role of intrinsic flexibility in signal transductioni · 2008-01-15 · role of intrinsic...

ZRT

ARTICLE IN PRESSYJMBI-60038; No. of pages: 12; 4C: 2, 3, 4, 5, 8

doi:10.1016/j.jmb.2007.12.016 J. Mol. Biol. (2007) xx, xxx–xxx

Available online at www.sciencedirect.com

Role of Intrinsic Flexibility in Signal TransductionMediated by the Cell Cycle Regulator, p27Kip1

Charles A. Galea1, Amanda Nourse2, Yuefeng Wang1,Sivashankar G. Sivakolundu1, William T. Heller3

and Richard W. Kriwacki1,4⁎

1Department of StructuralBiology, St. Jude Children'sResearch Hospital, 332 NorthLauderdale St., Memphis,TN 38105, USA2Hartwell Center forBioinformatics andBiotechnology, St. JudeChildren's Research Hospital,332 North Lauderdale St.,Memphis, TN 38105, USA3Center for StructuralMolecular Biology and ChemicalSciences Division, Oak RidgeNational Laboratory, Oak Ridge,TN 37831, USA4Department of MolecularSciences, University ofTennessee Health SciencesCenter, Memphis, TN 38163,USA

Received 12 September 2007;received in revised form27 November 2007;accepted 7 December 2007

*Corresponding author.DepartmentSt. Jude Children's Research HospitLauderdale St., Memphis, TN 38105,[email protected] used: Cdk, cyclin-d

intrinsically unstructured protein; Ntyrosine kinase; KID, kinase inhibitoheteronuclear single quantum coherheteronuclear nuclear Overhauser etransverse relaxation optimized spemolecular dynamics; AUC, analyticaSAXS, small-angle X-ray scattering.

0022-2836/$ - see front matter © 2007 E

Please cite this article as: Galea, C. A. etp27Kip1, J. Mol. Biol. (2007), doi:10.1016

p27Kip1 (p27), which controls eukaryotic cell division through interactionswith cyclin-dependent kinases (Cdks), integrates and transduces promito-genic signals from various nonreceptor tyrosine kinases by orchestrating itsown phosphorylation, ubiquitination and degradation. Intrinsic flexibilityallows p27 to act as a “conduit” for sequential signaling mediated bytyrosine and threonine phosphorylation and ubiquitination. While thestructural features of the Cdk/cyclin-binding domain of p27 are under-stood, how the C-terminal regulatory domain coordinates multistep sig-naling leading to p27 degradation is poorly understood. We show that the100-residue p27 C-terminal domain is extended and flexible when p27 isbound to Cdk2/cyclin A. We propose that the intrinsic flexibility of p27provides a molecular basis for the sequential signal transduction conduitthat regulates p27 degradation and cell division. Other intrinsicallyunstructured proteins possessing multiple sites of posttranslationalmodification may participate in similar signaling conduits.

© 2007 Elsevier Ltd. All rights reserved.

Keywords: cell cycle; cyclin-dependent kinase inhibitor; disordered protein;intrinsically unstructured protein; p27Kip1
Edited by P. Wright
of Structural Biology,al, 332 NorthUSA. E-mail address:

ependent kinase; IUP,RTK, nonreceptorry domain; HSQC,ence; hetNOE,ffect; TROSY,ctroscopy; MD,l ultracentrifugation;

lsevier Ltd. All rights reserve

al., Role of Intrinsic Flexibili/j.jmb.2007.12.016

Introduction

The cyclin-dependent kinase inhibitor p27Kip1

(p27)1–3 is a small intrinsically unstructured protein(IUP)4 that regulates cell proliferation throughinteractions with cyclin-dependent kinases (Cdks).5

For example, in G1 phase, an initially high level ofp27 blocks progression from G1 to S phase of thecell division cycle by inhibiting Cdk2/cyclin A andCdk2/cyclin E.6,7 The level of p27, which is con-trolled by translational regulation8 and ubiquitina-

d.

ty in Signal Transduction Mediated by the Cell Cycle Regulator,

http://dx.doi.org/10.1016/j.jmb.2007.12.016


mailto:http://[email protected]

Fig. 1. Structure of the p27/Cdk2/cyclin A complex.(a) Schematic view prepared with the program PyMOL[http://pymol.sourceforge.net] of full-length p27 boundto Cdk2/cyclin A showing the solvent-accessible surfacefor Cdk2 (cyan) and cyclin A (magenta). The kinaseinhibitory domain (KID) (grey) and C-terminal domain(yellow) of p27 are illustrated as ribbons. The structureshownwas obtained at 7.3 ns during theMD trajectory. (b)2-D 1H–15N HSQC spectrum of p27-C (green) overlaidwith the 2-D 1H–15N TROSY spectrum of 2H/15N-p27bound to unlabeled Cdk2/cyclin A (red) where over-lapping resonances are colored yellow. (c) 2-D 1H–15NHSQC spectrum of p27-KID (blue) overlaid with the 2-D1H–15N TROSY spectrum of 2H/15N-p27 bound tounlabeled Cdk2/cyclin A (red) where overlapping reso-nances are colored cyan.

2 p27 Kip1 Flexibility Mediates Signal Transduction

ARTICLE IN PRESS

tion-dependent proteolysis,8,9 must drop for Cdk2complexes to be fully activated and for cell divisionto progress. Ubiquitination of p27 at the G1/S tran-sition is regulated by a two-step mechanism thatinvolves phosphorylation of p27, first on tyrosine 88(Y88) by one of several nonreceptor tyrosine kinases(NRTKs) and second on threonine 187 (T187) byCdk2.10

Enhanced ubiquitination-mediated degradationof p2711 is common in human tumors12 and isassociated with poor clinical prognosis.13 p27 isubiquitinated through at least two pathways: (1)free, unphosphorylated p27 is ubiquitinated byKPC114,15 and (2) p27 bound to Cdk2/cyclin com-plexes is ubiquitinated by SCF/Skp2.12,16 Up-regulation of the two-step p27 phosphorylationmechanism by the oncogenic NRTKs, Bcr-Abl andSrc,10,17 was recently shown to promote SCF/Skp2-mediated p27 loss in chronic myelogenousleukemia10 and breast cancer cells,17 respectively.While the general features of the two-stepphosphorylation mechanism are understood, thefull extent to which the unusual structuralfeatures of p27—an intrinsically unstructuredprotein4—contribute to the mechanism is poorlyunderstood.The intrinsic flexibility of p27 allows its interac-

tions with Cdk2/cyclin A and SCF/Skp2 to bemodulated by phosphorylation of tyrosine andthreonine residues at opposite ends of the p27polypeptide chain. Prior to phosphorylation, Y88,which is found at the C-terminal end of the socalled “kinase inhibitory domain” (KID) of p27 thatbinds and inhibits Cdk2 (Fig. 1a), is lodged in theATP binding pocket of Cdk2 thereby inhibitingcatalysis by blocking access to ATP. However, dueto putative dynamics involving the 310 helixcontaining Y88 fluctuating into and out of theATP binding pocket,18 Y88 within p27/Cdk2/cyclin A complexes is accessible for phosphoryla-tion by NRTKs.10 Following phosphorylation—step1 of the two-step mechanism—Y88 and the entire310 helical segment of the KID are ejected from theATP binding pocket of Cdk2. While p27 remainsbound to Cdk2/cyclin A, ejection of phosphory-lated Y88 and the 310 helix restores significantcatalytic activity, allowing Cdk2 to phosphorylateT187 within the C-terminus of the same p27molecule that is bound to Cdk2 (step 2 of thetwo-step mechanism). The unimolecular nature ofstep 2 requires that the C-terminus of p27 “foldsback” to allow T187 to be phosphorylated byCdk2.10 While much is known about the structuraland dynamic features of the KID of p27 both in thefree state4,19 and when bound to Cdk2/cyclin A,18

detailed information on these properties for theC-terminal region of p27 is not available. Herein wereport results from a number of techniques thatfully describe the structural features of full-lengthp27 and explain how T187 can be phosphorylatedby Cdk2 via the pseudo unimolecular mechanism.These results highlight the importance of the in-trinsic flexibility of p27 in mediating the multistep

Please cite this article as: Galea, C. A. et al., Role of Intrinsic Flexibilip27Kip1, J. Mol. Biol. (2007), doi:10.1016/j.jmb.2007.12.016

signal transduction pathway that controls cell divi-sion in eukaryotes.

Results

NMR studies of the C-terminus of p27 in thep27/Cdk2/cyclin A ternary complex

p27 is composed of an N-terminal domain (p27-KID, residues 22–104) that binds and inhibits Cdk/cyclin complexes and a C-terminal domain (residues105–198, p27-C) that contains several sites ofposttranslational modification, including T187 (Fig.1a). While the structure of p27-KID in the freestate4,20 and bound to Cdk2/cyclin A10,18 have beencharacterized in detail, little information is availableon the structure of p27-C and how this domainorchestrates the two-step phosphorylation mechan-ism. We probed the structure and dynamics of thep27 C-terminus using NMR spectroscopy, first bystudying the isolated C-terminal domain (p27-C)


http://pymol.sourceforge.net

http://pymol.sourceforge.net


Fig. 2. Conformational flexibility of the C-terminaldomain of p27 in the p27/Cdk2/cyclin A complex.Superposition of p27/Cdk2/cyclin A structures at varyingtime intervals during a 12.7-ns MD simulation (5 ns,magenta; 6.1 ns, yellow; 7.2 ns, light green; 8.3 ns, cyan;9.4, blue; 10.5 ns, orange; 11.6 ns, red; 12.7 ns, dark green).The solvent-accessible surface for Cdk2/cyclin A iscolored grey and p27KID is colored white. The figurewas generated using the program PyMol. (b) Summary ofaverage ψ and ϕ torsion angle values for residues in the C-terminal domain of p27 over the course of the MDsimulation. Error bars indicate RMSD values for ψ and ϕtorsion angles during the simulation. (c and d) Electro-static repulsion between the C-terminal region of p27 andthe surface of the Cdk2/cyclin A complex. Electrostaticpotential surface generated with the programMOLMOL21

for the (c) p27/Cdk2/cyclin A complex and (d) Cdk2/cyclin A complex where p27 is illustrated as a ribbon andis colored according to Fig. 1. The structure shown wasobtained at 8.3 ns during the MD trajectory.

3p27 Kip1 Flexibility Mediates Signal Transduction

ARTICLE IN PRESS

and second by studying isotope-labeled p27 boundto Cdk2/cyclin A. The 2-D 1H–15N heteronuclearsingle quantum coherence (HSQC) spectrum of 15N-labeled p27-C (Fig. 1b, green) exhibited the limitedchemical shift dispersion that is typical of IUPs.Further, secondary 13Cα chemical shift values (Δδ13Cα) (Supplementary Fig. 1a) did not indicate theexistence of highly populated secondary structureand {1H}–15N heteronuclear nuclear Overhausereffect (hetNOE) values (Supplementary Fig. 1b)were almost exclusively negative, consistent withfrequent, random amide group fluctuations on thehigh picosecond–low nanosecond time scale. Inaddition, the chemical shift values of most amideswithin p27-C were very similar for the isolateddomain and this domain in the context of 2H/15N-p27 bound to Cdk2/cyclin A (Fig. 1b, red). Thestructural independence of p27-C was confirmed byshowing that resonances for p27-KID bound toCdk2/cyclin A (Fig. 1c, blue) were recapitulated inthe 2-D 1H–15N transverse relaxation optimizedspectroscopy (TROSY) spectrum of 2H/15N-p27/Cdk2/cyclin A (Fig. 1c, red). Together, these resultsindicated that p27-C in the p27/Cdk2/cyclin Acomplex is an independent, flexible domain thatlacks secondary structure. These NMR results pro-vide detailed insights into the average conformationand dynamics for most residues of p27-C within thebiologically active p27/Cdk2/cyclin A complex.However, chemical shift and hetNOE values reflectonly the local environment of nuclei within the poly-peptide backbone and do not define the global struc-ture of this domain.

Molecular dynamics computations

Based on our knowledge from NMR that p27-C isan independent, intrinsically unstructured domain,we used molecular dynamics (MD) computations tobetter understand the overall structure and dy-namics of p27-C in the context of the p27/Cdk2/cyclin A complex. During the 12.7-ns MD computa-tion, residues of p27 were only under the influenceof the Amber force field. This allowed residueswithin p27-C to explore energetically allowed con-formations constrained only by local, intra-p27interactions and longer range, intermolecular inter-actions with Cdk2/cyclin A. The use of a continuummodel of water, rather than explicit water, allowedefficient sampling of conformational space for thisrather large molecular system during the ∼13-nstrajectory. Interestingly, the segment of p27-Cimmediately following the KID protruded at anabrupt angle perpendicular to the surface of Cdk2/cyclin A and exhibited limited flexibility (Fig. 2a).The phi and psi torsion angles for residues in thisregion (residues 110–140) explored a more limitedrange of values than did those closer to theC-terminus (Fig. 2b). The abrupt protrusion of p27-C from the surface of the p27-KID/Cdk2/cyclin Acomplex may arise due to electrostatic repulsionof acidic amino acids in p27-C between residues110–140 and the largely negatively charged surface


of p27-KID/Cdk2/cyclin A proximal to the point ofprotrusion (Fig. 2c and d). This electrostatic guidingmay cause the more distal segment of p27-C(residues 150–198) to assume highly extended con-formations. These features give rise to large valuesof the end-to-end distance for complexes within theentire p27/Cdk2/cyclin A ensemble (average end-to-end distance, 138±9 Å; Supplementary Fig. 2).Based on the general absence of stable secondarystructure within p27-C throughout the MD trajec-tory, we conclude that the conformational propertiesof residues within p27-C in the ensemble of MDstructures are consistent with the NMR findings forthis segment.

Hydrodynamic analysis of p27/Cdk2/cyclin A

We used two techniques, analytical ultracentrifu-gation (AUC) and small-angle X-ray scattering(SAXS), to study the size and shape of the p27/Cdk2/cyclin A ternary complex and related findings



Fig. 3. Hydrodynamic studiesof the p27/Cdk2/cyclin A ternarycomplex. (a) Interference fringe dis-placed velocity data for p27/Cdk2/cyclin A were analyzed using thecontinuous sedimentation coeffi-cient distribution, c(s), model withthe c(s) distributions displayed infringes/S versus sedimentationcoefficient (s). Experiments wereconducted at 20 °C at a rotorspeed of 50,000 rpm and a startingprotein concentration of 2 mg/ml(24 μM). The s value of the complexwas determined as 3.98 S with abest-fit weight-average frictionalratio (f/f0)w of 1.63 and a calculatedmolecular mass of 87,603 Da. Thisanalysis was with regularization ata confidence level of p=0.7 and at aresolution of sedimentation coeffi-cients of n=100. The peak at ∼1.5 Smarked “p27” corresponds to asmall amount of unbound p27 andthat marked “aggr.” corresponds to

a small amount of unidentified protein aggregate. (b) Calculated sedimentation coefficients computed using the programHYDROPRO for p27/Cdk2/cyclin A structures from the MD trajectory. The red dotted line represents the standard valuebased the experimental s value and the blue dotted line represents the average of the calculated s values. (c) Contour plotof the two-dimensional c(s, f/f0) distribution calculated using SEDFITwith an equidistant f/f0 grid from 1.0 to 2.2 with 0.1steps, a linear s grid from 1 to 8 S, a resolution of 100 s values and regularization at one standard deviation. The differentlycolored contours represent c(s, f/f0) values from 0 fringes/S (white) to 1.0 fringes/S (red), with increasing colortemperature indicating larger values.


ARTICLE IN PRESS

from these two techniques to those from MD com-putations and NMR spectroscopy.

†www.analyticalultracentrifugation.com

AUC studies

Analysis of AUC data showed that p27/Cdk2/cyclin A sedimented as a monodisperse 1:1:1 specieswith an s value of 3.98 S, an (f/f0)w (weight-averagefrictional ratio, f/f0) value of 1.63 and a molar massof 87,603 Da, close to its theoretical molar mass of86,796 Da (Fig. 3a). The (f/f0)w value reflects sig-nificant extension of p27-C from the globular core ofthe p27/Cdk2/cyclin A complex. The monodispersenature of the ternary complex was confirmed bysedimentation equilibrium analysis, which showeda single species of mass 86,761 Da (SupplementaryFig. 3). We compared the experimental s value withthose calculated from explicit molecular structuresof p27/Cdk2/cyclin A taken at 100-ps intervalsfrom the equilibrated portion of the MD trajectory.To compare the experimental and theoretical values,the experimental value was converted to standardconditions (water solution at 20 °C), giving an s20,wvalue of 4.28 S. Theoretical s20,w values for p27/Cdk2/cyclin A were calculated using structuresfrom the MD calculations with the programHYDROPRO22,23 (Fig. 3b). The average s20,w valuethus obtained was 4.24±0.07 S. While the theoreticals20,w values reflected significant structural fluctua-tions, ranging from 4.05 to 4.35 S, the agreementobserved between the average of these values,


4.24±0.07 S, and the experimental s20,w value, 4.28S, indicated that the molecular structures compris-ing the MD ensemble accurately represented theactual molecular ensemble as regards to their hydro-dynamic properties based on AUC measurements.In addition, while p27/Cdk2/cyclin A sedimentswith a narrow distribution of s values, calculatingthe two-dimensional c(s,f/f0) distribution in theprogram SEDFIT† revealed that the peak at 4.00 Swas best fit using a distribution of f/f0 values (Fig.3c). This finding strongly suggests that conforma-tional fluctuations of p27-C within the ternary com-plex give rise to a distribution of molecular shapes,further strengthening the view that the ensemble ofMD structures accurately represents conformationalfeatures of p27/Cdk2/cyclin A.

SAXS studies

The reduced SAXS intensity profile for p27/Cdk2/cyclin A (Fig. 4a and b) was analyzed toprovide additional insights into the structure of theternary complex, especially with regard to the con-formation of p27-C. The data, which were collectedat a concentration of 5 mg/mL, displayed evidenceof slight aggregation at low q, which had limitedimpact on the data past q=0.025 Å−1. Data collectedat 2 mg/mL (not shown) were consistent with the


http://www.analyticalultracentrifugation.com



ARTICLE IN PRESS

data collected at 5 mg/mL, which indicates a lack ofconcentration-dependent effects. Performing a fitover the range q=0.025 to 0.040 Å−1 resulted in an Rgvalue of 40.6±4.8 Å. While this q range is beyond thenormally applicable Guinier region,27 the data werewell fit. The forward scatter was 0.332/cm, whichis ∼15% above the value expected based on theconcentration and molecular weight of the complexand indicates that most of the sample consists ofmonomers. Analysis of the data over the rangeq=0.025 to 0.070 Å−1 according to Debye,28 whichtreats the particle as an unfolded, Gaussian coil,yielded an Rg of 38.5±8.2 Å. The distance distribu-tion function P(r) was derived by fitting q valuesfrom 0.025 to 0.15 Å−1 (Fig. 4c). The value for themaximal diameter of the molecule (dmax) was 130±10 Å. The Rg determined from the P(r) fitting was40.2±4.8 Å. This value is in good agreement withboth the Guinier- and Debye-derived values andprovides confidence that the results of the analysesare reasonable. The P(r) curve suggests a somewhatextended particle.Intensity profiles calculated from explicit molecu-

lar structures of p27/Cdk2/cyclin A from the MDtrajectory were calculated using the programORNL_SAS29 and then compared with experimen-tal data over the range q=0.025 to 0.20 Å−1 (Fig. 4a).The contribution of p27-C within the p27/Cdk2/cyclin A ternary complex was assessed by compar-ing fits of the intensity profiles computed usingp27-KID/Cdk2/cyclin A,18 which lacks p27-C, andusing p27/Cdk2/cyclin A, in which p27-C wasmanipulated to be highly extended (p27EX), with theexperimental intensity profile (Fig. 4a). The χ2

values for these two fits were 0.595 and 0.590,


respectively. The calculated intensity profiles sug-gest that the p27EX bound to Cdk2/cyclin Acontributed little to the model intensity profileover this q range. More important, the intensityprofile calculated using the ensemble of p27/Cdk2/cyclin A structures generated by MD fitted theexperimental data much better than did thosecalculated using the two individual ternary complexstructures (χ2 =0.298). The improved fit to experi-mental data was most apparent for q values from0.05 to 0.10 Å−1, where the data and ensemble profilehad a flattened character compared to the other twocomputed intensity profiles that is more consistentwith the shape of the data (Fig. 4b). The apparentshift of the other two profiles results from the scalingused to compare the model intensity profiles to thedata by ORNL_SAS, which minimizes the χ2 valuethrough the use of a scaling constant and a baselineshift. If the model profiles were made to be moreconsistent in this range, then other portions of thecurve would be fit less well, resulting in a higher χ2

value. Further, the inherent slope of the modelprofiles of the truncated complex and the p27EX

complex in that particular region would remain lesstrue to the character of the data than the intensityprofile from the ensemble of structures. This q rangecorresponds to distances ranging from ∼60 to∼120 Å that are impacted to a lesser degree by theslight aggregation observed than the lower q values,such as those used for the Guiner fitting.27 Thesedistances are also likely to be most impacted by theensemble of p27 conformations that are significantlymore compact than the p27EX, which will onlycontribute slightly to the intensity profile in thisrange due to the much greater length of the

Fig. 4. The p27/Cdk2/cyclin Acomplex adopts an extended con-formation in solution. (a) SAXS datafor the p27/Cdk2/cyclin A complex(○). Profiles calculated with theprogram ORNL_SAS using p27-KID/Cdk2/cyclin A (Protein DataBank ID 1JSU) (black), p27/Cdk2/cyclin A in which p27 was fullyextended (p27EX) (red), and theMD-generated ensemble of p27/Cdk2/cyclin A structures (green). (b)Expanded view of the SAXS dataregion used for quantitative com-parisonwith calculated data. (c) P(r)derived from the SAXS data usingthe program GNOM.24 (d) Twoorthogonal views of a surface repre-sentation of the consensus envelopegenerated by GA_STRUCT25 fromSAXS data of the p27/Cdk2/CyclinA complex. The best fit of represen-tative MD-generated structures ofp27/Cdk2/cyclin A (shown astubes) to the envelope determinedby SUPCOMB26 is illustrated in thefigure. The color scheme is the sameas used in Fig. 1a and the figure wasgenerated using PyMOL.




ARTICLE IN PRESS

structure. Beyond 0.10 Å−1, the p27-C in theensemble also produces additional interatomicdistances b60 Å that are not present in the othertwo structures, resulting in the differences betweenthe model intensity profiles. The Rg values deter-mined from the second moments of the model P(r)curves calculated for the truncated structure, thep27EX structure and the ensemble of structures were25.5, 61.4 and 32.8 Å, respectively. Taken together,the SAXS results suggest that the ensemble of MDstructures is a better representation of the structureof p27/Cdk2/cyclin A in solution than is theindividual structure containing p27EX, but that theactual ensemble in solution may contain states thatare more extended than the structures sampled byMD.The low-resolution consensus envelope model

was generated from the SAXS intensity profiledata using the program GA_STRUCT25 (Fig. 4d).The globular portion of the p27/Cdk2/cyclin Acomplex (∼80 Å long, ∼60 Å wide) could readily beaccommodated within one end of the computedelongated envelope (∼175 Å long, ∼60 Å wide),while the extended conformations of the p27C-terminus for structures generated during theMD simulation could be accommodated within theremaining volume of this envelope (Fig. 4d). It isimportant to note that the computed molecularenvelope is of uniform density and, therefore, thestructural agreement between this model and theensemble of molecular structures generated fromthe MD simulations must be interpreted qualita-tively.30 However, the general dimensions andelongated nature of the computed molecular envel-ope are in good agreement with the general shape ofthe superposed molecular structures (Fig. 4d).Because the computed molecular envelope is ofuniform density, it is not possible to differentiatebetween ordered and disordered regions of thecomplex based on the results of the modeling.

Discussion

The results of our study provide the first detailedinsights into the conformation of full-length p27when bound to Cdk2/cyclin A, the form of p27 thatblocks eukaryotic cell division by inhibiting thecatalytic activity of this and several other Cdk/cyclin complexes. The crystal structure of p27-KIDbound to Cdk2/cyclin A determined by X-raydiffraction18 provided the first insights into howp27 binds and inhibits Cdks. More recently, NMRspectroscopy studies have provided detailedinsights into how the Cdk2/cyclin A-bound con-formation of p27-KID is altered by phosphorylationof Y88.10 These studies showed that the C-terminusof p27 does not interact extensively with the Cdk/cyclin complex, even though it plays an importantrole in the regulation of p27-mediated cell cyclearrest. We and others have previously reported thatCD spectropolarimetry analysis showed that full-length p27 is intrinsically unstructured.4,31 While the


N-terminal KID of free p27 contained regions oftransient structure, the C-terminal domain (p27-C,residues 105–198) lacked detectable secondarystructure. In addition, analysis by NMR spectro-scopy showed that, in the absence of its Cdk/cyclinbinding partners, the N- and C-termini of p27 do notinteract.4 However, these studies did not provide adetailed understanding of the structure of p27-Ceither in isolation or when the KID of full-length p27is bound to the Cdk/cyclin complex.In the present study, we have used a variety of

techniques to gain a comprehensive understandingof the structure of p27-C. First, we studied averagesecondary structure and picosecond–nanosecondtime scale dynamics of individual residues ofp27-C using NMR spectroscopy. Δδ 13Cα valuesfor residues of p27-C indicated a general lack ofsecondary structure, while most backbone amidemoieties exhibited negative hetNOE values, demon-strating that long-range interactions between differ-ent regions within p27-C do not occur.Second, based on our past experience using un-

restrained MD computations to explore the inherentconformational properties of p27-KID,20 we usedMD to explore the conformational features of p27-Cwhen p27 was bound to Cdk2/cyclin A. Surpris-ingly, p27-C remained quite highly extended duringthe trajectory, rather than assuming more compactconformations as might be expected. Analysis of phiand psi torsion angles showed that conformationalfreedom was limited for residues in p27-C that wereclose to the surface of Cdk2/cyclin A, and thatRMSD values for these torsion angles increasedtoward the extreme C-terminus furthest away fromCdk2/cyclin A. Visual inspection of the p27/Cdk2/cyclin A complex indicated that this restrictedmotion may be due to electrostatic repulsionbetween the molecular surface of Cdk2/cyclin Aand p27-C that “steers” the p27 polypeptide back-bone away from the globular core of Cdk2/cyclin A.Finally, torsion angles and RMSD values observedfor p27-C during the MD trajectory were consistentwith Δδ 13Cα values determined by NMR spectro-scopy. The overall general agreement of resultsobtained by NMR spectroscopy and MD regardingthe conformation of p27-C allowed us to use evenlysampled molecular structures from the MD trajec-tory to represent the solution conformation of full-length p27/Cdk2/cyclin A.The third approach employed to characterize the

conformation of p27-C was to use AUC and SAXS toanalyze the hydrodynamic properties of the p27/Cdk2/cyclin A complex. These techniques alloweddirect measurement of hydrodynamic propertiesthat reflect the average, global conformation of thisprotein complex in solution, in contrast to the localnature of chemical shift and hetNOE values deter-mined by NMR spectroscopy. We showed thatmolecular structures obtained during the MDtrajectory, in which p27-C is highly extended, wereconsistent with both AUC sedimentation velocityand SAXS data by calculating s values (AUC) andintensity profiles (SAXS) using an ensemble of MD-




ARTICLE IN PRESS

generated molecular structures. The agreementbetween computed and experimentally determinedvalues from AUC and SAXS strongly suggested thatthe MD structures are a true representation of thesolution structure of p27/Cdk/cyclin A and theextent to which the p27-C portion of this complexsamples conformational space.Two types of posttranslational modification,

phosphorylation and ubiquitination, regulate theCdk inhibitory activity and degradation of p27 ineukaryotic cells. These modifications occur insequential order and include a two-step phosphor-ylation mechanism followed by ubiquitination. Thefirst step involves phosphorylation of Y88 byNRTKsand partial “reactivation” of Cdk2 within the p27/Cdk2/cyclin A complex.10 In the second step,reactivated Cdk2 phosphorylates T187 locatedtoward the C-terminus of the 100-residue-longC-terminal domain of p27. Subsequently, phos-phorylated T187 and several neighboring residuesare recognized by the SCF/Skp2 E3 ubiquitin ligaseand p27 is polyubiquitinated at currently unknownsites. This modification cascade finally leads tothe elimination of p27 through 26S proteasome-mediated degradation and, consequently, de-repres-sion of the major blockade to progression of the celldivision cycle from G1 to S phase. Although thedegradation mechanism described above appliesspecifically to the interaction of p27 with Cdk2/cyclin A, p27 binds to many different Cdk/cyclintargets in vivo and it is likely that this multistepmodification cascade regulates p27 within its com-plex with Cdk2/cyclin E and possibly within com-plexes comprising Cdk1 and either cyclin A orcyclin E.Here we have shown that the C-terminal domain

of p27 adopts a disordered and quite highly ex-tended conformation when p27 is bound to Cdk2/cyclin A. The p27 C-terminus functions as aregulatory domain that is structurally independentof the N-terminal domain, which itself binds tightlyand specifically to several different Cdk/cyclincomplexes involved in regulating the G1/S cellcycle transition. The C-terminal domain acts as aflexible “conduit” for signal transduction, whichdetects a specific signal from NRTKs (phosphotyr-osine 88) and, through the catalytic activity of Cdk2,mediates conversion of this signal into a secondsignal, phosphothreonine 187 (pT187). The pT187signal, in turn, is functionally transduced throughpolyubiquitination (of p27) by SCF/Skp2 in apT187-dependent manner that signals the 26Sproteasomal degradation of p27. While not sampledin our brief MD trajectory, our previous biochemicaldata10 support the view that T187 within the p27C-terminus frequently approaches the Cdk2 activesite within individual ternary complexes. Theresults of the structural analysis reported hereinsupport this model by showing that p27-C lacksdiscreet structure and is highly flexible. Whileextended in the MD trajectory, p27-C appears tobe sufficiently unrestrained to occasionally sampleconformations in which T187 “folds back” to en-


counter the Cdk2 active site (Fig. 5). Based on ourobservation that the surface of Cdk2/cyclin A nearthe ATP binding site bears negative charge, wepropose that electrostatic interactions between thissurface and a cluster of positively charged residuesat the p27 C-terminus adjacent to T187 facilitateinteractions between T187 and the Cdk2 substratebinding site. If Y88 of p27 is phosphorylated whenthese interactions occur, then the oncogenic signalfrom NRTKs can be transduced through phosphor-ylation of T187 and subsequent ubiquitination anddegradation of p27. If Y88 is not phosphorylated, nosignal is transduced and Cdk2/T187 encounters arecatalytically unproductive. The extended natureand flexibility of the 100-residue-long p27 C-terminus makes it possible for T187 to constantlysense the signaling status of Y88 (i.e., absence orpresence of phosphorylation). In addition to flex-ibility within p27-C being critical for signalingalong this conduit, we previously discussed theimportance of Y88 being able to fluctuate betweenATP binding pocket-bound and solvent-exposedconformations so that Y88 is accessible for phos-phorylation by NRTKs.10 More recently, Y88 and asecond tyrosine residue of p27, tyrosine 74 (Y74),were shown to be phosphorylated by the NRTK, Src,in breast cancer cells.17 This observation suggeststhat not only does the Y88-containing segment of p27experience conformational fluctuations, but that thesegment containing Y74, within the β-hairpin that issandwiched against the N-terminal β-sheet lobe ofCdk218 (Fig. 1), also experiences conformationalfluctuations that make its side chain available forphosphorylation by Src. Phosphorylation of Y74 bySrc may disrupt local interactions between p27 andCdk2, which may in turn enhance the accessibility ofY88 to NRTKs and its subsequent phosphorylation.This proposal is supported by data showing thatmutation of Y74 to glutamic acid disrupts interac-tions between p27-KID andCdk2/cyclinA (Y.W. andR.K., unpublished data). Thus, this signal transduc-tion conduit may have evolved to allow signalamplification through multiple tyrosine phosphor-ylation events, one at Y88,which partially reactivatesCdk2, and a second at Y74, which amplifies theinitial signal. More important, coupling of theseelements within this signal transduction conduitentirely depends on the ability of p27 to samplemultiple conformations, even when bound toCdk2/cyclin A. Finally, the flexibility and solventaccessibility of the segment of p27 containingphosphorylated T187 allow it to be readily recog-nized and bound by SCF/Skp2 and subsequentlyubiquitinated and degraded,32 thus terminating thesignaling pathway.We conclude that the extensive flexibility of p27,

which stems from its general character as anintrinsically unstructured protein, is of central im-portance in the multistep phosphorylation/ubiqui-tination signal transduction pathway that regulatescell division. Protein flexibility has been shown to becritical in other signaling pathways. For example,the ability of the Wiskott–Aldrich syndrome protein



Fig. 5. Schematic highlighting the importance of p27 flexibility in transducing NRTK signals that mediate p27degradation and cell cycle progression. The p27/Cdk2/cyclin A complex is colored according to Fig. 1a. Regions ofpositive and negative electrostatic potential are colored blue and red, respectively. Phosphorylation of Y88 within the 310helix of the kinase inhibitory domain of p27 by various NRTKs leads to its ejection from the ATP binding pocket of Cdk2and partial reactivation of Cdk2. This may be facilitated by electrostatic repulsion between negatively charged residues onthe surface of Cdk2/cyclin A and residues of p27. In addition, binding of the C-terminal region of p27 adjacent to T187 tothe surface of Cdk2/cyclin A may be enhanced by electrostatic attraction between negatively and positively chargedresidues on Cdk2/cyclin A and p27, respectively. This binding would enhance the Cdk2-mediated phosphorylation ofT187. Once phosphorylated, T187, which lies within the flexible C-terminal region of p27, is fully exposed and recognizedby SCF/Skp2, which polyubiquitinates p27.


ARTICLE IN PRESS

to undergo structural rearrangements in responseto allosteric signals is critical for remodeling of theactin cytoskeleton.33 Further, substrates of kinasesand phosphatases are often specifically targeted forcatalysis not only through specific interactions ator near the catalytic site, but also through remotedocking interactions mediated by short recognitionelements;34,35 these two types of recognition sites areoften connected by intrinsically unstructured poly-peptide segments.34–36 p27, both a Cdk inhibitor andsubstrate, exhibits these latter features. p27 specifi-cally recognizes Cdk/cyclin complexes throughinteractions mediated by multiple short recognitionelements connected by flexible polypeptide seg-ments. For example, the RxLFG motif at theN-terminus of the p27 KID binds specifically tocyclin A, a second region within the KID (residues60–82) binds specifically to the N-terminal β-sheetlobe of Cdk2, the Y88 region binds in the ATPbinding pocket of Cdk2 and, finally, the region nearT187 binds the substrate binding site of Cdk2.Polypeptide flexibility between the multiple recog-nition elements in p27, and those in many otherexamples,34 affords the potential to interact withmultiple targets. This concept was first proposed asthe basis for the ability of p21Waf1, a p27 homolog, topromiscuously inhibit several different Cdk/cyclincomplexes that regulate cell division.36 While pre-vious studies established the importance of flex-ibility in molecular recognition by kinase substratesand inhibitors, the coupling of posttranslationalmodifications within the multiple, flexible short


recognition elements of p27 affords a level of sig-naling complexity that has not previously been ob-served. Flexible regions containing tyrosines 74 and88 coordinate responses to signaling by NRTKs,which are integrated and transduced throughpartial reactivation of Cdk2, phosphorylation ofT187 within the flexible C-terminus and recognitionof this flexible phosphothreonine-specific recogni-tion element by SCF/Skp2, which polyubiquitinatesp27. The end point of this signaling conduit isrecognition and degradation of polyubiquitinatedp27 by the 26S proteasome. The lack of tertiarystructure and intrinsic flexibility within p27 allowsits many short recognition elements to orchestratesignaling along this conduit. We propose that manyof the thousands of other eukaryotic IUPs known toplay roles in signaling and regulation do so throughsimilarly complex, multistep signaling pathwaysand that the molecular events that transmit thesemolecular signals depend on intrinsic protein flex-ibility. Utilizing intrinsic protein flexibility in thisway allows the transduction of multiple signalswithin a single protein complex due to the ability ofdifferent motifs within one polypeptide to “com-municate” with each other. A challenge for thefuture is to identify the sequence signatures of theindividual recognition elements within IUPs thatcomprise multistep signaling pathways (e.g., Y88 asa target of NRTKs, T187 as a target of Cdk2, andunknown sites as targets for ubiquitination by SCF/Skp2) and to begin to decipher the complex sig-naling pathways that they control.




ARTICLE IN PRESS

Materials and Methods

Protein preparation

Human Cdk2 (phosphorylated on T160), truncatedcyclin A (residues 173–432), and full-length human p27and fragments containing residues 22–104 (p27-KID) and105–198 (C-terminal domain, termed p27-C) wereexpressed in Escherichia coli and purified using establishedprocedures.37,38 Protein purity and identity were con-firmed by SDS-PAGE and mass spectrometry. Isotope-labeled p27 (2H/15N), p27-KID (2H/15N) and p27-C(13C/15N) were prepared in Mops-based minimalmedia39 using established procedures.10 After the indivi-dual proteins were combined, the ternary complexes, p27/Cdk2/cyclin A and p27-KID/Cdk2/cyclin A, were pur-ified using gel filtration chromatography (Superdex 200,Amersham-Pharmacia) in 20 mM buffer (Hepes for SAXSor Tris for AUC), pH 7.5, 300 mM NaCl and 5 mM DTT.10

NMR spectroscopy

Samples of p27-KID/Cdk2/cyclin A and p27/Cdk2/cyclin A were exchanged into NMR buffer [20 mMpotassium phosphate, pH 6.5, 50 mM arginine, 8% (v/v)2H2O, 5 mM DTT and 0.02% (w/v) sodium azide] andconcentrated to 0.3 mM by ultrafiltration (Centricon units,Amicon). 2-D 1H–15N HSQC spectra for p27, p27-KID andp27-C and 2-D 1H–15N TROSY spectra for p27-KID/Cdk2/cyclin A and p27/Cdk2/cyclin Awere recorded at35 °C using a Bruker Avance 800-MHz spectrometerequipped with a cryogenically cooled TCI probe. Back-bone resonance assignments for 13C/15N-labeled p27-Cwere made through the analysis of constant time (CT)HNCA,40 CT-HN (CO)CA,41 CT-HNCACB42 and CT-HN(CO)CACB43 spectra recorded at 25 °C in 20 mMpotassium phosphate, pH 6.5, 50 mM NaCl, 8% (v/v)2H2O, 5 mM DTT and 0.02% (w/v) sodium azide using aBruker Avance 600-MHz spectrometer equipped with acryogenically cooled TCI probe. {1H}–15N hetNOE valueswere determined from 2-D HSQC spectra recorded at600 MHz with and without 1H presaturation at 25 °C.Spectra were processed using NMRPipe software44 andanalyzed using Felix software (Accelerys, Inc.). For allspectra, the 1H dimension was referenced to external TSPand the 13C and 15N dimensions were referencedindirectly.45 Secondary 13Cα chemical shift values (Δδ13Cα) were calculated by subtracting sequence-dependentrandom coil values from the experimental values.46

Analytical ultracentrifugation

AUC experiments were performed using a Proteome-Lab XL-I analytical ultracentrifuge with an eight-hole An-50 Ti rotor (Beckman Coulter, Fullerton, CA). The partialspecific volume (at 20 °C), hydration and molecularweight for the proteins based on their amino acid com-position were calculated as well as the density andviscosity of the buffer (50 mM sodium phosphate, pH7.0, 150 mM NaCl, 5 mM DTT) using the computerprogram SEDNTERP‡. For the sedimentation velocityexperiments the loading volume, 400 μL, was identical forthe reference and sample chambers of the double-sector

‡www.bbri.org/rasmb/rasmb.html


centrepiece. Following a 3-h temperature equilibration at20 °C at rest, the rotor was accelerated to 50,000 rpm andrefractive index profiles were recorded in 1-min intervalswith the Rayleigh interference optical system. Interferencefringe displacement profiles were analysed with thesoftware SEDFIT† using the continuous sedimentationcoefficient distribution c(s) model and the two-dimen-sional size-and-shape distribution c(s,f/f0) model. The c(s)distribution is based on a single weight-average frictionalratio (f/f0)w of all the sedimenting species and the c(s,f/f0)model on the differential distribution of sedimentationcoefficients and frictional ratios.30–33 The positions of themeniscus and bottom, as well as time-invariant and radialnoises, were fitted. The s value of the protein was deter-mined by integration of the main peak of c(s). A two-dimensional size–shape distribution, c(s,f/f0) (s distribu-tion versus f/f0 distribution) was calculated with the samevelocity data and represented as a contour plot.Sedimentation equilibrium data were recorded over

30 h at a rotor temperature of 4 °C at increasing speeds of16,000 to 24,000 rpm using the long column technique.47

Protein at a concentration of 7.7 μM (150–160 μL) wasloaded into double-sector centrepieces and absorbancedistributions were recorded at 280 and 250 nm in 0.001-cmradial intervals with 10 replicates for each point. Globalleast squares modeling was performed at multiplewavelengths and rotor speeds with the software SED-PHAT† using a single-species model.

Molecular dynamics simulations

Molecular dynamics simulations were performed withthe AMBER 9 molecular modeling suite48 using 16processors on an IBM BladeServer Linux cluster. TheAMBER force field was utilized as previously described,20

using the generalized Born solvent model with a saltconcentration of 0.1 M and a nonbonded interaction cutoffof 22 Å. A starting model of full-length p27 bound toCdk2/cyclin A was constructed as follows. A molecularmodel of residues 96–198 of p27 in an extended, randomconformation was linked via a peptide bond to residue 95of p27-KID in the structure of p27-KID/Cdk2/cyclin A(Protein Data Bank ID 1JSU) using Insight II software(Accelerys). The starting structure was energy minimized,equilibrated using MD for 100 ps, and then an MDtrajectory was computed for 12.7 ns in the presence of aweak harmonic potential of 0.5 kcal mol−1 Å−2 on Cdk2and cyclin A in order to maintain the starting structure.However, an independent MD computation performedwithout this weak restraint did not show substantialchanges in the conformations of cyclin A and Cdk2. TheMD trajectory was analyzed using the ptraj module ofAMBER. Analysis of the coordinate RMSD fluctuationsfor Cdk2 and cyclin A using the ptraj module ofAMBER showed that the system was equilibratedwithin 5 ns. Molecular structures taken at 0.1-nsintervals from 5 to 12.7 ns were analyzed with regardto their conformation and were used to computehydrodynamic parameters for comparison with AUCand SAXS data. Sedimentation coefficient values for thisensemble of structures were calculated using theprogram HYDROPRO.22,23

Small-angle X-ray scattering

SAXS data were collected using the Center for StructuralMolecular Biology 4 m SAXS instrument, described


http://www.bbri.org/rasmb/rasmb.html



ARTICLE IN PRESS

previously.49 Data for samples andbackgrounds, consistingof buffer solution, were collected and averaged over a totalof 19 and 12 h, respectively. The total time included shorterconsecutive exposures (1 h) to test for time-dependentaggregation indicative of radiation damage; none wasfound. Data were reduced, scaled into absolute units(1/cm) and azimuthally averaged to the 1-D intensity pro-file I(q) versus q, where q=4πsin(θ)/λ, according topreviously published procedures.49 2θ is the scatteringangle from the direct beam and λ is the X-ray wavelength(1.542 Å).

SAXS and modeling

Data were analyzed according to Guinier and Fournet27

and Debye28 for radius of gyration, Rg, which assume acompact structure and a disordered Gaussian chainconformation, respectively. The structure of the complexis expected to contain both compact ordered regions anddisordered regions, so neither model is truly ideal for thisparticular scattering particle. The two methods shouldprovide comparable estimates of Rg. The data were alsoanalyzed for the pair–distance distribution function P(r)and maximum linear dimension, dmax, using the programGNOM.24 P(r) and I(q) are related through the Fouriertransform in Eq. (1).

P rð Þ ¼ 12p2

Z l

0dq qrð ÞI qð Þsin qrð Þ ð1Þ

The P(r) fitting also provides a secondary measure of Rg,which is the second moment of P(r). Low-resolution abinitio models and the consensus envelope model of thep27/Cdk2/cyclin A complex were generated using theprogram GA_STRUCT.25 To compare the measuredintensity against single, high-resolution structures andthe MD-generated ensemble of structures, the programORNL_SAS29 was used to calculate the expected SAXSintensity from the atomic coordinates. The quality of the fitof an intensity profile calculated from a structure wasevaluated using the reduced χ2 parameter.50 To calculatethe scattering from the ensemble of structures generatedduring MD calculations, intensities were calculated forstructures taken from 5 to 12.7 ns at 0.1-ns intervals. Thescattered intensity from the ensemble of MD structureswas calculated as a superposition of the individualintensities.30 The quality of fit of the ensemble intensityprofile to the experimental data was also determinedusing the reduced χ2 parameter.

Acknowledgements

The authors acknowledge Dr. Peter Schuck(National Institutes of Health, Bethesda, MD) forhelpful discussion on the analysis of analyticalcentrifugation data and Yiming Mo (Oak RidgeNational Laboratory, Oak Ridge, TN) for assistancein collecting SAXS data. This workwas supported bythe American Lebanese Syrian Associated Charities(ALSAC), National Cancer Institute (2R01CA082491to R.W.K.), and a Cancer Center (CORE) Supportgrant (5P30CA021765, St. Jude Children's ResearchHospital). SAXS studies were supported by the


Oak Ridge Center for Structural Molecular Bio-logy (KP1102010) of the Office of Biological andEnvironmental Research of the U.S. Departmentof Energy, under contract DE-AC05-00OR22725with Oak Ridge National Laboratory, managedand operated by UT-Batelle, LLC. The submittedmanuscript has been authored by a contractor ofthe U.S. Government under contract DE-AC05-00OR22725. Accordingly, the U.S. government re-tains a nonexclusive royalty-free license to publishor reproduce the published form of this contribu-tion, or allow others to do so, for U.S. governmentpurposes.

Supplementary Data

Supplementary data associated with this articlecan be found, in the online version, at doi:10.1016/j.jmb.2007.12.016

References

1. Hengst, L., Dulic, V., Slingerland, J. M., Lees, E. &Reed, S. I. (1994). A cell cycle-regulated inhibitor ofcyclin-dependent kinases. Proc. Natl Acad. Sci. USA,91, 5291–5295.

2. Polyak, K., Kato, J. Y., Solomon, M. J., Sherr, C. J.,Massague, J., Roberts, J. M. & Koff, A. (1994). p27kip1,a cyclin-Cdk inhibitor, links transforming growthfactor-beta and contact inhibition to cell cycle arrest.Genes Dev. 8, 9–22.

3. Toyoshima, H. & Hunter, T. (1994). p27, a novelinhibitor of G1 cyclin–Cdk protein kinase activity, isrelated to p21. Cell, 78, 67–74.

4. Lacy, E. R., Filippov, I., Lewis, W. S., Otieno, S., Xiao,L., Weiss, S. et al. (2004). p27 binds cyclin–CDKcomplexes through a sequential mechanism involvingbinding-induced protein folding.Nat. Struct. Mol. Biol.11, 358–364.

5. Morgan, D. O. (1995). Principles of CDK regulation.Nature, 374, 131–134.

6. Hengst, L. & Reed, S. I. (1998). Inhibitors of the Cip/Kip family. Curr. Top. Microbiol. Immunol. 227, 25–41.

7. Sherr, C. J. & Roberts, J. M. (1999). CDK inhibitors:positive and negative regulators of G1-phase progres-sion. Genes Dev. 13, 1501–1512.

8. Hengst, L. & Reed, S. I. (1996). Translational control ofp27Kip1 accumulation during the cell cycle. Science,271, 1861–1864.

9. Pagano, M., Tam, S. W., Theodoras, A. M., Beer-Romero, P., Del Sal, G., Chau, V. et al. (1995). Role ofthe ubiquitin–proteasome pathway in regulatingabundance of the cyclin-dependent kinase inhibitorp27. Science, 269, 682–685.

10. Grimmler, M., Wang, Y., Mund, T., Cilensek, Z.,Keidel, E. M., Waddell, M. B. et al. (2007). Cdk-inhibitory activity and stability of p27(Kip1) are di-rectly regulated by oncogenic tyrosine kinases. Cell,128, 269–280.

11. Bloom, J. & Pagano, M. (2003). Deregulated degrada-tion of the cdk inhibitor p27 and malignant transfor-mation. Semin. Cancer. Biol. 13, 41–47.

12. Bloom, J., Pagano, M., Ababou, M., Dutertre, S.,Lecluse, Y., Onclercq, R. et al. (2003). Deregulateddegradation of the cdk inhibitor p27 and malignant




ARTICLE IN PRESS

transformation ATM-dependent phosphorylationand accumulation of endogenous BLM protein inresponse to ionizing radiation; a comparison of cellcycle markers in well-differentiated lobular and ductalcarcinomas MNNG-transformed Bloom syndromeB-lymphoblastoids for the detection of Hodgkin'slymphoma-associated antigen in 2D Westerns. Semin.Cancer. Biol. 13, 41–47.

13. Slingerland, J. & Pagano, M. (2000). Regulation of thecdk inhibitor p27 and its deregulation in cancer. J. CellPhysiol. 183, 10–17.

14. Kamura, T., Hara, T., Matsumoto, M., Ishida, N.,Okumura, F., Hatakeyama, S. et al. (2004). Cytoplas-mic ubiquitin ligase KPC regulates proteolysis of p27(Kip1) at G1 phase. Nat. Cell Biol. 6, 1229–1235.

15. Kotoshiba, S., Kamura, T., Hara, T., Ishida, N. &Nakayama, K. I. (2005). Molecular dissection of theinteraction between p27 and Kip1 ubiquitylation-promoting complex, the ubiquitin ligase that regulatesproteolysis of p27 in G1 phase. J. Biol. Chem. 280,17694–17700.

16. Sabile, A., Meyer, A. M., Wirbelauer, C., Hess, D.,Kogel, U., Scheffner, M. & Krek, W. (2006). Regulationof p27 degradation and S-phase progression by Ro52RING finger protein. Mol. Cell Biol. 26, 5994–6004.

17. Chu, I., Sun, J., Arnaout, A., Kahn, H., Hanna, W.,Narod, S. et al. (2007). p27 phosphorylation by Srcregulates inhibition of cyclin E-Cdk2. Cell, 128,281–294.

18. Russo, A. A., Jeffrey, P. D., Patten, A. K., Massague, J.& Pavletich, N. P. (1996). Crystal structure of thep27Kip1 cyclin-dependent-kinase inhibitor bound tothe cyclin A–Cdk2 complex. Nature, 382, 325–331.

19. Bienkiewicz, E. A., Adkins, J. N. & Lumb, K. J. (2002).Functional consequences of preorganized helicalstructure in the intrinsically disordered cell-cycle in-hibitor p27(Kip1). Biochemistry, 41, 752–759.

20. Sivakolundu, S. G., Bashford, D. & Kriwacki, R. W.(2005). Disordered p27(Kip1) exhibits intrinsic struc-ture resembling the Cdk2/cyclin A-bound con-formation. J. Mol. Biol. 353, 1118–1128 (Epub 2005Sep 20).

21. Koradi, R., Billeter, M. & Wuthrich, K. (1996).MOLMOL: a program for display and analysis ofmacromolecular structures. J. Mol. Graphics, 14, 51–55,29–32.

22. Garcia de la Torre, J. (2001). Building hydrodynamicbead-shell models for rigid bioparticles of arbitraryshape. Biophys. Chem. 94, 265–274.

23. Garcia De La Torre, J., Huertas, M. L. & Carrasco, B.(2000). Calculation of hydrodynamic properties ofglobular proteins from their atomic-level structure.Biophys. J. 78, 719–730.

24. Svergun, D. I. (1992). Determination of the regulariza-tion parameter in indirect-transform methods usingperceptual criteria. J. Appl. Crystallogr. 25, 495–503.

25. Heller, W. T., Krueger, J. K. & Trewhella, J. (2003).Further insights into calmodulin–myosin light chainkinase interaction from solution scattering and shaperestoration. Biochemistry, 42, 10579–10588.

26. Kozin, M. B. & Svergun, D. I. (2001). Automatedmatching of high- and low-resolution structuralmodels. J. Appl. Crystallogr. 34, 33–41.

27. Guinier, A. & Fournet, G. (1955). Small-Angle Scatter-ing of X-rays, John Wiley & Sons, New York.

28. Debye, P. (1947). Molecular-weight determination bylight scattering. J. Phys. Colloid Chem. 51, 18–32.

29. Tjioe, E. & Heller, W. T. (2007). ORNL_SAS: softwarefor calculation of small-angle scattering intensities of


proteins and protein complexes. J. Appl. Crystallogr. 40,782–785.

30. Heller, W. T. (2005). Influence of multiple well definedconformations on small-angle scattering of proteins insolution. Acta Crystallogr., Sect. D: Biol. Crystallogr. 61,33–44.

31. Bienkiewicz, E. A., Adkins, J. N. & Lumb, K. J. (2002).Functional consequences of preorganized helicalstructure in the intrinsically disordered cell-cycleinhibitor p27(Kip1). Biochemistry, 41, 752–759.

32. Hao, B., Zheng, N., Schulman, B. A., Wu, G., Miller,J. J., Pagano, M. & Pavletich, N. P. (2005). Structuralbasis of the Cks1-dependent recognition of p27(Kip1)by the SCF(Skp2) ubiquitin ligase. Mol. Cell, 20, 9–19.

33. Buck, M., Xu, W. & Rosen, M. K. (2004). A two-stateallosteric model for autoinhibition rationalizes WASPsignal integration and targeting. J. Mol. Biol. 338,271–285.

34. Remenyi, A., Good, M. C. & Lim, W. A. (2006).Docking interactions in protein kinase and phospha-tase networks. Curr. Opin. Struct. Biol. 16, 676–685.

35. Neduva, V. & Russell, R. B. (2005). Linear motifs: evolu-tionary interaction switches. FEBS Lett. 579, 3342–3345.

36. Kriwacki, R. W., Hengst, L., Tennant, L., Reed, S. I. &Wright, P. E. (1996). Structural studies of p21(waf1/cip1/sdi1) in the free and Cdk2-bound state: Con-formational disorder mediates binding diversity. Proc.Natl Acad. Sci. USA, 93, 11504–11509.

37. Grimmler, M., Wang, Y., Mund, T., Cilensek, Z.,Keidel, E. M., Waddell, M. B. et al. (2007). Cdk-inhibitory activity and stability of p27Kip1 are directlyregulated by oncogenic tyrosine kinases. Cell, 128,269–280.

38. Wang, Y., Filippov, I., Richter, C., Luo, R. & Kriwacki,R. W. (2005). Solution NMR studies of an intrinsicallyunstructured protein within a dilute, 75 kDa eukar-yotic protein assembly; probing the practical limits forefficiently assigning polypeptide backbone reso-nances. ChemBioChem, 6, 2242–2246.

39. Neidhardt, F. C., Bloch, P. L. & Smith, D. F. (1974).Culture medium for enterobacteria. J. Bacteriol. 119,736–747.

40. Grzesiek, S. & Bax, A. (1992). Improved 3D triple reso-nance NMR techniques applied to a 31 kDa protein.J. Magn. Reson. 96, 432–440.

41. Grzesiek, S., Dobeli, H., Gentz, R., Garotta, G.,Labhardt, A. M. & Bax, A. (1992). 1H, 13C, and15N NMR backbone assignments and secondarystructure of human interferon-gamma. Biochemistry,31, 8180–8190.

42. Wittekind, M. & Mueller, L. (1993). HNCACB, a high-sensitivity 3D NMR experiment to correlate amide-proton and nitrogen resonances with the alpha- andbeta-carbon resonances in proteins. J. Magn. Reson. B,101, 201–205.

43. Yamazaki, T., Lee,W.,Arrowsmith,C.H.,Muhandiram,D. R. & Kay, L. E. (1994). A suite of triple resonanceNMR experiments for the backbone assignment of 15N,13C, 2H labeled proteins with high sensitivity. J. Am.Chem. Soc. 116, 11655–11666.

44. Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, G.,Pfeifer, J. & Bax, A. (1995). NMR Pipe: a multi-dimensional spectral processing system based onUNIX pipes. J. Biomol. NMR, 6, 277–293.

45. Cavanagh, J., Fairbrother, W. J., III, Palmer, A. G. &Skelton, N. J. (1996). Protein NMR Spectroscopy.Academic Press, New York.

46. Schwarzinger, S., Kroon, G. J., Foss, T. R., Chung, J.,Wright, P. E. & Dyson, H. J. (2001). Sequence-




ARTICLE IN PRESS

dependent correction of random coil NMR chemicalshifts. J. Am. Chem. Soc. 123, 2970–2978.

47. Vistica, J., Dam, J., Balbo, A., Yikilmaz, E., Mariuzza,R. A., Rouault, T. A. & Schuck, P. (2004). Sedimenta-tion equilibrium analysis of protein interactions withglobal implicit mass conservation constraints andsystematic noise decomposition. Anal. Biochem. 326,234–256.

48. Case, D. A., Darden, T. A., Cheatham, I. T. E.,


Simmerling, C. L., Wang, J., Duke, R. E. et al. (2006).AMBER 9. University of California, San Francisco.

49. Woodward, J. D., Pickel, J. M., Anovitz, L. M., Heller,W. T. & Rondinone, A. J. (2006). Self-assembledcolloidal crystals from ZrO2 nanoparticles. J. Phys.Chem. B, 110, 19456–19460.

50. Taylor, J. R. (1997). An Introduction to Error Analysis:The Study of Uncertainties in Physical Measurements, 2ndedit., University Science Books, Sausalito, CA.



role of intrinsic flexibility in signal transductioni · 2008-01-15 · role of intrinsic...

Documents