical series of inhibitors, suchDFG-OUT Src binders were reported recently(22–24). These reports overturn one of the keyguiding models of selective kinase inhibition.One series of DFG-OUT Src binders, in fact,reveals a distinct contact with a Phe in Src,which is one of the few distinguishable residuesbetween Src (Phe) and Abl(Tyr) (25). Thus, asingle hydroxyl group difference facilitates a

776 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


closed conformation in the STI-571 structurewith Abl, which is less stable in Src. A currentformulation of the conformational flexibilitybasis for inhibitor selectivity appears still toinvolve amino acid differences between closelyrelated kinases that result in stable conforma-tions that are compatible with drug binding.Importantly, some chemical series contain par-ticular heterocyclic structures that exploit suchsmall differences between kinases, whereas oth-ers capable of binding the same conformationalstates do not distinguish between such small dif-ferences. The lessons from STI-571 highlightthe interplay between small-molecule chemicalseries and amino acid differences in kinases andhow promiscuity or selectivity can result.

Selective Inhibition by Li+ IonsThe dominant chemical class of kinase in-hibitors are heterocyclic compounds (summa-rized in Figure 4). However, lithium ion is theone remarkably simple inhibitor of the kinaseGSK-3β (26). The fact that a metal ion couldregulate kinase activity could have been antic-ipated from the requirement for Mg2+ in theATP-catalyzed reaction. Lithium apparentlycompetes with an active-site Mg2+ in GSK-3β, disrupting the catalytic reaction (27). At500 μM Mg2+, Li+ (but not K+) blocks GSK-3β kinase activity with an IC50 of 2 mM (28).The remarkable aspect of the kinase inhibitoryactivity of Li+, however, is its selectivity for thekinase GSK-3β, as Li+ poorly inhibits closely

Type I inhibitors Type II inhibitors

Inhibitor inducesactive conformation

Inhibitor inducesDFG-OUT inactive

conformation

ATP competitive, reversible (e.g., Staurosporine, SB202190)

X-ray crystallographic

modelDiagram

ATP competitive, irreversible (e.g., FMK, HKI-272)

SH

C-lobe

N-lobe

Hin

ge αC

A loop

D of DFG

MgATP

Substrate

Inhibitorscaffold

Electro-philicgroup

O

Figure 4Binders inside the ATP pocket. Summary of active-site modulators: representation of the kinase domain with conserved structuralelements highlighted, depicting known mechanisms of protein kinase inhibition. Purple small molecules represent general inhibitorscaffolds (direct the inhibitors to the ATP pocket), and yellow represents an electrophilic group (makes the inhibitors irreversible uponreaction with a Cys in the target kinase).


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


related kinases. As the only approved drug fortreatment of bipolar disorder, Li+ is an impor-tant clinical agent, even though the basis for itsefficacy is still not clearly understood. The inhi-bition of GSK-3β is considered to be one of theprime candidates for the basis of the Li+ mech-anism of action (although GSK-3β is likely notthe only or sole target for Li+) (29, 30).

Chemical Genetic Kinase Inhibitors

Highly selective kinase inhibitors for particularkinases are rapidly becoming available, thoughthe pace of discovery for such important tools isslow compared with the vast size of the proteinkinase family (518 human kinases). A methodfor developing a selective inhibitor for everykinase in the genome that has been establishedrelies on the use of protein engineering to alter

the ATP active site of the kinase of interest(31). The altered ATP site is then targeted bya common inhibitor that does not bind to anywild-type enzyme (Figure 5). This approachrelies on the semiconserved nature of the gate-keeper residue across the yeast, worm, fly, andhuman kinomes. There are no kinases in theseorganisms with a small Gly or Ala gatekeeperresidue. This natural absence of a small gate-keeper provides a window of selectivity to beintroduced at will. Inhibitors such as 1NM-PP1and 1NA-PP1, which are based on the PP1 se-ries of Thr selective kinase inhibitor, bind onlyto kinases with an engineered gatekeeper (Glyand/or Ala) and not to the wild-type kinases.Introduction of a Gly gatekeeper kinase in placeof the corresponding wild-type kinase into thechromosome produces a cell or organism inwhich 1NM-PP1 is a monoselective inhibitor

All naturally occurring kinases(518 in humans)

GatekeeperGly, Ala

Gatekeeper Thr

Active-site Cys

Rsk

1,2,

4

11110

O

F

N

N NN

N

PP1 FMK1NMPP1

1NAPP1

NH2N

N N

NH2

N

N N

NH2

NN

N N

NH2

HO

Figure 5Venn diagram for naturally occurring (inside gray box) and engineered kinase mutants (red dot). A threoninegatekeeper residue ( green circle) provides a broad selectivity filter for a subset of kinases that can be targetedby inhibitors such as PP1. Bioinformatic searches identified only three kinases with both a Thr gatekeeperand particular active-site cysteine (Rsk1, Rsk2, Rsk4) ( yellow), enabling selective inhibition by FMK. Anykinase in gray can be moved into the red circle through mutagenesis of the gatekeeper to a glycine or alanineresidue, rendering them sensitive to inhibitors that do not inhibit wild-type kinases, such as 1NMPP1 or1NAPP1. Red regions of 1NMPP1 and 1NAPP1 fill the space of small gatekeeper pockets (Gly, Ala) butclash with naturally occurring gatekeepers (most often Thr, Phe, or Leu).

778 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


of a single kinase of interest. An importantcontrol for the specificity of the inhibitor isthe parallel analysis of the compound in awild-type-only cell or organism to ensure thatthe effects observed are due to inhibition of thekinase of interest. To date, this approach hasbeen applied to more than 50 different kinasesin several cell lines including Saccharomycescerevisiae, Caenorhabditis elegans, and mice.

As increasingly more kinase inhibitors aredeveloped and they become used as tools touncover new cell biology, the question ofspecificity for the target of interest versus otheroff-target kinases becomes central. In responseto this challenge, a chemical genetic approachhas been developed and used to address thespecificity of SB203580 as well as an Aurorakinase inhibitor VX-680 (32–34). The methodis particularly useful in answering the questionof whether a specific cellular effect induced bya particular inhibitor requires inhibition of aparticular kinase target. The development of anAurora A mutant (T217D) resistant to a clinicalAurora inhibitor MLN8237 was used to greateffect in deciphering whether the antiprolifer-ative effect of MLN8237 was due to inhibitionof Aurora A (34). Expression of drug-resistantAurora A protected cells from MLN8237,providing strong support that Aurora A and notAurora B was the key antiproliferative targetin the clinic. This experiment also proves thatother potential off-targets outside of the Aurorafamily are not sufficient to block proliferation.

There are several reasons to use the chem-ical genetic approaches described above fortarget validation and biological discovery.Chemical genetics is often the first step inrational drug design, as one can assess theinfluence of modulating a particular target witha specific inhibitor without going through thestandard drug discovery process. Techniquessuch as RNAi, which diminish protein abun-dance, are often not predictive of therapeutics,which change protein activity (35).

Finding selective inhibitors is a dauntingchallenge, and nowhere is this more evidentthan in the protein kinase family. The problemwith genetic manipulation is that the changes

are irreversible. The problem with usingdrugs is that they can be nonspecific. Chemicalgenetics combines the best of both worlds—theselectivity of a genetic perturbation with thecontrol of a pharmacological agent. Manysignificant discoveries about the mechanismof protein function have been elucidated usingchemical genetics (see below).

Structural Bioinformatics–BasedDesign

Rational design of highly selective kinaseinhibitors is remarkably challenging, as the in-hibition of GSK-3β by lithium highlights. Onecould not predict from the structure of GSK3β

that it would contain a uniquely sensitive metalbinding site compared with other kinases.Despite this challenge, a highly selectiveinhibitor of the p90RSK kinases was discov-ered using a structural bioinformatics–basedapproach (36). A search was carried out lookingfor two selectivity filters of inhibitor bindingacross the mammalian kinome (depicted inFigure 5). Eleven of the 518 kinases withan active-site cysteine residue that could betargeted by an electrophile bearing inhibitorwere identified by sequence alignment. As asecond selectivity filter, a small Thr gatekeeperresidue was used, which reduced the 11 kinasesfurther to only the p90RSK1,2,4 enzymes. Byappending an electrophile onto a Thrgatekeeper–directed kinase scaffold (PP1), ahighly p90RSK selective inhibitor, FMK wasproduced. This combination of reversible andirreversible, cysteine-based inhibitor/kinaserecognition has also been exploited effectivelyin the epidermal growth factor receptor(EGFR) family of kinases (37). It is likely thatnonconserved cysteines in kinases will continueto provide important selectivity handles fornew inhibitors.

Structural Bioinformatics–BasedTargeting of Parasite Kinases

The chemical genetic approach for inhibi-tion of kinases, which relies on mutating the


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


gatekeeper residue in the kinase of interest toa glycine in the model organism, was never en-visioned as a potential drug discovery platformbecause of the absolute absence of gatekeeperGly kinases across the kinomes typically stud-ied. The rapid expansion of sequence informa-tion from diverse organisms recently revealedthat the food-borne parasite Toxoplasma gondiicontains a single kinase, TgCDPK1, with a nat-ural Gly gatekeeper (38). This essential kinasehas a naturally occurring enlarged ATP pocket,making the parasite uniquely sensitive to 3MB-PP1. To prove that TgCDPK1 was the singlekinase sensitive to 3MB-PP1, a TgCDPK1Gly->Met parasite was constructed. The mutantwas found to be resistant to killing by the com-pound, confirming the structural bioinformaticprediction of the 3MB-PP1 target within theparasite. This study may provide an entry pointinto developing 3MB-PP1-like compounds fortreatment of parasites that fortuitously containessential Gly gatekeeper kinases.

KINASE INHIBITORS THAT BINDOUTSIDE OF THE ATP POCKET

Protein-Based Naturally OccurringSelective Kinase Inhibitors

Despite the remarkable selectivity achieved bycertain ATP site inhibitors, the ultimate abilityto generate a selective inhibitor of every kinaseis limited by targeting this widely conservedsite. How then can small molecules achieveselective inhibition of each kinase? Are therenatural examples of highly selective inhibitors,and if so how do they target a single kinase? Theprotein substrate binding site of protein kinasesis by definition a site of variability betweenfamily members because kinases do not phos-phorylate a common substrate. Highly selectivekinase inhibitors based on the substrate peptidesequence provide important clues as to possibileinhibitors outside of the ATP pocket. Pseu-dosubstrate sequences with an Ala substitutedfor the phosphorylatable Ser serve as potentinhibitors of the kinase PKA (Figure 6a) (39).Regions outside of the phosphoacceptor site

also provide surfaces or pockets for recognitionas exemplified by the CDK-cyclin inhibitorp27Kip1 (40) and the EGFR inhibitor MIG6(41). A 69-amino acid region of p27 binds tothe CDK-cyclin complex and wraps around theN-terminal subdomain of CDK2 continuing tofold into the interface between the two kinasesubdomains until a tyrosine hydroxyl groupmakes H-bond contacts with ATP-interactingresidues (Figure 6b). Thus, p27 containsfeatures of ATP and non-ATP site inhibitors.

MIG6 is an endogenous regulator of theEGFR. This receptor and its related familymembers are critical targets in cancer therapy.Inhibitors, including Lapatinib and Iressa, thattarget the EGFR active site have found clinicalapplications in breast and nonsmall cell lungcancers (Table 1). Structural analysis of EGFRhas demonstrated that the fully ON form of theenzyme is more than a single molecule work-ing alone but rather consists of an asymmetricdimer, in which one protomer (the “activator”)induces an active conformation in the secondprotomer (42). The activator protomer utilizesa surface of its C-lobe to bind to the N-lobe ofthe second protomer and, in doing so, stabilizesa conformation in which the catalytic efficiencyof the second protomer is enhanced (Figure 6c)(42).

Through this unique assembly of kinase do-mains, the activator of the EGFR dimer playsa role that is strikingly similar to the cyclins,which are known kinase regulators that do notshow any structural similarity to a kinase do-main (42). Cyclins positively modulate the ac-tivity of their CDK partners upon binding tothe N-lobe interface. Thus cyclins and EGFRhave converged on a similar solution to mod-ulate positively the catalytic output of kinasedomain-binding partners.

The finding that one member of theactive EGFR dimer does not use catalyticfunction may explain the prevalence of thecatalytically inactive EGF family memberHER3. Interestingly, HER3 functions as apositive regulator of EGF signaling (43).More recently, catalytic activity for HER3 hasbeen demonstrated despite several active-site

780 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


PKA-ATP-PKI1ATP

CDK2-cyclin A-p27KIP1JSU

EGFR asymmetric dimer2GS6

EGFR-MIG62RF9

Activatorprotomer

c d

a b

Figure 6Examples of naturally occurring protein kinase inhibitors. (a) The PKA inhibitor PKI ( yellow spheres) bindsthe C-lobe of PKA to block substrate binding. (b) P27Kip ( yellow spheres) binds across the CDK-cyclincomplex and overlaps partially with the ATP-binding pocket. (c) The EGFR asymmetric dimer. In thiscomplex, one EGFR kinase domain ( gray semitransparent surface) binds and stabilizes an active conformationwithin a second EGFR kinase domain (multicolor ribbons) (PDB ID: 2GS6). (d ) The inhibitor MIG6 ( yellowspheres) blocks the formation of the EGFR dimer (PDB ID: 2RF9).

Pseudokinase:members of the kinasefamily in which one ormore of the conservedcatalytic residues arelacking

substitutions (44). However, the relevance ofthis activity still remains undetermined (45).Furthermore, HER3 belongs to a larger subsetof all protein kinases termed pseudokinases.It is estimated that approximately 10% of allhuman kinases are pseudokinases, as they have

evolved kinase functions that are independentof catalysis (46). Similar to HER3, pseudo-kinases may represent true conformationalswitches, with phosphorylation activity a ves-tigial remnant hidden within their biologicalfunction.


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Allosteric site:in kinases, a bindingpocket that does notoverlap with the ATPactive site

The finding that the endogenous EGFRinhibitor, MIG6, blocks the formation of theasymmetric dimer through competition andbinding to the C-lobe of the activator protomer(Figure 6d) further confirms the importanceof the asymmetric dimer in EGFR regulation.Additional studies will undoubtedly be aimedtoward generating synthetic modulators of theEGFR dimerization interface: The biologicalsolutions exemplified by MIG6 and HER3 setthe precedent for ones that can act as inhibitorsand others that may behave like activators.

These natural peptide–based examples ofhighly specific kinase inhibitors reveal newbinding pockets on kinases for inhibitor dis-covery. Despite these prevalent natural exam-ples, few small drug-like molecules have beenreported to bind to the kinase substrate site, andno molecules with crystallographic characteri-zation are known.

So far we have discussed the chemicaltractability of targeting the two substrate pock-ets found in all kinases. What about allostericsites outside of these substrate pockets? Kinasesserve as highly regulated machines that respondto multiple input stimuli through the regula-tion of cellular localization, monomer to dimertransitions, and various second-messengerinputs. Small molecules that bind to allostericpockets have tremendous potential for selec-tive kinase inhibition because these sites arehighly divergent across the kinome. Relativeto active-site binders, allosteric modulatorshave proven difficult to find. Most chemicallibraries do not contain mimics that match thediversity of biological inhibitors or activatorsfor kinases. Designing allosteric drugs can bevery challenging even when crystallographicdata are available, whereas the guidelinesfor making an ATP-competitive drug for akinase are much better understood. Despite thedifficulty in finding allosteric modulators, thereare many reasons why they could be extremelyuseful. By moving outside the highly conservedATP-binding pocket, target specificity imme-diately becomes much more facile to achieve.Perhaps more importantly, though, small

molecules that bind outside the ATP pocketcan differ from ATP-competitive drugs in howthey modulate a target and thus can revealnovel or unappreciated aspects about the targetand its biological function. The following threeexamples highlight small-molecule regulatorsthat bind the kinase domain outside the ATP-binding pocket. These small molecules openavenues for novel modes of kinase modulation.

An Allosteric Inhibitor Competesfor Myristate

While screening a diverse library of hetero-cyclic kinase inhibitors, researchers serendip-itously found a novel regulator for Bcr-Abl(47): GNF-2 was shown to compete with amyristate lipid binding site on the kinase do-main of Abl that is normally engaged throughan intramolecular interaction with cAbl’s ownmyristolated N terminus. In cell-culture mod-els, GNF-2 is an extremely selective inhibitorof cells expressing Bcr-Abl, but not of parentalcells lacking the oncogene. The crystal struc-ture of an Abl-GNF2-Imatinib complex con-firmed a GNF2 binding site within a myristate-binding pocket formed in part through a uniquestructural element, termed helix αI, foundwithin the Abl kinase domain (Figure 7a) (47).That helix αI forms an integral part of theGNF2 binding pocket could at least partiallyexplain the inhibitor’s exquisite selectivity.

In the context of full-length cAbl, engage-ment of the N-terminal myristoyl modificationwithin the kinase domain functions as a latchfor the assembly of the SH2-SH3 units on thekinase domain, maintaining the kinase in an in-active configuration. Remarkably, even in theabsence of the inhibitory myristoyl latch thatis missing from Bcr-Abl, GNF-2 binding inthe kinase domain modulates the conformationof the ATP-binding site within Abl. The con-nection between the myristate-binding pocketand ATP-binding site was further supportedby the finding that GNF2 binding within Ablcould increase the sensitivity of Abl to ATP-competitive inhibitors (47). The pathway for

782 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


coupling the distinct myristoyl- and ATP-binding pockets likely arose as part of Abl’s ownautoregulatory mechanism. GNF2 shows thatthis communication network can be co-optedthrough a small-molecule drug.

An Interdomain Inhibitor of AKT

The assay for identifying a modulator of a tar-get can be critically important. GNF2 wouldnot have been identified as a “hit” for Abl inthe same way as Imatinib if a recombinant formof the purified Abl kinase domain was utilizedin the initial screen. The same could be saidfor the allosteric inhibitor termed AKTi, whichwas originally identified by Merck scientists(48). This compound emerged from a libraryof 270,000 molecules as a potent inhibitor ofan AKT construct representing the native pro-tein and containing the kinase domain of AKTlinked directly to its regulatory PH domain.AKTi is essentially inactive against AKT frag-ments consisting of only the kinase domain,suggesting an absolute requirement for the PHdomain for AKTi inhibition. While this resultmay have suggested direct binding of AKTi tothe PH domain, more recent biophysical anal-ysis suggests AKTi engages a pocket on thekinase domain of AKT, but one that is regu-lated through an intramolecular PH domain—kinase domain interaction (49). More recentanalysis identified mutations at Trp80 withinAKT that block the inhibitory properties ofAKTi (50). Recently an improved analog withgreater specificity, MK2206, has been reported(51). Future studies will be required to showthe binding site for AKTi within a crystal-lized form of full-length AKT and to deter-mine the structural basis for how AKTi influ-ences AKT activation by upstream regulators,including PDK1.

Kinase Activation via Binding of thePIF Pocket

In the next example, a similar communicationpathway between an allosteric site and the ATP

cAbl:GNF-2 complex3K5V

PDK1:ATP:PS483HRF

GNF-2 PS48a b

HNN

N

Cl

CF3

CO2H

NH2

O

O

Myristate-bindingpocket

PIF-bindingpocket

Figure 7Synthetic modulators that bind within kinase regulatory sites. (a) GNF-2 incomplex with Abl (PDB ID 3K5). GNF-2 binds within a myristate-bindingpocket on the Abl kinase domain to inhibit the enzyme. (b) PDK1 in complexwith ATP and a hydrophobic motif analog PS48 (PDB ID 3HRF). PS48 bindswithin the PIF-binding pocket to activate PDK1.

pocket is modulated, but here, instead of re-sulting in kinase inhibition, the small moleculeactivates its kinase target.

Within the context of the cell, PDK1activation is complex and involves the action ofa number of upstream regulators. The processincludes localization to the cell membrane,binding to the PI3K product PIP3, and phos-phorylation of the activation loop and a secondregulatory site, termed the hydrophobic motif(HM). The HM resides outside the kinasedomain within a C-terminal extension andoften includes two phenylalanine residuesflanking the phospho-site. PDK1, like its ownsubstrate AKT and other members of theAGC kinase family, requires phosphorylation


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


within the HM for maximal kinase activity.Structural analysis of several AGC kinaseshas demonstrated that, upon phosphorylation,the HM docks back onto the kinase within aregion on the N-lobe termed the PIF pocket.In doing so, the HM stabilizes the activeconformation of PDK1. In some instances, thephosphorylated HM comes from the kinasein which it binds, providing a mechanism forkinase autoregulation. In other cases, the HMcomes from a substrate, providing a mechanismby which a downstream substrate may activateits upstream kinase.

Based on the understanding of HM regula-tion for AGC kinases, a computational dock-ing study was initiated leading to the identifica-tion of PS48, a small molecule with the featuresof the HM and containing two phenyl ringsbridged by a charged carboxylate functioning asa phospho-mimetic. A recent crystal structureof PS48 bound to PDK1 shows the compounddocked within the activating PIF pocket (52)(Figure 7b). In vitro PS48 upregulates PDK1activity to a level that is similar to a phosphory-lated HM, suggesting that the small molecule isa true mimic of PDK1’s own regulatory appara-tus. Furthermore, through mutational analysis,a residue involved in translating PS48 bindinginto PDK1 kinase activation was mapped.

It remains to be shown whether PS48 func-tions in cells. The charged carboxylate of PS48belies its use as a traditional cellular reagent.However, the discovery of PS48 could lead tofuture compounds or derivatives that may formthe basis for modulating AGC kinases withincells. Very recently a benzoazepoin-2-one wasreported as an allosteric regulator of PDK1,also targeting the PIF pocket (53). Reagentssuch as PS48 could provide a means to iden-tify non-PIF-dependent PDK1 substrates(presumably PS48 would activate PDK1 butcompete for PIF-binding substrates). Thesereagents would also allow activation of PDK1and possibly bypass the normal upstreamsignaling components. In doing so, it wouldbe possible to perturb PDK1 function acutely,while leaving other effectors of RTK-Ras-PI3Ksignaling silent.

IN VITRO INHIBITORS BECOMEACTIVATORS IN CELLS

In the above examples (summarized inFigure 8) it is perhaps not surprising that thesites exploited for synthetic ligands are withinpockets on the kinase domain that overlapwith regions shared by endogenous regulators.GNF2 and PS48 directly mimic the binding in-teractions of known inhibitors and activators,respectively. What is surprising and could nothave been predicted in advance are the small-molecule regulators that repurpose the activesite and switch on the signaling properties ofa kinase target. The following examples high-light cases in which an ATP-competitive kinaseinhibitor in vitro behaves as a positive regulatorof kinase function in vivo.

Bypassing the Kinase Activity toActivate an RNase Domain

The unfolded protein response (UPR) pathwayis an evolutionarily conserved stress responsepathway that matches protein synthesis withprotein-folding capacity in the endoplasmicreticulum (ER). The most ancient memberof the UPR pathway, Ire1, is a bifunctionalkinase-ribonuclease enzyme and a transmem-brane protein with an ER lumenal domain thatdirectly senses unfolded proteins. Activation ofthe unusual RNase domain in Ire1 initiates anonconventional splicing reaction of an mRNAcoding for the transcription factor Hac1 (Xbp1in mammals) to produce its mature protein andultimately the UPR. Efforts to study this non-conventional signaling pathway using chemicalgenetics and by mutation of the Ire1 gatekeeperresidue from Leu to Gly produced a 50% lossof Ire1 catalytic function in a cellular assay(54). Paradoxically, addition of the Ire1-Alamutant specific inhibitor, 1NM-PP1, rescuedthe lost Ire1 function. Why did addition of theIre1 inhibitor not block the UPR and furtherreduce Ire1 function? The loss of Ire1 functionfrom the gatekeeper mutation suggests thatthe catalytic activity of Ire1’s kinase domain isessential for UPR activation. How then could

784 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Pseudosubstrate inhibitors(e.g., PKI, p27KIP)

O

Abl inhibitor GNF2 PDK1 activator PS48 AKTi inhibitor

O

PHdo

mai

n

Type III inhibitors(e.g., PD318088)

Kinase-ATP substrate(uninhibited)

(+)

(–)

C-lobe

N-lobeH

ing

e αC

A loop

D of DFGNon-ATP competitvebinder

MgMATP

Substrate

Figure 8Summary of protein kinase modulators that bind outside the ATP pocket: representation of the kinasedomain with conserved structural elements highlighted, depicting the known mechanisms of non-ATPcompetitive kinase modulation. The non-ATP competitive binders are depicted in yellow.

a kinase inhibitor actually activate a kinasepathway?

Through a number of detailed experiments,the emergent model suggests that, in the caseof Ire1, activation of the RNase domain doesnot require kinase activity per se, but rather canbe achieved by simply binding an alternativeligand (an inhibitor) within the ATP pocket ofthe kinase domain (Figure 9a). Structural datasuggest that binding of the inhibitor triggersoligomerization of the receptor into a largeordered helix-like structure that juxtaposesIre1 kinases, thereby leading to activation ofthe RNase domain. Why is kinase activityretained through natural selection? If we pre-sume that ATP-site conformation controls theconformation of the kinase domain (evidencesupporting this comes from many active andinactive kinase crystal structures), then thecell needs some mechanism to control theaffinity of the Ire1 kinase for endogenousligands such as ADP and ATP (55). The Ire1kinase transautophosphorylates itself; this

could provide the necessary nucleotide-affinitytuning device. Thus, oligomerization by un-folded proteins could provide a mechanism fortransautophosphorylation, which then providesfor higher occupancy by ATP or ADP, leadingto RNAse domain activation. The high-affinity“inhibitor” of Ire1 provides a way to dissociatethe intricate mechanism of natural kinasefunction. The ability of an inhibitor to bypassIre1 function also provides a new appreciationof kinase functions that can be controlled bydeveloping ligands for the ATP pocket.

Ligand Occupancy of the ATP SiteTemplates PKC Maturation

A step of kinase regulation that is oftenoverlooked is maturation following synthesisand protein folding but prior to pathwaystimulation and full activation of the kinase.The kinase PKCε contains several phospho-rylation sites that are required for propermaturation of the enzyme, including the HM


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


RNase

Ire1

sig

nal

ing

Inhibitor

Akt

res

po

nse

InhibitorR

AF

kin

ase

acti

vity

Inhibitor

pAKT

AKTcat

PH d

omai

n

PH domain

PIP3

P

308

473

Membrane

Membrane

(+)

a Inhibitor binding activates a bifunctional enzyme

b Inhibitor binding induces kinase localization and hyperphosphorylation

c Inhibitor binding induces transactivation of kinase dimers

AKTinhibitorbinding

AKT membranelocalization andphosphorylation

Ire1 activationthrough

inhibitor binding

Ire1 oligomerizationRNase activation

Active RAF dimer:inhibitor bound

protomer activatespartner in trans

Inactive RAF dimer:kinase activity from

both protomersblocked

Inactive RAFmonomer

P

Sensor ofER stress

Kinaseinhibitor

Transducer

Figure 9ATP competitive inhibitors as kinase agonists. (b) The bifunctional kinase Ire1 contains a cytoplasmic kinase domain linked directly to aribonuclease (RNase) domain. Inhibitor binding within the kinase domain of Ire1 facilitates RNase activation. The dose dependency ofinhibitor displays a sigmoidal relationship with respect to Ire1 signaling. (b) The kinase AKT translocates to the membrane of cells tobind the PI3K product PIP3, leading to phosphorylation of two regulatory sites, T308 and S473, which are markers for AKTactivation. AKT hyperphosphorylation (pAKT) occurs in an inhibitor-dependent manner upon occupancy of the AKT active site, butleading to reciprocal inhibition of downstream phosphorylation through AKT catalysis (AKTcat). (c) The catalytic activity of BRAF istriggered upon binding of two BRAF molecules to form a dimeric structure. Small molecules that bind one BRAF molecule facilitateformation of the dimer, leading to transactivation of the second Raf molecule. High concentrations of inhibitor effectively blocksignaling from both Raf molecules. Raf displays a bell-shaped dependency on inhibitor concentration within cells, consistent with themodel shown to the left of the graph.

786 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


site S729, introduced previously in the contextof another AGC kinase, PDK1. In the processof studying a kinase-dead mutant of PKCε,K437M, researchers noticed that the mutantdid not mature analogously to the wild type asmeasured by the absence of phosphorylationof HM site S729 (56). This observation hadbeen previously attributed to the need fortrue kinase autophosphorylation activity. Yet,addition of a general PKC inhibitor related tostaurosporine, Bim 1, led to HM phosphoryla-tion, despite the lack of enzyme activity by theinhibitor-occupied kinase—thus challengingthe assumption that autophosphorylationactivity is required for PKC maturation. Useof an Ala gatekeeper mutant of PKCε and1NAPP1, allowing for specific inhibition of justPKCε, affirmed the finding with the kinase-dead K437M mutant. The common featureof K437M and M486A was the weakened nu-cleotide affinity of the mutants owing to partialdisruption of the ATP site. The Bim 1 or themore specific 1NAPP1 inhibitor apparently by-passes the need for nucleotide binding duringthe maturation of PKCε, which results in theenzyme adopting an appropriate conformationfor phosphorylation by priming kinases.

Inhibitor Hijacking of AKT Activation

Inhibitor studies with another AGC kinase,AKT, provided an additional example of howan inhibitor could induce an unexpected stateof the target kinase. A-443654, a very potentinhibitor of AKT developed at Abbott Labs,blocks phosphorylation in cells of the AKTsubstrate, GSK-3β (57). Whereas that resultwas expected, the kinase inhibitor also quiteunexpectedly induced hyperphosphorylationof the HM S473 and activation loop T308phosphorylation of AKT. A chemical geneticsystem for studying this paradoxical activationrevealed, similar to the case with PKCε, thatinhibitor binding served to “prime” AKT toadopt the activated state. Because A-443654 isan extremely nonselective inhibitor, a deriva-tized version of A-443654 (Pr-INDz) was

generated that does not bind to any wild-typekinases, but does inhibit Akt containing agatekeeper Ala mutation. In combination withanalog-sensitive alleles of AKT1, -2, and -3 ex-pressed heterologously within cells, Pr-INDzwas used to show that inhibitor binding to AKTinduced AKT’s own hyperphosphorylationwithin cells. The basis for inhibitor bindingin a target leading to hyperphosphorylationof the very same target could be distinguishedmechanistically from hyperphosphorylationresulting from inhibition of negative-feedbackregulators of AKT. The key experiment wasuse of a kinase-dead and analog-sensitive alleleof AKT that was receptive to inhibitor binding,but by virtue of the kinase-dead mutationcould not signal to any AKT substrates. Thedouble AKT mutant was similarly found tobecome hyperphosphorylated upon addition ofPr-INDz. The experiment effectively ruled outcoincidental inhibition of targets other thanAKT or feedback upregulation as the basisfor AKT hyperphosphorylation. The chemicalgenetic studies suggested that A-443654 likelyalso leads to upregulation of endogenousAKT phosphorylation through a similarmechanism.

In the case of inhibitor activation of AKTpriming, the kinases responsible for AKT phos-phorylation are localized at the cell membrane,whereas the unprimed AKT kinase is cytoso-lic. During normal growth-factor stimulation,the cytosolic AKT is localized to the cell mem-brane via its PH-domain binding of PIP3, pro-duced by PI3K. Yet, how did AKT come to bephosphorylated by membrane-bound kinasesTORC2 and PDK1, even in the absence ofPI3K activation and growth factor stimulation?Immunofluorescence studies of AKT follow-ing A-443654 treatment revealed that the in-hibitor induced AKT membrane localizationand that addition of a PI3K inhibitor abrogatedthis drug-induced membrane localization (58).Thus, the inhibitor of AKT induced a rear-rangement of the AKT PH domain to increaseits affinity sufficiently so as to bind the low basallevel of PIP3 at the membrane (Figure 9b).


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


This form of inhibitor-induced AKT primingwas referred to as inhibitor hijacking to denotethe ability of the inhibitor to bypass the needfor complex growth-factor activation of multi-ple enzymes to produce doubly phosphorylatedAKT308/473PP.

In the PKCε and AKT cases, the binding ofa ligand to the ATP site leads to hyperphospho-rylation of the kinases on the natural primingsites necessary for protein stability and activity.The biochemical basis for such an effect couldcome from enhancement of upstream kinaseactivity or a protection from phosphatase-mediated dephosphorylation. Support for thelatter mechanism comes from in vitro studiesof dephosphorylation of PKCδ where the pres-ence of ATP-Mg2+ greatly suppressed proteinphosphatase 2A–mediated dephosphorylation(59). Further support for this mechanism comesfrom the rapid loss of AKT hyperphospho-rylation upon inhibitor wash-out from cells.Thus, both in vitro and in vivo experimentsseem to suggest ligand binding in the ATPpocket protects kinase phosphorylation sitesfrom dephosphorylation (58). The interplayof phosphatases and kinases in controlling thesteady-state level of specific phosphorylationevents is necessary to truly understand the basesfor many highly regulated phosphorylationevents in cells. The absence of specific phos-phatase inhibitors makes these studies currentlydifficult.

The preceding examples of ligand-basedmodulation of Ire1-RNase activity andPKC/AKT drug-induced hyperphosphoryla-tion have one important feature in common:The kinase ligand inhibits kinase activity. Thisseems completely self-evident because the lig-and competes for the ATP site, preventing thephosphodonor from binding to the active site.In fact, it seems impossible to imagine any casein which an ATP-competitive kinase inhibitorcould enhance substrate phosphorylation. Anunexpected observation from cellular studieswith RAF kinase inhibitors revealed this couldactually be occurring, although determinationof the mechanism required almost 10 years tobe uncovered (60).

A RAF Inhibitor That StimulatesRAF Activity

RAF phosphorylates and activates MEK. RAFhas been an attractive and important drugtarget because it is mutated in 6% of all cancersand is immediately downstream of K-RAS,the most frequently mutated oncogene incancer. Many different chemical series ofRAF inhibitors have been developed with aclinically approved drug sorafenib (Nexavar).In addition, a preclinical compound, PLX4720,provides good benchmarks. The three isoformsof RAF are A-, B-, and C-RAF, which are alltargeted by current RAF inhibitors. Cancermutations occur in B-RAF; V600E is thepredominant form in melanoma. In V600EB-RAF mutant cells, PLX4720 is a potentinhibitor of MEK phosphorylation. As aresult of this potent inhibition, a clinical RAFinhibitor related in structure to PLX4720,PLX4032, has shown tumor regression in90% of patients with BRAF(V600E) mutation(61). In studies of cells with other pathway-activating mutations, such as K-RASG12V,but with wild-type BRAF kinase, the effect ofthe Raf inhibitors is paradoxically to activateMEK phosphorylation. This is an intriguingobservation first made in 1999 by Hall-Jacksonet al. (60), although at that time no mechanismwas uncovered. Strikingly, during the clinicaltrial of PLX4032, approximately one-third ofmelanoma patients treated at the maximumtolerated dose of PLX4032 developed a differ-ent cancer, squamous cell carcinoma (61). Themechanistic basis for RAF-inhibitor-inducedMEK phosphorylation has implications forunderstanding not only RAS-RAF-MEK sig-naling, but also the drug-induced developmentof secondary cancer in patients.

The key observation relating to RAF-inhibitor-induced MEK phosphorylation wasthe bell-shaped dose-response curve, such thatmaximal stimulation occurred at intermediatedoses and the activation could be suppressedat high dose (62, 63). A model consistent withthese data is one in which the inhibitor bindingto a RAF monomer induces RAF dimerization

788 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


(Figure 9c). In the dimeric state, with oneRAF protomer containing the inhibitor andthe other protomer remaining inhibitor free,the RAF kinase activity is significantly above thelevel achieved by the monomeric form of thekinase. At the high level of inhibitor concentra-tion when both protomers of the RAF dimer areoccupied by inhibitor, the activity is completelyinhibited, which would explain the high-dosesuppression of activation. Other models thatcould explain the phenomenon involve negativefeedback between RAF isoforms (64). If BRAFactivity negatively regulates CRAF phosphory-lation of MEK, then a BRAF-selective inhibitorwould lead to activation of CRAF and an in-crease in MEK phosphorylation.

To distinguish between these models, itwas necessary to develop completely isoform-selective inhibitors of RAF using chemicalgenetics, because the majority of RAF in-hibitors were nonselective between B- andC-RAF isoforms. By introduction of a non-conserved Cys residue into the CRAF kinaseit was possible to develop a selective inhibitorof just CRAFS248C (62). To distinguishbetween models in which a negative-feedbackloop existed and one in which the inhibitorinduced dimerization and activation of theprimary target of the inhibitor, a kinase-deadversion of CRAF was used. Cells expressing akinase-dead/inhibitor-sensitive allele of CRAF(S248C/D486N) showed no MEK phospho-rylation until the addition of an allele-specificinhibitor induced dimerization with wild-typeCRAF (drug insensitive) and increased MEKphosphorylation (62). As proof of the require-ment for dimerization, a mutation known todisrupt RAF-RAF dimerization, R401A, wasshown to prevent the drug-induced activationof MEK phosphorylation in this system. Thesestudies suggest that a small-molecule inhibitorof a kinase that is regulated by a monomer-dimer transition can actually be activated bythe inhibitor (62). Specifically for the RAF in-hibitor case, these studies highlight the strikingdifference in signaling between a BRAFV600Ecell (MEK inhibited) versus a wild-type BRAFcell (MEK activated) by the BRAF inhibitors.

The clinical implications of these findings forpatients with KRAS- or BRAF-driven tumorswill produce one of the major areas of clinicaland basic research in molecularly targetedtherapy in the coming decade.

Kinase-Inactivating Mutationsand Cancer

If kinase inhibitors can induce signaling byinducing conformational states that transacti-vate inhibitor-free kinase molecules in cells, itseems plausible that some somatic mutations inkinases driving cancer may paradoxically inacti-vate the catalytic activity of a kinase. In kinome-wide unbiased screens for kinase mutations incancer, a large number of such inactivating mu-tations have been uncovered (Figure 10) (65).For example, in the mutant forms of DAPK3,HCK, and LYN, the conserved Mg2+-bindingaspartate that is required for catalysis hasbeen found mutated to residues that are notnegatively charged and thus cannot chelate theessential Mg2+. This has also been observed inBRAF where kinase-inactivating mutations areknown to be driver mutations and inactivate thesame aspartate BRAF D593V (66). Much workstill needs to be done to demonstrate that suchsomatic mutants are indeed driver mutationsand not merely passengers induced by genomicinstability with no benefit to the cancer cell.

A challenge for drug discovery is under-standing how to create a drug for an inactivekinase that may cause cancer. At first glance,it makes no sense to inhibit a dead enzyme tostop signal transduction. However, if the mu-tationally inactivated kinase signals by bindingto an active wild-type kinase that carries outthe catalytic function, then two drug strategiesemerge. First, an inhibitor of the mutation-ally activated kinase will also inhibit the wild-type kinase, thus blocking the pathway. In thiscase, the binding of the inhibitor to the inac-tive mutant kinase is not relevant to the ac-tivity of the inhibitor. Second, an inhibitor ofthe mutant (dead) kinase may be designed tolock the kinase in a state that is not compatiblewith transactivation of the wild-type enzyme.


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


BR

AF

RIGSGSFGTV

G4

69

AIn

act

ive

MLC

KVLGGGRFGQV

G6

01

EIn

act

ive

AU

RK

CPLGKGKFGNV

G1

8E

Ina

ctiv

eM

AP

3K

10

IIGVGGFGKV

G1

07

EIn

act

ive

ST

K3

8L

VIGRGAFGEV

G9

9A

Ina

ctiv

eS

TK

32

BAIGKGSFGKV

G3

5E

Ina

ctiv

eM

ST

4RIGKGSFGEV

G3

6W

Ina

ctiv

eGxGxxG

Act

ivat

ion

seg

men

t mu

tati

on

s

BR

AF

DFGLATVK-----------SRWSGSHQ--FEQL--SG---SILWMAPE

D5

94

VIn

act

ive

DA

PK

3DFGIAH-------------KIEAGN-----EFKNIFG---TPEFVAPE

D1

61

NIn

act

ive

HC

KDFGLAR-------------VIEDN------EYTAREGAKFPIKWTAPE

D3

78

GIn

act

ive

LYN

DFGLAR-------------VIEDN------EYTAREGAKFPIKWTAPE

D3

85

YIn

act

ive

FY

NDFGLAR-------------LIEDN------EYTARQGAKFPIKWTAPE

G4

10

RIn

act

ive

EP

HA

3DFGLSR-------------VLEDDPEA---AYTTR-GGKIPIRWTSPE

G7

66

EIn

act

ive

MA

PK

8DFGLART------------AGTSFMMT---PYVVT------RYYRAPE

G1

71

S

Ina

ctiv

eC

DK

11

DMGFARLF-----------NSPL

KPLADLDPVVVT------FWYRAPE

G1

75

SIn

act

ive

MA

ST

20

5DFGLSKMGLMSLTTNLYEGHIEKDARE--FLDKQVCG---TPEYIAPE

G6

55

AIn

act

ive

Sg

K4

94

DFGLSR-------------HVPQGA-----QAYTICG---TLQYMAPE

R2

91

C?

CD

KL2

DFGFARTL-----------AAPGEVYT---DYVAT------RWYRAPE

R1

49

Q?

PS

KH

2DFGLA-----------YSGKKSGDW-----TMKTLCG---TPEYIAPE

K2

12

I?

PR

KC

B1

DFGMCK-----------E-NIWDGV-----TTKTFCG---TPDYIAPE

V4

96

M?

CD

K8

DMGFARLF-----------NSPLKPLADLDPVVVT------FWYRAPE

D1

89

N?

KS

R2

DFGLFSISG----------VLQAGRRE--DKLRIQNG---WLCHLAPE

R8

55

H?

FG

FR

3DFGLAR-------------DVH-NLD----YYKKTTNGRLPVKWMAPE

K6

50

E?

DY

RK

1B

DFGSSC-------------QLGQRIY----QYIQS------RFYRSPE

Q2

75

R?

PA

K7

DFGFCA-------------QVSKEVP----KRKSLVG---TPYWMAPE

V6

04

I?

EP

HA

2DFGLSR-------------VLEDDPEA---TYTTS-GGKIPIRWTAPE

G7

77

S?

FLT

1DFGLAR-------------DIYKNPD----YVRKGDT-RLPLKWMAPE

L10

61

V?

PA

K3

DFGFCA-------------QITPEQS----KRSTMVG---TPY

WMAPE

T4

25

S?

FG

FR

1DFGLAR-------------DIH-HID----YYKKTTNGRLPVKWMAPE

V6

64

L?

SG

K2

DFGLCK-----------EG-VEPED-----TTSTFCG---TPEYLAPE

E2

59

K?

NT

RK

3DFGMSR-------------DVY-STD----YYRVGGHTMLPIRWMPPE

R7

21

F?

We

e1

bDLGHAT-------------SINKP------KVEEG-----DSRFLANE

R3

98

H?

PT

K2

DFGLSR-------------YMEDS------TYYKASKGKLPIKWMAPE

A6

12

V?

ALS

2C

R7

DFGLARAK-----------SIPSQTYS--SEVVTLWYRPPDALLGATE

E2

25

D?

DFG

APE

P-l

oo

p m

uta

tio

ns

Mu

tati

on

Pre

dic

ted

cat

alyt

ic a

ctiv

ity

Mu

tati

on

Pre

dic

ted

act

ivit

y

Figu

re10

An

alig

nmen

tofk

inas

esfo

und

mut

ated

inva

riou

sca

ncer

s.So

mat

icm

utat

ions

foun

dw

ithin

prot

ein

kina

ses

from

dive

rse

hum

anca

ncer

sar

elis

ted

toth

eri

ghto

fthe

sequ

ence

s.Fo

rB

RA

F,in

whi

chm

ultip

lem

utat

ions

have

been

foun

d,th

elis

ted

mut

antr

efer

sto

asu

bstit

utio

nat

the

first

posi

tion

high

light

ed.M

utat

ions

pred

icte

dto

dim

inis

hki

nase

cata

lytic

func

tion

are

liste

d.T

heP

loop

,whi

chis

impo

rtan

tfor

the

coor

dina

tion

ofA

TP

phos

phat

esan

dca

taly

sis

(con

sens

usG

XG

XX

G),

and

the

activ

atio

nse

gmen

t,w

hich

incl

udes

Mg-

coor

dina

ting

resi

dues

and

span

sth

eD

FGan

dA

PE

mot

ifs,a

resh

own.

Yel

low

high

light

sm

utat

ions

atre

sidu

esth

atar

eco

nser

ved

and

requ

ired

for

phos

phor

ylat

ion.

Blu

ehi

ghlig

hts

mut

atio

nsat

nonc

onse

rved

resi

dues

.The

figur

eis

adap

ted

bype

rmis

sion

from

Mac

mill

anP

ublis

hers

Ltd

.,N

atur

e,co

pyri

ght2

007

(65)

.

790 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Agonist: kinase ligandthat activates kinaseactivity

Antagonist: kinaseinhibitor that blockskinase activity

This is a plausible strategy using the varioustype II and type III kinase inhibitors. Impor-tantly, such conformation-specific kinase in-hibitors were originally designed to overcomethe challenge of achieving high selectivity, butnow they may be used as specific inducers ofinactive conformations in their targets. Theuse of such different classes of inhibitors to in-duce different cellular conformations of kinaseswould be predicted to have profoundly differentsignaling outputs. The broader implication ofthese principles is the transition in thinking ofkinase inhibitors as pure antagonists of kinasesand the appreciation that there can be ATP-sitebinders that are agonists of kinase signaling. Byapplying such principles to kinase drug discov-ery there may be new opportunities for devel-oping drugs that have improved activity and en-hanced therapeutic indices for treating a widevariety of diseases in which kinases are involved.

It is intriguing that a large proportion ofmutant kinases identified in cancer genomes,as well as many naturally occurring kinases,lack essential conserved catalytic residues. Theoccurrence of these mutations point to theATP pocket as an important site of regula-tion. Future efforts will involve investigatingthe properties of these atypical mutant kinasesin greater detail (67). Are they truly inactive?Do they bind ATP? Have they evolved distinctmechanisms for catalyzing phospho-transfer?Can their function be manipulated with small-molecule ligands? Recent structural analyses ofpseudokinases have revealed intriguing modesof regulation (45, 68–70) and catalysis (44, 71).As putative regulators of cell signaling networks(72, 73), chemical discovery efforts for mu-tant pseudokinases promise to be rich with new

breakthroughs in this burgeoning area of kinaseresearch.

SUMMARY

The high frequency of kinase mutations in driv-ing various cancers and the prominence of ki-nases in immune cell signaling have broughtkinases to the front of drug discovery effortsin cancer and autoimmune disease treatments.This attention has brought a wealth of chemi-cal discovery yielding highly diverse drug-likeligands that target multiple conformations andsurfaces of kinases. As these inhibitors are usedin cells, model organisms, and patients, re-searchers have gained a wealth of informationabout kinase signaling that was largely invisi-ble to genetic approaches for the study of ki-nases. As the field progresses we will likely seeadvances in the exploitation of conformation-specific inhibitors and paradoxical activators ofkinase signaling to tailor new medicines. Theremay be a limit to the number of single-kinasedrug targets that can be effectively inhibited toachieve patient responses. The highly degen-erate and adaptable kinase networks in diseasecells pose a challenge to highly selective in-hibitors because cells can dynamically adapt tothe presence of such molecules. Another fron-tier for kinase inhibitor development is ratio-nal polypharmacology in which small moleculeswith specific combinations of protein kinasetargets or protein and lipid kinase targets aredeveloped (74, 75). Such designed moleculesmay exhibit enhanced efficacy and less toxicityif they match the spectrum of kinases involvedin disease cells and minimally perturb normalcell signaling.

SUMMARY POINTS

1. Protein kinases are the critical drivers of many diseases and represent a large class of drugtargets.

2. Small-molecule inhibitors of kinases have diverse structures and bases for selectiveinhibition.


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


3. ATP-binding inhibitors can be made selective on the basis of nonconserved features ofeven closely related kinases.

4. The gatekeeper residue is a central determinant of kinase inhibitor selectivity.

5. Kinase inhibitors can paradoxically activate kinases.

6. Kinase inhibitors can activate signaling pathways.

7. Mutations that block kinase catalytic function can trigger signaling pathways in disease.Many kinase-inactivating mutations have been found in cancers and may represent afuture class of drug targets.

FUTURE ISSUES

1. Researchers must determine how to design selective kinase inhibitors, inhibitors thatbind outside of the ATP site, and inhibitors of mutant kinases that are inactive enzymes.

2. The repertoire of kinase agonists and antagonists for interrogating biological pathwaysand for generating novel therapeutics must be expanded.

3. New methods in which to assay catalytic and conformational protein kinase propertiesso that novel modulators can be identified need to be developed.

4. An entire class of kinase inhibitors not discussed in this review include those that de-rive their efficacy from simultaneously targeting multiple kinases—these have their ownchallenges but exhibit “emergent” properties in their ability to perturb multiple kinasepathways.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings thatmight be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

We thank Nick Hertz, Jonathan Ostrem, and Chao Zhang for helpful comments on themanuscript. We also apologize for any omissions in referencing the many important contributionsto the kinase inhibitor field.

LITERATURE CITED

1. Cohen P. 2002. Protein kinases—the major drug targets of the twenty-first century? Nat. Rev. Drug Discov.1:309–15

2. Liu Y, Gray NS. 2006. Rational design of inhibitors that bind to inactive kinase conformations. Nat. Chem.Biol. 2:358–64

3. Shen K, Hines AC, Schwarzer D, Pickin KA, Cole PA. 2005. Protein kinase structure and function analysiswith chemical tools. Biochim. Biophys. Acta 1754:65–78

4. Tamaoki T, Nomoto H, Takahashi I, Kato Y, Morimoto M, Tomita F. 1986. Staurosporine, a potentinhibitor of phospholipid/Ca++-dependent protein kinase. Biochem. Biophys. Res. Commun. 135:397–402

792 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


5. Karaman MW, Herrgard S, Treiber DK, Gallant P, Atteridge CE, et al. 2008. A quantitative analysis ofkinase inhibitor selectivity. Nat. Biotechnol. 26:127–32

6. Lawrie AM, Noble ME, Tunnah P, Brown NR, Johnson LN, Endicott JA. 1997. Protein kinase inhibitionby staurosporine revealed in details of the molecular interaction with CDK2. Nat. Struct. Biol. 4:796–801

7. Graziani Y, Erikson E, Erikson RL. 1983. The effect of quercetin on the phosphorylation activity of therous sarcoma virus transforming gene product in vitro and in vivo. Eur. J. Biochem. 135:583–89

8. Lyall RM, Zilberstein A, Gazit A, Gilon C, Levitzki A, Schlessinger J. 1989. Tyrphostins inhibit epidermalgrowth factor (EGF)-receptor tyrosine kinase activity in living cells and EGF-stimulated cell proliferation.J. Biol. Chem. 264:14503–9

9. Lee JC, Laydon JT, McDonnell PC, Gallagher TF, Kumar S, et al. 1994. A protein kinase involved inthe regulation of inflammatory cytokine biosynthesis. Nature 372:739–45

10. Eyers PA, Craxton M, Morrice N, Cohen P, Goedert M. 1998. Conversion of SB 203580-insensitive MAPkinase family members to drug-sensitive forms by a single amino acid substitution. Chem. Biol. 5:321–28

11. Cuenda A, Rouse J, Doza YN, Meier R, Cohen P, et al. 1995. SB 203580 is a specific inhibitor of a MAPkinase homologue which is stimulated by cellular stresses and interleukin-1. FEBS Lett. 364:229–33

12. Hanke JH, Gardner JP, Dow RL, Changelian PS, Brissette WH, et al. 1996. Discovery of a novel, potent,and Src family-selective tyrosine kinase inhibitor. J. Biol. Chem. 271:695–701

13. Fry DW, Kraker AJ, McMichael A, Ambroso LA, Nelson JM, et al. 1994. A specific inhibitor of theepidermal growth factor receptor tyrosine kinase. Science 265:1093–95

14. Dudley DT, Pang L, Decker SJ, Bridges AJ, Saltiel AR. 1995. A synthetic inhibitor of the mitogen-activated protein kinase cascade. Proc. Natl. Acad. Sci. USA 92:7686–89

15. Alessi DR, Cuenda A, Cohen P, Dudley DT, Saltiel AR. 1995. PD 098059 is a specific inhibitor of theactivation of mitogen-activated protein kinase kinase in vitro and in vivo. J. Biol. Chem. 270:27489–94

16. Ohren JF, Chen H, Pavlovsky A, Whitehead C, Zhang E, et al. 2004. Structures of human MAP kinasekinase 1 (MEK1) and MEK2 describe novel noncompetitive kinase inhibition. Nat. Struct. Mol. Biol.11:1192–97

17. Druker BJ, Tamura S, Buchdunger E, Ohno S, Segal GM, et al. 1996. Effects of a selective inhibitor ofthe Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat. Med. 2:561–66

18. Doyle DA, Morais Cabral J, Pfuetzner RA, Kuo A, Gulbis JM, et al. 1998. The structure of the potassiumchannel: molecular basis of K+ conduction and selectivity. Science 280:69–77

19. Schindler T, Bornmann W, Pellicena P, Miller WT, Clarkson B, Kuriyan J. 2000. Structural mechanismfor STI-571 inhibition of abelson tyrosine kinase. Science 289:1938–42

20. Knight ZA, Gonzalez B, Feldman ME, Zunder ER, Goldenberg DD, et al. 2006. A pharmacological mapof the PI3-K family defines a role for p110alpha in insulin signaling. Cell 125:733–47

21. Seeliger MA, Nagar B, Frank F, Cao X, Henderson MN, Kuriyan J. 2007. c-Src binds to the cancer drugimatinib with an inactive Abl/c-Kit conformation and a distributed thermodynamic penalty. Structure15:299–311

22. Dar AC, Lopez MS, Shokat KM. 2008. Small molecule recognition of c-Src via the Imatinib-bindingconformation. Chem. Biol. 15:1015–22

23. Seeliger MA, Ranjitkar P, Kasap C, Shan Y, Shaw DE, et al. 2009. Equally potent inhibition of c-Src andAbl by compounds that recognize inactive kinase conformations. Cancer Res. 69:2384–92

24. Simard JR, Kluter S, Grutter C, Getlik M, Rabiller M, et al. 2009. A new screening assay for allostericinhibitors of cSrc. Nat. Chem. Biol. 5:394–96

25. Hubbard SR. 2002. Protein tyrosine kinases: autoregulation and small-molecule inhibition. Curr. Opin.Struct. Biol. 12:735–41

26. Klein PS, Melton DA. 1996. A molecular mechanism for the effect of lithium on development. Proc. Natl.Acad. Sci. USA 93:8455–59

27. Ryves WJ, Dajani R, Pearl L, Harwood AJ. 2002. Glycogen synthase kinase-3 inhibition by lithium andberyllium suggests the presence of two magnesium binding sites. Biochem. Biophys. Res. Commun. 290:967–72

28. Bain J, Plater L, Elliott M, Shpiro N, Hastie CJ, et al. 2007. The selectivity of protein kinase inhibitors:a further update. Biochem. J. 408:297–315


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


29. Williams RS, Cheng L, Mudge AW, Harwood AJ. 2002. A common mechanism of action for threemood-stabilizing drugs. Nature 417:292–95

30. King J, Keim M, Teo R, Weening KE, Kapur M, et al. Genetic control of lithium sensitivity and regulationof inositol biosynthetic genes. PLoS ONE 5:e11151

31. Bishop AC, Ubersax JA, Petsch DT, Matheos DP, Gray NS, et al. 2000. A chemical switch for inhibitor-sensitive alleles of any protein kinase. Nature 407:395–401

32. Eyers PA, Van Den IP, Quinlan RA, Goedert M, Cohen P. 1999. Use of a drug-resistant mutant of stress-activated protein kinase 2a/p38 to validate the in vivo specificity of SB 203580. FEBS Lett. 451:191–96

33. Scutt PJ, Chu ML, Sloane DA, Cherry M, Bignell CR, et al. 2009. Discovery and exploitation of inhibitor-resistant aurora and polo kinase mutants for the analysis of mitotic networks. J. Biol. Chem. 284:15880–93

34. Sloane DA, Trikic MZ, Chu ML, Lamers MB, Mason CS, et al. Drug-resistant aurora A mutants forcellular target validation of the small molecule kinase inhibitors MLN8054 and MLN8237. ACS Chem.Biol. 5:563–76

35. Knight ZA, Shokat KM. 2007. Chemical genetics: where genetics and pharmacology meet. Cell 128:425–30

36. Cohen MS, Zhang C, Shokat KM, Taunton J. 2005. Structural bioinformatics-based design of selective,irreversible kinase inhibitors. Science 308:1318–21

37. Fry DW, Bridges AJ, Denny WA, Doherty A, Greis KD, et al. 1998. Specific, irreversible inactivation ofthe epidermal growth factor receptor and erbB2, by a new class of tyrosine kinase inhibitor. Proc. Natl.Acad. Sci. USA 95:12022–27

38. Lourido S, Shuman J, Zhang C, Shokat KM, Hui R, Sibley LD. 2010. Calcium-dependent protein kinase1 is an essential regulator of exocytosis in Toxoplasma. Nature 465:359–62

39. Knighton DR, Zheng JH, Ten-Eyck LF, Ashford VA, Zuong NH, et al. 1991. Crystal structure of thecatalytic subunit of cyclic adenosine monophosphate regulated protein kinase. Science 253:407–14

40. Russo AA, Jeffrey PD, Patten AK, Massague J, Pavletich NP. 1996. Crystal structure of the p27Kip1cyclin-dependent-kinase inhibitor bound to the cyclin A-Cdk2 complex. Nature 382:325–31

41. Zhang X, Pickin KA, Bose R, Jura N, Cole PA, Kuriyan J. 2007. Inhibition of the EGF receptor by bindingof MIG6 to an activating kinase domain interface. Nature 450:741–44

42. Zhang X, Gureasko J, Shen K, Cole PA, Kuriyan J. 2006. An allosteric mechanism for activation of thekinase domain of epidermal growth factor receptor. Cell 125:1137–49

43. Riese DJ 2nd, van Raaij TM, Plowman GD, Andrews GC, Stern DF. 1995. The cellular response toneuregulins is governed by complex interactions of the erbB receptor family. Mol. Cell Biol. 15:5770–6

44. Shi F, Telesco SE, Liu Y, Radhakrishnan R, Lemmon MA. ErbB3/HER3 intracellular domain is compe-tent to bind ATP and catalyze autophosphorylation. Proc. Natl. Acad. Sci. USA 107:7692–97

45. Jura N, Shan Y, Cao X, Shaw DE, Kuriyan J. 2009. Structural analysis of the catalytically inactive kinasedomain of the human EGF receptor 3. Proc. Natl. Acad. Sci. USA 106:21608–13

46. Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. 2002. The protein kinase complement ofthe human genome. Science 298:1912–34

47. Zhang J, Adrian FJ, Jahnke W, Cowan-Jacob SW, Li AG, et al. 2010. Targeting Bcr-Abl by combiningallosteric with ATP-binding-site inhibitors. Nature 463:501–6

48. Barnett SF, Defeo-Jones D, Fu S, Hancock PJ, Haskell KM, et al. 2005. Identification and characterizationof pleckstrin-homology-domain-dependent and isoenzyme-specific Akt inhibitors. Biochem. J. 385:399–408

49. Calleja V, Laguerre M, Parker PJ, Larijani B. 2009. Role of a novel PH-kinase domain interface inPKB/Akt regulation: structural mechanism for allosteric inhibition. PLoS Biol. 7:e17

50. Green CJ, Goransson O, Kular GS, Leslie NR, Gray A, et al. 2008. Use of Akt inhibitor and a drug-resistant mutant validates a critical role for protein kinase B/Akt in the insulin-dependent regulation ofglucose and system A amino acid uptake. J. Biol. Chem. 283:27653–67

51. Liu P, Cheng H, Roberts TM, Zhao JJ. 2009. Targeting the phosphoinositide 3-kinase pathway in cancer.Nat. Rev. Drug Discov. 8:627–44

52. Hindie V, Stroba A, Zhang H, Lopez-Garcia LA, Idrissova L, et al. 2009. Structure and allosteric effectsof low-molecular-weight activators on the protein kinase PDK1. Nat. Chem. Biol. 5:758–64

794 Dar · Shokat

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


53. Wei L, Gao X, Warne R, Hao X, Bussiere D, et al. 2010. Design and synthesis of benzoazepin-2-oneanalogs as allosteric binders targeting the PIF pocket of PDK1. Bioorg. Med. Chem. Lett. 20:3897–902

54. Papa FR, Zhang C, Shokat K, Walter P. 2003. Bypassing a kinase activity with an ATP-competitive drug.Science 302:1533–37

55. Korennykh AV, Egea PF, Korostelev AA, Finer-Moore J, Zhang C, et al. 2009. The unfolded proteinresponse signals through high-order assembly of Ire1. Nature 457:687–93

56. Cameron AJM, Escribano C, Saurin AT, Kostelecky B, Parker PJ. 2009. PKC maturation is promoted bynucleotide pocket occupation independently of intrinsic kinase activity. Nat. Struct. Mol. Biol. 16:624–30

57. Luo Y, Shoemaker AR, Liu X, Woods KW, Thomas SA, et al. 2005. Potent and selective inhibitors ofAkt kinases slow the progress of tumors in vivo. Mol. Cancer Ther. 4:977–86

58. Okuzumi T, Fiedler D, Zhang C, Gray DC, Aizenstein B, et al. 2009. Inhibitor hijacking of Akt activation.Nat. Chem. Biol. 5:484–93

59. Srivastava J, Procyk KJ, Iturrioz X, Parker PJ. 2002. Phosphorylation is required for PMA- and cell-cycle-induced degradation of protein kinase Cdelta. Biochem. J. 368:349–55

60. Hall-Jackson CA, Eyers PA, Cohen P, Goedert M, Boyle FT, et al. 1999. Paradoxical activation of Rafby a novel Raf inhibitor. Chem. Biol. 6:559–68

61. Flaherty K, Puzanov I, Sosman J, Kim K, Ribas A, et al. 2009. Phase I study of PLX4032: proof of conceptfor V600E BRAF mutation as a therapeutic target in human cancer. J. Clin. Oncol. 27:15s

62. Poulikakos PI, Zhang C, Bollag G, Shokat KM, Rosen N. 2010. RAF inhibitors transactivate RAF dimersand ERK signalling in cells with wild-type BRAF. Nature 464:427–30

63. Hatzivassiliou G, Song K, Yen I, Brandhuber BJ, Anderson DJ, et al. 2010. RAF inhibitors prime wild-typeRAF to activate the MAPK pathway and enhance growth. Nature 464:431–35

64. Heidorn SJ, Milagre C, Whittaker S, Nourry A, Niculescu-Duvas I, et al. 2010. Kinase-dead BRAF andoncogenic RAS cooperate to drive tumor progression through CRAF. Cell 140:209–21

65. Greenman C, Stephens P, Smith R, Dalgliesh GL, Hunter C, et al. 2007. Patterns of somatic mutationin human cancer genomes. Nature 446:153–58

66. Wan PTC, Garnett MJ, Roe SM, Lee S, Niculescu-Duvaz D, et al. 2004. Mechanism of activation of theRAF-ERK signaling pathway by oncogenic mutations of B-RAF. Cell 116:855–67

67. Kornev AP, Taylor SS. 2009. Pseudokinases: functional insights gleaned from structure. Structure 17:5–768. Fukuda K, Gupta S, Chen K, Wu C, Qin J. 2009. The pseudoactive site of ILK is essential for its binding

to alpha-Parvin and localization to focal adhesions. Mol. Cell 36:819–3069. Scheeff ED, Eswaran J, Bunkoczi G, Knapp S, Manning G. 2009. Structure of the pseudokinase VRK3

reveals a degraded catalytic site, a highly conserved kinase fold, and a putative regulatory binding site.Structure 17:128–38

70. Zeqiraj E, Filippi BM, Deak M, Alessi DR, van Aalten DM. 2009. Structure of the LKB1-STRAD-MO25complex reveals an allosteric mechanism of kinase activation. Science 326:1707–11

71. Mukherjee K, Sharma M, Urlaub H, Bourenkov GP, Jahn R, et al. 2008. CASK functions as a Mg2+-independent neurexin kinase. Cell 133:328–39

72. Boudeau J, Miranda-Saavedra D, Barton GJ, Alessi DR. 2006. Emerging roles of pseudokinases. TrendsCell Biol. 16:443–52

73. Rajakulendran T, Sicheri F. Allosteric protein kinase regulation by pseudokinases: insights from STRAD.Sci. Signal. 3:pe8

74. Apsel B, Blair JA, Gonzalez B, Nazif TM, Feldman ME, et al. 2008. Targeted polypharmacology: discoveryof dual inhibitors of tyrosine and phosphoinositide kinases. Nat. Chem. Biol. 4:691–99

75. Knight ZA, Lin H, Shokat KM. 2010. Targeting the cancer kinome through polypharmacology. Nat. Rev.Cancer 10:130–37


Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.

BI80-FrontMatter ARI 16 May 2011 14:13

Annual Review ofBiochemistry

Volume 80, 2011Contents

Preface

Past, Present, and Future Triumphs of BiochemistryJoAnne Stubbe � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �v

Prefatory

From Serendipity to TherapyElizabeth F. Neufeld � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

Journey of a Molecular BiologistMasayasu Nomura � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �16

My Life with NatureJulius Adler � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �42

Membrane Vesicle Theme

Protein Folding and Modification in the MammalianEndoplasmic ReticulumIneke Braakman and Neil J. Bulleid � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �71

Mechanisms of Membrane Curvature SensingBruno Antonny � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 101

Biogenesis and Cargo Selectivity of AutophagosomesHilla Weidberg, Elena Shvets, and Zvulun Elazar � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 125

Membrane Protein Folding and Insertion Theme

Introduction to Theme “Membrane Protein Folding and Insertion”Gunnar von Heijne � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 157

Assembly of Bacterial Inner Membrane ProteinsRoss E. Dalbey, Peng Wang, and Andreas Kuhn � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 161

β-Barrel Membrane Protein Assembly by the Bam ComplexChristine L. Hagan, Thomas J. Silhavy, and Daniel Kahne � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 189

vii

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Transmembrane Communication: General Principles and Lessonsfrom the Structure and Function of the M2 Proton Channel, K+

Channels, and Integrin ReceptorsGevorg Grigoryan, David T. Moore, and William F. DeGrado � � � � � � � � � � � � � � � � � � � � � � � � 211

Biological Mass Spectrometry Theme

Mass Spectrometry in the Postgenomic EraBrian T. Chait � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 239

Advances in the Mass Spectrometry of Membrane Proteins:From Individual Proteins to Intact ComplexesNelson P. Barrera and Carol V. Robinson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 247

Quantitative, High-Resolution Proteomics for Data-DrivenSystems BiologyJurgen Cox and Matthias Mann � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 273

Applications of Mass Spectrometry to Lipids and MembranesRichard Harkewicz and Edward A. Dennis � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 301

Cellular Imaging Theme

Emerging In Vivo Analyses of Cell Function UsingFluorescence ImagingJennifer Lippincott-Schwartz � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 327

Biochemistry of Mobile Zinc and Nitric Oxide Revealedby Fluorescent SensorsMichael D. Pluth, Elisa Tomat, and Stephen J. Lippard � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 333

Development of Probes for Cellular Functions Using FluorescentProteins and Fluorescence Resonance Energy TransferAtsushi Miyawaki � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 357

Reporting from the Field: Genetically Encoded Fluorescent ReportersUncover Signaling Dynamics in Living Biological SystemsSohum Mehta and Jin Zhang � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 375

Recent Advances in Biochemistry

DNA Replicases from a Bacterial PerspectiveCharles S. McHenry � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 403

Genomic and Biochemical Insights into the Specificity of ETSTranscription FactorsPeter C. Hollenhorst, Lawrence P. McIntosh, and Barbara J. Graves � � � � � � � � � � � � � � � � � � � 437

viii Contents

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Signals and Combinatorial Functions of Histone ModificationsTamaki Suganuma and Jerry L. Workman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 473

Assembly of Bacterial RibosomesZahra Shajani, Michael T. Sykes, and James R. Williamson � � � � � � � � � � � � � � � � � � � � � � � � � � � � 501

The Mechanism of Peptidyl Transfer Catalysis by the RibosomeEdward Ki Yun Leung, Nikolai Suslov, Nicole Tuttle, Raghuvir Sengupta,

and Joseph Anthony Piccirilli � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 527

Amyloid Structure: Conformational Diversity and ConsequencesBrandon H. Toyama and Jonathan S. Weissman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 557

AAA+ Proteases: ATP-Fueled Machines of Protein DestructionRobert T. Sauer and Tania A. Baker � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 587

The Structure of the Nuclear Pore ComplexAndre Hoelz, Erik W. Debler, and Gunter Blobel � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 613

Benchmark Reaction Rates, the Stability of Biological Moleculesin Water, and the Evolution of Catalytic Power in EnzymesRichard Wolfenden � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 645

Biological Phosphoryl-Transfer Reactions: Understanding Mechanismand CatalysisJonathan K. Lassila, Jesse G. Zalatan, and Daniel Herschlag � � � � � � � � � � � � � � � � � � � � � � � � � � � 669

Enzymatic Transition States, Transition-State Analogs, Dynamics,Thermodynamics, and LifetimesVern L. Schramm � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 703

Class I Ribonucleotide Reductases: Metallocofactor Assemblyand Repair In Vitro and In VivoJoseph A. Cotruvo Jr. and JoAnne Stubbe � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 733

The Evolution of Protein Kinase Inhibitors from Antagoniststo Agonists of Cellular SignalingArvin C. Dar and Kevan M. Shokat � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 769

Glycan Microarrays for Decoding the GlycomeCory D. Rillahan and James C. Paulson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 797

Cross Talk Between O-GlcNAcylation and Phosphorylation:Roles in Signaling, Transcription, and Chronic DiseaseGerald W. Hart, Chad Slawson, Genaro Ramirez-Correa, and Olof Lagerlof � � � � � � � � � 825

Regulation of Phospholipid Synthesis in the YeastSaccharomyces cerevisiaeGeorge M. Carman and Gil-Soo Han � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 859

Contents ix

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.


Sterol Regulation of Metabolism, Homeostasis, and DevelopmentJoshua Wollam and Adam Antebi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 885

Structural Biology of the Toll-Like Receptor FamilyJin Young Kang and Jie-Oh Lee � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 917

Structure-Function Relationships of the G Domain, a CanonicalSwitch MotifAlfred Wittinghofer and Ingrid R. Vetter � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 943

STIM Proteins and the Endoplasmic Reticulum-PlasmaMembrane JunctionsSilvia Carrasco and Tobias Meyer � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 973

Amino Acid Signaling in TOR ActivationJoungmok Kim and Kun-Liang Guan � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1001

Mitochondrial tRNA Import and Its Consequencesfor Mitochondrial TranslationAndre Schneider � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1033

Caspase Substrates and Cellular RemodelingEmily D. Crawford and James A. Wells � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1055

Regulation of HSF1 Function in the Heat Stress Response:Implications in Aging and DiseaseJulius Anckar and Lea Sistonen � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1089

Indexes

Cumulative Index of Contributing Authors, Volumes 76–80 � � � � � � � � � � � � � � � � � � � � � � � � � �1117

Cumulative Index of Chapter Titles, Volumes 76–80 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1121

Errata

An online log of corrections to Annual Review of Biochemistry articles may be found athttp://biochem.annualreviews.org/errata.shtml

x Contents

Ann

u. R

ev. B

ioch

em. 2

011.

80:7

69-7

95. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

Uni

vers

ity o

f C

alif

orni

a -

San

Fran

cisc

o U

CSF

on

06/3

0/19

. For

per

sona

l use

onl

y.

the evolution of protein kinase inhibitors from ...shokatlab.ucsf.edu/pdfs/21548788.pdfshown is the...

Documents