doi:10.1016/j.jmb.2011.06.032 J. Mol. Biol. (2011) xx, xxx–xxx
Contents lists available at www.sciencedirect.com
Journal of Molecular Biologyj ourna l homepage: ht tp : / /ees .e lsev ie r.com. jmb
Enzyme Inhibition by Allosteric Capture of anInactive Conformation
Gregory M. Lee1, Tina Shahian2, Aida Baharuddin1,Jonathan E. Gable3 and Charles S. Craik1⁎1Department of Pharmaceutical Chemistry, University of California, San Francisco, CA 94158-2280, USA2Graduate Group in Biochemistry and Molecular Biology, University of California, San Francisco, CA 94158-2280,USA3Graduate Group in Biophysics, University of California, San Francisco, CA 94158-2280, USA
Received 4 May 2011;received in revised form15 June 2011;accepted 16 June 2011
Please cite this article as: Lee, G. M. etdoi:10.1016/j.jmb.2011.06.032
All members of the human herpesvirus protease (HHV Pr) family are activeas weakly associating dimers but inactive as monomers. A small-moleculeallosteric inhibitor of Kaposi's sarcoma-associated herpesvirus protease(KSHV Pr) traps the enzyme in an inactive monomeric state where theC-terminal helices are unfolded and the hydrophobic dimer interface isexposed. NMR titration studies demonstrate that the inhibitor binds toKSHV Pr monomers with lowmicromolar affinity. A 2.0-Å-resolution X-raycrystal structure of a C-terminal truncated KSHV Pr–inhibitor complexlocates the binding pocket at the dimer interface and displays significantconformational perturbations at the active site, 15 Å from the allosteric site.NMR and CD data suggest that the small molecule inhibits humancytomegalovirus protease via a similar mechanism. As all HHV Prs arefunctionally and structurally homologous, the inhibitor represents a class ofcompounds that may be developed into broad-spectrum therapeutics thatallosterically regulate enzymatic activity by disrupting protein–proteininteractions.
Proteins exhibit conformational selection that canhave significant effects on their activity. In somecases, proteins exhibit intrinsic disorder as part oftheir regulatory mechanism.1,2 Processes of confor-mational selection, in the form of protein–proteininteractions (PPIs), are governed by the internal
56
57
58
59
60
61
62
63
64
65
66
ss: [email protected] 2; HCMV Pr,HV Pr, humannuclear singlesi's sarcoma-B, Protein Data Bank;, humanethyl sulfoxide.
ublished by Elsevier Ltd.
al., Enzyme Inhibition by All
motions of the individual subunits of the complex.Nonideal interactions within these macromolecularcomplexes may result in misregulation of functionand, eventually, to disease. Conversely, optimalbinding between protein partners may actuallyenhance a preexisting condition. For a betterunderstanding of these relationships, a recentupsurge of attention has been paid to linking proteindynamics and regulation of PPIs toward disease anddrug discovery.3–6 As a result, PPIs have gainedmore traction as targets for therapeutics.7,8
Until recently, the most successful mediators ofPPIs had been antibody or peptide based.9 Thoughquite powerful, both modalities have their liabilities.A growing number of examples of preclinicalcompounds that target PPIs have focused on smallmolecules that mimic protein structural motifs, suchas α-helices.10,11 Although large, relatively flathydrophobic surfaces devoid of deep pockets or
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
crevices are a key feature of PPIs, recent studies haveindicated that these small-molecule therapeuticsmay only need to bind a small subset of the interfaceresidues, termed the “hot spot.”12 Notably, bioin-formatic analyses have indicated the presence ofaromatic residues at these hot spots, with trypto-phan the most commonly occurring.13
PPIs play a significant role in the activity ofhuman herpesvirus proteases (HHV Prs). Twoarchetypal members of the HHV Pr family includeKaposi's sarcoma-associated herpesvirus protease(KSHV Pr) and human cytomegalovirus protease(HCMV Pr). As with other structurally and func-tionally homologous HHV Pr family members, bothKSHV Pr and HCMV Pr exist in equilibriumbetween an inactive monomeric state and an active,weakly associating dimeric state. The proteolyticallyactive dimer is critical for the viral life cycle. Theinterface of all HHV Pr dimers consists of twoα-helices (helix 5, one from each monomer), which
Fig. 1. Domain diagram of KSHV Pr. (a) Linear domain diaTrp109 (red), catalytic residues (cyan), and the conformationaindicated by yellow or blue asterisks. (b) The dimer interfaceomitted for clarity. The active site (cyan), the inhibitor-bindinand blue balls) are indicated as in (a) (see also Movie S1). (c)first-generation lead inhibitor of KSHV Pr.14
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
buries an approximate 2500-Å2 hydrophobic surfaceon each partner monomeric unit. Each monomercontains a noncanonical Ser-His-His catalytic triadand an accompanying substrate binding pocketlocated approximately 15–20 Å from the dimerinterface (Fig. 1a and b).15
Previous structural studies performed on KSHVPr suggested that the trigger for the concentration-dependent dimer formation is a disorder-to-ordertransition of the C-terminal residues.15,16 Notably,single point mutations of a key residue within helix5 influence this equilibrium: Met197 to Asp (M197D)results in an inactive obligate monomer,15 whileMet197 to Leu (M197L) stabilizes the dimer.17
NMR-based chemical shift perturbation mappingand hydrogen–deuterium exchange experiments per-formed on the KSHV Pr obligate monomer demon-strated that residues 191–230, which constitute helices5 and 6 in the dimer, are conformationally dynamic.15
An obligate monomeric version of KSHV Pr in which
gram of KSHV Pr displaying the positions of the hot spotlly dynamic C-terminus (gray). C-terminal truncations areof a KSHV Pr monomer (2PBK). The partner monomer isg hot spot Trp109 (red), and the truncation sites (yellowThe chemical structure of DD2, an optimized analog of a
steric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
helix 6 was “stapled” to the core structure via anengineered disulfide bond displayed enhanced enzy-matic activity in the oxidized state relative to thereduced state.15 This result suggested that thepositioning of helix 6 is also critical for stabilizingthe active dimeric conformation of KSHV Pr.In light of these observations, we focused our
efforts toward discovering small-molecule ligandsthat allosterically regulate HHV Pr activity bydisrupting dimerization. Since proteins and en-zymes are able to sample multiple preexistingconformational states,6,18 one such inhibitory mech-anism is to capture HHV Prs in their inactivemonomeric forms. Regulating this conformationalswitch, which is believed to be conserved across allHHV Prs, represents an unexploited pathway forthe development of potential broad-spectrum ther-apeutics against herpesviruses. In contrast to thesuccessful antivirals used for AIDS treatment thattarget proteolytic activity of human immunodefi-ciency virus (HIV) protease, no active-site inhibitorsthat target HHV Prs have been successfully devel-oped as clinical therapeutics. Successful disruptionof the herpesvirus protease dimer offers the poten-tial to resurrect a promising family of drug targetsthat were deemed undruggable due to the limitedefficacy of active-site-directed inhibitors.19–21
To validate this approach, we previously demon-strated that addition of a 30-residue helical peptideabolished protease activity by disrupting KSHV Prdimerization.17 Potential KSHV Pr inhibitors fromsmall-molecule helical mimetic libraries were thenscreened via a high-throughput fluorescence-basedactivity assay.14 One of the lead candidates, dimerdisruptor 2 (DD2) (Fig. 1c), resulted from chemicaloptimization of an initial screening hit and exhibitedan IC50 value of 3.1±0.2 μM against KSHV Pr. DD2is a diaryl-substituted 4-(pyridine-2-amido) benzoicacid. Initial 13C- and 15N-based NMR titrationmapping studies indicated that DD2 disruptsdimerization by binding at or near Trp109, whichis located in the center of the hydrophobic dimerinterface.14 This led to a proposed “monomer trap”model of inhibition: DD2 alters the KSHV Prmonomer–dimer equilibrium by capturing a preex-isting inactive monomer and shifting the populationof conformers from the active dimeric state.In this study, we answer two key questions.
(1) How does a small molecule alter the conforma-tion of a dimeric enzyme in order to trap an inactivemonomeric conformation? (2) Is this inhibitorymechanism applicable to other members of a familyof related enzymes? In particular, we use aC-terminal truncation variant of monomeric KSHVPr to characterize HHV Pr–DD2 interactions viaNMR spectroscopy and X-ray crystallography. Byemploying this truncated variant, we confirm thatKSHV Pr–DD2 binding occurs, even in the absenceof the conformationally dynamic C-terminus. We
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Alldoi:10.1016/j.jmb.2011.06.032
also report the first crystallographic structure of anallosterically inhibited HHV Pr monomer to date.Finally, using the structurally and functionallyhomologous HMCV Pr, we demonstrate that, ingeneral, dimer dissociation is a viable allostericroute toward inhibiting HHV Pr activity.
Results and Discussion
Truncation of the C-terminal helices does notaffect the core KSHV Pr structure
Residues 191–230 constitute helices 5 and 6 in theKSHV Pr dimer but are conformationally dynamicand only partially structured in the case of the KSHVPr M197D obligate monomer.15 Truncation of themonomeric KSHV Pr sequence to its native autolysissite, S204 (henceforth,Δ; Fig. 1a and b), displayed nosignificant chemical shift perturbations with respectto the full-length obligate monomer in the 15–>N–1Hheteronuclear single quantum coherence (HSQC)spectra.16 This suggested that the core structure ofthe KSHV Pr monomer (residues 1–191) remainsintact, even with the absence of helix 6. Determiningthe minimal construct required for the correct fold ofthe monomeric protease core was deemed necessaryfor optimizing both NMR and crystallographicstudies. Other constructs were therefore designedto remove portions of the C-terminal helices (Fig. 1aand b and Movie S1): KSHV Pr M197D Δ209(residues 1–209; Δ209), KSHV Pr Δ196 (residues1–196; Δ196), and KSHV Pr Δ191 (residues 1–191;Δ191). As with the Δ construct, size-exclusionchromatogram elution profiles and 15N–1H HSQCspectra indicate that the resulting truncations areexpressed as monomers, with the core proteasestructure intact and well folded (Fig. S1). Since DD2binds to the obligate monomer,14 these newlycreated truncations can be used to simplify furthersmall-molecule inhibitor binding studies.
Isoleucine δ1-methyl groups act as KSHV Prdimer interface binding probes
Previous NMR studies of the KSHV Pr M197Dmonomer reported the hydrophobic dimer interfaceanchored by Trp109 as the DD2 binding hot spot.14
However, significant backbone amide peak broad-ening was observed in the 15N–1H HSQC spectraduring the titration. This result is indicative ofconformational or chemical exchange on an inter-mediate NMR timescale and precludes quantifica-tion of the binding effect. Subsequently, the non-branched δ1-methyl groups of isoleucine residueswere used as NMR binding probes. Eachmonomericunit of the KSHV Pr dimer structure [Protein DataBank (PDB) accession code 2PBK]22 contains 10
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
isoleucine residues, which can be separated into five“zones” in relation to their distance from Trp109(Fig. 2a). The 13C–1H HSQC spectrum of selective
Fig. 2. KSHV Pr–DD2 titration. (a) The distribution of isoleustructure, based upon 2PBK. The view is a 90° rotation aboutseparate into five zones with respect to distance from Trp109,spectra of the KSHV Pr M197D (b) and Δ196 (c) constructs wspectral overlays display apo (black) and N5 molar equivalentblue box), and Ile71 (continuous black box) are used as the binblack;Δ196, broken red) represent the average apparent Kd valas a bar chart (e). HSQC titration spectra and binding curves
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
[13C–1H methyl]Ile-Leu-Val (ILV)-labeled KSHV PrM197D displays 10 relatively well-dispersedδ1-methyl isoleucine resonances (Fig. S2). In order
cine δ1-methyl groups on a full-length KSHV Pr monomerthe horizontal axis, relative to that in Fig. 1b. Isoleucinesas indicated by colored balls. The 13C–1H HSQC titrationith DD2, focusing on the isoleucine δ1-methyl region. Thes of DD2 (red). Ile44 (continuous blue box), Ile105 (dottedding probes. (d) The binding curves (M197D, continuousues calculated for the three Ile probes and are summarizedfor the M197D-I201V construct are displayed in Fig. S4.
steric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
t1:36Residues in most favored regions 90.0t1:37Residues in additional allowed regions 8.7t1:38Residues in generously allowed regions 0.6t1:39Residues in disallowed regions 0.7t1:40t1:41Mean B-factor (Å2)t1:42Protein 46.4t1:43Inhibitorst1:443NJQ A 197 (DD2-A) 31.0t1:453NJQ B 198 (DD2-B) 29.0t1:463NJQ B 199 (DD2-C, bridging molecule) 33.6t1:47Water molecules 27.0
Numbers in parentheses represent the highest-resolution shell. t1:48
ALS, Advanced Light Source. t1:49a Rsym=(∑|(I− ⟨I⟩)|/∑I), where ⟨I⟩ is the average intensity of
multiple measurements. t1:50b Rwork=(∑|(Fo−Fc)|/∑|Fo|). t1:51c Rfree=Rwork based on ∼1000 (at least 10%) of reflections
excluded from refinement t1:52d Calculated using PROCHECK.24 t1:53
5KSHV Pr – DD2 Binding Modes
to assign the resonance peaks, we performed Ile-to-Val mutations on the M197D and Δ196 constructs.As with the truncations, both 13C–1H (Fig. S3) and15N–1H (data not shown) HSQC spectra of theIle-to-Val mutants indicate no significant structuralperturbation in the KSHV Pr core.As expected, Ile44 and Ile105, both located within
5 Å of the KSHV Pr dimer interface, display thelargest chemical shift resonance perturbations upontruncation of the C-terminal residues (Fig. S2). Ile71is located within 10 Å of the KSHV Pr dimerinterface and exhibits moderate resonance pertur-bations. Conversely, isoleucine residues locatedgreater than 10 Å from the dimer interface exhibitlittle or no resonance perturbations. In addition,overlap of the Ile71–Ile206 and Ile105–Ile201 reso-nances is observed in the Δ209 13C–1H HSQCspectrum but not in the corresponding Δ196spectrum (Fig. S2). Although the Δ191 and Δ196spectra exhibit no differences (data not shown), theΔ191 construct displayed greater tendencies toaggregate in solution under the current NMRconditions. As a result, DD2 binding studies wereperformed using the Δ196 construct, in comparisonwith the full-length M197D obligate monomer.
DD2 binds to KSHV Pr in the presence andabsence of the dynamic C-terminus
Addition of N5 molar equivalents of DD2 to theILV-labeled M197D obligate monomer inducedsignificant chemical shift perturbations of the isoleu-cine residues located at the KSHV Pr dimer interface(Fig. 2b). Here, the methyl resonances of theisoleucine residues most proximal to the hot spotTrp109 (Ile44, Ile71, and Ile105) were the mostaffected. Not unexpectedly, resonances of the C-terminal isoleucines (Ile201, Ile206, and Ile222) arealso perturbed, as these are the residues that wouldtransiently interact with the KSHV Pr dimer interfacein the absence of DD2. Because the Ile105 and Ile201resonances of KSHV Pr M197D overlapped, theKSHV Pr M197D-I201V construct was also examined(Fig. S4). All isoleucine peaks in the M197D-I201Vconstruct displayed similar resonances as those inthe “wild-type” obligate monomer. Moreover, theabsent Ile201 peak in the M197D-I201V constructallows for easier determination of the Ile105 chemicalshift resonances. DD2 titration spectra acquired onthe Δ196 truncation (Fig. 2c) display the samegeneral patterns as observed for the full-lengthKSHV Pr monomer analogs.Ile44, Ile71, and Ile105 were chosen as NMR
reporter probes due to their proximity of less than10 Å to the hot spot Trp109. The Hill equation wasused to perform a nonlinear regression curve fitanalysis of the δ1-methyl group chemical shiftresonances as a function of total ligand concentration(Fig. 2d). The apparent Kd values from these three
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Alldoi:10.1016/j.jmb.2011.06.032
reporter residues were then averaged to obtain anestimate of DD2 binding. The resulting NMR-basedbinding curves indicate that DD2 binds with equalaffinity to the M197D (Kd,app=5.5±3.5 μM) andM197D-I201V (Kd,app=5.8±2.1 μM) constructs (Fig.2e). Notably, no binding events were observed forany of the KSHV Pr constructs using surface plasmonresonance or isothermal titration calorimetry (datanot shown). DD2 appears to bind to the Δ196construct with lower affinity (Kd,app=13.0±2.0 μM).
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
However, this observation may be a reflection of thepresence (M197D or M197D-I201V variants) orabsence (Δ196 construct) of C-terminal residuestransiently interacting with Trp109. Collectively, theNMR titration data clearly demonstrate that muta-tions or deletions of residues within the conforma-tionally dynamic C-terminus have no significanteffect toward DD2 binding affinity.
The crystal structure of the KSHV PrΔ196–DD2 complex
In parallel with the NMR-based DD2 titrationstudies, a 2.0-Å-resolution crystal structure of theKSHV Pr–DD2 complex was obtained (PDB acces-sion code 3NJQ). To date, there have been noreported NMR or X-ray crystallographic structuresof an allosterically inhibited HHV Pr monomer. In
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
general, the majority of the published HHV Prstructures have been in their respective apo oractive-site inhibited dimeric forms, as illustrated bythe apo (PDB accession code 1FL1)23 and covalentlybound peptide-phosphonate-inhibited (2PBK)22 di-meric KSHV Pr structures. Prior attempts tocrystallize the full-length uninhibited KSHV PrM197D obligate monomer proved fruitless, possiblydue to the conformationally dynamic C-terminalresidues disrupting the formation of a stable crystallattice. Crystallization trials using the KSHV PrM197D, Δ209, and Δ191 constructs yielded noappreciable crystals in either the presence or theabsence of DD2. However, the Δ196 construct in thepresence of excess molar equivalents of DD2 yieldedsmall cube-like crystals that were used to acquire a2.0-Å-resolution X-ray diffraction data set (Table 1and Table S1). Notably, no crystals were observedfor the apo Δ196 under similar reservoir conditions.In contrast to the solution state, where all the
KSHV Pr constructs are monomeric in the NMRspectra, the Δ196–DD2 complex crystallizes as anasymmetric pair of KSHV Pr Δ196 monomerscontaining three DD2 molecules (Fig. 3). Previouslypublished structures of the full-length KSHV Prdimers22,23 contain two symmetrical monomerscentered about a C2 rotation axis. While the dimerinterface of proteolytically active KSHV Pr consistsof interfacial α-helices and a hydrophobic surfacecentered on Trp109, the Δ196–DD2 complex forms adimer on the distal side of the molecule with respectto Trp109. The dimer interface of the Δ196–DD2complex buries a total surface area of approximately1800 Å2, which mostly consists of hydrophilicresidues, with the crystal structure containing anumber of water molecules between the two
Fig. 3. Structural comparison of the apo and DD2-inhibited KSHV Pr monomers. (a) The dimer structure ofpeptide-phosphonate-inhibited KSHV Pr (2PBK). Thecatalytic residues (cyan) are located ∼15–20 Å from thedimer interface. The interfacial helix 5 and the followinghelix 6 (monomer A, red; monomer B, green) are displayed.Helix 1 of monomer A (blue) and helix 1 of monomer B(orange) also form a portion of the dimer interface and arealigned in an antiparallel orientation with respect to eachother. (b) The structure of the KSHV Pr Δ196–DD2 complex(3NJQ) crystallizes as an asymmetric dimer, with dimer-ization occurring on the opposite face with respect to 2PBK.DD2 molecules bind to the hydrophobic surface normallyoccupied by helix 5. Monomer A of the complex containsone DD2 molecule (pose 1, green carbons), while monomerB contains two DD2 molecules (pose 2, magenta carbons;pose 3, cyan carbons). The truncated C-terminal residues ofthe Δ196 constructs (monomer A, red; monomer B, green)are also indicated. Helix 1 of monomer A (blue) and helix 1of monomer B (orange) are oriented end on end withrespect to each other. Below each structure are cartoonrepresentations of the monomeric units, with the wedgesrepresenting the active site and stars representing the DD2molecules.
steric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
monomers. Both monomers in the asymmetric unitare conformationally similar with respect to oneanother and to residues 1–196 of the previouslypublished KSHV Pr dimer structures (1FL1 and2PBK), with overall RMSDs of less than 1.0 Å forbackbone and heavy-atom overlays (Table S2).Comparing a monomeric unit of the KSHV Prdimer and the Δ196–DD2 complex structures re-veals two significant differences: the formation of anallosteric DD2 binding pocket and the conforma-tional perturbation of the active site.
The hot spot tryptophan 109 acts as a hinge tocreate the DD2 binding pocket
One of the major differences in the backboneconformation between the “apo” and the DD2-bound states is apparent in the α1–α2 loop (residues87–99) and helix 2 (residues 100–110) (Fig. 4 andMovie S2). In the case of the enzymatically activeKSHV Pr dimer, the Trp109 indole ring adopts a“closed” conformation in each monomer, forminga relatively flat hydrophobic surface that interactswith the two interfacial helices. Dimerization isstabilized by the Met197 and Ile201 side chains ofthe partner monomer, forming intermolecularhydrophobic interactions with Trp109 (Fig. 4aand b).14,22 Conversely, the Trp109 indole ring
Fig. 4. Comparison of the apo and DD2-bound KSHV Pr cryconsists of two helices from each monomeric unit (helix 5, tanactive site (cyan) via the C-terminal helix 6 and occlude the Trp5 of monomer B (orange) form hydrophobic interactions with Tmonomer B (Fig. S5) exhibit independent DD2 binding pocketsopen form (d) (see also Movie S2).
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Alldoi:10.1016/j.jmb.2011.06.032
adopts an “open” conformation in each monomerof the Δ196–DD2 complex, creating a hydrophobiccavity that acts as the DD2 binding pocket (Fig. 4cand d).
The DD2 binding pocket of the KSHV PrΔ196–DD2 complex
The DD2 binding pocket is dominated byaliphatic residues located in several secondary-structure motifs. Residues within the β2–β3 loop(residues 44–52), helix 1 (residues 73–91), helix 2(residues 100–110), the β6–β7 loop (residues 139–148), and the C-terminus (residues 189–196) areinvolved (Fig. 5 and Fig. S5). With the exception ofresidues 189–196, the backbone conformations ofthe DD2 binding pockets are similar in the twoasymmetric Δ196–DD2 monomers. Minor fluctua-tions in the orientation of the Trp109 side-chainindole ring and in the loop located between helix 1and helix 2 suggest that the binding pocket may beable to sample other conformational states, even inthe presence of the small-molecule inhibitor.Notably, the asymmetric dimer contains three
DD2 poses. While monomer A (Fig. 5) contains onlyone DD2 molecule (DD2-A) in the binding pocket,monomer B (Fig. S5) displays two DD2 molecules(DD2-B and DD2-C). Inspection of the two
stal structures. (a) The dimer interface of KSHV Pr (2PBK), monomer A; light blue, monomer B), which stabilize the109 (red). (b) The Met197 and Ile201 side chains from helixrp109 of monomer A. BothΔ196–DD2monomer A (c) andin which the Trp109 side-chain indole ring (red) adopts an
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
Fig. 5. The DD2 binding pocket. (a) The hydrophobic DD2 binding pocket is composed of aliphatic residues from theβ2–β3 loop (yellow), helix 1 and the α1–α2 loop (blue), helix 2 (red), the β6–β7 loop (orange), and the C-terminus(magenta). DD2 (green carbons) is shown as a space-filling model. (b) Stereo view of DD2 (green) within theΔ196 bindingpocket of monomer A, in relation to the hot spot Trp109 (red) and the Ile44 and Ile105 reporter groups (yellow). Alsodisplayed are buried aliphatic residues of helix 1 (blue), helix 2 (red), and the C-terminus (magenta) that compose thebinding pocket. The mesh represents the 2Fo−Fc 1 σ electron density map. (c) The same with (b) with the proteinbackbone ribbons displayed. Comparable views of monomer B are displayed in Fig. S5.
8 KSHV Pr – DD2 Binding Modes
monomers indicates that DD2-A and DD2-B arelocated within the DD2 binding pocket. Positioningof the DD2-A and DD2-B molecules within thebinding pocket is offset by an approximate 80°rotation about the vertical axis of the “backbone”atoms (Fig. 6). The differences in the rotation axisand backbone positioning are due to the DD2-Cmolecule displacing DD2-B. DD2-C is located
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
outside the hydrophobic cavity and appears to bea consequence of a crystal contact, acting as a“bridge” molecule between adjoining unit cells. Theaverage B-factor values observed for the three DD2molecules (31.0 Å2, 29.0 Å2, and 33.6 Å2), as well asthat of residues within close proximity (≤5 Å) tothe inhibitor (33.4 Å2), are lower than the overallmean B-factor of the complex (46.4 Å2). The specific
steric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
Fig. 6. The conformations of KSHV-Pr-bound DD2 in situ. (a) DD2-A (green) and DD2-B (magenta) within the Δ196–DD2 complex are overlaid with helix 5 of the KSHV Pr partner monomer B (cyan). Orange balls represent the non-branched methyl groups of Met197 and Ile201. The DD2 side chains match the relative positions observed for those of thenative Met197 and Ile201 and are inserted into the hydrophobic pocket vacated by the Trp109 indole ring. The proteinmonomer structures of the Δ196–DD2 complexes and the apo KSHV Pr dimer are omitted for clarity. (b) An overlay ofΔ196–DD2monomers with their constituent DD2molecules. Monomer A (gray) contains DD2-A (green), while monomerB (dark blue) contains DD2-B (magenta) and DD2-C (cyan). DD2-C is an artifact of crystal packing and is situated outsidethe DD2 binding pocket. Major conformational differences in the DD2 binding pocket are only observed for theC-terminal residues (Ser191–Leu196). (c) An overlay of Δ196–DD2 monomers with their respective DD2 moleculesdisplaying the side chains of the residues constituting the DD2 binding pocket. The benzyl side chains of DD2-A andDD2-B are in close proximity to Phe76, while the cyclohexyl ring interacts with Ile105, Leu106, and Leu110.
9KSHV Pr – DD2 Binding Modes
individual B-factors are listed in Table 1 and TableS3. In addition, each of the DD2 molecules exhibitsoccupancies of 1.0 in the 2Fo−Fc difference map(Fig. 5 and Fig. S5) and Fo−Fc omit map (Fig. S6).These results suggest that, upon DD2 binding, theside chains of the hydrophobic binding pocketsample less conformational mobility relative to therest of the structure.The overall RMSD values of the three conformers
suggest that the individual DD2 structures aresimilar (Table S4). However, visual inspection oftheir structural overlays indicates that poses 1 and 2are more closely related to each other than to DD2-C(Fig. S7). The DD2 benzyl and cyclohexyl-methyle-nyl “side-chain” groups adopt the same relativeelevations and positions within the KSHV Prbinding pocket as the Met197 and Ile201 side chainsof the partner monomer in the active dimeric
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Alldoi:10.1016/j.jmb.2011.06.032
enzyme (Fig. 6a). This suggests that the DD2 “sidechains” mimic the i→ i+4 positioning inherent withside chains of residues that adopt a helical confor-mation. Moreover, the DD2 side chains of both DD2-A and DD2-B are enveloped by the hydrophobiccavity formed upon the Trp109 indole ring adoptingthe open position (Fig. 5 and Fig. S5). Notably, theDD2-A and DD2-B side chains occupy the sameconformational space (Fig. 6). In this case, thecyclohexyl rings are coplanar but offset by onecarbon atom, while the plane of the benzyl rings isrotated at an approximate 40° angle. Selectedinteratomic distances between the DD2 side-chaincarbons and the residues within the Δ196 bindingpocket are reported in Table S5.Further examination of the DD2 conformations
may also explain the structure–activity relationshipsin the helical mimetic library used in the initial
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
fluorescence-based activity screening. Of our origi-nal optimized mimetic library, only DD2 containedan amide group in the ortho position with respect tothe nitrogen atom of the pyridinyl ring.14 Thestacking positions of the cyclic DD2 side-chaingroups prevent the amide proton and the pyridinylnitrogen from interacting with the protease, eitherdirectly or via a water molecule (Fig. S7). Intera-tomic distances and backbone dihedrals measuredfrom the crystallographic poses of DD2 suggest thatthe amide proton and pyridinyl nitrogenmay form aweak intramolecular hydrogen bond (Table S4).Such interactions may help stabilize the conforma-tion necessary for efficient binding to KSHV Pr.Further alterations to the DD2 backbone, such asthe addition of polar functional groups, would berequired to enhance solubility and, perhaps,increase efficacy toward the HHV Prs.
Loss of helix 6 disrupts the conformation of theKSHV Pr active site
The loss of helix 6 results in a dramatic confor-mational perturbation of the KSHV Pr active site,located ∼15–20 Å from the dimer interface (Fig. 7
Fig. 7. Structural perturbation of the KSHV Pr active siterepresents an apo state of KSHV Pr. The catalytic triad (H46,stabilizing arginine residues (R142 and R143, red) are displayeand α-helix 0 (yellow), the β1–α0 loop (dark green), and the β6–(c and d) The conformation of the apo-state active-site residcomplex (3NJQ). The Arg142 and Arg143 side chains (red) aconformation while in complex with DD2. In the DD2-boundtriad (cyan) and disrupts the substrate binding pocket (see als
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
and Movie S3). The most noticeable change isobserved for the conformation of β-strand 1(residues 5–13) and is propagated through theβ1–α0 loop (residues 14–27) and helix 0 (residues27–33). In the case of the KSHV Pr dimer, β-strand 1points outward from the structural core of theenzyme, allowing the β1–α0 loop and the followinghelices to adopt a conformation suitable forsubstrate binding and proteolytic cleavage (Fig. 7aand b). Although the β1–α0 loop plays a minimalrole in substrate recognition, specifically the S3pocket,22 its primary function is to help form one ofthe ridges of the substrate binding pocket. In thecase of the Δ196–DD2 complex, the C-terminalportion of β-strand 1 adopts a nearly orthogonalconformation with respect to its correspondingposition in the KSHV Pr dimer. This results in arotation of the β1–α0 loop into a position thatcollapses the substrate binding pocket and occludesthe catalytic triad (Fig. 7c and d).Another striking difference in this conformational
rearrangement is the positioning of Arg142 andArg143 (Fig. 7). These two arginine residues areconserved among all members of the HHV Pr familyand are known to have important roles in stabilizing
upon DD2 binding. (a and b) The active site of 2PBKH134, and S114, cyan) and the conserved oxyanion-hole-d as sticks. Also highlighted are the positions of β-strand 1β7 loop (orange). Residues 197–230 are omitted for clarity.ues displays clear differences relative to the Δ196–DD2dopt a closed conformation in the apo state but an openstate, the β1–α0 loop (dark green) occludes the catalytico Movie S3).
steric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
Fig. 8. CD spectra of DD2 titrations with KSHV Pr andHCMV Pr. The CD spectra of ∼3 μM (a) KSHV Pr and (b)HCMV Pr in the presence of 0 μM (black), 30 μM (red), and80 μM (blue) DD2. Estimated fractional helicity (fH) valuesderived from the mean residue ellipticity of the 222-nmband are listed in the insets and indicate loss of helicalcontent with increasing molar equivalents of DD2. Loss ofhelicity is a strong indication of HHV Pr dimer disruption.
11KSHV Pr – DD2 Binding Modes
the structure of the oxyanion hole.25 In the case ofKSHV Pr, both the backbone amide and the side-chain guanidino groups of Arg142 and Arg143stabilize peptide substrate interactions within theactive site via a series of intermolecular hydrogenbonds.22 The KSHV Pr dimer structure displays theArg142 and Arg143 side chains pointing inwardtoward the catalytic residues and forming a ridge ofthe substrate binding pocket (Fig. 7a and b). Weterm this the “closed” state in conjunction with theclosed Trp109 indole ring. Alternatively, in thestructure of the Δ196–DD2 complex, the side chainsof both arginine residues point outward toward thespace originally occupied by helix 6 (Fig. 7c and d).We term this the “open” conformational state,analogous to the open Trp109 indole ring.An examination of the average B-factor values
indicates that the residues of the catalytic triad(33.6 Å2) are smaller than that of the overallstructure (46.4 Å2), but those of the β1–α0 loop(56.6 Å2) are larger. In addition, the average B-factorvalues for residues of monomer A in the loopcontaining Arg142 and Arg143 are also significantlylarger (59.1 Å2) than the overall average value. Theindividual B-factor values corresponding to theseresidues are listed in Table S6. This suggests ageneral allosteric mechanism of HHV Pr inhibition:that DD2 binding at the dimer interface enhancesmobility of the oxyanion-hole-stabilizing loop,allowing the active site to be conformationallydestabilized. Notably, no significant differences inthe average B-factor values are observed for residues138–149 of monomer B (35.2 Å2). However, this mayreflect the presence of the bridging DD2 molecule inmonomer B restricting conformational mobility ofthe Cys138–Val149 loop.These observations point to the importance of a
structured helix 5 in propagating conformationalchanges within KSHV Pr. Here, a DD2 moleculecaptures an inactive KSHV Pr conformation bybinding to the hydrophobic patch in the areanormally occupied by helix 5, disrupting interac-tions between the protease structural core andhelices 5 and 6. If helix 5 was located at the dimerinterface, helix 6 would be in a position to stericallyforce the side chains of Arg142 and Arg143 inwardtoward the structural core of the enzyme, allowingthe formation of the active site and oxyanion hole.The loss of helix 6 likely allows the loop containingArg142 and Arg143 to sample multiple conforma-tional states, effectively destabilizing both theoxyanion hole and the substrate binding pocket.
DD2 inhibits HCMV Pr by binding to thedimer interface
As all eight HHV Prs are structurally andfunctionally homologous,25 we speculated thatDD2 acts upon HCMV Pr in the same manner as
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Alldoi:10.1016/j.jmb.2011.06.032
upon KSHV Pr. Using a fluorescence-based substratecleavage kinetics assay previously described forKSHV Pr,14 we redetermined the in vitro dissociationconstant for the HCMV Pr monomer–dimer equilib-rium as 1.3±0.1 μM. This value is nearly identicalwith that reported for KSHV Pr (Kd=1.7 μM) undersimilar conditions,14,26 indicating that both HHV Prsare weakly associating dimers. Activity assays alsoindicate that DD2 exhibits 4-fold weaker inhibitionagainst HCMV Pr relative to KSHV Pr, with an IC50value of 12.8±1.1 μM. In addition, circular dichro-ism (CD) spectra of the wild-type HCMV Pr exhibitloss of helical content upon addition of excessmolar equivalents of DD2 (Fig. 8), suggesting thatthe C-terminal helices are unfolded.To further probe DD2 interactions with HCMV Pr,
we acquired HSQC spectra of selective [13C–1Hmethyl]isoleucine-labeled samples. Two HCMV Pr
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
constructs homologous to the KSHV Pr monomersmentioned above were engineered. The first was anHCMV Pr obligate monomer (L222D) that exhibitsan approximate 25-fold reduction in specific prote-ase activity relative to the wild-type sequence (datanot shown). The second was an HCMV Pr trunca-tion (Δ221) that mimics the KSHV Pr Δ196construct. In the case of the full-length HCMV Probligate monomer, two isoleucines (Ile61 and Ile96)are positioned at the dimer interface, while the third(Ile231) is located in the C-terminal region. Tyr128
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
was identified as the potential hot spot aromaticresidue of HCMV Pr (Fig. 9a). 13C–1H HSQC spectraacquired on the selectively labeled HCMV Pr L222Dfull-length obligate (Fig. 9b) and HCMV Pr Δ221truncated (Fig. 9c) monomers indicate that DD2binds to the dimer interface. Both Ile61 and Ile96resonances are affected upon addition of DD2, withIle61 exhibiting more extensive peak broadening.Importantly, as in the case of KSHV Pr, DD2 canbind to the C-terminal truncated HCMV Pr analogbut exhibits greater peak broadening and smallerchemical shift perturbations even at higher relativeDD2 concentrations. The extensive peak broadeningsuggests a weaker binding affinity of DD2 to HCMVPr, relative to KSHV Pr. Collectively, the fluores-cence activity assay and the CD and NMR spectrasuggest that DD2 binds to HCMV Pr in astructurally homologous pocket as KSHV Pr. Theapparent weaker inhibition of HCMV Pr by DD2may be the result of the smaller Tyr aromatic moietycreating a smaller hydrophobic pocket relative toKSHV Pr. Although the HCMV Pr Ty128 and theKSHV Pr Trp109 residues are out of register by oneposition in the primary sequences, their side chainsoccupy the same general space on the surface of thedimer interface (Fig. 10). With further chemicaloptimization of DD2, these results provide apromising pathway for developing broad-spectrumallosteric inhibitors for all eight HHV Prs.
610
611
612
613
614
615
616
617
618
619
620
621
622
623
Discussion
From a drug discovery standpoint, focusing on aconformationally dynamic target and capturing aninactive state in order to influence regulatorypathways are an appealing concept.4,6 In particular,regulating PPIs via allosteric mechanisms hasrecently gained traction. Allosterically targeting thesubunit interface of dimeric complexes, such as theinterdigitated β-sheets of HIV protease, has becomean important goal due to increased incidence ofresistance toward active-site therapeutics. Recentreports have noted the development of peptidicHIV protease dimer disruptors,27,28 including analkylated tripeptide that “sequesters” a monomer.29
Fig. 9. HCMV Pr–DD2 titration data. (a) The threeisoleucine δ1-methyl groups in human CMV Pr arelocalized at the dimer interface and color coded withrespect to distance to Tyr128, as indicated. Helix 5 (tan),the active site (cyan), and Tyr128 (red) are also displayed.Tyr128 is homologous to Trp109 of KSHV Pr. The 13C–1HHSQC spectra of selective [13C–1H methyl]isoleucine-labeled CMV Pr L222D obligate monomer (b) and Δ221truncation (c) in the presence of 0 molar equivalent (black)and 16 molar equivalents (red) of DD2 indicate that DD2binds at the dimer interface. Both Ile61 and Ile96 areputatively assigned; Ile231 was assigned by the loss of theresonance in the Δ221 truncation.
steric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
Fig. 10. Structural homology of the HHV Pr “hot spots.” (a) Representative X-ray crystallographic structures of thethree structurally homologous HHV Pr subfamilies (HSV-2 Pr, 1AT3; HCMV Pr, 1CMV; and KSHV Pr, 2PBK), with theactive site (cyan) and interfacial helix 5 and following helix 6 (tan) as indicated. Hot spot aromatic residues (red) arelocated at the center of the hydrophobic dimer interface and are potential target sites for small-molecule inhibitors thatdisrupt PPIs. α-Subfamily HHV Prs: HSV-1, herpes simplex virus 1; HSV-2, herpes simplex virus 2; and VZV, VarcellaZoster virus. β-Subfamily HHV Prs: HCMV, human cytomegalovirus; HHV-6, human herpesvirus 6; and HHV-7, humanherpesvirus 7. γ-Subfamily HHV Prs: KSHV, Kaposi's sarcoma-associated herpesvirus; EBV, Epstein–Barr virus. (b)Backbone overlay of KSHV Pr (2PBK, gray) and HSV-2 Pr (1AT3, cyan), focusing on the dimer interface hot spot region.The side chains of Trp109 (KSHV Pr, red) and Tyr124 (HSV-2 Pr, yellow) are displayed in space-filling mode. Helices 5and 6 are omitted for clarity. (c) Backbone overlay of KSHV Pr (2PBK, gray) and HCMV Pr (1CMV, cyan), focusing on thedimer interface hot spot region. The side chains of Trp109 (KSHV Pr, red) and Tyr128 (HCMV Pr, yellow) are displayed inspace-filling mode. Helices 5 and 6 are omitted for clarity.
13KSHV Pr – DD2 Binding Modes
Importantly, small-molecule peptide mimetics havebeen shown to disrupt other PPIs. Perhaps themost well documented are the helical mimeticssuch as the Nutlins that inhibit p53/MDM230
interactions and the multi-aryl compounds thatdisrupt calmodulin/smMLCK31 and Bcl-xL/Bak
32,33
interactions. Multi-aryl and multi-cyclic compoundswere also recently discovered to regulate CREBKID/CBP KIX domain34,35 and human survivinhomodimer36 interactions.We have validated a methodology in which small
molecules are used to allosterically inhibit an activeenzyme by targeting a conformationally dynamicregion. In the case of KSHV Pr, this conformation-ally dynamic region also governs the PPIs thatinfluence the inactive monomer/active dimer equi-librium. Taken together, the NMR and X-raycrystallography data indicate a monomer trap
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Alldoi:10.1016/j.jmb.2011.06.032
model of KSHV Pr inhibition:14 that DD2 binds toand captures a preexisting inactive monomer,thereby shifting the monomer–dimer equilibriumtoward the inactive conformational state. SinceHHV Prs are known to play critical roles in thelate lytic stage of the viral life cycle,25 inhibitingprotease function at this stage would be key to thedevelopment of novel therapeutics targeting her-pesviruses. Although promising results are ob-served in vitro for two of the eight members of theHHV Pr family examined thus far, modifications toDD2 or the discovery of new helical mimeticscaffolds with improved pharmacokinetic and phar-macodynamic properties is necessary for furtherdevelopment. In particular, DD2 did not display aninhibitory effect in preliminary viral infectivitystudies (data not shown) using a stably KSHV-infected SLK219 endothelial cell line.37–39 These
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
results are explained by poor solubility and cellpermeability of DD2 (Table S7), and we arecurrently exploring different chemical modificationsto the DD2 scaffold in order to address these issues.As the first structure of an allosterically inhibited
HHV Pr monomer to date, our KSHV Pr Δ196–DD2complex provides valuable insights. First, confor-mational perturbations in the crystal structure of theinactive complex relative to the active KSHV Prdimer suggest a reason that previous attempts atdeveloping an active-site inhibitor have failed.Specifically, the substrate binding pocket of theactive site may be too shallow and too dynamic for acompetitive inhibitor to bind, as evidenced by thelack of efficacious inhibitors that target the activesite. Second, the Δ196–DD2 structure, focusing onthe DD2 binding pocket, can be used as a templatefor in silico docking studies in future efforts todiscover other allosteric KSHV Pr inhibitors, includ-ing those that exhibit more pharmacologicallyfavorable characteristics than DD2. Third, andmore importantly, although members of the HHVPr family have low sequence homology, they arestructurally (Table S8) and functionally highlyhomologous. Inspection of the available HHV Prstructures (Fig. 10) reveals that all eight proteaseshave a potential hot spot aromatic residue at thedimer interface. This region could act as a target sitefor more potent allosteric, pan-specific inhibitorsthat trap the other HHV Prs in their respectiveinactive monomeric states. Allosteric inhibition as amethod to regulate conformationally dynamic com-plexes such as PPIs remains a mostly unexploitedpathway. By examining KSHV Pr–DD2 interactions,we have utilized this route as a potential solutiontoward developing therapeutics that regulate high-value, but previously intractable, targets.
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
Materials and Methods
Protease truncation and mutagenesis
The KSHV Pr M197D-S204G sequence15,17 was thetemplate used to generate genes encoding the Δ209,Δ196,andΔ191 truncation constructs. Genes encoding all KSHVPr M197D Ile→Val single point mutations were synthe-sized using the unoptimized sequences of KSHV PrM197D or KSHV Pr Δ196 as templates. Briefly, codonswithin the gene sequence representing isoleucine (ATA,ATC, or ATT) were changed to those encoding valine(GTA, GTC, or GTT). All gene sequences were designed atthe University of California, San Francisco, and synthe-sized/purchased from GeneArt, Inc. The wild-typeHCMV Pr sequence40 was the template for site-directedmutagenesis for the HMCV Pr L222D construct, per-formed using the QuikChange Site-Directed MutagenesisKit (Stratagene). The wild-type HCMV Pr sequence alsowas the template for generating the HCMV Pr ΔL221truncation sequence using the pET11a expression vector
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
(Novagen). All resulting genes were verified throughDNA sequencing.Primers (Integrated DNA Technologies, Inc.) encoding
All KSHV Pr protein samples were expressed in M9minimal media (NMR) or Luria broth (X-ray crystallogra-phy) and purified as previously described.14,16,22 SolubleHCMV Pr was expressed and purified as describedpreviously41 with the following modifications. Followingovernight expression at 16 °C, cells were pelleted and lysedby sonication, 0.5 s pulse/1 s recovery for 5 min, followedby 3 s pulse/3 s recovery for 2 min. After centrifugation at30,000g for 45min, soluble fractionswere passed over a 1-mlHisTrap FF column (GE Healthcare Life Sciences) using thefollowing buffers: wash/binding (50 mM NaPi, 300 mMNaCl, 25 mM imidazole, and 5 mM β-mercaptoethanol,pH 8.0) and elution (50 mM NaPi, 300 mM NaCl, 300 mMimidazole, and 5 mM β-mercaptoethanol). Eluted His-tagged proteins were buffer exchanged into 25 mM NaPi,150 mM NaCl, and 5 mM β-mercaptoethanol and furtherpurified over a Superdex75 (26/60) size-exclusion chroma-tography column (GE Healthcare Life Sciences).Starting protein concentrations for all assays described
below were measured using a Nanodrop 2000c UVspectrophotometer (Thermo Scientific). Selective [13C–1Hmethyl] isotopic labeling of the isoleucine, leucine, andvaline residues was achieved by adding 100 mg l−1
[dimethyl-13C2]-α-ketoisovaleric acid (Cambridge IsotopeLaboratories) and 50 mg l−1 [methyl-13C1]-α-ketobutyricacid (Sigma) to otherwise unlabeled M9 minimal media1 h prior to IPTG induction.42
Calculation of protein sample concentrations
Starting protein concentrations for all assays describedbelow were measured using a Nanodrop 2000c UVspectrophotometer (Thermo Scientific) at 280 nm using thefollowing extinction coefficients: KSHV Pr M197D andKSHV Pr M197D-I201V, ɛ=23,950 M−1 cm−1; KSHV PrΔ209, Δ196, Δ191, and all Δ196 Ile→Val mutants,
steric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
ɛ=22,460 M−1 cm−1; HCMV Pr wild type and HCMV PrL222D, ɛ=28,420M−1 cm−1; HCMVPrΔ221, ɛ=23,950M−1
cm−1.
846847848849850851
852
Protease kinetic analysis
Kinetic proteolysis assays were performed on HCMV Prvariants using the same methodology as previouslydescribed for KSHV Pr studies.14 Monomer–dimer Kd(N=3) and IC50 (N=4) values reported herein representthe average and standard deviation.
853
854855856857858859860861862863864865866
867
868
869
870
871
872
873
874
875
876
CD analysis
Protease was allowed to equilibrate with or withoutDD2 for 1 h at room temperature in buffer solution(25 mM sodium phosphate buffer, 75 mM NaCl, 1 mMβ-mercaptoethanol, and 0.8% dioxane, pH 8). Dioxanewasused to solubilize the DD2 stocks in place of dimethylsulfoxide (DMSO), which absorbs far-UV CD wave-lengths. Spectra were obtained using a 2-mm-path-lengthcuvette on a Jasco J-715 spectropolarimeter with a finalprotein concentration of ∼3 μM. Data were acquired at aconstant temperature of 25 °C with the followingparameters: accumulations, 3; scan rate, 50 nm min−1;data pitch, 0.1 nm; response, 4 s; bandwidth, 20 nm; andstandard sensitivity. Both buffer with DD2 and thatwithout DD2 were used to acquire background spectra.Final CD data spectra are reported as the mean residueellipticity (deg cm2 dmol−1 residue−1). Fractional helicity(fH) values were estimated from the 222-nm band:43
u222;max = −44; 000 + 250Tð Þ 1 −2:5N
� �ð1Þ
where θ222,max is the calculated mean residue ellipticityvalue for a theoretical 100% helical polypeptide of Nresidues, collected at temperature T (°C). At 25 °C: KSHVPr (N=229 residues), θ222,max=−37,338 deg cm2 dmol−1
All spectra were acquired on cryoprobe-equippedBruker Avance 500-MHz or 800-MHz NMR spectrometersat 12 °C or 27 °C. Typical NMR samples used for the13C–1H HSQC titration studies consisted of ∼0.01–0.02 mM selective [13C–1H methyl]ILV-labeled protein in0.45 ml buffer. Sample preparation, including bufferconditions and DD2 titrations, and spectral acquisitionparameters were as previously described.14 Each titrationstudy contained at least one repeat acquisition point. Alldata were processed using NMRPipe44 and analyzedusing Sparky.45
892
893
894
895
896
897
898
Calculation of NMR-based apparent Kd values
Chemical shift data for the δ1-methyl groups of Ile44,Ile71, and Ile105 were converted to frequency values.Chemical shift perturbation values versus resonances
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Alldoi:10.1016/j.jmb.2011.06.032
corresponding to the apoprotein are expressed as com-bined 13C–1H frequency perturbations (Δωobs):
where ΔδH and ΔδC correspond to the 1H and 13Cchemical shift perturbations versus the apoprotein, respec-tively. specfreqH and specfreqC correspond to the spectralfrequencies of 1H and 13C, respectively.Titration data points from the 13C–1H HSQC spectra
were fit to a modified Hill equation46 using Matlab (TheMathworks, Inc.).
DNobs =DNmax × L½ �NTKd + L½ �NT
ð3Þ
where Δωmax is the maximum frequency perturbation,corresponding to a fully saturated ligand-bound state. [L]Tis the total ligand concentration, and N is the Hillcoefficient. Estimated values for Δωmax, N, and Kd werecalculated from a nonlinear regression fitting of Eq. (2),using a grid search methodology to minimize χ2 errors.Final estimated apparent Kd values and errors reported inthis paper are averages and standard deviations calculat-ed for Ile44, Ile71, and Ile105. Estimated experimentalerrors were based upon repeat data points acquiredduring the NMR titrations and were propagated through-out the calculations.
X-ray crystallography data acquisition and structuredetermination
KSHV Pr Δ196 stock solutions used for X-raycrystallography consisted of 25 mM Tris (pH 8.0),150 mM KCl, 0.1 mM ethylenediaminetetraacetic acid,and 2 mM DTT. Crystals were grown at 25 °C with thehanging-drop vapor diffusion method, with DD2 ingreater than 5-fold molar excess with respect to Δ196.The reservoir solution consisted of 2 M (NH4)2SO4, 0.2 MNaCl, 0.05 M sodium cacodylate (pH 7.0), 0.1 M urea,and 0.1 M sodium acetate. Following 1 day of incubation,1 μl of 14 M β-mercaptoethanol was added to thereservoir. After 14–21 days, crystals appeared as a smallcube measuring 0.03 mm×0.05 mm×0.05 mm. A 2.0-Å-resolution X-ray diffraction data set was collected at theLawrence Berkeley National Laboratory Advanced LightSource Beamline 8.3.1, using a crystal flash-cooled to100 K in mother liquor with 20% glycerol as thecryoprotectant.Diffraction images were processed using DENZO and
SCALEPACK from the HKL-2000 suite.47 The resultingstructure was solved by molecular replacement withPhaser48 and using residues 1–196 of a monomeric unitof the phosphonate-inhibited KSHV Pr dimer (2PBK)22 asthe template search model. The resulting structure modelwas a dimer in an asymmetric unit and was subjected tomultiple rounds of restrained refinement and isotropic B-factor minimization with REFMAC49 and Coot50 (pre-refinement: R-factor=45.6%, Rfree=49.5%; post-refine-ment: R-factor=20.3%, Rfree=24.8%).All structural figures and animations within this article
and the Supplementary Materials were created using
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
PyMOL 1.2 (Schrödinger, LLC). Morph calculations(Movies S2 and 3) were performed using the YaleMorph Server†.
DD2 solubility measurements
Test compounds were serially diluted from 10,000 μMto 625 μM in DMSO and placed in columns 1–5 and 7–11of a 96-well polypropylene plate (Costar 3365). Columns 6and 12 were filled with DMSO as the background. Fromeach well, 5 μl was transferred into the 96-well disposableUV-Star plate (Greiner Bio-One). Acetonitrile (97.5 μl) andphosphate-buffered saline buffer (pH 7.4, 97.5 μl) wereadded to each well, and the plate was agitated for 30 minusing an IKAmicrotiter plate shaker. The UV spectra from200 to 500 nm were measured for all wells and subtractedfrom the DMSO background. Correlations betweenconcentrations and absorbance at 260, 280, and 300 nmwere determined as slopes. Then, 5 μl from each well ofthe polypropylene plate was added to a MultiScreenSolubility Filter Plate (Millipore) and diluted with 195 μl ofphosphate-buffered saline. The plate was agitated for 2 hand filtered into a 96-well disposable UV-Star plate, andthe UV absorbance at 260, 280, and 300 nmwas measured.The aqueous solubility (Amax filtrate/slope) was deter-mined for all three wavelengths, and values are given asthe means with 95% confidence intervals.
DD2 cell permeability measurements
All liquid-handling steps for the PAMPA assay wereperformed on a Biomek FX Laboratory AutomationWorkstation (Beckman Coulter) and analyzed by pION's(London, UK) PAMPA Evolution 96 Command Software.The PAMPA Evolution 96 Permeability Assay Kit includesthe Acceptor Sink Buffer, Double-Sink Lipid Solution, anda PAMPA sandwich plate, preloaded with magnetic disks.For each experiment, 4 μl of lipid was transferred onto thesupport membrane in the acceptor well, followed byaddition of 200 μl of Acceptor Sink Buffer (pH 7.4). Then,180 μl of diluted test compound (50 μM in system buffer atpH 7.4 starting from a 10-mM DMSO solution) was addedto the donor wells. The PAMPA sandwich plate wasassembled and placed on the Gut-Box and stirred for30 min. The distribution of the compounds in the donorand acceptor buffers (100 μl aliquot) was determined bymeasuring the UV spectra from 200 to 500 nm using theSpectraMax reader (Molecular Devices). The permeabilitycoefficient was determined using the maximum absor-bance from 200 to 500 nm:51
Pe =2:3VD
A t − tlag� �� � log10 1
1 − Rð Þ
CD tð ÞCD 0ð Þ
� �ð4Þ
where VD is the donor well volume (cm3), A is the filterarea (cm2), CD(0) is the sample concentration in the donorwell at time 0 (mol cm−3), CD(t) is the sample concentra-tion in the donor well at time t (mol cm−3), t is the intervalof time (s), tLAG is the lag time needed to reach steady-state
†http://www.molmovdb.org/
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
conditions (s), and R is the membrane retention (related tothe membrane/water partition coefficient). Standardsused were verapamil (Pe=1505×10
−6 cm s−1) as high-permeability standard, carbamazepine (Pe=150×10
− 6 cms−1) as medium-permeability standard, and ranitidine(Pe=2.3×10
− 6 cm s−1) as low-permeability standard. Thecompounds were measured in triplicate, and values aregiven as the mean values with 95% confidence intervals.
PDB accession numbers
Structural coordinates for the KSHV Pr Δ196–DD2complex have been deposited at the Brookhaven PDBwithaccession number 3NJQ.
Acknowledgements
This work was funded in part by grants from theNational Institutes of Health (grant 1R01-A1067423)and the HIV Accessory and Regulatory ComplexCenter (grant P50-GM082250). This research wasalso supported by a grant from the NationalInstitutes of Health, University of California SanFrancisco/Gladstone Institute of Virology & Immu-nology Center for AIDS Research (P30-A122763 toG.M.L.). We thank Prof. Don Ganem, Prof. R. KipGuy, Dr. Leggy Arnold, and Mr. Ernest Lam fortechnical support. We also thank Prof. John D. Grossand Dr. Ana Lazic for helpful discussions of themanuscript.
Supplementary Data
Supplementary data to this article can be foundonline at doi:10.1016/j.jmb.2011.06.032
996
997
998
References
1. Dyson, H. J. & Wright, P. E. (2005). Intrinsicallyunstructured proteins and their functions. Nat. Rev.,Mol. Cell Biol. 6, 197–208.
2. Boehr, D. D., Nussinov, R. & Wright, P. E. (2009). Therole of dynamic conformational ensembles in biomo-lecular recognition. Nat. Chem. Biol. 5, 789–796.
3. Zinzalla, G. & Thurston, D. E. (2009). Targetingprotein–protein interactions for therapeutic interven-tion: a challenge for the future. Future Med. Chem. 1,65–93.
4. Metallo, S. J. (2010). Intrinsically disordered proteinsare potential drug targets. Curr. Opin. Chem. Biol. 14,481–488.
5. Kar, G., Keskin, O., Gursoy, A. & Nussinov, R. (2010).Allostery and population shift in drug discovery.Curr. Opin. Pharmacol. 10, 715–722.
6. Lee, G. M. & Craik, C. S. (2009). Trapping movingtargets with small molecules. Science, 324, 213–215.
steric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
7. Arkin, M. R. & Wells, J. A. (2004). Small-moleculeinhibitors of protein–protein interactions: progressingtowards the dream. Nat. Rev., Drug Discov. 3, 301–317.
8. Berg, T. (2008). Small-molecule inhibitors of protein–protein interactions. Curr. Opin. Drug Discovery Dev.11, 666–674.
9. Yin, H. & Hamilton, A. D. (2005). Strategies fortargeting protein–protein interactions with syntheticagents. Angew. Chem., Int. Ed. Engl. 44, 4130–4163.
10. Hershberger, S. J., Lee, S. G. & Chmielewski, J. (2007).Scaffolds for blocking protein–protein interactions.Curr. Top. Med. Chem. 7, 928–942.
11. Ross, N. T., Katt, W. P. & Hamilton, A. D. (2010).Synthetic mimetics of protein secondary structuredomains. Philos. Trans.-Royal Soc., Math. Phys. Eng. Sci.368, 989–1008.
12. Clackson, T. & Wells, J. A. (1995). A hot spot ofbinding energy in a hormone–receptor interface.Science, 267, 383–386.
13. Ma, B. & Nussinov, R. (2007). Trp/Met/Phe hot spotsin protein–protein interactions: potential targets indrug design. Curr. Top. Med. Chem. 7, 999–1005.
14. Shahian, T., Lee, G. M., Lazic, A., Arnold, L. A.,Velusamy, P., Roels, C. M. et al. (2009). Inhibition of aviral enzyme by a small-molecule dimer disruptor.Nat. Chem. Biol. 5, 640–646.
15. Nomura, A. M., Marnett, A. B., Shimba, N., Dotsch, V.& Craik, C. S. (2005). Induced structure of a helicalswitch as a mechanism to regulate enzymatic activity.Nat. Struct. Mol. Biol. 12, 1019–1020.
16. Nomura, A. M., Marnett, A. B., Shimba, N., Dotsch, V.& Craik, C. S. (2006). One functional switch mediatesreversible and irreversible inactivation of a herpesvi-rus protease. Biochemistry, 45, 3572–3579.
17. Shimba,N.,Nomura,A.M.,Marnett,A.B.&Craik, C. S.(2004). Herpesvirus protease inhibition by dimerdisruption. J. Virol. 78, 6657–6665.
18. del Sol, A., Tsai, C. J., Ma, B. & Nussinov, R. (2009).The origin of allosteric functional modulation: multi-ple pre-existing pathways. Structure, 17, 1042–1050.
19. Ogilvie, W., Bailey, M., Poupart, M. A., Abraham, A.,Bhavsar, A., Bonneau, P. et al. (1997). Peptidomimeticinhibitors of the human cytomegalovirus protease. J.Med. Chem. 40, 4113–4135.
20. Borthwick, A. D., Crame, A. J., Ertl, P. F., Exall, A. M.,Haley, T. M., Hart, G. J. et al. (2002). Design andsynthesis of pyrrolidine-5,5-trans-lactams (5-oxohex-ahydropyrrolo[3,2-b]pyrroles) as novel mechanism-based inhibitors of human cytomegalovirus protease.2. Potency and chirality. J. Med. Chem. 45, 1–18.
21. Gopalsamy, A., Lim, K., Ellingboe, J. W., Mitsner, B.,Nikitenko, A., Upeslacis, J. et al. (2004). Design andsyntheses of 1,6-naphthalene derivatives as selectiveHCMV protease inhibitors. J. Med. Chem. 47,1893–1899.
22. Lazic, A., Goetz, D. H., Nomura, A. M., Marnett, A. B.& Craik, C. S. (2007). Substrate modulation of enzymeactivity in the herpesvirus protease family. J. Mol. Biol.373, 913–923.
23. Reiling, K. K., Pray, T. R., Craik, C. S. & Stroud, R. M.(2000). Functional consequences of the Kaposi'ssarcoma-associated herpesvirus protease structure:regulation of activity and dimerization by conservedstructural elements. Biochemistry, 39, 12796–12803.
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Alldoi:10.1016/j.jmb.2011.06.032
24. Laskowski, R. A., MacArthur, M. W., Moss, D. S. &Thornton, J. M. (1993). PROCHECK: a program tocheck the stereochemical quality of protein structures.J. Appl. Crystallogr. 26, 283–291.
25. Tong, L. (2002). Viral proteases. Chem. Rev. 102,4609–4626.
26. Pray, T. R., Reiling, K. K., Demirjian, B. G. & Craik, C. S.(2002). Conformational change coupling the dimeriza-tion and activation of KSHV protease. Biochemistry, 41,1474–1482.
27. Hwang, Y. S. & Chmielewski, J. (2005). Developmentof low molecular weight HIV-1 protease dimerizationinhibitors. J. Med. Chem. 48, 2239–2242.
28. Frutos, S., Rodriguez-Mias, R. A., Madurga, S.,Collinet, B., Reboud-Ravaux, M., Ludevid, D. &Giralt, E. (2007). Disruption of the HIV-1 proteasedimer with interface peptides: structural studies usingNMR spectroscopy combined with [2-13C]-Trp selec-tive labeling. Biopolymers, 88, 164–173.
29. Bannwarth, L., Rose, T., Dufau, L., Vanderesse, R.,Dumond, J., Jamart-Gregoire, B. et al. (2009). Dimerdisruption and monomer sequestration by alkyltripeptides are successful strategies for inhibitingwild-type and multidrug-resistant mutated HIV-1proteases. Biochemistry, 48, 379–387.
30. Hu, C. Q. &Hu, Y. Z. (2008). Small molecule inhibitorsof the p53-MDM2. Curr. Med. Chem. 15, 1720–1730.
31. Orner, B. P., Ernst, J. T. & Hamilton, A. D. (2001).Toward proteomimetics: terphenyl derivatives asstructural and functional mimics of extendedregions of an alpha-helix. J. Am. Chem. Soc. 123,5382–5383.
32. Kutzki, O., Park, H. S., Ernst, J. T., Orner, B. P., Yin, H.& Hamilton, A. D. (2002). Development of a potentBcl-x(L) antagonist based on alpha-helix mimicry. J.Am. Chem. Soc. 124, 11838–11839.
33. Petros, A.M., Dinges, J., Augeri, D. J., Baumeister, S. A.,Betebenner,D.A., Bures,M.G. et al. (2006).Discovery ofa potent inhibitor of the antiapoptotic protein Bcl-xLfrom NMR and parallel synthesis. J. Med. Chem. 49,656–663.
34. Best, J. L., Amezcua, C. A., Mayr, B., Flechner, L.,Murawsky, C. M., Emerson, B. et al. (2004). Identifi-cation of small-molecule antagonists that inhibit anactivator: coactivator interaction. Proc. Natl Acad. Sci.USA, 101, 17622–17627.
35. Bates, C. A., Pomerantz, W. C. & Mapp, A. K. (2011).Transcriptional tools: small molecules for modulatingCBP KIX-dependent transcriptional activators. Bio-polymers, 95, 17–23.
36. Wendt, M. D., Sun, C., Kunzer, A., Sauer, D., Sarris,K., Hoff, E. et al. (2007). Discovery of a novel smallmolecule binding site of human survivin. Bioorg. Med.Chem. Lett. 17, 3122–3129.
37. Bechtel, J. T., Liang, Y., Hvidding, J. & Ganem, D.(2003). Host range of Kaposi's sarcoma-associatedherpesvirus in cultured cells. J. Virol. 77, 6474–6481.
38. Vieira, J. & O'Hearn, P. M. (2004). Use of the redfluorescent protein as a marker of Kaposi's sarcoma-associated herpesvirus lytic gene expression. Virology,325, 225–240.
39. Chandriani, S. & Ganem, D. (2010). Array-basedtranscript profiling and limiting-dilution reversetranscription-PCR analysis identify additional latent
osteric Capture of an Inactive Conformation, J. Mol. Biol. (2011),
genes in Kaposi's sarcoma-associated herpesvirus.J. Virol. 84, 5565–5573.
40. Loveland, A. N., Chan, C. K., Brignole, E. J. & Gibson,W. (2005). Cleavage of human cytomegalovirus prote-ase pUL80a at internal and cryptic sites is not essentialbut enhances infectivity. J. Virol. 79, 12961–12968.
41. Brignole, E. J. & Gibson, W. (2007). Enzymaticactivities of human cytomegalovirus maturationalprotease assemblin and its precursor (pPR, pUL80a)are comparable: [corrected] maximal activity of pPRrequires self-interaction through its scaffolding do-main. J. Virol. 81, 4091–4103.
42. Goto, N. K., Gardner, K. H., Mueller, G. A., Willis,R. C. & Kay, L. E. (1999). A robust and cost-effectivemethod for the production of Val, Leu, Ile (delta 1)methyl-protonated 15N-, 13C-, 2H-labeled proteins. J.Biomol. NMR, 13, 369–374.
43. Wallimann, P., Kennedy, R. J. & Kemp, D. S. (1999).Large circular dichroism ellipticities for N-templatedhelical polypeptides are inconsistent with currentlyaccepted helicity algorithms. Angew. Chem., Int. Ed.Engl. 38, 1290–1292.
44. Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, G.,Pfeifer, J. & Bax, A. (1995). NMRPipe: a multidimen-sional spectral processing system based on UNIXpipes. J. Biomol. NMR, 6, 277–293.
Please cite this article as: Lee, G. M. et al., Enzyme Inhibition by Allodoi:10.1016/j.jmb.2011.06.032
45. Goddard, T. D. & Kneller, D. G. (1999). Sparky 3.University of California, San Francisco, CA.
46. Ramstad, T.,Hadden, C. E.,Martin, G. E., Speaker, S.M.,Teagarden, D. L. & Thamann, T. J. (2005). Determinationby NMR of the binding constant for the molecularcomplex between alprostadil and alpha-cyclodextrin.Implications for a freeze-dried formulation. Int. J. Pharm.296, 55–63.
47. Otwinowski, Z., Minor, W. & Carter, C. W., Jr. (1997).Processing of X-ray diffraction data collected inoscillation mode. Methods Enzymol. 276, 307–326.
48. McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D.,Winn, M. D., Storoni, L. C. & Read, R. J. (2007). Phasercrystallographic software. J. Appl. Crystallogr. 40,658–674.
49. Skubak, P., Murshudov, G. N. & Pannu, N. S. (2004).Direct incorporation of experimental phase informa-tion in model refinement. Acta Crystallogr., Sect. D:Biol. Crystallogr. 60, 2196–2201.
50. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan,K. (2010). Features and development of Coot.Acta Crystallogr., Sect. D: Biol. Crystallogr. 66,486–501.
