of April 9, 2019.This information is current as

and Multiple ConformationsKinase 4 Structures Reveal Novel Features Cutting Edge: IL-1 Receptor-Associated

Jim W. Barnett and Michelle F. BrownerSimon W. Lee, Stan Tsing, Linghao Niu, Kyung W. Song, Andreas Kuglstatter, Armando G. Villaseñor, David Shaw,

http://www.jimmunol.org/content/178/5/2641doi: 10.4049/jimmunol.178.5.2641

2007; 178:2641-2645; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/178/5/2641.full#ref-list-1

, 5 of which you can access for free at: cites 23 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2007 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on April 9, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on April 9, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

CUTTING EDGE

IMMUNOLOGY

THE OFJOURNAL

Cutting Edge: IL-1 Receptor-Associated Kinase 4Structures Reveal Novel Features and MultipleConformationsAndreas Kuglstatter,1 Armando G. Villasenor, David Shaw, Simon W. Lee, Stan Tsing,Linghao Niu, Kyung W. Song, Jim W. Barnett, and Michelle F. Browner

IL-1R-associated kinase (IRAK)4 plays a central role ininnate and adaptive immunity, and is a crucial compo-nent in IL-1/TLR signaling. We have determined the crys-tal structures of the apo and ligand-bound forms of hu-man IRAK4 kinase domain. These structures revealseveral features that provide opportunities for the design ofselective IRAK4 inhibitors. The N-terminal lobe of theIRAK4 kinase domain is structurally distinctive due to aloop insertion after an extended N-terminal helix. Thegatekeeper residue is a tyrosine, a unique feature of theIRAK family. The IRAK4 structures also provide insightsinto the regulation of its activity. In the apo structure, twoconformations coexist, differing in the relative orientationof the two kinase lobes and the position of helix C. In thepresence of an ATP analog only one conformation is ob-served, indicating that this is the active conformation. TheJournal of Immunology, 2007, 178: 2641–2645.

I nterleukin-1R-associated kinase (IRAK)42 is a serine/thre-onine-specific protein kinase that is essential for innate (1)and adaptive (2) immune response. It plays a central role

in IL-1 and TLR signaling pathways (3, 4) and protection frombacterial infections (4). Upon IL-1 stimulation, IRAK4 andIRAK1 colocalize at the IL-1R and both kinases bind to theadaptor protein MyD88. IRAK1 is phosphorylated by IRAK4and subsequently by autophosphorylation. PhosphorylatedIRAK1 dissociates from the receptor and associates withTNFR-associated factor 6 to initiate downstream signaling cas-cades, resulting in activation of NF-�B, p38, and JNK mi-togen-activated protein kinases. Given the central role ofIRAK4 in Toll-like/IL-1R signaling and immunological pro-tection, IRAK4 inhibitors have been implicated as valuabletherapeutics in inflammatory diseases, sepsis, and autoim-mune disorders (5).

IRAK4 belongs to the IRAK family of kinases, which consistsof IRAK1, IRAK2, IRAK-M, and IRAK4 (3). They all contain

an N-terminal death domain followed by a linker of unknowndomain structure and a kinase domain. IRAK4 is the onlyIRAK family member that does not have a C-terminal exten-sion. Within the kinase complement of the human genome, theIRAK4 kinase domain is relatively distinct with 38% sequenceidentity to the kinase domain of IRAK1 and �32% sequenceidentity to other human kinases. The crystal structure of theIRAK4 death domain was recently reported (6), but no struc-tural information has been available for the kinase domain ofany of the IRAK family members.

We have solved the crystal structures of IRAK4 kinase do-main in its apo form, complexed with the nonhydrolysableATP analog AMPPNP, and complexed with the pan-specifickinase inhibitor staurosporine. These structures providevaluable insight into regulation of IRAK4 kinase activity andform the framework for the design of potent and selectiveIRAK4 inhibitors.

Materials and MethodsProtein production

To determine the most suitable boundary of the IRAK4 kinase domain for mo-lecular structure determination, sequential 1-aa N-terminal truncations (resi-dues 150–165) and 3-aa C-terminal truncations (residues 417–453) weremade from IRAK4 cDNA. Truncated IRAK4 constructs were generated byoverlap-extension PCR in which (from the N terminus) a 6-His tag, TEV pro-tease spacer, and TEV cleavage site were encoded. PCR products were sub-cloned into the baculovirus transfer vector pVL1392 (BD Biosciences), andDNA sequence was verified. Correct constructs were cotransfected into Sf9cells, and the produced virus was plaque-purified to generate clonal popula-tions. Titered, plaque-purified virus was used to determine that a multiplicity ofinfection of 0.3 with a harvest time of 72 h was optimal for protein expression.An N-terminal truncation in which amino acid residues 160 through 460 ofIRAK4 are expressed was one candidate identified to possess good expressionand purification properties.

For large-scale purification of IRAK4 kinase domain, pellets from 10 L ofinfected Sf9 cell were resuspended in 600 ml of lysis buffer (50 mM HEPES(pH 7.5), 300 mM NaCl, 10 mM 2-ME, 12 Complete protease inhibitor tab-lets (Roche Applied Science), 5 mg of DNase per 100 mg of cells). Cells werethen lysed with a microfluidizer (Microfluidics) at �9,000 psi on ice. The lysatewas centrifuged at 25,000 � g, and the supernatant was applied to a 50-mlTalon Superflow metal affinity column (BD Biosciences) pre-equilibrated in

Roche Palo Alto LLC, Palo Alto, CA 94304

Received for publication November 20, 2006. Accepted for publication January 10, 2007.

The costs of publication of this article were defrayed in part by the payment of page charges.This article must therefore be hereby marked advertisement in accordance with 18 U.S.C.Section 1734 solely to indicate this fact.1 Address correspondence and reprint requests to Andreas Kuglstatter, RochePalo Alto LLC, 3431 Hillview Avenue, Palo Alto, CA 94304. E-mail address:andreas.kuglstatter@roche.com

2 Abbreviation used in this paper: IRAK, IL-1R-associated kinase.

Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00

www.jimmunol.org

by guest on April 9, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

column buffer A (50 mM HEPES (pH 7.5), 300 mM NaCl, 10% (v/v) glyc-erol, 10 mM 2-ME, and 20 mM imidazole). The bound protein was eluted withbuffer B (buffer A with 100 mM imidazole (pH 7.5)) and concentrated to 5mg/ml. The 6-His tag was cleaved overnight at 4°C with AcTEV protease (In-vitrogen Life Technologies) in 0.3% (w/v) n-octyl-�-D-glucopyranoside, �20mM imidazole, 1� stock buffer, 1 mM DTT, and 2,000 U AcTEV. The re-action mixture was passed over a 5-ml Ni-HP column (GE Healthcare) pre-equilibrated in buffer A, and the flow-through was concentrated to 5–10 mlwith an Amicon Ultra-15 concentrator (Millipore). IRAK4 kinase domain wasfinally purified on a Superdex 200 26/60 column (GE Healthcare) and con-centrated to 13 mg/ml.

Enzymatic assay

IRAK4 kinase domain activity was measured by phosphorylation of the IRAK1peptide substrate KKARFSRFAGSSPSQSSMVAR (Anaspec) using[�-33P]ATP (Amersham). The enzyme reaction was conducted in 25 mMHEPES (pH 7.5), 2 mM DTT, 150 mM NaCl, 20 mM MgCl2, 0.001%Tween 20, and 0.1% bovine �-globulin (Sigma-Aldrich). For ATP Km deter-minations, initial rates of enzymatic reactions were measured using 4 nM en-zyme in the presence of 4.5 mM peptide substrate and 0.1–6.4 mM ATP. TheKm of the peptide substrate was determined using 6.4 mM ATP, 4 nM enzyme,and 0.25–4.5 mM peptide. The inhibition potency of staurosporine was de-termined by measuring the activity of 2 nM IRAK4 kinase domain in the ab-sence and presence of 10 inhibitor concentrations at 0.25 Km concentrations ofboth ATP and the peptide substrate.

Crystallization, data collection, and processing

The IRAK4 kinase domain at a concentration of 13 mg/ml (50 mM HEPES(pH 7.5), 10% (v/v) glycerol, 300 mM NaCl, 10 mM 2-ME, and 0.02% (w/v)octyl-�-D-glucopyranoside) was incubated for 2 h at 4°C either with 1 mMstaurosporine (Sigma-Aldrich) and 5% (v/v) DMSO or with 12 mM AMPPNP(Sigma-Aldrich). The protein was crystallized by vapor diffusion in hangingdrops at 20°C. A total of 0.7 �l of protein solution was mixed with 0.7 �l of 2.3M sodium malonate, 100 mM sodium acetate (pH 5.0), and 10 mM DTTresulting in a final pH of 7.0. Crystals of space group C2 (a � 147Å, b � 139Å,c � 89Å, � � 124°) appeared within 3–10 days and grew to a maximum size of500 � 300 � 200 �m3. Because IRAK4 crystals grown after incubation withAMPPNP showed no ligand density at the ATP binding site, they were soakedwith saturated solution of AMPPNP for 6 h.

X-ray diffraction data of apo and AMPPNP-bound IRAK4 kinase domainwas collected at the synchrotron beam line 9-1 of the Stanford SynchrotronRadiation Laboratory (Palo Alto, CA) using an ADSC Quantum 315 CCDdetector. Data of the IRAK4-staurosporine complex was collected at beam line5.0.3 of the Advanced Light Source (Berkeley, CA) using an ADSC Quantum210 CCD detector. All diffraction images were processed with DENZO, andthe intensities were scaled with SCALEPACK (7).

Molecular replacement, structure refinement, and model building

The search model for molecular replacement is based on the crystal structure ofthe proto-oncogene serine/threonine-protein kinase B-RAF (Ref. 8; ProteinData Bank accession no. 1UWH). Regions of very low sequence similarity weredeleted, and the remaining amino acids were exchanged to the IRAK4 sequenceusing the program XSAE (C. Broger, F. Hoffmann-La Roche, Basel, Switzer-land). Molecular replacement was performed with the program PHASER (9).The solution found was an IRAK4 tetramer in space group C2. The model wasrefined against the experimental data, and electron density maps were calculatedusing REFMAC5 (10). The structure model was built with the graphics soft-ware MOLOC (11) and validated with PROCHECK (12). Illustrations of thefinal IRAK4 crystal structures were created using PYMOL (Delano Scientific).

Results and DiscussionOverall structure and ligand binding

We have determined the x-ray crystal structures of IRAK4 ki-nase domain in its nonliganded form as well as complexed withthe natural microbial alkaloid staurosporine and the nonhydro-lysable ATP analog AMPPNP (Table I). The IRAK4 kinase do-main protein selected for crystallization is catalytically active(ATP Km � 650 �M, substrate Km � 1100 �M, kcat � 420min�1). IRAK4 kinase domain crystals contain four proteinchains in the asymmetric unit. Although several structural dif-ferences are observed between the four proteins in the asymmet-ric unit and between the different ligand complexes at the levelof individual amino acids and positioning of secondary struc-

ture elements, their overall architectures are conserved. IRAK4kinase domain possesses the typical two-lobe structure. The N-terminal lobe consists of a five-stranded antiparallel �-sheet, he-lix C, and a nonprototypic N-terminal �-helix (helix B) thatpacks against the �-sheet (Fig. 1). In addition, a unique loop ispresent in the IRAK4 kinase domain between the N-terminalhelix and the first �-strand, which is significantly shortened bythis insertion. The C-terminal kinase lobe of IRAK4 is com-prised of five larger and two shorter �-helices, one �-hairpin,the activation loop, and several additional loop structures. TheN- and C-terminal lobes are connected by the so-called hingesequence, which partially defines the binding site for ATP andATP-competitive kinase inhibitors.

Recently, three autophosphorylation sites in the activationloop of the IRAK-4 kinase domain have been identified:p-T342, p-T345, and p-S346 (13). Mutations at these posi-tions reduce IRAK-4 kinase activity significantly. We have in-dependently identified the same three phosphoresidues in elec-tron density maps of insect cell-expressed IRAK4 kinasedomain (Fig. 2A). p-T345 represents the prototypical phos-phoresidue required for the activation of many protein kinases

Table I. Statistics for x-ray data processing and model refinement

Apo Staurosporine AMPPNP

PDB accession code 20IB 20IC 20IDResolution limits (Å) 50.0–2.0 50.0–2.4 50.0–2.3Completeness (%)a 92.6 (62.3) 99.6 (96.7) 99.9 (99.9)I/�(I)a 9.1 (2.2) 16.0 (2.1) 8.9 (2.9)Rsym (%)a,b 9.3 (51.0) 7.1 (51.1) 11.2 (62.3)Rcryst/Rfree (%)c 22.1/27.0 22.3/27.8 24.1/29.5Rmsd bond length (Å) 0.008 0.011 0.012Rmsd bond angles (°) 1.1 1.8 1.4

a Number in parentheses are values for the highest of 10 resolution shells. Rmsd,Root mean square deviation.

b Rsym � �hkl ��I�–I�/�hkl �I�.c Rcryst � �hkl ��Fo�–Fc�/�hkl �Fo�. Rfree is calculated the same way using a random

5% test set of reflections.

N

C

αB

αC

αD αE

αF

αG

αH

αI

αJ

β1β2

β3β4

β5

β6β7

FIGURE 1. Staurosporine-bound IRAK4 kinase domain. Green, Gly-richloop; cyan, helix C; magenta, activation loop; brown, helix B; red, loop uniqueto IRAK4; yellow, staurosporine. Amino acids not resolved in the electron den-sity maps are outlined as dots.

2642 CUTTING EDGE: CRYSTAL STRUCTURES OF HUMAN IL-1R-ASSOCIATED KINASE 4

by guest on April 9, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

(14). It interacts with the R334 side chain and the p-T346backbone amine. In addition, the phosphate forms water-me-diated hydrogen bonds with the backbone carbonyl groups ofM344 and R347. p-T342 and p-S346 are nonprototypical sitesof phosphorylation. p-T342 is coordinated by the side chains ofT365, K367, and K441, whereas p-S346 is highly solvent ex-posed and its phosphate group does not make an intramolecularinteraction.

The pan-kinase inhibitor staurosporine (IC50 � 4 nM) bindsto the hinge loop of IRAK4 as observed in other kinases. Thelactam acceptor oxygen and donor nitrogen form hydrogenbonds to the M265 backbone amine and the V263 backbonecarbonyl group, respectively (Fig. 2B). The lipophilic bulk ofstaurosporine makes numerous hydrophobic and van derWaals’ interactions with the predominantly lipophilic ATP-binding pocket of IRAK4. In addition, one of the two stauro-sporine phenyl rings is forming an offset stacked aromatic in-teraction with the phenyl ring of the hinge residue Y264, whilethe other phenyl ring is forming an edge-to-face aromatic inter-action with the phenyl ring of the gatekeeper residue Y262.

Electron density maps of the crystal structure of IRAK4 ki-nase domain complexed with the nonhydrolysable ATP analogAMPPNP are less well defined than the staurosporine or apostructures. Presumably, this is a consequence of the high AMP-PNP concentration required to achieve sufficient ligand occu-pancy. Nevertheless, the AMPPNP molecule could unambigu-ously be identified (Fig. 2C). AMPPNP adopts the standardposition with the adenine hydrogen bonding to the M265backbone amine and the V263 backbone carbonyl group. Nohydrogen bond is observed between either of the two ribose hy-droxyl groups and the protein. It is possible that this missing

interaction contributes to the relative high ATP Km (650 �M)observed for IRAK4 kinase domain. Because the electron den-sity of the AMPPNP phosphates is of limited quality, their in-teractions with the protein and solvent molecules could not beresolved unambiguously.

Unique N-terminal extension and gatekeeper residue

The primary challenge in designing therapeutic kinase inhibi-tors is to achieve the desired selectivity, because off-target kinaseinhibition may result in adverse effects. Therefore, identifica-tion of unique features for a kinase drug target is critical forsuccessful drug design. The N-terminal kinase lobe of IRAK4has two defining features not seen in previously reported kinasestructures: a unique N-terminal extension and tyrosine as gate-keeper residue.

In addition to the five-stranded antiparallel �-sheet and helixC, the N-terminal kinase lobe of IRAK4 contains an N-termi-nal extension of so far unknown function that is unique toIRAK4 as judged by a sequence alignment of 518 human ki-nases, including the other IRAK family members (15). It startswith an amphiphilic �-helix (helix B, aas 169–176) that makesextensive hydrophobic interactions with lipophilic side chainsof amino acids forming the �-sheet (Fig. 1). This �-helix is fol-lowed by three motifs that have been described as helix-cappingelements: 1) an ST motif (Ref. 16; aas 177–181), which is char-acterized by hydrogen bonds from the T177 hydroxyl group tothe F180 backbone amine and from the T177 backbone car-bonyl group to the D181 backbone amine; 2) an ASX motif(17, aas 181–183) defined by the hydrogen bond between theD181 side chain carbonyl group and the R183 backboneamine; and 3) a highly solvent-exposed Schellman loop (Ref.18; aas 184–189) characterized by hydrogen bonds between thebackbone amines of G188/G189 and the backbone carbonylgroups of P184/I185, respectively (Fig. 3). C-terminal to theIRAK4 loop continues the prototypic Gly-rich loop, which isformed by the first two strands of the five-stranded �-sheet. TheSchellman loop, in particular I185, is adjacent to the ATP bind-ing site, which makes it an ideal interaction partner for IRAK4selective kinase inhibitors.

p-S346

p-T345

p-T342

R334

R347

M344

V343

T365K367

K441

B CY262

Y262

Y264Y264

V263 V263

M265 M265

P266P266

N267 N267G268 G268

S269S269

A

FIGURE 2. Fo-Fc electron density omit maps contoured at 3�. Selected hy-drogen bonds are indicated as dashes. Activation loop residues (A), staurospor-ine (B), and AMPPNP (C) are shown in yellow, and contacting amino acids areshown in gray.

T177

R183

G188G189

I185

P184

N178

D181

V187

S186

E182

N179

F180

Schellman loop

Gly-rich loop

helix B

ST motif

ASX motif

FIGURE 3. N-terminal region unique to IRAK4. Hydrogen bonds thatcharacterize the ST, ASX, and Schellman motifs are indicated as dashes.

2643The Journal of Immunology

by guest on April 9, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Immediately N-terminal to the hinge loop (aas 263–268) isthe so-called gatekeeper residue, Y262 in IRAK4. The size andflexibility of the side chain of this residue generally determineswhether an additional lipophilic subpocket, referred to as the“back pocket,” is present adjacent to the ATP binding site (19).The most abundant gatekeeper residue in kinases is methio-nine, followed by leucine, threonine, and phenylalanine. Asjudged by primary sequence alignments of human kinase do-mains, tyrosine occupies the gatekeeper position only in thefour members of the IRAK family. The IRAK4 crystal struc-tures reported in this study are the first structural representa-tives for kinases with tyrosine as a gatekeeper residue. The bulkyside chain of the gatekeeper Y262 is oriented toward the con-served E233 forming a hydrogen bond with the glutamate sidechain (Fig. 4). Therefore, the back pocket of IRAK4 is blockedby the gatekeeper tyrosine and not accessible for ligands. How-ever, ligand interactions with the hydroxyl group of the gate-keeper residue Y262 should be considered in the design ofIRAK family selective inhibitors.

Two kinase domain conformations and implications on IRAK4regulation

Many kinases are activated by phosphorylation of activationloop residues, either by upstream kinases or autocatalytically. Ithas also been shown for several kinases that a hinge movementof helix C away from the ATP binding site is involved in switch-ing theses enzymes from their catalytically active to an inactivestate (14). This process results in disruption of an essential saltbridge between an invariant lysine on strand 3 of the antiparallel�-sheet and an invariant glutamate on helix C. Several differentmechanisms have been described that trigger the position of he-lix C and therefore the activation state of the kinase. Examplesinclude the intermolecular interaction between helix C of cy-clin-dependent kinase and cyclin (20, 21), and the intramolec-ular interaction between helix C of c-Src and the c-Src SH3domain (22, 23).

The IRAK4 apo-form crystal structure shows two distinctprotein conformations in the tetramer comprising the crystal-lographic asymmetric unit. These conformations differ signifi-cantly in their relative orientation between their N- and C-ter-minal lobes. The regions of the protein primarily affected by the

lobe rotation of IRAK4 are the Gly-rich loop, the unique N-terminal extension, and helix C (Fig. 4). The overall effect ofthis rotation on helix C is a similar hinge movement as de-scribed above for other kinases. We therefore refer to the twoIRAK4 kinase domain conformations as “helix C-in” and “helixC-out.” The positioning of helix C in IRAK4 has the oppositeeffect on the lysine-glutamate salt bridge (Fig. 4) comparedwith other kinases. The K213-E233 interaction in IRAK4 isonly present in the helix C-out position. In the helix C-in po-sition, the K213 side chain is disordered and the carboxyl groupof E233 forms a hydrogen bond with the backbone amine ofF350, which is part of the conserved DFG sequence that marksthe N-terminal end of the activation loop.

In the IRAK4-AMPPNP structure, all four kinase domainspresent in the crystallographic asymmetric unit adopt the helixC-in conformation. As observed in the apo structure, in thisconformation the K213-E233 salt bridge is not formed. Pre-sumably, in the absence of ATP analog both kinase domainconformations coexist in solution. Binding of AMPPNP be-tween the N- and C-terminal kinase lobes shifts this equilib-rium to the helix C-in conformation. The presence of AMP-PNP at the ATP binding site was achieved by soakingAMPPNP into apo crystals, therefore the transition from thehelix C-out to the helix C-in conformation can occur in thecrystal lattice. The fact that AMPPNP binds only to the helixC-in conformation of IRAK4 suggests that this represents theactive form of the kinase, despite the missing K213-E233 saltbridge (Fig. 4). It is possible that the presence of the hydroxylgroup of the unique Y262 gatekeeper side chain affects the localhydrogen bond pattern requirements for nucleotide binding.

The mechanism of IRAK4 activation in cells is not well un-derstood. In analogy with other kinases, it seems likely thatIRAK4 phosphoryl transfer activity is regulated both by activa-tion loop phosphorylation and by switching between the twoobserved (or more) kinase domain conformations (Fig. 4). Inthe case of IRAK4, this switch could be triggered by intramo-lecular interaction between kinase domain and its death do-main or by intermolecular interactions with adaptor proteinslike MyD88. The crystal structures of IRAK4 kinase domainpresented in this study form the basis for the design of experi-ments that decipher these interactions, and therefore howIRAK4 mediates IL-1/TLR signaling.Note Added in Proof. After submission of this manuscript, Wanget al. (24) published the crystal structure of IRAK4 kinase domaincomplexed with staurosporine and a benzimidazole inhibitor.

AcknowledgmentsWe acknowledge Virginia L. Rath (Reciprocal Space Consulting, LLC, Oak-land, CA) for x-ray data collection.

DisclosuresThe authors have no financial conflict of interest.

References1. Suzuki, N., S. Suzuki, G. S. Duncan, D. G. Millar, T. Wada, C. Mirtsos, H. Takada,

A. Wakeham, A. Itie, S. Li, et al. 2002. Severe impairment of interleukin-1 and Toll-like receptor signalling in mice lacking IRAK-4. Nature 416: 750–756.

2. Suzuki, N., S. Suzuki, D. G. Millar, M. Unno, H. Hara, T. Calzascia, S. Yamasaki,T. Yokosuka, N. J. Chen, A. R. Elford, et al. 2006. A critical role for the innate im-mune signaling molecule IRAK-4 in T cell activation. Science 311: 1927–1932.

3. Janssens, S., and R. Beyaert. 2003. Functional diversity and regulation of differentinterleukin-1 receptor-associated kinase (IRAK) family members. Mol. Cell. 11:293–302.

Lys213

Tyr262

Glu233

Phe330

Hinge

Gly-rich loop

Irak4-unique loopHelix C-in

Helix C-out

FIGURE 4. Overlay of the two protein conformations observed in theIRAK4 apo structure. The superposition is based on the C� atoms of the C-terminal kinase lobe. Amino acids not resolved in the electron density maps areoutlined as dots. In the helix C-in conformation the K213 side chain isdisordered.

2644 CUTTING EDGE: CRYSTAL STRUCTURES OF HUMAN IL-1R-ASSOCIATED KINASE 4

by guest on April 9, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

4. Picard, C., A. Puel, M. Bonnet, C. L. Ku, J. Bustamante, K. Yang, C. Soudais,S. Dupuis, J. Feinberg, C. Fieschi, et al. 2003. Pyogenic bacterial infections in humanswith IRAK-4 deficiency. Science 299: 2076–2079.

5. Wietek, C., and L. A. ONeill. 2002. IRAK-4: a new drug target in inflammation,sepsis, and autoimmunity. Mol. Interv. 2: 212–215.

6. Lasker, M. V., M. M. Gajjar, and S. K. Nair. 2005. Cutting edge: molecular structureof the IL-1R-associated kinase-4 death domain and its implications for TLR signaling.J. Immunol. 175: 4175–4179.

7. Otwinowski, Z., and W. Minor. 1997. Processing of x-ray diffraction data collected inoscillation mode. In Macromolecular Crystallography, Part A, Vol. 276.C. W. J. Carter and R. M. Sweet, eds. Academic Press, New York, pp. 307.

8. Wan, P. T., M. J. Garnett, S. M. Roe, S. Lee, D. Niculescu-Duvaz, V. M. Good,C. M. Jones, C. J. Marshall, C. J. Springer, D. Barford, and R. Marais. 2004. Mech-anism of activation of the RAF-ERK signaling pathway by oncogenic mutations ofB-RAF. Cell 116: 855–867.

9. McCoy, A. J., R. W. Grosse-Kunstleve, L. C. Storoni, and R. J. Read. 2005. Likeli-hood-enhanced fast translation functions. Acta Crystallogr. D. Biol. Crystallogr. 61:458–464.

10. Murshudov, G. N., A. A. Vagin, and E. J. Dodson. 1997. Refinement of macromo-lecular structures by the maximum-likelihood method. Acta Crystallogr. D. Biol. Crys-tallogr. 53: 240–255.

11. Gerber, P. R. 1992. Peptide mechanics: a force field for peptides and proteins workingwith entire residues as small unites. Biopolymers 32: 1003–1017.

12. Laskowski, R. A., M. W. MacArthur, D. S. Moss, and J. M. Thornton. 1993. PRO-CHECK: a program to check the stereochemical quality of protein structures. J. Appl.Crystallogr. 26: 283–291.

13. Cheng, H., T. Addona, H. Keshishian, E. Dahlstrand, C. Lu, M. Dorsch, Z. Li,A. Wang, T. D. Ocain, P. Li, et al. 2007. Regulation of IRAK-4 kinase activity viaautophosphorylation within its activation loop. Biochem. Biophys. Res. Commun. 352:609–616.

14. Huse, M., and J. Kuriyan. 2002. The conformational plasticity of protein kinases. Cell109: 275–282.

15. Manning, G., D. B. Whyte, R. Martinez, T. Hunter, and S. Sudarsanam. 2002. Theprotein kinase complement of the human genome. Science 298: 1912–1934.

16. Wan, W. Y., and E. J. Milner-White. 1999. A recurring two-hydrogen-bond motifincorporating a serine or threonine residue is found both at �-helical N termini and inother situations. J. Mol. Biol. 286: 1651–1662.

17. Wan, W. Y., and E. J. Milner-White. 1999. A natural grouping of motifs with anaspartate or asparagine residue forming two hydrogen bonds to residues ahead in se-quence: their occurrence at �-helical N termini and in other situations. J. Mol. Biol.286: 1633–1649.

18. Schellman, C. 1980. The �L conformation at the ends of helices. In Protein Folding.R. Jaenicke, ed. Elsevier, Amsterdam, pp. 53.

19. Noble, M. E., J. A. Endicott, and L. N. Johnson. 2004. Protein kinase inhibitors:insights into drug design from structure. Science 303: 1800–1805.

20. De Bondt, H. L., J. Rosenblatt, J. Jancarik, H. D. Jones, D. O. Morgan, andS. H. Kim. 1993. Crystal structure of cyclin-dependent kinase 2. Nature 363:595–602.

21. Jeffrey, P. D., A. A. Russo, K. Polyak, E. Gibbs, J. Hurwitz, J. Massague, andN. P. Pavletich. 1995. Mechanism of CDK activation revealed by the structure of acyclinA-CDK2 complex. Nature 376: 313–320.

22. Sicheri, F., I. Moarefi, and J. Kuriyan. 1997. Crystal structure of the Src family ty-rosine kinase Hck. Nature 385: 602–609.

23. Xu, W., A. Doshi, M. Lei, M. J. Eck, and S. C. Harrison. 1999. Crystal structures ofc-Src reveal features of its autoinhibitory mechanism. Mol. Cell. 3: 629–638.

24. Wang Z., J. Liu, A. Sudom, M. Ayres, S. Li, H. Wesche, J. P. Powers, and N. P.Walker. 2006. Crystal structures of IRAK-4 kinase in complex with inhibitors: aserine/threonine kinase with tyrosine as a gatekeeper. Structure 14: 1835–1844.

2645The Journal of Immunology

by guest on April 9, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from