Chapter 14 - Biophysical and Biochemical Investigations of ... · CHAPTER FOURTEEN Biophysical and Biochemical Investigations of dsRNA-Activated Kinase PKR ... Figure 14.1 Regulation

C H A P T E R F O U R T E E N

Biophysical and BiochemicalInvestigations of dsRNA-ActivatedKinase PKR

Sean� A.� McKenna,�*� Darrin� A.� Lindhout,*� Takashi� Shimoike,�*,†

and� Joseph� D.� Puglisi*,�‡

Contents1. Introduction 374

2. Expression and Purification of PKR 376

3. RNA Synthesis 379

4. Phosphorylation Assays 380

5. Measuring RNA–Protein Stabilities 383

6. Monitoring the Association State of PKR 386

7. NMR Spectroscopy 388

8. In Vitro Translation Assays 392

9. Conclusions 393

References 394

AbstractProtein kinase RNA-activated (PKR) is a serine/threonine kinase that containsan N-terminal RNA-binding domain (dsRNA) and a C-terminal kinase domain.On binding viral dsRNA molecules, PKR can become activated and phosphorylatecellular targets, such as eukaryotic translation initiation factor 2a (eIF-2a). Phos-phorylation of eIF-2a results in attenuation of protein translation initiation. There-fore, PKR plays an integral role in the antiviral response to cellular infection. Herewe provide a methodological framework for probing PKR function by use ofassays for phosphorylation, RNA–protein stability, PKR dimerization, and in vitrotranslation. These methods are complemented by nuclear magnetic resonanceapproaches for probing structural features of PKR activation. Considerationsrequired for both PKR and dsRNA sample preparation are also discussed.

Methods in Enzymology, Volume 430 # 2007 Elsevier Inc.ISSN 0076-6879, DOI: 10.1016/S0076-6879(07)30014-1 All rights reserved.

* Department of Structural Biology, Stanford University School of Medicine, Stanford, California{ Department of Virology II, National Institute of Infectious Diseases, Musashi-Murayama, Tokyo, Japan{ Stanford Magnetic Resonance Laboratory, Stanford University School of Medicine, Stanford, California

373

1. Introduction

Ahallmark of viral replication is the dependence on the host–cell proteintranslation machinery for the production of viral proteins (Gale et al., 2000;Schneider andMohr, 2003). The ability to suppress translation at the initiationstage represents a crucial host–cell defense against viral attack in eukaryotes.Phosphorylation of the a-subunit of eukaryotic initiation factor 2 (eIF2a) atSer51 (Chong et al., 1992) inhibits the guanine nucleotide exchange activity ofeIF2B (Sonenberg and Dever, 2003) (Fig. 14.1A). The pool of active eIF2ternary complex is thus reduced, causing a global decrease of both viral andcellular protein synthesis. Phosphorylation of eIF2a is accomplished by anRNA-activated Ser/Thr protein kinase (PKR); PKR is activated throughautophosphorylation on association with double-stranded RNA (dsRNA)generated during the replicative cycle of both RNA- and DNA-based viralgenomes (Schneider and Mohr, 2003). That PKR is one of many interferon-induced proteins synthesized in response to viral infiltration only highlightsits central role in the antiviral response (Malmgaard, 2004). The detailedunderstanding of the mechanism of activation of PKR by viral dsRNA ineukaryotic cells represents an area of intense interest and will be the focus ofthis chapter.

eIF2B

GTP

GDP

GDP GTP

Translationinitiation

GDPGTP eIF2BMet-

tRNAi

Met-tRNAi

eIF2B

GDP

bg

a

bg

a bg

a bg

a

bg

a

bg

a

GDP

GDPGTP Translationinitiationinhibition

XP P

PKR-dsRNA

A

B

dsRBD1 dsRBD2 Kinase

Interdomainlinker

1 169 252 551

T446 T451

Figure 14.1 Regulation of translation initiation by PKR. (A) eIF2 ternary complex(a, b, g) is responsible for delivery of initiator Met-tRNA to the translation initiationcomplex in aGTP-dependent fashion.Activated PKRphosphorylates eukaryotic initia-tionfactor2a(eIF2a)atSer51;phosphorylationatthis site inhibitstheactivityofaguaninenucleotide exchange factor, eIF2B.The pool of active eIF2 ternary complex decreases,causing attenuation of translation initiation. (B) Domain organization of PKR.N-terminaldsRBDs,C-terminal kinase domain, and the interdomain linker are shown.Critical autophosphorylation sites (T446,T451) in thekinase domain are indicated.

374 Sean A. McKenna et al.

PKR is a 551-residue enzyme consisting of a N-terminal RNA-binding(dsRBD1/2, 19 kDa) and C-terminal kinase domain (PKRKD, 38 kDa)connected by an 80-residue unstructured interdomain linker (Clemens,1997; McKenna, 2007) (Fig. 14.1B). The modular domain architectureof PKR has allowed for the high-resolution structure determination andfunctional characterization of each of these domains individually. TheN-terminal domain contains tandem 70-residue dsRNA binding motifs(dsRBDs) that act as an intracellular sensor for dsRNA (Clemens, 1997).PKR is highly specific for dsRNA over other nucleic acid substrates (Galeand Katze, 1998); length and RNA structural features are thought to modu-late the interaction (Bevilacqua and Cech, 1996; Kim et al., 2006; McKennaet� al.,� 2006).� Each� dsRBD� adopts� a� canonical� a�!b!b�!b!a� fold� in� whichthe� a-helices� pack� against� an� antiparallel� b-sheet� (Nanduri� et� al�.,� 1998a,b).High-affinity PKR–nucleic acid interactions require both dsRBDs; individ-ually expressed dsRBDs do not interact with appreciable affinity (Kim et al.,2006; McCormack et al., 1994; McKenna et al., 2006). Conversely, theC-terminal Ser/Thr kinase domain is responsible for both phosphorylationactivity and substrate recognition. The crystal structure of the isolated kinasedomain in complex with eIF2a revealed insights into the ATP coordinationsite, substrate recognition site, and autophosphorylation sites required foractivation (Thr446 and T451) (Dar et al., 2005; Dey et al., 2005).

Although the modular nature of PKR has allowed for the detailedbiophysical characterization of individual domains, the central mechanisticquandary of how dsRNA leads to kinase domain autophosphorylationremains poorly understood in the context of the full-length protein.Although some groups have proposed an autoinhibitory model in whichthe latent form of PKR remains inactive through intramolecular associationbetween the N- and C-terminal domains (Gelev et al., 2006; Li et al., 2006;Nanduri et al., 2000; Vattem et al., 2001; Wu and Kaufman, 1997), bothourselves (McKenna, 2007) and others (Lemaire et al., 2006) have observedthat in the context of the full-length enzyme, no such interactions areobserved. What is well understood is that latent PKR undergoes autophos-phorylation on dsRNA binding by means of a bimolecular reaction mecha-nism in which dimerization between PKR molecules seems important(McKenna, 2007; Romano et al., 1998; Taylor et al., 1996). A potentialconnection between dsRNA sensing and kinase activity has been observed,because it seems that both dsRNA binding and PKR phosphorylation resultin increased self-affinity between PKR monomers.

Here, we present our methods to study PKR from a biophysical per-spective to answer central mechanistic questions. The expression and puri-fication of PKR, PKRmutants, and its individual domains will be discussed,as will in vitro transcription of viral RNAs. Various structural and biochemi-cal assays, including kinase phosphorylation kinetics, protein–RNA affi-nity determinations, dimerization status, low- and high-resolution structural

Investigations of dsRNA-activated Kinase PKR 375

approaches, and in vitro translational assays will be discussed. These ap-proaches should prove generally useful for those looking to study dynamickinases such as PKR or in the study of RNA-binding proteins and theirinterplay with translation.

2. Expression and Purification of PKR

Central to the biophysical characterization of any biomolecule is theproduction of a homogeneous and high-yield sample. The expression andpurification of recombinant human PKR in bacteria is particularly challeng-ing because of low yield, heterogeneous phosphorylation state, and non-specific cellular RNA binding. We present here our current protocol for thepurification of PKR and various PKR derivatives (Fig. 14.2A).

Full-length human PKR was subcloned into pET-29b(þ) vector withNdeI (50 end) and KpnI (30 end) such that it could be expressed and purifiedas C-terminal 6# His fusion protein under control of a T7 promoter. Boththis vector and a pET-43.1-based vector containing the sequence forl-phosphatase were cotransformed into the Escherichia coli (E. coli) strainBL21(DE3)-RIG (Stratagene) for expression. Cultures are grown at 37$ toA590 ¼ 0.6 to 0.8 in LB media supplemented with appropriate antibiotics,and then rapidly cooled on ice for 15 min. After cooling, expression of bothPKR and l-phosphatase is induced with isopropyl b-D-thiogalactopyrono-side (IPTG, 1 mM) for 18 h at 20$. We have observed that the combinationof cooling and induction at 20$ can significantly improve yield by reducingthe incorporation of PKR into inclusion bodies.

All subsequent steps were performed at 4$ to minimize proteolysis. Cellsare harvested by centrifugation at 4000 rpm (Beckman JLA 8.1000 rotor)and resuspended in 10 ml/L culture of disruption buffer (50 mM TRIS/Cl,pH 8.0, 1 M NaCl, 5% glycerol, 5 mM b-mercaptoethanol) and lysed bygentle sonication. After centrifugation at 17,000 rpm for 30 min (BeckmanJA 25.50 rotor), the supernatant is applied to a 2.5 ml/L culture Ni-NTAcolumn (Qiagen) preequilibrated with disruption buffer. Bound PKR iswashed with 100 column volumes of His-A buffer (50 mM TRIS/Cl, pH8.0, 1 M NaCl, 5 mM b-mercaptoethanol, 1 mM imidazole), an essentialstep in the removal nonspecifically bound cellular RNA.We have estimatedwithout the high salt wash, nearly 80% of PKR molecules are contaminatedwith nucleic acids, whereas with the high-salt wash only 5 to 10% remainbound. The bound fusion protein is then washed with 10 column volumesof His-B buffer (50 mM TRIS/Cl, pH 8.0, 300 mM NaCl, 5% glycerol,5 mM b-mercaptoethanol, 10 mM imidazole) before elution in 25 ml ofHis-elution buffer (50 mM TRIS/Cl, pH 8.0, 300 mM NaCl, 100 mMimidazole, 5 mM b-mercaptoethanol). The eluate is then loaded onto a


Co-expression of 6XHis-PKR andl-phosphatase

24 h

Sonication

Ni-NTA column

Size exclusion chromatography

Size exclusion chromatography

l-phosphatase treatment

1 h

3 h

3 h

12 h

3 h

Homogeneous sample

A B

PKRP

ATP

Elution volume (mL)

AfterNi-NTAcolumn

Afterl-phosphatase

treatment

A28

0

C

100 150 200 250

400

350

300

250

200

150

100

50

0100 150 200 250 300 350

Elution volume (mL)

A28

0

Figure 14.2 Expression andpurificationofhumanPKR. (A)Schematicoutlineof PKRsample preparation. Estimated time for each step is shown. (B) Purification of humanPKRby size exclusion chromatography. Elution profile obtained from the size-exclusionchromatography step on the Superdex 200 (26/60) column after either Ni-NTA column(top) or l-phosphatase treatment (bottom) are shown. Traces corresponding to both280 nm (solid) and 260 nm (dashed) are shown. (C) Superdex 200 26/60 elution profile ofpurified PKRP (15 mM) activated in the absence of dsRNA. PKRP and ATP peaks areindicated.


Superdex 200 26/60 size exclusion chromatography column (GEHealthcare)with a 50-ml Superloop (GEHealthcare) at 2 ml/min (50 mM TRIS/Cl, pH8.0, 300 mM NaCl, 5 mM b-mercaptoethanol), and eluted in 5-ml fractionsat a flow rate of 2 ml/min. The chromatographic profile typically revealsfour distinct peaks that correspond to RNA-bound PKR (&110 ml), PKR(&190 ml), and small molecular weight compounds (&280 ml) (Fig. 14.2B,top). This stage is critical, because RNA-contaminated PKR is completelyremoved from our sample. The monomeric PKR peak is pooled.

The final purification step involves the removal of phosphorylationheterogeneity from our sample. Although PKR coexpression with l-phos-phatase dephosphorylates most of our PKR population, we have observedthat without complete dephosphorylation, constitutive kinase activation inthe absence of dsRNA activator can occur. Therefore, the PKR pool issubsequently treated with lPPase (25 kDa, New England Biolabs) over-night at 4$ in the presence of 2 mM MnCl2 to ensure complete dephos-phorylation. We have confirmed the dephosphorylation state by bothactivation assays and NMR analysis (McKenna, 2007). The sample is thentreated with an excess of ethylenediaminetetraacetic acid (EDTA) to chelatedivalent cations and reapplied onto a Superdex 200 26/60 size exclusionchromatography column by means of a 50-ml Superloop at 2 ml/min(50 mM TRIS/Cl, pH 7.5, 100 mM NaCl, 5 mM b-mercaptoethanol),and eluted in 5-ml fractions at a flow rate of 2 ml/min (Fig. 14.2B, bottom).Samples are then concentrated (Vivaspin 20, 10K MWCO, PES mem-brane) as desired. The 6# His fusion can be removed, if desired, byincubation overnight at 4$ with Thrombin (Novagen), although the taggedand untagged versions appear functionally equivalent in all assays examined.Protein purity and yield were assessed by SDS-PAGE analysis (4 to 20%TRIS HCl gels), Bradford assay (Bio-Rad) and/or A280 values by use of acalculated molar extinction coefficient. Typical yields are approximately5 to 10 mg per initial liter of liquid culture. Full-length PKR is typicallystored at 4$ at low concentrations (<2 mM) to avoid potential RNA-independent activation and is stable if periodically supplemented withb-mercaptoethanol. Freeze/thaw cycles of PKR significantly affect theactivity of the enzyme.

PKR offers the advantage that each of its domains can be expressedand purified independently, although activity cannot be restored by addingthe domains in trans, even in the presence of the interdomain linker(McKenna, 2007). The C-terminal kinase domain (PKRKD, residues252 to 551) is expressed and purified in an identical manner to the full-length protein, although the extensive high-salt wash can be reducedsignificantly, because this construct does not associate with bacterialRNA. The N-terminal dsRBD1/2 (residues 1 to 169), dsRBD1 (residues1 to 99), and dsRBD2 (residues 106 to 169) constructs are expressedand purified in an identical manner to the full-length protein, with the


following exceptions: (1) neither rapid cooling before induction nor growthat 20$ is necessary; (2) after induction, the dsRBD constructs are grown at37$ for 3 h before harvesting; and (3) a Superdex 75 26/60 size exclusioncolumn is typically used to achieve maximal separation from RNA-boundprotein.

Phosphorylated PKR can be generated by incubating purified full-lengthhuman PKR at high concentrations (>15 mM) in the presence of ATP(1 mM) and MgCl2 (2 mM) at 30$ for 2 h in a buffered solution containingTRIS HCl (pH 7.5), 100 mM NaCl, and 5 mM b-mercaptoethanol. Phos-phorylated PKR is then purified in the same buffer on a Superdex 26/60 sizeexclusion column (Fig. 14.2C). It is important to note that PKR pro-duced in this manner appears homogeneously phosphorylated (McKenna,2007), whereas bacterially expressed PKR that has not been treated withl-phosphatase is not.

3. RNA Synthesis

Central to the study of the PKR activation is the production and use ofbiologically relevant dsRNA ligands for the kinase. Whereas many groupshave chosen to use poly I ' C (Sigma) as the model RNA ligand, its largesize and broad molecular weight distribution are unfavorable for bothmechanistic and high-resolution structural studies. In addition to canonicalA-form dsRNA, many RNAs, including 30-UTR of a-tropomyosin, inter-feron-g mRNA, human hepatitis d agent RNA, adenovirus VAI RNA,Epstein–Barr virus EBER-1 and -2 RNAs, and HIV TAR, regulate PKRactivation (Ben-Asouli et al., 2002; Bommer et al., 2002; Maitra et al., 1994;Osman et al., 1999; Spanggord and Beal, 2001; Vuyisich et al., 2002). Allknown viral activators of PKR are distorted dsRNA helices, possessingstructural features beyond simple duplex, such as bulges and hairpin loops.Length and RNA structural features are also thought to modulate theinteraction between the dsRBDs and dsRNAs (Bevilacqua and Cech,1996; Kim et al., 2006; McKenna et al., 2006). Therefore, the validation ofa mechanism of PKR activation should include a panel of viral dsRNAmolecules. The favored approach for dsRNA ligand production is in vitrotranscription by use of T7 polymerase. The design and transcription ofhigh-yield RNA samples for biophysical studies have been recently dis-cussed in depth elsewhere (Kim, 2007; Lukavsky and Puglisi, 2004), and,therefore, we will provide only a brief overview of the procedure here(Fig. 14.3A).

A PCR product is generated that contains, in sequential order, the T7RNA polymerase promoter, desired RNA template sequence to be tran-scribed, and restriction endonuclease site for vector linearization (Fig. 14.3B).


The PCR product is then ligated into a pUC119 high copy plasmid, whichis used as the template for in vitro transcription. After restriction enzyme(BstZ17I or BsaI) treatment, the linearized plasmid template is used forin vitro transcription by use of T7 polymerase under standard conditions(Kim et al., 2007). After 2 to 3 h of incubation at 37$, pyrophosphateprecipitate is pelleted (3000g # 10 min), and the reaction quenched byadding EDTA. An equal volume of phenol/chloroform is added to thesupernatant, mixed and centrifuged (3000g # 10 min). In this process, therestriction enzyme, which is carried over with the linearized plasmid DNA,and the T7 RNA polymerase are removed from the reaction mixture,because only the aqueous phase is retained.

After three phenol/chloroform extractions, the aqueous phase (upper) isdirectly loaded on a desalting column (10DG column, Bio-Rad) to removephenol and a significant amount of remaining NTPs. After application to adesalting column, the eluted transcription reaction is directly loaded on a sizeexclusion column without requiring any further treatments (i.e., precipita-tion or concentration). Milligram quantities of pure RNA can be preparedfrom transcription in a single day.

4. Phosphorylation Assays

To probe the mechanism of PKR activation, we developed simpleassays to assess kinase activity quantitatively. The general method used32P-gATP as a probe after the kinase reaction by incorporation of the labelinto protein. PKR is a Ser/Thr protein kinase that undergoes trans-autophosphorylation on productive interaction with dsRNA.

A

B

In vitrotranscription

Phenol/chloroformextraction

Desaltingcolumn

Size exclusionchromatography

3 h RNAsample

3 h1 h1 h

AAGCTTAATACGACTCACTATAGGGGGGGGGGGGCGAGACCGAATTCTTCGAATTATGCTGAGTGATATCCCCCCCCCCCCGCTCTGGCTTAAG

5!3!

HindIII BsaI EcoRI

T7 promoter Templateregion

3!5!

Figure 14.3 Design and transcription of RNA ligands. (A) Schematic outline of RNAsample preparation. Estimated time for each step is shown. (B) SyntheticDNAtemplateused for in vitro transcription of RNA ligand. Restriction sites for vector ligation (Hin-dIII, EcoRI) and plasmid linearization (BsaI) are indicated.T7 promoter and hypotheti-cal RNA template regions are boxed. Note that the bottom strand (30^50) serves as thetemplate for synthesis of the desiredRNA.


Quantitation of the number of autophosphorylation sites on PKR hasvaried greatly, indicating that between 1 and 15 sites exist (Lemaire et al.,2005; McKenna et al., 2006). However, there is general agreement thatphosphorylation at T446 and T451 seem to be sufficient for ultimateactivation of the kinase activity and substrate recognition (Dar et al.,2005; Dey et al., 2005; McKenna et al., 2007). On phosphorylation,PKR is capable of phosphorylating its target substrates; the best character-ized being the a-subunit of eIF2B on Ser51 (Chong et al., 1992;Sonenberg and Dever, 2003). Therefore, phosphorylation assays typicallyfollow the incorporation of radioactivity into either PKR itself, or onto atarget substrate such as eIF2a.

Autophosphorylation assays are performed on PKR (100 nM) in 50 mMTRIS/Cl (pH 7.5), 100 mM NaCl, 1 mM ATP, 2 mM MgCl2, andsupplemented 1 mCi of 32P-gATP. We typically add activator dsRNA infourfold molar excess at this protein concentration. Submicromolar concen-trations for both PKR and RNA are chosen to minimize RNA-independentactivation that has been observed at higher concentrations. Reactions areincubated at 30$ for 0 to 120 min, and quenched with 5# SDS-PAGE loadmix. Proteins are separated on a 4 to 20% SDS-PAGE gel (Bio-Rad), dried for30 min at 80$, and autoradiographed (GEHealthcare) to detect the incorpora-tion of 32P into PKR. Band intensities can be quantitated by use of the ImageQuant software program (GE Healthcare) or ImageJ (NIH) (Abramoff et al.,2004). From reaction setup to quantitation typically takes <24 h.

Numerous applications of this assay have been used successfully to probethe mechanism of PKR kinase activity. First, autophosphorylation assaysserve as a quality control for PKR preparations; any PKR that remainsassociated with bacterial RNA or remains phosphorylated will result indsRNA-independent autophosphorylation and reduce the validity of results.To test our preparations, we perform autophosphorylation assays in thepresence and absence of a standard dsRNA activator (HIV-TAR; Kimet al., 2006) and ensure that no incorporation of 32P is observed into PKRunless dsRNA addition has occurred (Fig. 14.4A, lanes 1 and 2). Eachpreparation of full-length PKR is tested in this manner before use in bothbiochemical and structural studies to ensure a trustworthy sample.

The dsRNA concentration dependence on PKR autophosphorylationcan be examined (Fig. 14.4A). In this variation of the basic autophosphor-ylation assay, the concentration of PKR is fixed (100 nM), whereas theconcentration of dsRNA activator is increased from zero to a vast molarexcess.

The time-dependence of autophosphorylation yields kinetic parametersfor PKR. The standard reaction containing [g-32P]ATP, PKR, and dsRNAactivator is incubated for a series of time points (0 to 120 min), quenched withEDTA, separated by denaturing SDS-PAGE, and quantified by autoradiogra-phy as described previously. A sigmoidal buildup of autophosphorylated PKR


is typically observedwith respect to time,with a lag phase beforemaximal ratesof autophosphorylation (Lemaire et al., 2005, 2006; McKenna et al., 2006)(Fig. 14.4B). These kinetics are characteristic of an autocatalytic process, andas such are well fit by a second order, bimolecular mechanism. By obtainingkinetic data at different PKR–dsRNA complex concentrations, global fittingof these data simultaneously is possible, resulting in extraction of kineticparameters (including apparent KM and kcat) (Wang and Wu, 2002).

Phosphorylated PKR (PKRP) can be used as a kinase capable of trans-phosphorylating PKR, PKR mutants, and PKR derivatives (Fig. 14.4C).These experiments remove the need for RNA activators completely, and,

A

C

B

0 0.5 210.2 43dsRNA

(fold-excess):

32P-PKR

32P-PKR

Time (min)

32P-

PKR

(mM

)

++

+

+

−

+++

+−−

+ +−−

−−+ +TAR:

PKR:PKRP:

0.16

0.12

0.08

0.04

00 20 40 60 80 100 120

Figure 14.4 PKR phosphorylation assays. (A) Purified PKR (100 nM) was incubatedwith increasing amounts of HIV-TAR RNA activator in the presence 32P-gATP at 30$for 90 min and resolved on denaturing SDS-PAGE gels. Gels were dried and autoradio-graphed to quantitate the extent of PKR autophosphorylation. (B) Progress curve ofPKR autophosphorylation in the presence of equimolarTAR (100 nM) and 32P-gATP at30$. Samples were quenched at the time point indicated and quantitated as in (A).(C) Trans-autophosphorylation assays in which mixtures of PKRP (100 nM), TAR(300 nM), PKR (300 nM), or purified PKR-TAR complex (300 nM, boxed) were incu-bated in the presence of [g-32P]ATP at 30$ for 15 min. The reactions were quenched,resolved, and quantified as in (A).


therefore, one can dissect the ability of PKR to serve as a substrate fortrans-phosphorylation in various states that would not otherwise bepossible (i.e., mutations, bound to activator or inhibitory dsRNAs). Trans-autophosphorylation of PKR by PKRP occurs at a 20-fold faster ratethan dsRNA-mediated autophosphorylation (McKenna et al., 2007) and,therefore, requires only short incubation times (10 min) before quenching.Short reaction times also limit potential side reactions when examiningRNA-bound PKR derivatives for their ability to serve as substrates fortrans-autophosphorylation. We have observed that full-length PKR isequivalently able to serve as a substrate when free or bound to activatordsRNA molecules, but is not competent when bound to viral dsRNAinhibitors. Whereas mutations to the ATP coordination site (PKRK296R)result in wild-type phosphorylation levels, mutations at either Thr446 orThr451 in the kinase domain attenuate phosphorylation.

Last, substrate phosphorylation can be incorporated into any of the assayspreviously described to probe the ultimate kinase activity of PKR. Typicalsubstrates have included eIF2a and (histone 2A) H2A. Rates of substratephosphorylation can be significantly faster than autophosphorylation ratesand should be taken into account when selecting time points.

5. Measuring RNA–Protein Stabilities

The kinase activity of PKR is directly modulated by the interaction withdsRNAmolecules (Ben-Asouli et al., 2002; Bommer et al., 2002;Maitra et al.,1994; Osman et al., 1999; Spanggord and Beal, 2001; Vuyisich et al., 2002),and, therefore, methods to probe RNA–protein stabilities are paramount.Highly quantitative techniques are preferable to differentiate affinities betweenvarious RNA ligands and assess the contribution that individual domains inPKR make to the RNA–protein interaction. These stability experiments canthen be correlated with activation assays discussed in the previous sectionto gain insight into the coupling of RNA binding and kinase activation. Wehave used isothermal titration calorimetry (ITC) and native gel shifts to studyRNA–protein thermodynamics.

ITC uses stepwise injections of one reagent into a highly sensitivecalorimetric cell containing a second reagent to accurately measure thechange in enthalpy, DH (Ababou and Ladbury, 2006; Velazquez-Campoyet al., 2004). Because the heat evolved during each injection is proportionalto complex formation, the equilibrium binding constant (KA) is determined.From these results, an entire suite of thermodynamic parameters can becalculated (DH,KD, DG, DS) as can the stoichiometry (n) of the interaction.Other approaches such as surface plasmon resonance (SPR) or analyticalultracentrifugation are also useful techniques but measure only bindingaffinity. SPR is further hampered by the requirement for one reagent to


be surface anchored, which can introduce experimental artifacts. Therefore,ITC is a quantitaive technique well suited to characterize RNA–proteinstabilities in the PKR system.

A VP-ITC microcalorimeter (Microcal) has been used to analyze thethermodynamics of PKR–RNA complexes. The instrument is typically cali-brated by injecting a 10% methanol solution (syringe, 320 ml capacity) intowater (sample cell, 1.4 ml capacity) as a control to ensure a stable baseline andrepetitive evolution of heat. The core of the ITC instrumentation is anextremely sensitive calorimeter, and, therefore, even small sample imperfec-tions can create uninterpretable results. All samples are centrifuged in atabletop centrifuge extensively and degassed under vacuum immediatelybefore use. The sample cell is also routinely cleaned with detergent (10%Contrad-70) to eliminate solid precipitates. Furthermore, extreme care istaken to eliminate small air bubbles from both the sample cell and syringe.We recommend that the sample cell contain the RNA ligand (&5 to 10 mMRNA, in 50 mMTRIS/Cl, pH7.5, 100 mMNaCl) and that the concentratedprotein solution (&50 to 200 mM, identical buffer) is loaded into the syringe.Sample cell and syringe contents are chosen to maintain an excess of RNAover PKR to prevent the formation of nonspecific PKR–RNA aggregates. Itshould also be noted that the concentrations required in both sample cell andsyringe will vary on the basis of the specific interaction being examined.Titrations are performed at 30$, such that the results can be correlated withactivation assays, DLS experiments, and NMR studies performed at the sametemperature. A standard experiment involves 25 to 30 injections of 10 ml (250to 300 ml total) into the sample cell containing RNA, and, therefore, signifi-cant amounts of both protein and RNA are required. Titration curves are fitby a nonlinear least squares method with a model for one or two binding sitesby use of Microcal Origin (version 5.0) to extract thermodynamic parameters.Outlying data points (typically the first two to three acquired) are discardedbefore analysis. In all cases involving PKR–RNA interactions a single high-affinity site and a second weak affinity nonspecific site were reported, and,therefore, data were fit with a model for two binding sites (Kim et al., 2006;McKenna et al., 2006). Reported results are from the high-affinity binding siteonly. From these data, the changes in entropy (DS ) and free energy (DG) werecalculated by use of the following equations:

DG ¼ !RT lnK

DG ¼ DH ! TDS

where R is the universal molar gas constant, and T is the temperature inKelvin. The accuracy of curve fitting results is highly dependent on theaccuracy of the reagent concentrations; accurate concentration determinationis likely the largest source of error in the determination of thermodynamic


parameters by ITC. For reliable results, we recommend performing a singleITC measurement in at least triplicate.

ITC has been used successfully to delineate the thermodynamic con-tributions that both specific structural features in RNA ligands (i.e., loops,bulges, length) and the contributions that each domain of PKRmakes to theRNA–protein stability (i.e., dsRB1, dsRBD2, kinase domain, interdomainlinker) (Kim et al., 2006; McKenna et al., 2006, 2007). Furthermore, ITChas also been used to distinguish between specific and nonspecific bindingevents on the basis of measured complex stability. Reported dissociationconstants typically fall in the range of 50 to 150 nM for specific PKR–dsRNA complexes and are generally two orders of magnitude higher whennonspecific binding is occurring. A representative example of the use ofITC on a PKR-dsRNA sample is shown in Fig. 14.5A.

ITC

∆H

Molar ratio

kcal

/mol

dsR

BD1/

2mc

al/s

0 1 2 3

−2

−6

−10

0.4

0.0

−0.4

−0.8

A

B

− dsR

BD1/

2

PKR

RBD

1R

BP2

TAR

ITC

∆H

Molar ratio

kcal

/mol

dsR

BD1/

2mc

al/s

0 1 2 3

−2

−6

−10

0.4

0.0

−0.4

−0.8

Figure 14.5 MeasuringRNA^PKRcomplex stability. (A) ITC thermograms for titra-tion of a dsRNA ligand (left) or dsRNA negative control (right) (10 mM, sample cell)with dsRBD1/2 (200 mM, syringe), with the heat evolved shown on a per injection basis(top). ITC titration curve for the same reaction is also shown (bottom), where the heatevolved per injection is shown as a function of the molar ratio of dsRBD1/2 to RNA.Experiments were performed at 30$. (B) Native gel mobility shift experiment for HIVTARdsRNA (1 mM) binding to PKRand domain derivatives (1 mM) on 5%TBEgels.


Unfortunately, ITC is both sample and time intensive. Native gel shiftmobility assays, in which the RNA ligand of interest is mixed with proteinunder nondenaturing conditions, and RNA-containing complexes areresolved on a native gel, suffer from neither of these limitations and, therefore,provide an extremely useful complementary technique. Highly similar disso-ciation constants have been measured by use of the gel shift mobility assay andITC method on PKR–RNA complexes (Kim et al., 2006; McKenna et al.,2006). The negatively charged phosphate backbone of dsRNA makes itthe ideal ligand for native gel shift assays; addition of PKR binding partnersshifts the dsRNA band to a high molecular weight species, whereas no shiftis observed for unstable complexes. Although the inverse experiment ispossible, PKR does not migrate significantly into native protein gels unless itis phosphorylated.

Native gel shift mobility assays have proven extremely important in deli-neating regions in both PKR and dsRNA ligands required for complexformation and maintaining the stability of these complexes ( Jammi andBeal, 2001; Kim et al., 2006; McCormack and Samuel, 1995; McKennaet al., 2006, 2007; Schmedt et al., 1995; Zheng and Bevilacqua, 2000)(Fig. 14.5B). Both RNA and protein samples are prepared in a suitable buffer(50 mM TRIS/Cl, pH 7.5, 100 mMNaCl) and incubated at room tempera-ture for 10 min. Samples are then mixed with load mix (0.02% bromophenolblue, 0.01 xylene cyanol FF, 10% glycerol in 1XTBE), and loaded onto anondenaturing TBE gel (available commercially from Bio-Rad, or can beprepared in house). Sample running buffer contains 0.5#TBE (50 mMTRISbase, 41.5 mM boric acid, and 1 mM EDTA, final pH 8.3). Electrophoresis isperformed at 60 V at 4$, so as tominimize sample heating. The band detectionmethod required is dictated primarily by the concentration range used; gelscan be stained with 0.1% toluidine blue solution at micromolar concentra-tions, SybrGreenII fluorescent dye (Molecular Probes Inc.) at submicromolarconcentrations, or radiolabeled RNA for subnanomolar complex concentra-tions. It should be noted that only the latter two approaches are quantitative,whereas toluidine blue is used only for comparative purposes. By varying theprotein concentration at a fixed dsRNA concentration, macroscopic dissocia-tion constants can be calculated for the interaction between protein and RNAligand by fitting data to the following equation: fraction bound ¼ [PKR]/([PKR] þ KD) ( Jammi and Beal, 2001).

6. Monitoring the Association State of PKR

The activation of PKR is a bimolecular process, and multiple lines ofevidence support the importance of dsRNA in modulating PKR self-affinity,and ultimately kinase activation (Dar et al., 2005; Dey et al., 2005; Gabel et al.,2006; Lemaire et al., 2005; McKenna et al., 2007; Wu and Kaufman, 2004).


In this respect, dynamic light scattering (DLS) has proven an extremelyvaluable technique to monitor the association state of PKR.

The time dependence of the laser light scattered from a very small regionof solution is measured in DLS experiments (Schurr, 1977; Wilson, 2003).The rate of diffusion of molecules in and out of the region being studied isrelated to their diffusion coefficients and, therefore, hydrodynamic radius ofthe particles doing the scattering. The polydispersity of the sample is alsoassessed during fitting of the data, which reflects the uniformity of thesample. These measurements make DLS an ideal technique for the moni-toring of association state during PKR activation and as a screening tech-nique before high-resolution structural studies of RNA–protein complexeswhere monodispersity is favored.

DLS experiments are performed by loading a buffered sample of interestinto a small cuvette (<50 ml). We have used a DynaPro-801 molecularsizing instrument (Protein Solutions Co.) equipped with the DYNAMICS(version 6.0) software package to extract the hydrodynamic radius, apparentmolecular weight, and polydispersity on the basis of the autocorrelationanalysis of scattered light intensity data. It should be noted that under thisregimen, average values for species with similar tumbling times are reported(i.e., PKR and PKR-PKR dimer). Because DLS measures scattering phe-nomena, particulates or bubbles present in the sample cell are problematic.All samples are passed through a 0.22-mMMillex-GV syringe filter (Millipore)to remove particulates, and the sample cuvette is blown with compressed airbefore use. We recommend a total of 50 data points be collected at 10-secintervals and that outlying data be excluded from the use of the recom-mended statistical analyses in the DYNAMICS 6.0 software. We havesuccessfully performed experiments that use PKR concentrations rangingfrom 0.5 to 10 mg/ml, with the upper limit likely reflecting the solubilitylimit for PKR. The temperature at which the experiment is conductedcan be chosen at will; we typically select a standard temperature of 30$ forconsistency with other techniques. Samples should be preequilibrated atthis temperature before filtration; a significant increase in temperature onplacing the sample in the DLS apparatus can introduce bubbling in thecuvette during the DLS experiment.

Two examples of the types of experiments that can be performed by use ofDLS are shown in Fig. 14.6. Time-resolved experiments in which a singlesample is followed for an extended period time can reveal changes tothe hydrodynamic radius on the activation of PKR. Sample preparation andpreequilibration are particularly important for these experiments, because theyspend a comparatively long period of time in the DLS cuvette. Far simpler ismonitoring the dimerization of PKR at various concentrations; individualsamples are prepared for a species of interest at each concentration. In this case,DLS measurements are nicely supplemented by performing size exclusionchromatography and native gel shifts on identical complexes.


7. NMR Spectroscopy

The activation of PKR is a highly dynamic process governed primarilyby transient protein–RNA interactions (KD &0.1 mM) and protein–proteinself-association (KD &20 to 500 mM). A powerful tool in the characteriza-tion of PKR activation would be the high-resolution structure determina-tion of the full-length kinase in various states of activation: free, bound todsRNA, phosphorylated, and bound to substrate. Structure determinationof individual domains has been achieved (Dar et al., 2005; Nanduri et al.,1998b), but as of yet, the full-length protein and RNA–PKR complexeshave not been amenable to high-resolution structural studies. Fortunately,nuclear magnetic resonance (NMR) is particularly well suited to the studyof dynamic, weakly interacting systems in solution (Bonvin et al., 2005; Gaoet al., 2004; Zuiderweg, 2002) and has been extremely usefully in mappingregions of association between of PKR–RNA and PKR–PKR complexesat lower resolution.

The basis of the NMR experiment is that magnetically active nuclei (1H,13C, 15N) oriented in a strong magnetic field absorb radiation at specificradiofrequencies dictated by their local chemical environment. The resulting

B

[PKR] (mM)

PKRP

PKR-TAR

PKR

Time (min)

App

aren

t Mr (

kDa)

App

aren

t Mr (

kDa)

A110 130

120

110

100

90

800 10 20 30 40 50 60 70 80

105

100

95

90

85

800 20 40 60 80 100 120

PKR-TAR

PKR-TAR + ATP/Mg

PKRK296R-TAR + ATP/Mg

Figure 14.6 Dynamic light scattering of RNA^PKRcomplexes. (A) Time-dependentmolecular weight determination of PKR-TAR (diamonds), PKR-TARplus ATP/MgCl2(circles), orK296Rplus ATP/MgCl2 (squares, dashed line) byDLS. Equimolar complexes(2 mM) were incubated in the cuvette for the time specified before acquisition. Each datapoint was repeated in triplicate. Error bars have been omitted for clarity, but reflect lessthan 5%of the value observed in all cases. (B)Concentration-dependentdimerization ofPKRwas examined by determining the molecular weight at the specified concentrationof PKR,PKRP,orPKR-TAR.Eachdatapointwas repeated in triplicate.Errorbars are notshownforclarity, buterrorswere typically<5%of thevalue shown.


spectral lines, commonly referred to as chemical shift, are therefore exquisitelysensitive to changes local environment on perturbations to the system (i.e.,ligand binding, dimerization) for 15N, Ca, and Cb nuclei (Gao et al., 2004).These observations form the basis for a very useful technique known aschemical shift mapping; one molecule is isotopically enriched with NMR-active nuclei, whereas a binding partner is usually unlabeled. Comparison ofchemical shift perturbations before and after interaction allow for the detectionof potential binding sites, if resonance assignments are available. These resultsare enhanced when high-resolution structures of the unbound species areavailable, because binding sites can be mapped onto the molecular surface.

Mapping the surfaces of interaction on protein (PKR) is accomplishedthrough the 1H-15N heteronuclear single quantum coherence (HSQC)experiment. This experiment essentially provides a two-dimensional finger-print of each backbone amide and nitrogen-containing side chains in a 15N-isotopically enriched protein sample. On addition of unlabeled bindingpartner, chemical shift perturbations occur and residue-specific informationabout the interaction is gained (Fig. 14.7). Conversely, the 1D 1H imino

15N

(ppm

)

15N

(ppm

)

1H (ppm) 1H (ppm)

A B

Q365K136

E29

K261I503

L25 L30

Figure 14.7 NMRof 15N2H-PKR in the presence and absence of TARRNA. (A) 15N-TROSYHSQC spectral overlay of unbound PKR1-551 (black) andTARbound PKR1-551 (grey). NMRspectral datawere obtained by use of aVarianUnity INOVA 800-MHzspectrometer at 30$ and acquired by use of the sensitivity-enhanced gradient pulsescheme in Biopack (Varian Inc.). (B) Detailed viewof the spectra shown in (A), demon-strating chemical shift perturbations only to specific RNA binding resonances withinPKR1-551onTARbinding.


proton resonances (10 to 15 ppm range) are used to map the regions ofRNA ligand used for interaction with PKR. Because only the iminoprotons of G and U nucleotides involved in base-pair interactions can beobserved (Furtig et al., 2003), the imino proton spectra provides a finger-print of each of the base pairs in the RNA stem-loop structures. Anadditional bonus is that labeled RNA is not required; 1H resonances arisingfrom imino protons are usually nonoverlapping with those from protein(Fig. 14.8).

Isotopically enriched protein necessary for NMR studies is typicallyexpressed in supplemented M9 minimal media containing 1 mM MgSO4,0.1 mM CaCl2, 90 mM FeSO4, 0.01% (w/v) yeast extract, 1 mg thiamine-HCl, 1mg biotin, 2 g of D-glucose (or 13C-glucose), and 1 g NH4Cl (or15N-NH4Cl) per liter of liquid media. Deuterated protein, which reducessignal complexity and its increases its intensity in an NMR spectrum, isproduced by replacing H2O with the desired D2O/H2O ratio; no precon-ditioning of the cells in deuterated media is required for optimal growth.Purification is identical to that for unlabeled protein. NMR samples areprepared to a volume of approximately 250 ml (90% H2O, 10% D2O) in asuitable buffered saline solution in 5-mm Shigemi tubes (Shigemi Inc.).Samples are prepared by mixing components at low concentration(<10 mM) and concentrating to the desired concentration with a centrifugalconcentration device (Vivascience). When protein–RNA complexes areexamined, we also ensure that a slight excess of RNA is present to minimizethe formation of nonspecific interactions. Our best samples are typicallyprepared when these complexes are purified away from unbound compo-nents by size exclusion chromatography before NMR analysis.

A necessary precursor to chemical shift mapping is the complete assign-ment of backbone amide 1HN-15N cross-peaks in the 1H-15N HSQCNMR spectra of protein and imino resonances for RNA samples. Suchresults are already available for the dsRBDs (Nanduri et al., 1998a) andkinase domain (Gelev et al., 2006) of PKR. Resonance assignment repre-sents the rate-limiting step of this method and its full discussion are beyondthe scope of this chapter.

In the context of isolated domains, a weak interaction between dsRBD2and the kinase domain has been identified by use of 1H-15N HSQCs (Gelevet al., 2006; Li et al., 2006; Nanduri et al., 2000), although in the context ofthe full-length protein, no such changes in chemical shifts are observed(McKenna et al., 2007). The surfaces of interaction on both PKR and avariety of dsRNA ligands has also been mapped by use of chemical shiftperturbation (Kim et al., 2006; McKenna et al., 2006). Studies of this naturein the context of the full-length protein have allowed for novel insights intothe communication between the dsRBDs and kinase domain on RNAbinding and phosphorylation.


U13U38

U46G28

G12

U10, G36, U42

G3, G16, G44

U14

G21

G52, G2

G26G54

U8, G43

U50G53

G11

U13U38

U46G28

U10

G12, G36, U42

G3, G44

G16U14

G21, G52

G2G26G54G43

U8

U50

G53G11

ppm10.010.511.011.512.012.513.013.514.014.5

ppm10.010.511.011.512.012.513.013.514.014.5

U40

TAR (~20 kDa)

dsRBD1/2-TAR (~40 kDa)

NMR� approaches� other� than� chemical� shift� mapping� have� also� been� andwill� be� useful� for� probing� the� activity� of� PKR–RNA� complexes.� Theorientation� of� dsRNA� ligands� relative� to� the� dsRBDs� has� been� establishedby� incorporating� site-specific� spin-labeling� into� each� of� the� dsRBDs� indi-vidually� (I.� Kim,� personal� communication,� May� 2006).� These� approachesuse� spin� labels� (commonly� TEMPO,� a� stable� free� radical� nitroxide)incorporated� onto� cysteine� residues� in� proteins� or� the� termini� of� RNAmolecules� and� look� for� the� broadening� of� resonances� between� the� labeland� NMR� nuclei.� Because� the� effectiveness� of� broadening� is� proportional� tor!�6� distance,� resonances� spatially� close� to� the� spin-label� will� preferentiallybroaden� (Zuiderweg,� 2002).� Backbone� and� side� chain� relaxation� studies� byNMR� have� also� been� used� effectively� to� gain� insights� into� the� inherentflexibility of PKR, particularly in the dsRBDs (Barnwal et al., 2006;Nanduri et al., 2000). Taken together, it is clear that NMR approaches tostudy structural characteristics of PKR have provided major insights. Withadvances in NMR methods and segmental labeling strategies, we are opti-mistic that further information will be gained on this complex system.

8. In Vitro Translation Assays

Translation initiation is attenuated by PKR activation through thephosphorylation of eIF2a at Ser51 (Sonenberg and Dever, 2003). Establish-ing a link between mechanistic/structural observations made and actualeffects on the protein synthesis machinery provides an added layer ofconfidence to reported results. In vitro translation assays are a rapid, straight-forward method to establish this connection.

The basis for in vitro translation assays is a HeLa cell lysate that contains all ofthe necessary protein translation machinery (Fig. 14.9). The use of a reporterconstruct that contains a 50-capped luciferase mRNA (cap-Luc), the activity(intensity of luminescence from substrate) of cap-Luc is determined by theluciferase assay system (Promega) by use of an appropriate luminometer.Preparation of HeLa cell extracts is discussed elsewhere (Otto and Puglisi,2004). Pellets (approximately 1 # 106 cells) from a 1L HeLa S3 (the NationalCell Culture Center) culture are suspended in 1.25 ml of hypotonic buffer in a5-ml-Dounce homogenizer. After 10 min, 225 ml hypertonic buffer is added,

Figure 14.8 Binding interface determined by imino proton titration NMR experi-ments. Imino proton regions of 1D 1H NMR spectra of free HIV TARRNA (top) andcomplex with dsRBD1-dsRBD2 (bottom). Each resonance peak corresponds to aspecific base-paired G or U nucleotide and can, therefore, be used to probe regions ofinteraction with binding partners. Assignments are indicated by numbers above theresonance peaks.


and cells are lysed by 20 rapid strokes of the homogenizer. The lysate is thenspun and the supernatant retained.

Inhibition of cap-dependent translation by PKR can be determined byadding purified PKR (70 nM) and RNA ligands (50 nM) to the lysatecontaining cap-Luc (50 nM). Reporter RNA is preincubated at 65$ for3 min, and then immediately cooled in ice-cold water before addition. TheHeLa S3 S10 lysate (50% vol) is supplemented with 0.5 U/ml RNaseinhibitor, SUPERase'In (Ambion), 25 mM amino acid mixture, complete(Promega), 60 mM KCl (for cap-dependent translation), and 1 mM MgCl2(required for activation of PKR). Together the translation reaction isincubated at 30$ for 60 min before luciferase activity assay.

The phosphorylation state of PKR and target substrates can also bemonitored from the same in vitro translation assay to correlate these datawith translational competency. An approach that has proven useful is toperform immunoblot analysis by use of site-specific antibodies for detection.Phosphorylated or total (phosphorylated and unphosphorylated) PKR andeIF2a are detected by use of the following antibodies: anti-PKR (preparedfrom rabbit serum raised against the purified kinase domain of PKR) fordetection of total PKR, sc-16565-R (Santa Cruz Biotechnology) for detec-tion of PKRphosphorylation at Thr446, ab5359-50 (abcam) for detection oftotal eIF2a, and ab4837-50 for detection of eIF2a phosphorylation at Ser51.

9. Conclusions

We have outlined an ensemble of experimental approaches to probethe mechanistic features of PKR, a kinase central to translation initiationregulation on viral infection of cells. PKR is a highly dynamic protein thatundergoes ligand binding, dimerization, phosphorylation, and ultimately

m7G Luc. mRNA

Translation + Luciferase substrate

Luminescence

PKR dsRNA

S10 Lystae

Figure 14.9 In vitro translation assay to monitor PKR activity. S10 HeLa cell lysatescontaining the protein translationmachinery are incubated in the presence of 50-cappedluciferase mRNA. If translation proceeds efficiently, addition of luciferase substratesgenerates luminescence, which can be detected and quantitated. Addition of PKR anddsRNA ligands allows for themonitoring of their effects on translation.


kinase activation. A panel of different methods was presented to investigateeach aspect of PKR activity. Reliable, homogeneous PKR sample prepara-tion is key to the usefulness of these approaches, and methods for expressingand purifying the enzyme were presented. Although high-resolution struc-tures of PKR in various states is an ultimate goal in the field, lowerresolution structural approaches should prove extremely useful in delineat-ing the mechanism whereby RNA binding is couple to kinase activationand subsequent translational inhibition.

REFERENCES

Ababou, A., and Ladbury, J. E. (2006). Survey of the year 2004: Literature on applications ofisothermal titration calorimetry. J. Mol. Recognit. 19, 79–89.

Abramoff, M. D., Magelhaes, P. J., and Ram, S. J. (2004). Image processing with ImageJ.Biophotonics Int. 11, 36–42.

Barnwal, R. P., Chaudhuri, T. R., Nanduri, S., Qin, J., and Chary, K. V. (2006). Methyldynamics for understanding hydrophobic core packing of dynamically different motifs ofdouble-stranded RNA binding domain of protein kinase R. Proteins 62, 501–508.

Ben-Asouli, Y., Banai, Y., Pel-Or, Y., Shir, A., and Kaempfer, R. (2002). Humaninterferon-gamma mRNA autoregulates its translation through a pseudoknot that acti-vates the interferon-inducible protein kinase PKR. Cell 108, 221–232.

Bevilacqua, P. C., and Cech, T. R. (1996). Minor-groove recognition of double-strandedRNA by the double-stranded RNA-binding domain from the RNA-activated proteinkinase PKR. Biochemistry 35, 9983–9994.

Bommer, U. A., Borovjagin, A. V., Greagg, M. A., Jeffrey, I. W., Russell, P., Laing, K. G.,Lee, M., and Clemens, M. J. (2002). The mRNA of the translationally controlled tumorprotein P23/TCTP is a highly structured RNA, which activates the dsRNA-dependentprotein kinase PKR. RNA 8, 478–496.

Bonvin, A. M., Boelens, R., and Kaptein, R. (2005). NMR analysis of protein interactions.Curr. Opin. Chem. Biol. 9, 501–508.

Chong, K. L., Feng, L., Schappert, K., Meurs, E., Donahue, T. F., Friesen, J. D.,Hovanessian, A. G., and Williams, B. R. (1992). Human p68 kinase exhibits growthsuppression in yeast and homology to the translational regulator GCN2. EMBO J. 11,1553–1562.

Clemens, M. J. (1997). PKR—a protein kinase regulated by double-stranded RNA. Int. J.Biochem. Cell Biol. 29, 945–949.

Dar, A. C., Dever, T. E., and Sicheri, F. (2005). Higher-order substrate recognition ofeIF2alpha by the RNA-dependent protein kinase PKR. Cell 122, 887–900.

Dey, M., Cao, C., Dar, A. C., Tamura, T., Ozato, K., Sicheri, F., and Dever, T. E. (2005).Mechanistic link between PKR dimerization, autophosphorylation, and eIF2alphasubstrate recognition. Cell 122, 901–913.

Furtig, B., Richter, C., Wohnert, J., and Schwalbe, H. (2003). NMR spectroscopy of RNA.Chembiochem 4, 936–962.

Gabel, F., Wang, D., Madern, D., Sadler, A., Dayie, K., Daryoush, M. Z., Schwahn, D.,Zaccai, G., Lee, X., and Williams, B. R. (2006). Dynamic flexibility of double-strandedRNA activated PKR in solution. J. Mol. Biol. 359, 610–623.

Gale, M., Jr., and Katze, M. G. (1998). Molecular mechanisms of interferon resistancemediated by viral-directed inhibition of PKR, the interferon-induced protein kinase.Pharmacol. Ther. 78, 29–46.


Gale, M., Jr., Tan, S. L., and Katze, M. G. (2000). Translational control of viral geneexpression in eukaryotes. Microbiol. Mol. Biol. Rev. 64, 239–280.

Gao, G., Williams, J. G., and Campbell, S. L. (2004). Protein-protein interaction analysis bynuclear magnetic resonance spectroscopy. Methods Mol. Biol. 261, 79–92.

Gelev, V., Aktas, H., Marintchev, A., Ito, T., Frueh, D., Hemond, M., Rovnyak, D.,Debus, M., Hyberts, S., Usheva, A., et al. (2006). Mapping of the auto-inhibitoryinteractions of protein kinase r by nuclear magnetic resonance. J. Mol. Biol. 364, 352–363.

Jammi, N. V., and Beal, P. A. (2001). Phosphorylation of the RNA-dependent proteinkinase regulates its RNA-binding activity. Nucleic Acids Res. 29, 3020–3029.

Kim, I., Liu, C.W., and Puglisi, J. D. (2006). Specific recognition of HIV TARRNA by thedsRNA binding domains (dsRBD1-dsRBD2) of PKR. J. Mol. Biol. 358, 430–442.

Kim, I., McKenna, S. A., Puglisi, E. V., and Puglisi, J. D. (2007). Rapid purification ofRNAs using Fast Performance Liquid Chromatography (FPLC). RNA 13, 1–6.

Lemaire, P. A., Lary, J., and Cole, J. L. (2005). Mechanism of PKR activation: Dimer-ization and kinase activation in the absence of double-stranded RNA. J. Mol. Biol. 345,81–90.

Lemaire, P. A., Tessmer, I., Craig, R., Erie, D. A., and Cole, J. L. (2006). Unactivated PKRexists in an open conformation capable of binding nucleotides. Biochemistry 45,9074–9084.

Li, S., Peters, G. A., Ding, K., Zhang, X., Qin, J., and Sen, G. C. (2006). Molecular basis forPKR activation by PACT or dsRNA. Proc. Natl. Acad. Sci. USA 103, 10005–10010.

Lukavsky, P. J., and Puglisi, J. D. (2004). Large-scale preparation and purification ofpolyacrylamide-free RNA oligonucleotides. RNA 10, 889–893.

Maitra, R. K., McMillan, N. A., Desai, S., McSwiggen, J., Hovanessian, A. G., Sen, G.,Williams, B. R., and Silverman, R. H. (1994). HIV-1 TAR RNA has an intrinsic abilityto activate interferon-inducible enzymes. Virology 204, 823–827.

Malmgaard, L. (2004). Induction and regulation of IFNs during viral infections. J. InterferonCytokine Res. 24, 439–454.

McCormack, S. J., Ortega, L. G., Doohan, J. P., and Samuel, C. E. (1994). Mechanism ofinterferon action motif I of the interferon-induced, RNA-dependent protein kinase(PKR) is sufficient to mediate RNA-binding activity. Virology 198, 92–99.

McCormack, S. J., and Samuel, C. E. (1995). Mechanism of interferon action: RNA-binding activity of full-length and R-domain forms of the RNA-dependent proteinkinase PKR—determination of KD values for VAI and TAR RNAs. Virology 206,511–519.

McKenna, S. A., Kim, I., Liu, C.W., and Puglisi, J. D. (2006). Uncoupling of RNA bindingand PKR kinase activation by viral inhibitor RNAs. J. Mol. Biol. 358, 1270–1285.

McKenna, S. A., Lindhout, D. A., Kim, I., Liu, C. W., Gelev, V. M., Wagner, G., andPuglisi, J. D. (2007). Molecular framework for the activation of PKR. J. Biol. Chem. 285,11474–11486.

Nanduri, S., Carpick, B., Yang, Y., Williams, B. R., and Qin, J. (1998a). 1H, 13C, 15Nresonance assignment of the 20 kDa double stranded RNA binding domain of PKR.J. Biomol. NMR 12, 349–351.

Nanduri, S., Carpick, B. W., Yang, Y., Williams, B. R., and Qin, J. (1998b). Structure ofthe double-stranded RNA-binding domain of the protein kinase PKR reveals themolecular basis of its dsRNA-mediated activation. EMBO J. 17, 5458–5465.

Nanduri, S., Rahman, F., Williams, B. R., and Qin, J. (2000). A dynamically tuned double-stranded RNA binding mechanism for the activation of antiviral kinase PKR. EMBO J.19, 5567–5574.

Osman, F., Jarrous, N., Ben-Asouli, Y., and Kaempfer, R. (1999). A cis-acting element inthe 30-untranslated region of human TNF-alpha mRNA renders splicing dependent onthe activation of protein kinase PKR. Genes Dev. 13, 3280–3293.


Otto, G. A., and Puglisi, J. D. (2004). The pathway of HCV IRES-mediated translationinitiation. Cell 119, 369–380.

Romano, P. R., Garcia-Barrio, M. T., Zhang, X., Wang, Q., Taylor, D. R., Zhang, F.,Herring, C., Mathews, M. B., Qin, J., and Hinnebusch, A. G. (1998). Autophosphor-ylation in the activation loop is required for full kinase activity in vivo of human and yeasteukaryotic initiation factor 2alpha kinases PKR and GCN2. Mol. Cell Biol. 18,2282–2297.

Schmedt, C., Green, S. R., Manche, L., Taylor, D. R., Ma, Y., and Mathews, M. B. (1995).Functional characterization of the RNA-binding domain and motif of the double-stranded RNA-dependent protein kinase DAI (PKR). J. Mol. Biol. 249, 29–44.

Schneider, R. J., and Mohr, I. (2003). Translation initiation and viral tricks. Trends Biochem.Sci. 28, 130–136.

Schurr, J. M. (1977). Dynamic light scattering of biopolymers and biocolloids. CRC Crit.Rev. Biochem. 4, 371–431.

Sonenberg, N., and Dever, T. E. (2003). Eukaryotic translation initiation factors andregulators. Curr. Opin. Struct. Biol. 13, 56–63.

Spanggord, R. J., and Beal, P. A. (2001). Selective binding by the RNA binding domain ofPKR revealed by affinity cleavage. Biochemistry 40, 4272–4280.

Taylor, D. R., Lee, S. B., Romano, P. R., Marshak, D. R., Hinnebusch, A. G., Esteban, M.,and Mathews, M. B. (1996). Autophosphorylation sites participate in the activation of thedouble-stranded-RNA-activated protein kinase PKR. Mol. Cell Biol. 16, 6295–6302.

Vattem, K. M., Staschke, K. A., Zhu, S., andWek, R. C. (2001). Inhibitory sequences in theN-terminus of the double-stranded-RNA-dependent protein kinase, PKR, are impor-tant for regulating phosphorylation of eukaryotic initiation factor 2alpha (eIF2alpha).Eur. J. Biochem. 268, 1143–1153.

Velazquez-Campoy, A., Leavitt, S. A., and Freire, E. (2004). Characterization of protein-protein interactions by isothermal titration calorimetry. Methods Mol. Biol. 261, 35–54.

Vuyisich, M., Spanggord, R. J., and Beal, P. A. (2002). The binding site of the RNA-dependent protein kinase (PKR) on EBER1 RNA from Epstein-Barr virus. EMBO Rep.3, 622–627.

Wang, Z. X., and Wu, J. W. (2002). Autophosphorylation kinetics of protein kinases.Biochem. J. 368, 947–952.

Wilson, W.W. (2003). Light scattering as a diagnostic for protein crystal growth—a practicalapproach. J. Struct. Biol. 142, 56–65.

Wu, S., and Kaufman, R. J. (1997). A model for the double-stranded RNA (dsRNA)-dependent dimerization and activation of the dsRNA-activated protein kinase PKR.J. Biol. Chem. 272, 1291–1296.

Wu, S., and Kaufman, R. J. (2004). Trans-autophosphorylation by the isolated kinasedomain is not sufficient for dimerization or activation of the dsRNA-activated proteinkinase PKR. Biochemistry 43, 11027–11034.

Zheng, X., and Bevilacqua, P. C. (2000). Straightening of bulged RNA by the double-stranded RNA-binding domain from the protein kinase PKR. Proc. Natl. Acad. Sci. USA97, 14162–14167.

Zuiderweg, E. R. (2002). Mapping protein–protein interactions in solution by NMRspectroscopy. Biochemistry 41, 1–7.


Documents

Chapter 14 - Biophysical and Biochemical Investigations of ... · CHAPTER FOURTEEN Biophysical and Biochemical Investigations of dsRNA-Activated Kinase PKR ... Figure 14.1 Regulation