Site-directed Mutagenesis and Characterization of Uracil-DNAGlycosylase Inhibitor ProteinROLE OF SPECIFIC CARBOXYLIC AMINO ACIDS IN COMPLEX FORMATION WITH ESCHERICHIA COLIURACIL-DNA GLYCOSYLASE*

(Received for publication, December 18, 1996, and in revised form, May 19, 1997)

Amy J. Lundquist‡, Richard D. Beger§, Samuel E. Bennett‡¶, Philip H. Bolton§**,and Dale W. Mosbaugh‡‡‡§§

From the ‡Departments of Agricultural Chemistry and Biochemistry and Biophysics and the ‡‡Environmental HealthSciences Center, Oregon State University, Corvallis, Oregon 97331 and the §Chemistry Department, Wesleyan University,Middleton, Connecticut 06459

Bacteriophage PBS2 uracil-DNA glycosylase inhibitor(Ugi) protein inactivates uracil-DNA glycosylase (Ung)by acting as a DNA mimic to bind Ung in an irreversiblecomplex. Seven mutant Ugi proteins (E20I, E27A, E28L,E30L, E31L, D61G, and E78V) were created to assess therole of various negatively charged residues in the bind-ing mechanism. Each mutant Ugi protein was purifiedand characterized with respect to inhibitor activity andUng binding properties relative to the wild type Ugi.Analysis of the Ugi protein solution structures by nu-clear magnetic resonance indicated that the mutant Ugiproteins were folded into the same general conforma-tion as wild type Ugi. All seven of the Ugi proteins werecapable of forming a UngzUgi complex but varied con-siderably in their individual ability to inhibit Ung activ-ity. Like the wild type Ugi, five of the mutants formed anirreversible complex with Ung; however, the binding ofUgi E20I and E28L to Ung was shown to be reversible.The tertiary structure of [13C,15N]Ugi in complex withUng was determined by solution state multi-dimen-sional nuclear magnetic resonance and compared withthe unbound Ugi structure. Structural and functionalanalysis of these proteins have elucidated the two-stepmechanism involved in UngzUgi association and irre-versible complex formation.

The Bacillus subtilis bacteriophage PBS1 and -2 exhibit aunique genetic system that naturally contains uracil in place ofthymine in a double-stranded DNA genome (1, 2). Stable incor-poration of uracil residues into the phage DNA is achieved bythe substitution of dUTP for dTTP as precursor in DNA syn-

thesis and the concomitant inactivation of the host uracil-mediated base excision DNA repair pathway (2–4). To blockuracil-DNA repair and protect the uracil-containing phageDNA from degradation, an early phage gene (ugi)1 is expressedthat inhibits the B. subtilis uracil-DNA glycosylase. The aminoacid sequences of the PBS1 and -2 Ugi proteins appear to beidentical (5–7).

The PBS2 ugi gene encodes a small (9,474 dalton), mono-meric, heat stable protein of 84 amino acids that inactivatesuracil-DNA glycosylases from diverse biological sources (5, 8,9). The ugi gene product is an unusually acidic protein (12 Glu,6 Asp) with a pI 5 4.2 that migrates anomalously duringSDS-polyacrylamide gel electrophoresis (5, 10, 11). Ugi inacti-vates Ung by forming a tightly bound noncovalent complexwith 1:1 stoichiometry that is essentially irreversible underphysiological conditions (10, 12). Stopped-flow kinetic studiesof the Ugi interaction with Escherichia coli Ung indicate thatcomplex formation is accomplished through a two-step bindingreaction (12). In the initial step, the association between freeUgi and Ung is characterized by a rapid pre-equilibrium reac-tion with a dissociation constant (Kd) of 1.3 mM; the second step,the formation of an irreversible complex, is characterized bythe rate constant k 5 195 s21. Thus, UngzUgi complex forma-tion initiates with a “docking” interaction that facilitates opti-mal alignment between the two proteins. If correct alignmentbetween Ung and Ugi does not occur, a reversible associationwill transpire. If, however, proper alignment is achieved, thena “locked” complex quickly follows.

The secondary and tertiary structures of free Ugi were re-cently determined by solution state multi-dimensional NMRtechniques and found to include two a-helices and five anti-parallel b-strands as illustrated in Fig. 1 (13, 14). The fivecontiguous b-strands are connected by short loop regions toform an anti-parallel b-sheet. Analysis of the electrostatic po-tential of Ugi revealed several striking features (14). Seven ofthe 18 acidic amino acid residues (Glu-20, Asp-48, Glu-49,Asp-52, Glu-53, Asp-74, and Glu-78) come together to form aregion of high negative potential on one face of the protein.Each of the residues that form this electrostatic region or“knob” are located immediately adjacent to or terminate ab-strand. Two other acidic amino acid residues (Asp-40 and

* This work was supported by National Institutes of Health GrantsGM32823 and ES00210 (to D. W. M.) and NP-750 from the AmericanCancer Society (to P. H. B.). The NMR spectrometer was purchasedwith support from the National Science Foundation Grant BIR-95-12478 and by a grant from the Camille and Henry Dreyfus Foundation(to P. H. B.). This is Technical Report 11083 from the Oregon Agricul-tural Experimental Station. The costs of publication of this article weredefrayed in part by the payment of page charges. This article musttherefore be hereby marked “advertisement” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.

¶ Present address: Laboratory of Molecular Genetics, NIEHS, Re-search Triangle Park, NC 27709.

** To whom correspondence may be addressed: Dept. of Chemistry,Wesleyan University, Middleton, CT 06459. Tel.: 860-685-2668; Fax:860-685-2211.

§§ To whom correspondence should be addressed: Dept. of Agricul-tural Chemistry, Oregon State University, 1007 Agricultural & LifeSciences Bldg., Corvallis, OR 97331-7301. Tel.: 541-737-1797; Fax:541-737-0497.

1 The abbreviations used are: ugi and Ugi, bacteriophage PBS1 or 2uracil-DNA glycosylase inhibitor gene and protein, respectively; Ung,E. coli uracil-DNA glycosylase; NOE, nuclear Overhauser effect;NOESY, nuclear Overhauser effect correlation spectroscopy; BSA, bo-vine serum albumin; HSV-1, herpes simplex virus type-1; bp, basepair(s); kb, kilobase pair(s); TOCSY-HSQC, total correlation spectros-copy-heteronuclear multiple quantum correlation spectroscopy.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 272, No. 34, Issue of August 22, pp. 21408–21419, 1997© 1997 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org21408

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Asp-61) are also in juxtaposition to the end of b-strands; Glu-78and Glu-64 reside in the loop regions (14). The remaining sevennegatively charged residues are located in the a1-helix (Asp-6,Glu-9, and Glu-11) and a2-helix (Glu-27, Glu-28, Glu-30, andGlu-31). Both the a1- and a2-helix elements project away fromthe b-sheet and are located on potentially flexible arms of thepolypeptide (14). Furthermore, the a2-helix is longitudinallysegmented into a hydrophobic face and a negatively chargedface where the four glutamic acid residues protrude.

Several lines of evidence suggest that some of the negativelycharged amino acid residues of Ugi may act as a DNA mimicand mediate the interaction with Ung. First, UV-catalyzedcross-linking of oligonucleotide (dT)20 to the DNA-binding siteof Ung blocked Ugi from forming a UngzUgi complex (15).Second, the x-ray crystallographic structure of Ugi in complexwith human (16) and HSV-1 (17) uracil-DNA glycosylase re-veals that the interfacing surface of Ugi shares shape andelectrostatic complementarity to the DNA-binding groove of theenzyme (16, 17). Third, the negative electrostatic knob of Ugiexhibits an electrostatic potential of .6.6 kcal, which is similarto that generated by the negatively charged phosphate back-bone of DNA (14). Fourth, the recent x-ray structure of humanuracil-DNA glycosylase complexed with a 10-bp oligonucleotidecontaining a target GzU mispair reveals the DNA complexed atthe same site as Ugi (18). Fifth, charge neutralization by car-bodiimide-mediated adduction of Ugi carboxylic acid residuescaused a decrease in inhibitor protein activity (19). Finally,chemical adduction of specific glutamic acid residues (Glu-28and Glu-31) of Ugi located in the a2-helix correlated with theformation of an unstable UngzUgi complex (19).

Bennett et al. (12) suggested that after the Ung/Ugi associ-ation, the transition to the locked configuration may involve a

conformational change in either one or both proteins. Subse-quently, Sanderson and Mosbaugh (19) proposed that the lock-ing reaction is caused predominantly by a change in Ugi struc-ture. This position is supported by a comparison of the crystalstructures of free human and HSV-1 uracil-DNA glycosylasewith the structures of each enzyme in complex with Ugi (16, 17,20, 21). In both cases, the tertiary structure of the enzymeshows only minor structural changes. In contrast, a comparisonof the heteronuclear multiple quantum correlation spectra offree and bound [15N]Ugi indicates that many residues of Ugiundergo conformational change upon binding to Ung (14). Atpresent, the tertiary structure of the unbound Ugi protein insolution was determined solely by solution state NMR tech-niques (14). Also, a comparison of the solution structure of freeUgi with the crystal structure of Ugi complexed with either thehuman or HSV-1 uracil-DNA glycosylase demonstrates signif-icant structural changes occur in Ugi (14, 16, 17). A morecomplete understanding of the docking and locking reactionsmay well be gained by determining the solution state structureof Ugi in complex with Ung.

In the present report we (i) conduct site-directed mutagene-sis of seven acidic residues of Ugi; (ii) purify each mutant Ugiprotein to apparent homogeneity; (iii) characterize each mu-tant Ugi with regard to specific activity and UngzUgi complexstability and reversibility; (iv) determine the structural simi-larity between wild type Ugi and the Ugi mutant proteins usingNMR methods; (v) compare the free and complexed Ugi solu-tion structures; and (vi) model the interactions in the wild typeand mutant UngzUgi complexes.

EXPERIMENTAL PROCEDURES

Materials—Restriction endonucleases (EcoRI, EcoRV, HindIII, PstI,SacI, and XmnI), T4 polynucleotide kinase, T4 DNA polymerase, andT4 DNA ligase were purchased from New England Biolabs. Isopropyl-1-thio-b-D-galactopyranoside and ScaI were obtained from Life Tech-nologies, Inc. and AcyI came from Promega. [3H]Leucine and [35S]me-thionine were obtained from NEN Life Science Products; [3H]dUTP wasfrom Amersham Corp., and [13C]glucose and [15N]ammonium chloridewere from Cambridge Isotope Laboratories.

E. coli JM105 was provided by W. Ream (Oregon State University),and E. coli CJ236 was obtained from T. A. Kunkel (NIEHS). Epicuriancoli XL2-Blue ultracompetent cells, phagemid pBluescript II SK(2),and VCS-M13 helper phage were supplied by Stratagene. PlasmidpKK223-3 was obtained from Pharmacia Biotech Inc.; pZWtacl (22) andpSB1051 (12) were constructed as described previously. Oligonucleo-tides were synthesized using an Applied Biosystems 380B DNA Syn-thesizer by the Center for Gene Research and Biotechnology (OregonState University).

Site-directed Mutagenesis of the Uracil-DNA Glycosylase InhibitorGene—The first step in the site-directed mutagenesis procedure in-volved subcloning of the ugi gene into a pBluescript-based phagemid toproduce single-stranded DNA. The pZWtac1 EcoRI-HindIII restrictionfragment (726 bp) containing the ugi gene (Fig. 2) was inserted into thecorresponding EcoRI and HindIII sites of pBluescript II SK(2) using T4DNA ligase. The resulting phagemid (pAL) was transformed into E. coliJM105, plated on LB plates containing 100 mg/ml ampicillin, 40 mM

isopropyl-1-thio-b-D-galactopyranoside, and 40 mg/ml 5-bromo-4-chloro-3-indolyl b-D-galactopyranoside, after which pAL DNA was purifiedfrom white colonies. E. coli CJ236 (dut, ung) was then transformed withphagemid pAL and grown at 37 °C in 1.0 liter of 2 3 YT mediumsupplemented with 34 mg/ml chloramphenicol and 100 mg/ml ampicillin.Upon reaching a cell density of 108 cells/ml (1 A600 nm 5 8 3 108

cells/ml), uridine was added to a final concentration of 0.25 mg/ml;VCS-M13 helper phage was added at a multiplicity of infection equal to1.0, and incubation was continued at 37 °C for 1.5 h. Kanamycin (26mg/ml final concentration) was added to select for infected E. coli cellsand growth continued for an additional 5.5 h. The culture was centri-fuged at 7000 rpm for 15 min at 4 °C in a GSA (Sorvall) rotor, and thesupernatant fraction was processed to precipitate pAL phage with theaddition of 0.25 volume of a 15% PEG-8000 and 2.5 M NaCl solution.Phage DNA was isolated from the supernatant fraction following ex-tractions with phenol and chloroform:isoamyl alcohol (24:1) and precip-itation with ethanol (23). The precipitated DNA was centrifuged at 9500

FIG. 1. Tertiary structure of the uracil-DNA glycosylase inhib-itor protein and location of Glu and Asp residues. The tertiarystructure of Ugi determined by solution state NMR techniques (14) isshown with the 12 Glu residues in red and 6 Asp residues in yellow.Secondary structural elements include the a1-helix (Ser-5 to Lys-14),a2-helix (Glu-27 to Asn-35), b1-strand (Glu-20 to Met-24), b2-strand(Ile-41 to Asp-48), b3-strand (Glu-53 to Ser-60), b4-strand (Ala-69 toAsp-74), and b5-strand (Asn-79 to Leu-84).

Uracil-DNA Glycosylase Inhibitor Protein 21409

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

rpm for 20 min at 4 °C in a SA600 (Sorvall) rotor, air dried, andresuspended in 500 ml of TE buffer (10 mM Tris-HCl (pH 8.0), 1 mM

EDTA). This single-stranded uracil-substituted DNA that containedthe antisense ugi gene sequence was termed pALU(ss) DNA.

The second step involved in vitro primer extension of various oligo-nucleotides containing site-directed mutations in the ugi structuralgene. Deblocked/deprotected oligonucleotides were purified by Seph-adex G-25 chromatography and phosphorylated at the 59 end using ATPand T4 polynucleotide kinase as described previously (23). DefinedDNA primer/templates were constructed by annealing various oligonu-cleotides (20 pmol) to pALU(ss) DNA at a 3:1 (primer:template) ratio ina 100-ml volume as described (24, 25). Primer extension reaction mix-tures (100 ml) contained 20 mM Hepes-KOH (pH 8.0), 2 mM dithiothre-itol, 10 mM MgCl2, 2 mM ATP, 500 mM each of dATP, dCTP, dGTP, anddTTP, 0.5 mg/ml BSA, 6 units of T4 DNA polymerase, 200 units of T4DNA ligase, and 1.8 pmol of the heteroduplex pALU DNA primer/template. Following incubation for 5 min at 25 °C and then 2 h at 37 °C,a sample (10 ml) terminated with the addition of 1.5 ml of 0.1 M EDTAwas analyzed by 1% agarose gel electrophoresis to determine the extentof primer extension. Transformation of E. coli JM105 CCMB80 compe-tent cells was performed with 10 ml of the primer extension reactionmixture (26). Transformed bacterial colonies were grown on LB platescontaining 100 mg/ml ampicillin, and isolated colonies were subse-quently grown to saturation in 2 ml of 2 3 YT medium supplementedwith ampicillin. Plasmid DNA (pSugi) was isolated using the WizardMiniprep DNA purification technique (Promega). Isolated plasmids(pSugi) were analyzed by 1% agarose gel electrophoresis after threeindependent restriction endonuclease digestions as follows: (i) HindIII,(ii) HindIII plus EcoRI, and (iii) the appropriate novel restriction endo-nuclease for the site introduced into the pSugi DNA by site-directedmutagenesis (Fig. 2).

The third step of the procedure involved subcloning the ugi genescontaining site-directed mutations from pSugi to the pKK223-3 derivedoverexpression vector pZWtac1, replacing the wild type ugi gene. BothpSugi and pZWtac1 DNA (1 mg) were separately digested with excessHindIII and EcoRI, and the products were resolved by 0.8% agarose(low melting point) gel electrophoresis. Bands corresponding to the ugigene containing DNA fragment (726 bp) from pSugi and the 4.6-kbfragment from pZWtac1 were excised, DNA extracted, and purifiedusing glass milk (27). Samples containing the 726-bp and 4.6-kb EcoRI/HindIII DNA fragments were mixed, and the complementary ends werejoined using T4 DNA ligase. The ligation reaction mixture (10 ml)contained 50 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 10 mM dithiothreitol,1 mM ATP, 25 mg/ml BSA, 0.75 pmol of the 726-bp fragment, and 0.25pmol of the 4.6-kb DNA fragment. Following incubation for ;16 h at16 °C, the ligation mixture was used to transform either E. coli JM105or XL2-Blue competent cells (26). Transformed colonies were isolatedafter growth on LB plates supplemented with 100 mg/ml ampicillin and

grown overnight in 2 ml of 2 3 YT medium containing ampicillin.Plasmid DNA was then isolated using the Wizard Miniprep procedureand analyzed using restriction endonuclease digestion for HindIII,HindIII plus EcoRI, and the appropriate introduced restriction sites(Fig. 2) as described above. In the resulting plasmids (pKugi), themutant ugi gene was expressed under the control of the isopropyl-1-thio-b-D-galactopyranoside-inducible tac promoter.

DNA Sequence Analysis—Double-stranded pKugi DNA containingsite-directed mutations were used as templates for nucleotide sequenc-ing using the dideoxynucleotide chain termination method originallydescribed by Sanger et al. (28). Either primer FP/PKK that was com-plementary to the ugi sense strand at position 2143 to 2122 or primerIRPUGI that hybridized to the antisense strand at position 1379 to1401 was used to initiate DNA polymerase-mediated synthesis. DNAsequencing was conducted using an Applied Biosystems model 373Asequencer by the Center for Gene Research and Biotechnology (OregonState University).

Purification of Uracil-DNA Glycosylase and Inhibitor Protein—Puri-fication of fraction V [leucine-3H]Ung (13.5 cpm/pmol) was carried outsimilarly to that described by Bennett et al. (15). Nonradioactive Ung(fraction V) was purified from E. coli JM105/pSB1051 grown in LBmedium supplemented with 100 mg/ml ampicillin (19). Fraction IV[35S]methionine-labeled Ugi (40 cpm/pmol) and [13C,15N]Ugi were pu-rified as described by Sanderson and Mosbaugh (19) from E. coli JM105/pZWtac1 cultures (1–1.5 liters) grown in M9 minimal medium supple-mented with 100 mg/ml ampicillin, 10 mg/ml thiamine, and either 4.2nmol of [35S]methionine or 0.2% [13C]glucose and 0.17% [15N]ammo-nium chloride, respectively. Nonradioactive Ugi and the various site-directed mutants of Ugi were purified following the same procedure,except that bacterial growth occurred in LB medium containing 100mg/ml ampicillin.

Purification of [3H]UngzUgi Complexes—Wild type Ugi or varioussite-directed mutants of Ugi protein were mixed with [3H]Ung in bufferA (30 mM Tris-HCl (pH 7.4), 1 mM EDTA, 1 mM dithiothreitol, 5% (w/v)glycerol) containing 50 mM NaCl and incubated at 25 °C for 10 min andthen at 4 °C for 20 min. Following complex formation, each sample wasapplied to a DE52 cellulose column equilibrated in buffer A containing50 mM NaCl, washed with equilibration buffer, and step-eluted, asdescribed previously (19). Fractions (1 ml) were collected and sampleswere analyzed for 3H radioactivity. The [3H]UngzUgi complex was de-tected by 18% nondenaturing polyacrylamide gel electrophoresis, andfractions containing complex were pooled and concentrated using aCentriplus-10 (Amicon) concentrator.

Enzyme Assays—Uracil-DNA glycosylase inhibitor activity wasmeasured under previously described conditions (10). When appropri-ate, Ugi was diluted with IDB buffer (50 mM Tris-HCl (pH 8.0), 1 mM

EDTA, 1 mM dithiothreitol, 100 mM NaCl). One unit of uracil-DNAglycosylase inhibitor inactivates 1 unit of uracil-DNA glycosylase in the

FIG. 2. Strategy for oligonucleotide-directed mutagenesis of the ugi gene. The EcoRI/HindIII DNA fragment (726 bp) from pZWtac1contains the ugi structural gene with nucleotide position 11 starting the ATG codon and position 1255 terminating the TAA stop codon. Sevenoligonucleotides were synthesized as shown that are partially complementary to the antisense sequence of the ugi structural gene over the regiondepicted. The beginning and ending nucleotide positions are indicated. Noncomplementary bases were engineered at the positions (*) to introducesite-specific mutations that eliminated Glu or Asp residues by amino acid substitution and introduced endonuclease recognition sites into the ugigene. Endonuclease recognition sites are underlined, and the cleavage sites are indicated (∧∧). Each oligonucleotide was separately hybridized touracil-containing pALU(ss) DNA and served to initiate in vitro primer extension. Oligonucleotidedirected mutagenesis was conducted as describedby Kunkel et al. (23) with modifications indicated under the “Experimental Procedures.” The specific amino acid location and substitution producedby each oligonucleotide is indicated in the inlaid table along with the novel restriction endonuclease sites and cleavage location within the 726-bpDNA fragment.

Uracil-DNA Glycosylase Inhibitor Protein21410

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

standard reaction. Uracil-DNA glycosylase activity was similarly meas-ured except that Ugi was omitted (10). One unit of uracil-DNA glyco-sylase is defined as the amount that releases 1 nmol of uracil/h understandard reaction conditions.

Protein Measurements—Protein concentrations were determined byabsorbance spectroscopy using the molar extinction coefficients e280 nm

5 4.2 3 104 liter/mol cm (Ung) and e280 nm 5 1.2 3 104 liter/mol cm(Ugi). The concentration of [3H]UngzUgi complex was determined fromthe specific radioactivity of [3H]Ung (10).

Electrophoresis—Sodium dodecyl sulfate-polyacrylamide slab gelelectrophoresis was performed essentially as described by Laemmli (29)and modified by Bennett et al. (12).

Nondenaturing polyacrylamide slab gel electrophoresis was per-formed at 4 °C essentially as described by Sanderson and Mosbaugh(19) with resolving gels containing 18% acrylamide and 0.39% N,N9-methylenebis (acrylamide). The gel was immediately stained usingthe rapid stain procedure described by Reisner (30) and modified bySanderson and Mosbaugh (19).

Nondenaturing polyacrylamide tube gel electrophoresis was con-ducted using the same components in the resolving gel (9 3 0.6 cmdiameter) and stacking gel (1 cm) as described above. Following elec-trophoresis the resolving gel was either stained with Coomassie Bril-liant Blue G-250 or sliced horizontally into 3.1-mm sections, placed intoscintillation vials, dehydrated overnight, and solubilized in 500 ml ofH2O2 at 55 °C for 24–36 h as described by Sanderson and Mosbaugh(19). After complete solubilization, 5 ml of Formula 989 fluor wasadded and 3H and 35S radioactivity was measured by scintillationspectrometry.

Nuclear Magnetic Resonance Analysis—All of the wild type and mu-tant Ugi protein samples were concentrated to 7–18 mg/ml and dialyzedagainst NMR buffer containing 25 mM deuterated Tris, 0.2 mM EDTA,0.2 mM EGTA, and 100 mM NaCl, at pH 7.0. NOESY experiments werecarried out using a 150-ms mixing time at 25 °C on a Varian INOVA500 MHz NMR spectrometer. The spectral width in each dimension was7500 Hz. The final pulse in the NOESY experiment was replaced witha watergate gradient pulse sequence and also a gradient pulse beforeeach equilibration delay. The watergate sequence used a 1-ms z-direc-tion gradient followed by a 2.2-ms selective shaped pulse for water anda 180° pulse followed by another 2.2-ms selective shaped pulse andanother 1-ms z-direction gradient. A weak z-direction gradient wasapplied in the first half of t1 and its negative applied in the second halfof t1. Each NOESY experiment had 512 increments in t1 with anacquisition time of 137 ms. Each NOESY spectrum was transferred into1024 3 4096 points and weighted using shifted Gaussians along eachdimension.

Protein Structure Determination of [13C,15N]Ugi Complexed to Ung inSolution—Samples of [13C,15N]Ugi were complexed with unlabeled Ungas described previously (14). All of the NMR spectra were obtained withthe sample at 30 °C using a Varian Unityplus 400 spectrometerequipped with a Nalorac ID400 probe (31–34). Data were acquiredusing States-Haberkorn for the indirectly detected dimension and usingshifted Gaussians in the data processing along each dimension. Theresults of a 60-ms mixing time 15N TOCSY-HSQC spectrum were usedto identify spin systems. The experiment had an acquisition time of0.108 s and there were 128 increments of t1 and 24 increments of t2 foreach data set. The data were linearly predicted to 256 points in t1 and48 points in t2 before being Fourier-transformed into 512 3 128 3 1024points. The spectral widths were 5000 Hz for F1, 1500 Hz for F2, and5000 Hz for F3.

A 15N/1H NOESY-HMQC spectrum was recorded with a mixing timeof 200 ms and 16 transients per increment. There were 1024 points inF3 and an acquisition time of 0.108 s was used. There were 128 incre-ments in t1 and 20 increments in t2. The data were linearly predicted to256 points in t1 and 40 points in t2 before being Fourier-transformedinto 512 3 128 3 1024 points. The 15N NOESY-HMQC data was usedalong with the 15N TOCSY-HSQC to make the chemical shift identifi-cation with the 15N and HN chemical shifts of Ugi. These assignmentsled to 707 NOE constraints for structure determination. The spectralwidths were 5000 Hz for F1, 1500 Hz for F2, and 5000 Hz for F3.

A 13C TOCSY-HSQC-SE spectrum with a 40-ms mixing time wasused to group the spin systems for 13C-labeled atoms. The data werecollected with 16 transients per increment. The acquisition time was0.108 s. There were 256 increments of t1 for each of the complex datasets. The data were linearly predicted to 512 points in t1 before beingFourier-transformed into 512 3 1 3 1024 points. The spectral widthswere 5000 Hz in F1 and 12000 Hz in F2.

A 13C NOESY-HMQC with a mixing time of 150 ms was collectedwith 16 transients per increment. The acquisition time was 0.0832 s.

There were 128 increments in t1 and 48 in t2. The data were linearlypredicted to 228 points in t1 and 96 points in t2 before being Fourier-transformed into 512 3 256 3 1024 points. The spectral widths were5000 Hz in F1, 12000 Hz in F2, and 5000 Hz for F3. These assignmentsled to 325 NOE constraints. A two-dimensional 13C NOESY-HMQCwith a mixing time of 160 ms was collected with 16 transients perincrement. There were 330 increments in t1 with the offset set in themiddle of the aromatic region. Analysis of this NOE spectrum gave 35NOE constraints. The normalized Z4-score analysis of 1HN, 1Ha, 13Cb,and 15N chemical shifts for Ugi produced 106 f and c dihedral con-straints (35).

The constraints were grouped into strong, medium, and weak. Astrong NOE peak was constrained to 1.8 ,r ,4.0 Å, a medium NOEpeak was constrained to 2.1 , r , 4.5 Å, and a weak NOE peak wasconstrained to 2.4 , r , 5.0 Å during simulated annealing and refinedsimulated annealing protein structure determinations protocols. Oncethe secondary structure of Ugi in the complex was determined therewere 24 sets of hydrogen bonds that were used for a total of 48 con-straints. The hydrogen bond constrained the oxygen to amide proton tobe 1.8 , r , 2.5 Å and the oxygen to nitrogen distance to be 2.5 , r ,3.3 Å. The normalized Z4 score analysis of chemical shifts for Ugiproduced 106 f and c dihedral constraints (35).

The simulated annealing and refinements protocols followed thesame procedures as described for the structure of the free uracil-DNAglycosylase inhibitor protein (14) as were previously reported (36). Thesimulated annealing and refinement protocols were run on an IBM 3CTrunning X-PLOR 3.1 (37).

Protein Modeling—Starting with the HSV-1 uracil-DNA glyco-sylasezUgi complex co-crystal coordinates described by Savva and Pearl(6, 17), mutant Ugi forms in complex were generated by exchanging anindividual wild type Ugi amino acid with a mutant residue using theresidue replacement command in INSIGHT (BIOSYM). This is thermo-dynamically reasonable as all the mutant Ugi structures are similar tothat of the free Ugi structure as evidenced by their NOESY spectra.Complete free energy analysis of the transition from the free to thebound form of Ugi is not computationally feasible. Therefore, rigid bodyenergy minimizations were performed to determine a reasonable esti-mate of the DDE involved between mutant forms of Ugi and wild typeUgi when bound to Ung (38). These calculations do not take into accountthe energy differences involved in the structural conformation changesthat occur during binding to Ung. Rigid energy minimizations werethen executed using an IBM 3CT running X-PLOR 3.1 (36). The rigidenergy minimization procedure utilized all residues within 5 Å of thea2-helix and b1-strand of Ugi which included 40 residues of Ugi and 38residues of uracil-DNA glycosylase. A dielectric constant of 6.0 was usedto compensate for not using water with a cut-on distance of 6.0 Å and acutoff distance of 6.5 Å. Energy minimizations were conducted using atwo-step method. The first step involved 1000 iterations of rigid energyminimization with a large van der Waals radius but without consider-ing electrostatic forces. In the second step, 2000 iterations were con-ducted with both electrostatic interactions and normal van der Waalsradius influencing the structure. After the rigid energy minimizationsconverged, the minimized structure of the uracil-DNA glycosylasezUgicomplex emerged. Each individual unbound wild type and mutant Ugistructure was similarly generated. Interaction energies were calculatedby combining the van der Waals, electrostatic, and hydrogen bondenergies of the enzyme-inhibitor complex and unbound Ugi. Changes ininteraction energies, DEint, are defined as the difference in the interac-tion energies of the uracil-DNA glycosylasezUgi wild type and mutantcomplex. The differences in the change in the interactive energiesDDEint are defined by subtracting the difference of the DEint of the wildtype and mutant Ugi from the DEint of the mutant Ugi-containingcomplex.

RESULTS

Site-directed Mutagenesis of the Uracil-DNA Glycosylase In-hibitor Gene—To investigate the role of specific negativelycharged amino acid residues in the Ung/Ugi interaction, site-directed mutagenesis producing single amino acid substitu-tions was performed on the ugi gene. The specific sites andsubstitutions selected were based on knowledge of the 1.9-Åcrystal structure of Ugi complexed with human uracil-DNAglycosylase (16). Significant similarity exists between the hu-man and E. coli enzyme around the proposed sites of Ung/Ugiinteraction (Table I). Oligonucleotides were synthesized thatintroduced a codon change at seven Glu or Asp sites and a new

Uracil-DNA Glycosylase Inhibitor Protein 21411

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

restriction endonuclease cleavage site into the ugi gene asindicated in Fig. 2. To overproduce the mutant Ugi proteins,the EcoRI/HindIII DNA fragment containing the ugi structuralgene was subcloned into the overexpression vector pKK223–3producing a set of pKugi plasmids. Two methods were used toverify that the engineered mutations had been introduced intoeach pKugi DNA. First, restriction endonuclease digestionswere conducted to establish the presence of the newly intro-duced recognition site within the EcoRI/HindIII DNA (726 bp)fragment. Second, dideoxynucleotide chain termination DNAsequencing of double-stranded pKugi DNAs was performed(data not shown). For all mutants, the entire ugi gene wasbidirectionally sequenced and the results confirmed the de-signed nucleotide changes, exclusively.

Purification and Specific Activity of the Mutant Ugi Pro-teins—To facilitate characterization of the inhibitor activityexhibited by wild type Ugi and seven mutant Ugi proteins, eachprotein was overproduced using the appropriate pKugi vectorand purified according to Sanderson and Mosbaugh (19). Thepurity of Ugi from the final purification step (fraction IV) wasanalyzed using 20% SDS-polyacrylamide gel electrophoresis(Fig. 3A). As previously observed the electrophoretic mobility ofwild type Ugi was greater than that predicted for a 9474-daltonprotein (10). Each mutant Ugi protein migrated with a uniqueslower mobility with respect to wild type Ugi, consistent withthe elimination of a negatively charged residue by site-directedmutagenesis. However, these observations also imply that themutant Ugi proteins exhibit different propensities to bind SDSor adopt to different protein conformations during electro-phoresis, since each mutant protein carries the same charge.The specific activity of each purified Ugi protein was deter-mined under standard conditions (Fig. 3B). Ugi E20I was es-sentially void of inhibitory activity, displaying ;1% of the wildtype specific activity, whereas Ugi E78V was virtually unaf-fected, displaying ;105% activity. The four mutations in Glu

residues located in the a2-helix (E27A, E28L, E30L, and E31L)showed progressively decreased levels of activity with 95, 88,70, and 53% of control activity, respectively. Significant inac-tivation was also observed with the Ugi D61G protein whichshowed ;25% of wild type Ugi activity.

Ability of Mutant Ugi Proteins to Form a Complex withUng—To determine whether the mutant Ugi proteins wereable to form a UngzUgi complex, a 3-fold molar excess of Ungwas incubated with each Ugi protein under standard bindingconditions. The resultant UngzUgi complexes were resolvedfrom the component proteins by nondenaturing polyacrylamidegel electrophoresis (Fig. 4). As controls, free Ung, wild type Ugi,and a 3:1 ratio of UngzUgi were analyzed for comparison withmutant forms of free Ugi and UngzUgi complexes. Each mutantUgi protein migrated as a single band with a mobility similar tobut slightly slower than that of wild type Ugi. In each case, themutant Ugi proteins formed a UgizUng complex that also mi-grated slightly slower than the wild type complex. With theexception of Ugi E20I, it appeared that each mutant Ugi pro-tein formed a stable and complete complex with Ung since nofree Ugi was detected. In contrast, the appearance of some freeUgi E20I, less UngzUgi E20I complex, and a smear of proteinbetween the Ung and UngzUgi E20I complex bands suggestedthat Ugi E20I formed an unstable complex (Fig. 4, lane 5).

Relative Ability of Mutant and Wild Type Ugi Proteins toComplex with Ung—Competition experiments were conductedto determine the relative ability of each mutant Ugi protein toform a complex with Ung in the presence of wild type Ugi. Ungwas incubated with a 2-fold molar excess of a mixture of Ugiand/or Ugi E27A at various ratios. The proteins were thenresolved by nondenaturing polyacrylamide gel electrophoresisand detected by Coomassie Blue staining (Fig. 5A, lanes 1-6).Under these conditions, the UngzUgi E27A complexes wereonly partially resolved, whereas free Ugi and Ugi E27A wereseparated as independent bands. To quantitatively analyze the

TABLE IInteractions of Ugi amino acid residues with various uracil-DNA glycosylases

Ugi, aminoacid

Uracil-DNA glycosylasea

HSV-1b HumanbE. coli

Amino AcidcInteraction Amino acid Interaction Amino acid

b1-StrandGlu-20 H2 Ser-202 H2 Ser-169 Ser-88

WH1 Ser-302 Ser-270 Ser-189

a2-HelixGlu-27Glu-28 H1 Thr-280 H2 Ser-247 Ser-166

His-300 His-268 His-187WH2

Ser-305 Ser-273 Ser-192Glu-30Glu-31 Lys-306 SB Arg-276 Arg-195

Loop regionsAsp-61 H1 Arg-252 SB Lys-218 Ala-137Glu-78 Pro-303 Pro-271 Pro-190

Leu-304 Leu-272 Leu-191Ser-305 WHN Ser-273 Ser-192Lys-306 Val-274 Ala-193Val-307 Tyr-275 His-194

a Amino acid sequence alignment and position numbers of uracil-DNA glycosylase correspond to those described by Caradonna et al. (9). Theoriginal reference and GenBank accession number for each enzyme corresponding to HSV-1, herpes simplex virus type 1 (40) is X14112; UDG1,human (39) is X15653; and E. coli (41) is J03725.

b Interactions between PBS1 Ugi and HSV-1 uracil-DNA glycosylase have been described based on a 2.7-Å crystal structure of the complex (17),that involving PBS2 Ugi and human UDG1 were described from a 1.9-Å crystal structure (16).

c E. coli Ung amino acid residues corresponding to the aligned HSV-1 and human uracil-DNA glycosylase polypeptides are indicated (9).d Chemical interactions listed include the following: H1, a hydrogen bond between carboxylate and Thr backbone amide or Arg side chains; H2,

a pair of hydrogen bonds between the carboxylate and Ser backbone amide and side chain Og; WH1, water-mediated hydrogen bond between thecarboxylate and Ser-Og; WH2, water-mediated hydrogen bonds with His backbone amide and Ser-Og; WHN, hydrogen-bonded network withordered solvent molecules and backbone atoms of UDG1 residues; and SB, salt bridge.

Uracil-DNA Glycosylase Inhibitor Protein21412

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

ability of mutant Ugi proteins to compete with the wild typeinhibitor protein, similar experiments were conducted aftermixing each mutant Ugi with wild type [35S]Ugi and incubat-ing the mixtures with Ung. Following electrophoresis, 35S ra-dioactivity was detected in two bands that corresponded to[35S]Ugi free and in complex. Thus, the amount of [35S]Ugidetected in the complex band reflected the competitive abilityof the mutant inhibitor protein to stably associate with Ungwhile in the presence of wild type Ugi. As a control, variousratios of [35S]Ugi to Ugi (both wild type proteins) were mixedand analyzed by electrophoresis (Fig. 5B, black bars). Theamount of [35S]Ugi detected in the complex approximatelyequaled the amount expected based on equal competition be-tween the two inhibitor proteins and on the various ratiosbetween [35S]Ugi and Ugi. Thus, this result confirms the iden-tical nature of the two wild type inhibitor protein preparationsand validates the experimental design. After examining each ofthe mutant Ugi proteins for their ability to compete with[35S]Ugi for complex formation, it was observed that two mu-tant Ugi proteins (Ugi E20I and Ugi E28L) consistently showedan increased amount of [35S]Ugi in complex with Ung over thatpredicted by equal competition. Hence, Ugi E20I and Ugi E28Ldemonstrated a decreased ability to form complex with Ung inthe presence of wild type [35S]Ugi. The results also suggestedthat Ugi E20I competes very poorly with wild type Ugi sincethe maximum amount of [35S]Ugi based on the molar amount ofUng was found in complex for all ratios utilizing Ugi E20I.

Reversibility of the UngzUgi Complex with Various MutantUgi Proteins—To characterize further the nature of theUngzUgi complexes containing mutant Ugi proteins, we as-sessed the ability of wild type Ugi to displace mutant Ugi froma preformed complex. [3H]UngzUgi complexes were formed byindividually incubating either wild type or mutant Ugi with a3-fold molar excess of [3H]Ung and the complex species iso-lated by DEAE-cellulose chromatography. Purification of[3H]UngzUgi E27A is illustrated in Fig. 6A. Analysis of frac-tions across the peak by nondenaturing polyacrylamide gelelectrophoresis verified that .95% of [3H]Ung formed complexand that no detectable free [3H]Ung or Ugi was observed inthese fractions (Fig. 6A, inset). Stable preformed [3H]UngzUgicomplexes were isolated using this procedure for each mutantUgi protein (Fig. 6B) except Ugi E20I, which was unable toform a stable complex that could be purified (data not shown).

Competition experiments were conducted to determine ifwild type [35S]Ugi could exchange with mutant Ugi containedin the preformed [3H]UngzUgi complexes. Each complex prep-aration was incubated with a 10-fold molar excess of [35S]Ugi,and nondenaturing polyacrylamide gel electrophoresis was per-formed, as described above. If a mutant [3H]UngzUgi associa-tion was reversible, then wild type [35S]Ugi would exchangewith the mutant Ugi in complex; the amount of [35S]Ugi in thecomplex would reflect the amount of mutant Ugi exchanged. Asa control, wild type [35S]Ugi was incubated with preformed[3H]UngzUgi (wild type) complex; 6.8% of the [3H]Ung wasfound associated with [35S]Ugi in complex (Fig. 6C). This valuerepresents a background level when comparing results with themutant preformed complexes. Of the six mutant Ugi containedin preformed complexes, only Ugi E28L was significantly dis-placed by wild type [35S]Ugi (Fig. 6C). In this case, 50% of UgiE28L was replaced by [35S]Ugi, demonstrating that theUngzUgi E28L complex was reversible. The other preformedcomplexes containing mutant Ugi proteins showed slightlyabove background levels of wild type [35S]Ugi exchange (1.9%E27A, 0.5% E30L, 3.4% E31L, 1.4% D61G, and 1.8% E78V).Thus, in contrast to UngzUgi E28L the other mutant complexes

FIG. 3. Purity and specific activity of site-directed mutant Ugiproteins. A, SDS-polyacrylamide gel electrophoresis of purified mu-tant Ugi preparations. Nine samples (50 ml) each containing 3.6 mg offraction IV wild type or mutant Ugi protein were applied to a 20%polyacrylamide gel containing 0.1% SDS, and electrophoresis was con-ducted as described under “Experimental Procedures.” Protein bandswere visualized after staining with Coomassie Brilliant Blue G-250(Bio-Rad). The molecular weight standards for BSA, ovalbumin, glyc-eraldehyde-3-phosphate dehydrogenase, carbonic anhydrase, trypsino-gen, and trypsin inhibitor are indicated by arrows from top to bottom,respectively. The location of the tracking dye (TD) is indicated by anarrow. Lanes 1 and 9 contain wild type Ugi; lane 2, Ugi E20I; lane 3, UgiE27A; lane 4, Ugi E28L; lane 5, Ugi E30L; lane 6, Ugi E31L; lane 7, UgiD61G; and lane 8, Ugi E78V. B, determination of wild type and mutantUgi specific activity. Standard uracil-DNA glycosylase inhibitor assayswere performed on fraction IV Ugi preparations as described under“Experimental Procedures.” Relative specific activity was determinedby comparing each mutant Ugi activity to the control (wild type Ugi)that equaled 1.1 3 106 units/mg.

FIG. 4. Ability of mutant Ugi proteins to form complex with E.coli Ung. Wild type or various mutant Ugi proteins (96 pmol) weremixed with or without 288 pmol of Ung and the sample (45 ml) incu-bated at 25 °C for 10 min followed by 20 min at 4 °C to form complex.Each sample was then mixed with loading buffer, applied to a nonde-naturing 18% polyacrylamide gel, electrophoresis was carried out, andthe gel was stained with Coomassie Brilliant Blue G-250 as describedunder “Experimental Procedures.” Lanes 1 and 20 contained 288 pmolof Ung; lanes 2 and 19 contained 96 pmol of wild type Ugi; lanes 3 and18, wild type Ugi plus Ung; lanes 4, 6, 8, 10, 12, 14, and 16, variousmutant Ugi proteins; lanes 5, 7, 9, 11, 13, 15, and 17, contained mutantUgi plus Ung as indicated. The arrows indicate the location of Ung,UngzUgi, Ugi, and the tracking dye (TD).

Uracil-DNA Glycosylase Inhibitor Protein 21413

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

were irreversible, as is the wild type UngzUgi complex.Rate and Extent of Wild Type Ugi Exchange with the UngzUgi

E28L Complex—The reversible nature of the [3H]UngzUgiE28L complex was exploited to determine the rate of exchangewith wild type [35S]Ugi. Competition reaction mixtures con-taining the preformed [3H]UngzUgi E28L complex (0.6 nmol)and [35S]Ugi (6 nmol) were incubated at 25 °C for various timesto allow exchange before determining the amount of [35S]Ugithat resided with [3H]Ung in complex (Fig. 7, closed circles).The results indicated that rapid exchange occurred since 28%of the complex contained [35S]Ugi without incubation. Theamount of exchange increased with incubation time andreached a plateau after 120 min with ;75% of the Ugi E28Lexchanging with wild type [35S]Ugi. In contrast, the[3H]UngzUgi (wild type) complex showed no significant ex-change with [35S]Ugi after 240 min confirming the irreversiblenature of this association (Fig. 7, open circles). While the UgiE28L mutant was capable of forming a stable complex withUng, an irreversible complex was not achieved.

Solution State Structure of Mutant and Wild Type [15N]UgiProteins—NMR structural determinations were made to ana-lyze and compare the polypeptide structures of the wild typeand seven mutant Ugi proteins. Each Ugi protein showed one-dimensional proton spectra consistent with a well ordered andfolded structure (data not shown). The one-dimensional spectra

obtained on the samples in 2H2O also indicated that all eightUgi proteins exhibited about the same number of slowly ex-changing amide protons. In addition, the distribution of the

FIG. 5. Ability of mutant Ugi proteins to compete with wildtype [35S]Ugi for complex formation with E. coli Ung. A, sixcompetition reaction mixtures (70 ml) containing 314 pmol of Ung and628 pmol (total) of Ugi and/or Ugi E27A were prepared at molar ratiosof 100:0, 80:20, 60:40, 40:60, 20:80, and 0:100 ([35S]Ugi to Ugi E27A) forlanes 1–6, respectively. After the addition of Ung, the samples weremixed, incubated under standard complexing conditions, tracking dyewas added, and the samples were loaded onto nondenaturing 18%polyacrylamide tube gels as described under “Experimental Proce-dures.” Following electrophoresis at 4 °C, each gel was stained withCoomassie Brilliant Blue G-250. The direction of migration was fromtop to bottom, and the tracking dye (TD) is located by an arrow. Thelocation of the UngzUgi and UngzUgi E27A complex, free Ugi E27A, andUgi are indicated by arrows. B, eight sets of gels were prepared asdescribed in A except that [35S]Ugi (wild type) was mixed at variousmolar ratios, excluding the 0:100 ([35S]Ugi to Ugi) sample, with eitherwild type Ugi (control) or various mutant Ugi proteins, as indicated.After electrophoresis, each gel was horizontally sliced (3.1 mm), dried,solubilized, and analyzed for 35S radioactivity. The average amount of[35S]Ugi (wild type) that formed complex with Ung in duplicate compe-tition reactions was determined from the amount of 35S radioactivitydetected in the UngzUgi complex band. The expected amount of [35S]Ugito form complex when competing with wild type Ugi (control) is indi-cated (●).

FIG. 6. Ability of free [35S]Ugi to exchange with various mu-tant Ugi proteins in a preformed [3H]UngzUgi complex. A, forma-tion and purification of the [3H]UngzUgi E27A complex. A sample (300ml) containing 12 nmol of [3H]Ung and 36 nmol of Ugi E27A wasincubated under standard complexing conditions and applied to a DE52cellulose column (0.8 cm2 3 3 cm). The column was eluted with bufferA containing 150 mM and 250 mM NaCl (arrows) as described under“Experimental Procedures.” Fractions (1 ml) were collected, and sam-ples (100 m1) were analyzed for 3H radioactivity (●). Samples (40 ml)from fraction numbers 26–40 were analyzed on a nondenaturing 18%polyacrylamide slab gel (lanes 5–17, respectively). Electrophoresis wascarried out at 4 °C, and protein bands were visualized after stainingwith Coomassie Brilliant Blue G-250 as shown in the inset. Lane 1contained 260 pmol of [3H]Ung; lane 2, 640 pmol of Ugi E27A; lane 3,136 pmol of [3H]UngzUgi; and lane 4 contained a sample (4 ml) of the[3H]Ung/Ugi E27A mixture that was loaded onto the DE52 column. Thelocation of the tracking dye (TD) is indicated by an arrow. Peak frac-tions containing the [3H]UngzUgi E27A complex were pooled (bar) andconcentrated using a Centriplus-10 (Amicon) concentrator. B, purity ofthe various [3H]UngzUgi complexes. Samples (40 ml) containing 100pmol of [3H]Ung (lane E), [3H]UngzUgi (lanes 1 and 8), [3H]UngzUgiE27A (lane 2), [3H]UngzUgi E28L (lane 3), [3H]UngzUgi E30L (lane 4),[3H]UngzUgi E31L (lane 5), [3H]UngzUgi D61G (lane 6), and[3H]UngzUgi E78V (lane 7) were applied to a nondenaturing 18% poly-acrylamide slab gel, and electrophoresis was conducted at 4 °C. Thelocation of free Ung (arrow) and UngzUgi complexes (arrow) are indi-cated. C, ability of free [35S]Ugi to exchange with various mutant Ugiproteins in a preformed complex. Seven competition reactions (150 ml)containing 0.6 nmol of [3H]UngzUgi (wild type or mutant) and 6.0 nmolof [35S]Ugi (wild type) were incubated under standard complexing con-ditions and applied to a nondenaturing 18% polyacrylamide tube gel;electrophoresis was conducted,and gels were horizontally sliced (;3mm), dried, solubilized as indicated under “Experimental Procedures,”and analyzed for 3H and 35S radioactivity using a double isotope detec-tion technique. The amount of [3H]Ung (M) and [35S]Ugi (f) found inthe UngzUgi complex band is plotted for the seven competition reactionsas indicated. The background level of 35S radioactivity detected in thewild type [3H]Ungz[35S]Ugi complex was not subtracted from the mu-tant complexes. Each reaction was carried out in duplicate and repre-sents the average values.

Uracil-DNA Glycosylase Inhibitor Protein21414

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

amide proton chemical shifts was consistent with each mutantUgi containing a high percentage of b-structure, as is the casefor wild type Ugi (13, 14). Structural determinations were alsomade by comparing secondary structural NOE peaks from am-ide to amide and amide to a-NOESY spectra. The amide toamide NOESY spectra for wild type and mutant Ugi forms areshown in Fig. 8. Each mutant Ugi protein was found to containtwo a-helices and five b-strands identical to the secondarystructural elements as exhibited by the unbound wild type Ugiprotein. The NOESY spectra of each mutant Ugi was comparedwith that of the assigned wild type spectrum. Detailed exami-nation showed that Ugi E20I, D61G, and E78V have structuresthat are very similar to wild type Ugi. However, the chemicalshifts of many of the cross-peaks in the Ugi E20I spectrum aredistinct from those of the wild type protein. Analysis of theNOESY spectra for Ugi E27A, E28L, E30L, and E31L likewiseindicated close structural similarity to wild type Ugi, with theexception of the a2-helix length. Ugi E27A and E28L containedan a2-helix that was shorter at the N-terminal end, whereasUgi E30L and E31L exhibited a shorter a2-helix at the C-terminal end.

Solution State Structure of [13C,15N]Ugi Complexed toUng—To determine the structure of Ugi bound to Ung, a sam-ple of [13C,15N]Ugi (820 nmol) was combined with an excess ofunlabeled Ung (1220 nmol), and the Ungz[13C,15N]Ugi complex(1.27 mM) was prepared as described previously (14). NMRstructural determinations were made using 15N-TOCSY-HSQC, 15N/1H-NOESY-HMQC, 13C-TOCSY-HSQC, and 13C-TOCSY-HSQC-SE spectra, and the solution state structure of[13C,15N]Ugi complexed to Ung is shown in Fig. 9. A compari-son of free [15N]Ugi (Fig. 9A) to [13C,15N]Ugi bound to Ung (Fig.9B) indicates that significant structural change of Ugi occurredas a consequence of complex formation. The electrostatic sur-faces of the free (Fig. 9, C and E) and complexed Ugi protein(Fig. 9, D and E) were evaluated using the GRASP program

(38).

DISCUSSION

We have used site-directed mutagenesis to assess the role ofspecific negatively charged amino acids in Ugi activity. Threestructural domains of Ugi were targeted that included thebl-strand (E20I), the a2-helix (E27A, E28L, E30L, and E31L),and the loop regions joining the anti-parallel b-strands (D61G,E78V). To gain information about the structural changes in-duced by the specific amino acid replacements, NMR spectralanalysis was performed for each Ugi protein. The one-dimen-sional proton spectra of the wild type and mutant Ugi proteinsappeared to be quite similar indicating that each Ugi proteinfolded in much the same manner. The results of binding exper-iments indicated that each mutant Ugi protein remained capa-ble of associating with Ung and forming a UngzUgi complex.However, complex stability and reversibility was found to bealtered by some amino acid substitutions. These results sug-gest that none of the individual Ugi amino acids examined playan essential role in mediating Ung/Ugi binding. Rather, thenegatively charged residues act collectively to facilitate stablecomplex formation.

The specific chemical interactions that stabilize the UngzUgicomplex can be inferred from those in the x-ray crystallo-graphic structures identified of HSV-1 (17) and human (16)uracil-DNA glycosylasezUgi complexes. Such a comparison isjustified since E. coli Ung shares extensive amino acid homol-ogy with its HSV-1 and human counterparts (39), and bothco-crystal structures show significant structural similarity (16,17). The structure of Ugi in complex with Ung has been deter-mined by conventional solution state methods and found to beessentially the same as the crystal structure (16, 17). As indi-cated in Table I, the locations and types of interactions linkingUgi residues with either HSV-1 or human uracil-DNA glycosy-lase were found to be highly conserved. Amino acid sequencealignment of E. coli Ung to both the HSV-1 and human enzymerevealed identical or conservative substitutions at the sites ofUgi interaction. The ability of Ugi to perform DNA mimicry hasapparently capitalized on the conservation of Ung residueslocated in the highly conserved DNA-binding pocket (16, 18,21). This striking amino acid correspondence suggests thatsimilar interactions most likely mediate the Ugi associationwith all three uracil-DNA glycosylases examined here and pos-sibly others.

The Ugi E20I protein, although capable of forming a UngzUgicomplex, did not completely block Ung activity, presumablydue to an inability to form a stable complex with Ung. Theinstability of this association was evident from the dissociationdetected during nondenaturing polyacrylamide gel electro-phoresis, the inability to isolate a UngzUgi E20I complex byanion exchange chromatography, and the ineffectiveness of UgiE20I to compete with wild type Ugi for Ung binding. Theposition of Glu-20 is apparently stabilized by a pair of hydrogenbonds between the carboxylate side chain of the conservedSer-88 backbone amide and Og of E. coli Ung (Table I). Inaddition, a water-mediated hydrogen bond may also form be-tween Ugi Glu-20 and Ung Ser-189, as has been described forthe complex involving the HSV-1 enzyme (17). The loss of UgiE20I activity may be explained by a weakening of these inter-actions due to charge neutralization or peptide conformationalchange surrounding this key residue. Protein modeling indi-cated that the van der Waals energy dropped considerably (217kcal/mol); in this case, more than enough to stop the interac-tion. Taken together the results suggest that Ugi E20I forms afrail unlocked complex that fails to prevent Ung associationwith uracil-DNA.

The four mutations (E27A, E28L, E30L, and E31L) created

FIG. 7. Time course of the [35S]Ugi-induced displacement re-action of Ugi E28L from the preformed [3H]UngzUgi E28L com-plex. Six competition reaction mixtures (150 ml) containing 0.6 nmol of[3H]UngzUgi E28L and 6 nmol of [35S]Ugi were prepared in duplicate asdescribed in Fig. 6. Two control reactions containing 0.54 nmol of[3H]UngzUgi (wild type) and 5.4 nmol of [35S]Ugi were prepared. Eachreaction was incubated at 25 °C for the times indicated and then loadedonto nondenaturing 18% polyacrylamide tube gels. The unincubatedsample (0 min) was mixed on ice and electrophoresis initiated as rapidlyas possible. Following electrophoresis at 4 °C, the gels were sliced andprocessed as described under “Experimental Procedures.” The 3H and35S radioactivities were measured in each gel slice, and the percentageof the once [3H]UngzUgi E28L (●) or [3H]UngzUgi (E) complex thatcontained [35S]Ugi was determined by dividing the amount (nmol) of[35S]Ugi by the amount (nmol) of [3H]Ung found in the UngzUgi complexband and multiplying by 100.

Uracil-DNA Glycosylase Inhibitor Protein 21415

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

within the a2-helix had quite different effects on Ugi activity.While Ugi E27A retained near full inhibitor activity, the otherthree mutations caused a progressive reduction of Ugi-specificactivity (E28L . E30L . E31L) with Ugi E31L maintaining;50% wild type activity. The influence of these mutations was

particularly interesting since a major structural difference be-tween the free and complexed forms of Ugi involves the orien-tation of the a2-helix (Fig. 9, A and B). Furthermore, x-raycrystallographic studies have indicated that when in complexthe a2-helix and b1-sheet resides over the DNA-binding groove

FIG. 8. The NOESY spectra of wildtype Ugi and seven mutant Ugi pro-teins. The two-dimensional 500-MHz,150-ms mixing time NOESY spectra ofwild type Ugi (A) and Ugi E20I (B), E27A(C), E28L (D), E30L (E), E31L (F), D61G(G), and E78V (H) are shown. The regionshown contains signals primarily fromamide, aromatic, and amino protons andthe NOE cross-peaks between them.

Uracil-DNA Glycosylase Inhibitor Protein21416

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

and provides the majority of contacts between the enzyme andinhibitor (16, 17). Therefore, it was not surprising that UgiE27A activity was unaffected, since Glu-27 has not been impli-

cated in complex interaction (Table I). The results obtained forUgi E28L and E31L support the inference drawn from chemicalmodification that Glu-28 and/or Glu-31 play an important role

FIG. 9. Tertiary structure of free Ugi and Ugi bound to Ung. The tertiary structure of the solution state [15N]Ugi (A) was previouslydetermined by Beger et al. (14). The tertiary structure of [13C,15N]Ugi bound to E. coli Ung was determined by solution state multidimensionalNMR techniques as described under “Experimental Procedures.” Several secondary structural elements are highlighted in both free Ugi (A) andin complex (B) structures as follows: a1-helix (light blue); a2-helix (dark blue); bl-strand (red); b2-b5-strands (salmon); and the loop between b3-and b4-strands (yellow). The location of the N-terminal (N), C-terminal (C), Glu-28 (28), and Glu-3l (31) residues are also indicated. Theelectrostatic surfaces of the free (C) and complexed (D) forms of Ugi were generated using the program GRASP (38) as described previously (14).Structures A and B of the free and bound Ugi correspond to the same view as indicated in C and D, respectively. The bottom panel depicts Ugirotated 180°. The electrostatic potentials were calculated with a dielectic constant of 6.0 for the protein and 80.0 for the solvent. The ionic strengthof the solution was set to 0. Only the charges of the side chains of Lys, Asp, Asn, Glu, and Gln residues were used. The electrostatic potential cutoffwas set to 6.6 kcal/mol, and the regions with a negative potential of this magnitude are shown in red, and the regions with a positive potential ofthis magnitude are shown in blue.

Uracil-DNA Glycosylase Inhibitor Protein 21417

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

in promoting stable UngzUgi complex formation (19). Theunique ability of Ugi E28L to form a stable but reversiblecomplex when challenged with wild type Ugi indicates that thisresidue plays a critical role in forming the locked complex. LikeGlu-20, Glu-28 appears to form hydrogen bonds with a con-served Ser (Ser-166 of E. coli Ung) amide and the side chain Og(Table I). In addition, Glu-28 also forms water-mediated hydro-gen bonds to a universally conserved active site His backboneamide and Ser Og atom (His-187 and Ser-192 of E. coli Ung).We speculate that these contacts are responsible, at least inpart, for creating the irreversible nature of the UngzUgi com-plex. Additionally, the observation that Ugi E31L, like wildtype Ugi, formed an essentially irreversible complex with Ungargues that Glu-31 does not play a major role in the lockingreaction.

The two mutations in the loop regions connecting the con-secutive b-strands of Ugi provided distinctly different results.Ugi D61G caused ;75% reduction of activity, whereas UgiE78V showed a specific activity equivalent to wild type Ugi.The inability of the E78V mutation to affect activity may beexplained since Glu-78 resides within the electrostatic knobregion of Ugi that contains seven Glu or Asp residues (14). Theresults suggest that neutralizing the negative charge of Glu-78may have little effect on the overall inhibitory action due to therelatively small individual contribution of Glu-78. The involve-ments of Glu-61 in Ugi/Ung binding remains to be determined;however, the reduced activity of Ugi D61G was not attributedto a defective locking reaction.

Several lines of evidence have led to a proposal that free Ugiundergoes a conformational change during formation of theUngzUgi complex (14, 16, 17, 19–21). A direct demonstration ofthis change is evident by comparing the NMR solution struc-ture of free [15N]Ugi with that of [13C,15N]Ugi complexed to E.coli Ung (Fig. 9). Clearly, the tertiary structure of the free andbound Ugi are quite different; however, both forms of the pro-tein contain similar secondary structural elements (i.e. twoa-helices and five b-strands). The b2-b3-b4-b5 portion of theanti-parallel b-sheet remains generally unchanged in the twostructures and provides a focal point for comparison. The majortransition between these structures involves a collapse of thepolypeptide segments containing the a1- and a2-helix. In theunbound state, both helices extend away from the core of Ugi(14). We speculate that this Y-shaped structure may arise fromthe negative charge repulsion between the negative electro-

static knob and both the negatively charged a1- and a2-helices.Upon binding to Ung, the flexible arms containing the a1- anda2-helix reorient to allow the positioning of b1 and a2 over thepositively charged DNA-binding pocket of uracil-DNA glycosy-lase (16, 17). As a consequence, several other structuralchanges occur as follows: (i) the b1-strand becomes twisted; (ii)the a1- and a2-helix move toward the core of Ugi; and (iii) theloop between the b3- and b4-strands becomes slightly reori-ented. The orientation and negative charge of Glu-28 in theDNA-binding pocket mediates the formation of the locked com-plex and excludes DNA. The involvement of a Ugi structuralchange may explain the specificity exhibited by this inhibitorprotein toward uracil-DNA glycosylases acting through amechanism involving DNA mimicry.

The structural changes that occur during complex formationhave a pronounced effect on the electrostatics of Ugi (Fig. 9).The primary changes appear to result from the positions of thea1- and a2-helices relative to the rest of the protein. Thea1-helix is positioned behind the b-sheet of the complex struc-ture shown, and the a2-helix is positioned in front. There aresmaller changes of the b-strands. The helices appear to haverelatively few interactions with the rest of the protein in thefree form, and there may be no particularly large barriersbetween the free and bound conformations. The positioning inthe bound state of the a1-helix effectively covers the electro-static potential of the knob region that is exposed in the freeprotein. The position of the a2-helix in combination with themodest rearrangements of the b-strands gives rise to a largenegative electrostatic potential on the face that forms most ofthe contacts in the Ung complex. This suggests that electro-static interactions will play a considerable role in the complexand that the a2-helix appears to be involved in theseinteractions.

Molecular modeling studies were conducted using the co-crystal coordinates of the HSV-1 uracil-DNA glycosylasezUgicomplex and variations of the free and bound Ugi structure.The models only allowed differences in the position of theamino acid side chains corresponding to the mutant Ugi pro-teins. Information concerning the contributions to the energiesof complex stability was assessed for the mutations in theb1-strand and a2-helix. The modeling indicated that the com-plex containing Ugi E20I has by far the highest energy, con-sistent with the low activity of this protein. The modelingsuggested that Ugi E20I forms the same unbound protein

FIG. 10. Molecular modeling of theUgi E28L and E31L mutant proteinscomplexed with uracil-DNA glycosy-lase. The Ugi protein on the left repre-sents a partial polypeptide structure offree Ugi as previously determined byBeger et al. (14). The a2-helix and loopregion joining the b3- and b4-strands areshown with the location of Glu-28 andGlu-31 indicated. A portion of the HSV-1uracil-DNA glycosylasezUgi complex de-rived from the co-crystal coordinates de-scribed by Savva and Pearl (6, 17) is illus-trated on the right. Modeled structures ofUgi E28L and E31L (yellow) complexedwith HSV-1 uracil-DNA glycosylase (lightblue) were obtained as described under“Experimental Procedures” and areshown in the top and bottom middle, re-spectively. The changes in the location ofthe enzyme and inhibitor side chains inthe complexes containing Ugi E28L andE31L are depicted (red).

Uracil-DNA Glycosylase Inhibitor Protein21418

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

structure as wild type Ugi but that there are very unfavorable(;15 kcal) van der Waals interactions in the complex. In con-trast, Ugi E27A and E30L were found to have energies that areessentially identical to that of wild type Ugi in the complex.This is consistent with neither Glu-27 nor Glu-30 residuesparticipating in an interaction with the enzyme (Table I) andboth Ugi E27A and E30L showing only partially reduced activ-ity. Models of Ugi E28L and E31L were examined in an at-tempt to explain the reason that Ugi E28L was the only mutantprotein capable of forming a stable but reversible complex. Asshown in Fig. 10, the positions of the Leu side chains in bothmutant Ugi polypeptides were quite similar to the wild typeGlu, although they are shorter in length. Since both mutantproteins are structurally very similar to wild type, we infer thatthe absence of the carboxyl group precipitates the change inproperties of each mutant. The Leu side chain should not becapable of mediating the hydrogen bond interactions that sta-bilize the enzyme-inhibitor complex (Table I). Under this con-dition Ugi E28L appears capable of conducting the dockingreaction but not the locking reaction. Analysis of the energyterms showed that the electrostatics of the complexes with UgiE28L and E31L are ;5 and ;2 kcal, respectively, less favor-able than that of complex containing wild type Ugi. Both mu-tant proteins showed ;3 kcal less stability than wild type Ugiin complex based on van der Waals forces. Hence, Ugi E28Ldiffered from the other mutations in the a2-helix in that it notonly had the highest energy but was unfavorable in both elec-trostatic and van der Waals energy relative to the wild type Ugiin complex. This suggests that the locking reaction may involveboth the electrostatic potential and the hydrogen bond interac-tions of Glu-28.

This study has demonstrated that Ugi exists in three differ-ent conformational states (free, unlocked, and locked Ugi) dur-ing the binding reaction with Ung. The involvement of a sig-nificant structural transformation and role of Glu-20 andGlu-28 in mediating the locking reaction has been demon-strated. However, several issues remain to be investigatedconcerning the structure and function of Ugi during Ung com-plex formation. First, do individual amino acid residues play anessential role in the docking reaction? Second, what is the effectof various mutations on the kinetics of the Ung-Ugi interac-tion? Third, what is the structure of E. coli Ung when com-plexed with Ugi? Additional protein structural and biochemicalanalysis will be required to elucidate these important issues.

Acknowledgments—We thank Dr. Reg McParland and Anne-MarieGirard for conducting the nucleic acid sequencing and Barbara Robbinsfor conducting the oligonucleotide synthesis.

REFERENCES

1. Takahashi, I., and Marmur, J. (1963) Nature 197, 794–7952. Mosbaugh, D. W., and Bennett, S. E. (1994) Prog. Nucleic Acid Res. 48,

315–3703. Price, A. R., and Frato, J. (1975) J. Biol. Chem. 250, 8804–88114. Friedberg, E. C., Ganesan, A. K., and Minton, K. (1975) J. Virol. 16, 315–3215. Wang, Z., and Mosbaugh, D. W. (1989) J. Biol. Chem. 264, 1163–11716. Pearl, L. H., and Savva, R. (May 29, 1996) UDI Protein Data Bank Full Tables

Release 76, Vol. 1, Brookhaven National Laboratory, Upton, NY7. Savva, R., and Pearl, L. H. (1995) Proteins 22, 287–2898. Karran, P., Cone, R., and Friedberg, E. C. (1981) Biochemistry 20, 6092–60969. Caradonna, S., Ladner, R., Hansbury, M., Kosciuk, M., Lynch, F., and Muller,

S. (1996) Exp. Cell Res. 222, 345–35910. Bennett, S. E., and Mosbaugh, D. W. (1992) J. Biol. Chem. 267, 22512–2252111. Cone, R., Bonura, T., and Friedberg, E. C. (1980) J. Biol. Chem. 255,

10354–1035812. Bennett, S. E., Schimerlik, M. I., and Mosbaugh, D. W. (1993) J. Biol. Chem.

268, 26879–2688513. Balasubramanian, S., Beger, R. D., Bennett, S. E., Mosbaugh, D. W., and

Bolton, P. H. (1995) J. Biol. Chem. 270, 296–30314. Beger, R. D., Balasubramanian, S., Bennett, S. E., Mosbaugh, D. W., and

Bolton, P. H. (1995) J. Biol. Chem. 270, 16840–1684715. Bennett, S. E., Jensen, O. N., Barofsky, D. F., and Mosbaugh, D. W. (1994)

J. Biol. Chem. 269, 21870–2187916. Mol, C. D., Arvai, A. S., Sanderson, R. J., Slupphaug, G., Kavli, B., Krokan,

H. E., Mosbaugh, D. W., and Tainer, J. A. (1995) Cell 82, 701–70817. Savva, R., and Pearl, L. (1995) Nat. Struct. Biol. 2, 752–75718. Slupphaug, G., Mol, C. D., Kavli, B., Arvai, A. S., Krokan, H. E., and Tainer,

J. A. (1996) Nature 384, 87–9219. Sanderson, R. J., and Mosbaugh, D. W. (1996) J. Biol. Chem. 271, 29l70–29l8l20. Mol, C. D., Arvai, A. S., Slupphaug, G., Kavli, B., Alseth, I., Krokan, H. E., and

Tainer, J. A. (1995) Cell 80, 869–87821. Savva, R., McAuley-Hecht, K., Brown, T., and Pearl, L. (1995) Nature 373,

487–49322. Wang, Z., Smith, D. G., and Mosbaugh, D. W. (1991) Gene (Amst.) 99, 31–3723. Kunkle, T. A., Roberts, J. D., and Zakour, R. A. (1987) Methods Enzymol. l54,

367–38224. Longley, M. J., and Mosbaugh, D. W. (1989) Methods Mol. Cell. Biol. 1, 79–9425. Longley, M. J., and Mosbaugh, D. W. (1993) Methods Enzymol. 218, 587–60926. Hanahan, D., Jessee, J., and Bloom, F. R. (1991) Methods Enzymol. 204, 63-ll527. Bennett, S. E. (1995) Characterization of Escherichia coli Uracil-DNA Glyco-

sylase-Inhibtor Protein Interaction. Ph.D. thesis, Oregon State University28. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U. S. A.

74, 5463–546729. Laemmli, U. K. (1970) Nature 227, 680–68530. Reisner, A. H. (1984) Methods Enzymol. 104, 439–44131. Kessler, H., and Seip, S. (1994) in Two-Dimensional NMR Spectroscopy

(Croasman, W. R., and Carlson, R. M., eds) pp. 619–654, VCH Publishers,Inc., New York

32. LeMaster, D. M. (1994) Prog. NMR Spectroscopy 26, 349–37133. Hull, W. E. (1994) in Two-Dimensional NMR Spectroscopy (Croasman, W. R.,

and Carlson, R. M., eds) pp. 67–456, VCH Publishers, Inc., New York34. Dyson, H. J., and Wright, P. E. (1994) in Two-dimensional NMR Spectroscopy

(Croasman, W. R., and Carlson, R. M., eds) pp. 655–698, VCH Publishers,Inc., New York

35. Beger, R. D., and Bolton, P. H. (1997) J. Biomol. NMR, in press36. Nigles, M., Clore, G. M., and Gronenborn, A. M. (1988) FEBS Lett. 229,

3l7–32437. Brunger, A. T., Clore, G. M., Gronenborn, A. M., Saffrich, R., and Nigles, M.

(1993) Science 261, 328–33138. Kollman, P. A., and Merz, K. M. (1990) Acc. Chem Res. 23, 246–25239. Olsen, L. C., Aasland, R., Wittwer, C. U., Krokan, H. E., and Helland, D. E.

(1989) EMBO J. 8, 3121–312540. McGeoch, D. J., Dalrymple, M. A., Davison, A. J., Dolan, A., Frame, M. C.,

McNab, D., Perry, L. J., Scott, J. E., and Taylor, P. (1988) J. Gen. Virol. 69,1531–1574

41. Varshney, U., Hutcheon, T., and van de Sande, J. H. (1988) J. Biol. Chem. 263,7776–7784

Uracil-DNA Glycosylase Inhibitor Protein 21419

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from

MosbaughAmy J. Lundquist, Richard D. Beger, Samuel E. Bennett, Philip H. Bolton and Dale W.FORMATION WITH ESCHERICHIA COLI URACIL-DNA GLYCOSYLASE Protein: ROLE OF SPECIFIC CARBOXYLIC AMINO ACIDS IN COMPLEX

Site-directed Mutagenesis and Characterization of Uracil-DNA Glycosylase Inhibitor

doi: 10.1074/jbc.272.34.214081997, 272:21408-21419.J. Biol. Chem. 

  http://www.jbc.org/content/272/34/21408Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/272/34/21408.full.html#ref-list-1

This article cites 35 references, 13 of which can be accessed free at

by guest on March 24, 2018

http://ww

w.jbc.org/

Dow

nloaded from