doi:10.1016/j.jmb.2010.09.065 J. Mol. Biol. (2010) 404, 639–649

Contents lists available at www.sciencedirect.com

Journal of Molecular Biologyj ourna l homepage: ht tp : / /ees .e lsev ie r.com. jmb

Recognition of Nucleoside Monophosphate Substratesby Haemophilus influenzae Class C Acid Phosphatase

Harkewal Singh1, Jonathan P. Schuermann2, Thomas J. Reilly3,4,Michael J. Calcutt3 and John J. Tanner1,5⁎1Department of Chemistry, University of Missouri—Columbia, Columbia, MO 65211, USA2Northeastern Collaborative Access Team, Department of Chemistry and Chemical Biology, Cornell University,Ithaca, NY 14853, USA3Department of Veterinary Pathobiology, University of Missouri—Columbia, Columbia, MO 65211, USA4Veterinary Medical Diagnostic Laboratory, University of Missouri—Columbia, Columbia, MO 65211, USA5Department of Biochemistry, University of Missouri—Columbia, Columbia, MO 65211, USA

Received 17 August 2010;received in revised form27 September 2010;accepted 30 September 2010Available online8 October 2010

Edited by G. Schulz

Keywords:X-ray crystallography;phosphatase;nucleotidase;substrate recognition;steady-state kinetics

*Corresponding author. DepartmenUniversity of Missouri—Columbia, CUSA. E-mail address: tannerjj@missAbbreviations used: CCAP, class

rP4, recombinant P4; NMN, nicotinamononucleotide; NR, nicotinamidehaloacid dehalogenase; Pi, inorganicclass B acid phosphatase; PMEA, 9-methoxy)ethyl]adenine; PDB, Proteipolyethylene glycol.

0022-2836/$ - see front matter © 2010 E

The e (P4) phosphatase from Haemophilus influenzae functions in a vestigialNAD+ utilization pathway by dephosphorylating nicotinamide mononu-cleotide to nicotinamide riboside. P4 is also the prototype of class C acidphosphatases (CCAPs), which are nonspecific 5′,3′-nucleotidases localizedto the bacterial outer membrane. To understand substrate recognition by P4and other class C phosphatases, we have determined the crystal structuresof a substrate-trapping mutant P4 enzyme complexed with nicotinamidemononucleotide, 5′-AMP, 3′-AMP, and 2′-AMP. The structures reveal ananchor-shaped substrate-binding cavity comprising a conserved hydro-phobic box that clamps the nucleotide base, a buried phosphoryl bindingsite, and three solvent-filled pockets that contact the ribose and thehydrogen-bonding edge of the base. The span between the hydrophobic boxand the phosphoryl site is optimal for recognizing nucleoside monopho-sphates, explaining the general preference for this class of substrate. Thebase makes no hydrogen bonds with the enzyme, consistent with anobserved lack of base specificity. Two solvent-filled pockets flanking theribose are key to the dual recognition of 5′-nucleotides and 3′-nucleotides.These pockets minimize the enzyme's direct interactions with the ribose andprovide sufficient space to accommodate 5′ substrates in an anticonformation and 3′ substrates in a syn conformation. Finally, the structuressuggest that class B acid phosphatases and CCAPs share a common strategyfor nucleotide recognition.

© 2010 Elsevier Ltd. All rights reserved.

t of Chemistry,olumbia, MO 65211,

ouri.edu.C acid phosphatase;mideriboside; HAD,phosphate; CBAP,

[(R)-2-(phosphono-n Data Bank; PEG,

lsevier Ltd. All rights reserve

Introduction

The lipoprotein e (P4)1,2 is a 28-kDa outermembrane acid phosphatase from Haemophilusinfluenzae, a common commensal inhabitant of thehuman nasopharynx and the etiologic agent of localand invasive infections in humans, particularly inchildren.3,4 The enzyme is a major component ofthe outer membrane and is highly conservedamong H. influenzae strains. The high conservation

d.

Fig. 1. Ribbon representation of D66N complexed withNMN. The core and cap domains are shown in blue andpink, respectively. NMN is shown in gray. The yellowsphere represents Mg2+. Residues of the aromatic box areshown in green. This figure and others were created withPyMOL.21

640 Structures of rP4 Complexes

and outer membrane location of e (P4) havemotivated investigations of the enzyme as apotential vaccine component. Studies have shownthat recombinant P4 (rP4) and rP4 mutant enzymesare highly immunogenic, that anti-rP4 antibodiesexhibit bactericidal activity, and that immunizationof mice with rP4 reduces nasal colonization ofnontypeable H. influenzae strains.5–7

The main biological role of e (P4) is to catalyze theconversion of nicotinamide mononucleotide (NMN)into nicotinamide riboside (NR) as part of a vestigialNAD+ utilization pathway.8,9 H. influenzae lacks thefull repertoire of enzymes needed for the de novobiosynthesis of NAD+; therefore, it must obtain thisessential cofactor from the host. TheNAD+ utilizationpathway includes an uptake system that importsNAD+, NMN, and NR into the periplasm. Within theperiplasm, the NAD+ nucleotidase NadN catalyzesthe hydrolysis of NAD+ to generate NMN and AMP.NMN produced by NadN or imported by the uptakesystem is dephosphorylated to NR by e (P4). NadNalso hasNMN5′-nucleotidase activity, but e (P4) has ahigher efficiency for NMN and is thus thought to bethe major catalyst for the production of NR for thepathway.8 NR is then transported across the innermembrane into the cytosol by the NR-specificpermease PnuC, where it is converted into NAD+

by the bifunctional NR kinase/NMN adenylyltrans-ferase NadR.10 Although the biological function of e(P4) in NAD+ utilization is well established, thestructural basis for the recognition of NMN by P4 isunknown.The larger context for the research described here is

that e (P4) is the prototype of class C acid phospha-tases (CCAPs). First recognized by Thaller et al. in1998 as a family of related bacterial enzymes, CCAPsbelong to the DDDD superfamily of phosphohydro-lases and are defined at the primary structure level bythe conserved bipartite sequence motif of [IV]-[VAL]-D-[IL]-D-E-T-[VM]-L-X-[NT]-X(2)-Y and [IV]-[LM]-X(2)-G-D-[NT]-L-X-D-F (Asp residues of the DDDDmotif in boldface).11

In addition to e (P4),1,2,12–14 several CCAPs, includ-ing those from Elizabethkingia meningoseptica (OlpA15),Streptococcus equisimilis (LppC16), Helicobacter pylori(HppA17), and Clostridium perfringens,18,19 have beencharacterized to various degrees. CCAPs are dimericenzymes that exhibit phosphomonoesterase activityfor commonly used aryl phosphate substrates such asp-nitrophenyl phosphate and 4-methylumbelliferylphosphate. Among the physiologically relevant mole-cules that have been tested, the highest catalyticefficiencieshave been achieved with nucleoside 5′-monophosphates. CCAPs do not exhibit a strong basepreference among this class of substrates. OlpA,LppC,20 rP4 (vide infra), and the C. perfringensenzyme18 also exhibit activity with nucleoside 3′-monophosphate substrates, but with lower efficiencythan nucleoside 5′-monophosphate substrates. Thus,

available in vitro data suggest that CCAPs functionprimarily as nonspecific 5′,3′-nucleotidases.The structure of one CCAP (rP4) has been

determined.14 rP4 has a two-domain fold consistingof a core α/β domain (Fig. 1, blue) and an α-helicalcap domain (Fig. 1, pink). The fourAsp residues of theDDDDmotif are clustered around a Mg2+ at the baseof the active site (Fig. 1, yellow sphere). The coredomain fold indicates that rP4 belongs to the haloaciddehalogenase (HAD) structural superfamily.22

The structure of rP4 complexed with the inhibitortungstate provided insight into the identities of thenucleophile that attacks the substrate phosphorylgroup (Asp64), the residue that protonates theleaving group (Asp66), and the side chains thatstabilize the substrate phosphoryl group (Lys161and Thr124). However, the residues that interactwith the nonphosphoryl groups of substrates havenot been identified. Thus, the structural elementsthat enforce the preference for nucleoside monopho-sphates are unknown. Furthermore, the question ofhow CCAPs achieve the dual recognition of nucle-oside 5′-monophosphate and nucleoside 3′-mono-phosphate remains unanswered.Within this context, we initiated a structure-based

study of substrate recognition with the goals ofunderstanding how rP4 binds its known biologicalsubstrate (NMN) and,more generally, elucidating the

Table 1. Data collection and refinement statistics

NMN 5′-AMP 3′-AMP 2′-AMP Phosphate

Beamline 24-ID-C 4.2.2 4.2.2 4.2.2 24-ID-CWavelength 0.9792 1.00 1.00 1.00 0.9792Data collection resolution (Å) 50–1.35 (1.40–1.35) 50–1.55 (1.61–1.55) 50–1.85 (1.92–1.85) 50–1.90 (1.93–1.90) 50–1.40 (1.45–1.40)Number of observations 622,380 944,628 536,096 499,757 340,873Number of unique reflections 66,003 44,418 26,166 24,575 57,486Rmerge(I) 0.074 (0.603) 0.088 (0.427) 0.114 (0.439) 0.120 (0.534) 0.064 (0.557)Average I/σ 39.6 (4.3) 47.0 (5.9) 27.6 (4.35) 30.2 (3.47) 25.4 (2.8)Completeness (%) 99.7 (100) 100 (100) 99.9 (99.4) 99.9 (98.3) 95.9 (98.3)Redundancy 9.4 (9.4) 21.3 (19.0) 20.5 (14.3) 20.3 (13.0) 5.9 (5.4)Refinement resolution (Å) 50–1.35 (1.37–1.35) 50–1.55 (1.58–1.55) 50–1.85 (1.92–1.85) 50–1.90 (1.98–1.90) 50–1.40 (1.42–1.40)Rcryst 0.135 (0.211) 0.141 (0.159) 0.164 (0.177) 0.178 (0.196) 0.139 (0.207)Rfree

a 0.154 (0.245) 0.171 (0.219) 0.189 (0.202) 0.217 (0.245) 0.168 (0.277)Number of protein residues 247 247 247 247 247Number of protein atoms 1933 1938 1931 1912 1929Number of water molecules 274 280 231 173 256Average B-factor (Å2)

Protein 14.1 15.8 16.8 26.8 14.2Water 25.0 27.5 24.4 30.4 24.9Ligand 12.2 15.7 22.1 23.3 11.7Mg2+ 7.7 9.7 11.8 15.8 7.6

RMSDb

Bonds (Å) 0.006 0.005 0.006 0.006 0.005Angles (°) 1.17 1.04 1.00 1.02 1.03

Ramachandran plotc

Favored (%) 98.0 98.0 98.0 97.5 98.0Allowed (%) 2.0 2.0 2.0 2.5 2.0

Coordinate error (Å)d 0.13 0.17 0.18 0.21 0.14PDB code 3OCU 3OCV 3OCW 3OCX 3OCY

Values for the outer resolution shell are given in parentheses.a A common set of test reflections (5%) was used for the refinement of all structures.b Compared to the parameters of Engh and Huber.23c The Ramachandran plot was generated with RAMPAGE.24d Maximum-likelihood-based coordinate error estimate.

641Structures of rP4 Complexes

structural features of CCAPs that are responsible forrecognizing nucleoside monophosphate substrates.To this end, we have determined the high-resolutioncrystal structures of a substrate-trapping mutant ofrP4 complexed with NMN, 5′-AMP, 3′-AMP, and2′-AMP, as well as a structure of rP4 complexedwith the product inorganic phosphate (Pi) (Table 1).The structures and the accompanying kinetic dataprovide insight into the basis of the nucleotidaseactivities of P4 and other CCAPs.

Results

Structures of D66N complexed with NMN and5′-AMP

Crystal structures of rP4 complexedwithNMNand5′-AMP were determined to understand how theenzyme recognizes nucleoside 5′-monophosphatesubstrates. D66N, a mutant of rP4 in which theresidue that protonates the leaving group (Asp66) hasbeen changed to Asn, was used for the structuredetermination of enzyme–substrate complexes. Struc-tures of D66N complexed with NMN (Fig. 1) and 5′-AMP were determined at high-resolution limits of

1.35 Å and 1.55 Å, respectively (Table 1). The electrondensity maps at these resolutions allowed an unam-biguous determination of the conformations of thebound substrates, enumeration of enzyme–substrateinteractions, and identification of water moleculesinvolved in substrate binding (Fig. 2a and b).NMN and 5′-AMP bind in an anchor-shaped cavity

located at the junction of theα/β core domain and thecap domain, as shown for 5′-AMP in Fig. 3a. NMNand 5′-AMP adopt identical conformations whenbound to the enzyme (Fig. 4). Based on thenomenclature described by Saenger, the ribose ineach case displays an asymmetrical twist with majorC3′-endo pucker andminor C2′-exo pucker (

3T2).25 The

bases adopt an anti orientation about the glycosylbond.The substrate phosphoryl group occupies the well-

known phosphoryl binding site of HAD superfamilyenzymes (Fig. 4). In both structures, the phosphorylgroup interacts with the active site Mg2+, theconserved residues Thr124 and Lys161, and thebackbone N–H group of Asn66. These interactionsare identical with those observed in the structure ofrP4 complexed with Pi (Fig. 2e). Furthermore, thesubstrate phosphoryl sits above the nucleophilic Oatom of Asp64 in an orientation suggestive of

Fig. 2. Conformations of ligandsbound to rP4: (a) NMN, (b) 5′-AMP,(c) 3′-AMP, (d) 2′-AMP, and (e) Pi. Ineach panel, the cage (cyan for theligand; silver for protein side chains)represents a simulated-annealingσA-weighted F−Fc omit map con-toured at 3.0σ. Prior to map calcula-tion, the ligand, surroundingresidues, and water molecules wereremoved, and simulated-annealingrefinement was performed usingPHENIX.

642 Structures of rP4 Complexes

backside nucleophilic attack. In particular, the nucle-ophilic O atom is poised 3.0 Å from the P atom, andthe angle formed by the nucleophile, P, and O5′ is179°. Finally, the O atom of the scissile bond forms ahydrogen bondwith Asn66 (2.8 Å). This interaction isconsistent with Asp66 functioning as the acid thatprotonates the leaving group.The ribose moieties of NMN and 5′-AMP form

identical interactions with rP4 (Fig. 4). In eachstructure, the ribose is oriented such that the C2′–C3′locus contacts the pyrrole ring of Trp91, while theopposite edge of the ring (i.e., O4′) points towards alarge solvent-filled pocket on the right-hand side ofthe active site (Fig. 3a, pocket 1). The hydroxyl groupsof the ribose occupy a second smaller pocket on the

left side (Fig. 3a, pocket 2). Within pocket 2, the 2′-hydroxyl group forms a water-mediated hydrogenbond with Glu131 (Fig. 4).Finally, the bases of NMN and 5′-AMP bind in an

aromatic box formed by Phe86, Trp91, and Tyr221(Fig. 4). The first two residues are part of the helix–loop–helix substructure of the cap domain, whileTyr221 is located in the loop that follows the laststrand of the core domain (Fig. 1). Phe86 and Tyr221form the sides of the box, while Trp91 forms the floor.In both complexes, the base stacks in parallel betweenPhe86 and Tyr221, forming an aromatic sandwich.Trp91 contacts the C5–C6 locus of the NMN nicotin-amide and the pyrrole ring of adenine. The hydrogen-bonding groups of the bases are directed towards a

Fig. 3. Close-up view of the rP4active site emphasizing shape andsolvent content. The panels corre-spond to D66N complexed with (a)5′-AMP, (b) 3′-AMP, and (c) 2′-AMP.Three solvent-filled pockets are la-beled 1, 2, and 3. The core and capdomains are shown in blue andpink,respectively, and residues of thearomatic box are shown in green.The yellow sphere represents Mg2+.

643Structures of rP4 Complexes

solvent-filled cavity at the top of the active site (Fig.3a, pocket 3). As a result, neither base forms directhydrogen bonds with the enzyme, but there arewater-mediatedhydrogenbondswithGln79,Asn220,and Glu225.

Structures of D66N complexed with 3′-AMP and2′-AMP

Structures ofD66N complexedwith 3′-AMPand 2′-AMP were determined to understand how rP4 andother CCAPs accommodate nucleoside monopho-sphate substrates differing in the position of thephosphoryl group on the ribose. Both electron density

maps exhibited a strong feature representing thebound substrates. The quality of the 2′-AMP map(Fig. 2d) rivaled those of NMN and 5′-AMP.Although the 3′-AMP map exhibited a lower quality,the locations of the adenine ring, phosphoryl group,2′-hydroxyl, andO4′were unambiguous (Fig. 2c). Thedensity for the C5′–O5′ bond of the ribose was weakerand suggested the same conformation as in 2′-AMP.The weaker density for 3′-AMP also suggested thatthe occupancy of the ligand is less than 1.0; refinementof the occupancy resulted in a value of 0.81.The conformations of 2′-AMP and 3′-AMP differ

substantially from that of 5′-AMP (Fig. 5). In both 2′-AMP and 3′-AMP, the ribose adopts the C2′-endo

Fig. 4. Recognition of the nucleoside 5′-monophosphate substrates (a) NMN and (b) 5′-AMP (stereographic views). Inboth panels, the substrate is represented in yellow sticks, and Mg2+ is depicted as a yellow sphere. Secondary structuralelements of the core and cap domains are shown in blue and pink, respectively, and residues of the aromatic box areshown in green.

644 Structures of rP4 Complexes

pucker (2E), and the base is in a syn orientation. Wenote that the C2′-endo pucker is favored for synnucleosides.25 The syn conformation is stabilized byan intramolecular hydrogen bond between N3 andO5′ (2.9 Å in 3′-AMP and 2.8 Å in 2′-AMP). We notethat this type of interaction is commonly found insyn nucleotides.25 As observed with 5′-AMP, thebases of 2′-AMP and 3′-AMP bind in the aromaticbox, with the N6–N7 edge directed towards pocket 3(Fig. 3b and c). However, the ribose rings of 2′-AMPand 3′-AMP are oriented with O4′ pointing towardsthe left into pocket 2, which is the reverse of theorientation of the bound 5′ substrates (compare Fig.3b and c to Fig. 3a).Finally, there is a difference between the position

of the phosphoryl group of 2′-AMP and the positionof the phosphoryl group of the other substrates (Fig.5b). The phosphoryl group of 2′-AMP is shifted by

1.8 Å from the expected site such that the P atom is4.3 Å from the nucleophilic O atom of Asp64, whichis obviously not optimal for catalysis. As a result ofthe shift, the phosphoryl group does not form theexpected interactions with the conserved residuesThr124 and Lys161.

Kinetic characterization of rP4

The catalytic efficiencies (kcat/Km) of rP4 for certainnucleoside monophosphate substrates were estimat-ed using steady-state kinetic assays. NMN, 5′-AMP,3′-AMP, and 2′-AMP were used as substrates; theresults are summarized in Table 2. Among thesubstrates tested, the enzyme has the highestefficiency for NMN, followed by 5′-AMP, 3′-AMP,and 2′-AMP. The efficiency for NMN is only 2.5times that for 5′-AMP, suggesting that rP4 does not

Fig. 5. The active sites of D66N complexed with (a) 3′-AMP and (b) 2′-AMP (stereographic views). In both panels, thesubstrate is represented in yellow sticks, and Mg2+ is depicted as a yellow sphere. Secondary structural elements of thecore and cap domains are shown in blue and pink, respectively, and residues of the aromatic box are shown in green. In(b), the location of a normal phosphoryl binding site is indicated by the phosphate ion shown in lines and labeled Pi.

645Structures of rP4 Complexes

exhibit a strong preference for NMN over othernucleoside 5′-monophosphatases. Comparing thedata for the AMP substrates, we found that rP4exhibits only a 2-fold preference for 5′-AMP over 3′-AMP. 2′-AMP is the poorest substrate tested, with a20-fold lower efficiency than 5′-AMP. Thus, likesome other CCAPs, it is reasonable to classify rP4 as adual 5′,3′-nucleotidase.

Table 2. Kinetic parameters for rP4

Substrate Km (mM) kcat (s−1) kcat/Km (%)a

NMN 0.7±0.2 0.52±0.05 1005′-AMP 0.23±0.04 0.070±0.004 403′-AMP 0.8±0.2 0.10±0.01 202′-AMP 6±2 0.11±0.01 2

a Relative to NMN.

Discussion

Structural basis of substrate recognition

The main aim of our work is to provide structure-based insight into the substrate preference andpromiscuity of rP4 and CCAPs closely related torP4. CCAPs are somewhat selective in the sense thatthe in vivo substrates are thought to be nucleosidemonophosphates. For example, genetic and molecu-lar studies suggest that NMN is a biological substratefor e (P4).9 Although the in vivo substrates of otherCCAPs have not been similarly identified, available invitro kinetic data suggest nucleosidemonophosphatesas biologically relevant substrates. On the other hand,CCAPs are generally promiscuous with regard to theidentity of the base and whether the phosphoryl is

646 Structures of rP4 Complexes

attached at the 5′ position or the 3′ position of thesugar. For example, we have shown here that thecatalytic efficiency of rP4 for 5′-AMP is approximatelyhalf that of NMN and only twice that of 3′-AMP.Similar results have been reported for otherCCAPs.15,17,18

The basis for the preference for nucleoside mono-phosphates is evident from the D66N complexes.The aromatic box is well suited for binding thearomatic ring systems of nucleotides. In particular,the box provides two aromatic residues that stack inparallel, with the base forming a sandwich. Further-more, the span between the aromatic box and thephosphoryl binding pocket is optimal for nucleosidemonophosphates.The structures also provide insight into substrate

promiscuity with respect to the base. The base isaligned in the aromatic box such that the hydrogen-bonding groups point into a solvent-filled pocked. Asa result, there are no direct hydrogen bonds with theenzyme implying low base selectivity. Indeed, wehave shown here that the catalytic efficiency of rP4 for5′-AMP rivals that for NMN. These results imply thatP4 is not tuned to exclusively recognize NMN, andthat the enzyme may have other biological functionsbeyond NAD+ utilization, such as acquiring Pi fromnucleoside monophosphates found in the bacterium'senvironment.The D66N structures also shed light on how rP4

achieves the dual recognition of 5′-nucleoside mono-phosphate and 3′-nucleoside monophosphate, acharacteristic of some CCAPs. The 5′ substrates bindwith the base in an anti conformation, whereas 3′-AMP adopts a syn conformation. The somewhathigher catalytic efficiency of 5′ substrates likelyreflects the lower conformational energy of the anticonformation. The two solvent-filled pockets flankingthe ribose appear to be important for the enzyme'sability to bind both types of substrates. These pocketsminimize the enzyme's direct interactions with theribose and provide sufficient space to accommodatethe different ribose orientations. Indeed, the riboseoccupies the widest part of the active site (Fig. 3).Thus, the open active site of rP4 appears to underlieboth the weak base specificity and the dual recogni-tion of 5′-nucleoside monophosphate and 3′-nucleo-side monophosphate.The structure of D66N complexed with 2′-AMP

further demonstrates the ability of the active site tobind different nucleoside monophosphates. Thestructure is unusual in that the phosphoryl is shiftedaway from the catalytic Asp into a solvent-filledpocket and is thus not aligned optimally for catalysis.This conformation perhaps represents a nonproduc-tive complex. As discussed by Cornish-Bowden,nonproductive binding of the substrate is a form ofcompetitive inhibition, and the measured values ofVmax andKm are lower than expected by an unknownand typically immeasurable amount.26 Nevertheless,

Vmax/Km does provide a correct measure of thecatalytic properties of the enzyme in such cases.26 Wefound that the apparent catalytic efficiency of rP4 for2′-AMP is substantially lower (10–50 times) thanthose for the other substrates tested, indicating that itis a poor substrate. Regardless of whether theobserved conformation represents nonproductivesubstrate binding, the rP4/2′-AMPstructure providesanother illustration of how the open active site plays arole in binding different nucleoside monophosphateligands.

Connections with other CCAPs and class B acidphosphatases

Sequence conservation suggests that the rP4 com-plexes reported here are representative of otherCCAPs. The residues of the phosphate binding pocket(Asp64 Asp66, Lys161, and Thr124) are highlyconserved in the HAD superfamily, and, as expected,these residues are highly conserved amongCCAPs. Infact, Asp64, Asp66, and Lys161 are invariant amongthe CCAPs, and Thr124 appears as Ser (a conservativesubstitution) in some CCAPs. Residues of thearomatic box are also highly conserved. Tyr221appears to be universally conserved among CCAPs.An aromatic ring in the formof Phe, Tyr, Trp, orHis isalways present at the residue corresponding to rP4Phe86. Thus, all CCAPs appear to have residuescapable of forming an aromatic sandwich with thenucleotide base. Finally, the floor of the box (Trp91) isalmost invariant; substitution with Phe (a conserva-tive change) is observed in some CCAP sequences.Thus, we suggest that the structures reported hereprovide a model for understanding substrate recog-nition in other CCAPs.The rP4 structures also reveal a new relationship

betweenCCAPs and the related phosphatases knownas class B acid phosphatases (CBAPs). The D66N/5′-AMP complex is reminiscent of the structure of theCBAP AphA complexed with the 5′-AMP analog 9-[(R)-2-(phosphono-methoxy)ethyl]adenine (PMEA)[Protein Data Bank (PDB) code 2G1A27]. CBAPs alsobelong to the DDDD superfamily, and AphA isregarded as the prototype of the family. CBAPs have abipartite sequence motif that is similar to that ofCCAPs, and, likeCCAPs, class B enzymes also showapreference for nucleoside 5′-monophosphate andnucleoside 3′-monophosphate substrates. AlthoughrP4 and AphA share a common HAD superfamilycore domain, their sequences have negligible similar-ity outside of the bipartite sequence motif (13%overall identity), and their cap domains have differentfolds. The latter difference is related to the differentquaternary structures of the two enzymes (dimer forCCAPs and tetramer for CBAPs).Despite the different cap domain structures, AphA

and P4 exhibit commonalities in substrate recogni-tion. In particular, the adenine ring of PMEA packs

Fig. 6. Comparison of the active sites of D66N/5′-AMP(white) and AphA/PMEA (cyan). Residue numbers forthe aromatic boxes are listed as rP4/AphA. The yellowsphere represents Mg2+.

647Structures of rP4 Complexes

into a hydrophobic pocket that is similar to thearomatic box of rP4 (Fig. 6). The hydrophobic pocketof AphA consists of Leu71, Tyr193, and Phe56, whichare analogous to the rP4 aromatic box residues Phe86,Tyr221, and Trp91, respectively. Note also that theadenine rings have nearly identical orientations in thetwo structures. We thus suggest that CBAPs andCCAPs share a common strategy for nucleotiderecognition.

Materials and Methods

Subcloning and mutagenesis

Previous structural studies of rP4 used a recombinantenzyme lacking a polyhistidine affinity tag, but thepurification of that enzyme was inefficient,13 and crystalli-zationwas not highly reproducible. Therefore, for thiswork,an rP4 construct encoding the enzyme fused to a C-terminalhexahistidine tag was created to aid purification. The helgene was subcloned into pET20b using NcoI and XhoI sitessuch that the N-terminal signal sequence was replaced withthe pelB leader sequence from Erwinia chrysanthemi, and theN-terminal Cys of the mature protein was replaced withMet. As a result, the rP4 protein used here contains aC-terminal hexahistidine tag, is free of lipid modification,and is targeted to the Escherichia coli periplasm.A site-directed mutant of rP4 in which the residue that

protonates the leaving group (Asp66) is changed to Asn(D66N) was created to determine the crystal structures ofenzyme–substrate complexes. We note that an analogousstrategy has been used to trap substrate complexes ofother phosphatases.28,29 The mutation was introducedinto the aforementioned plasmid using the QuickChangekit (Stratagene) and confirmed by DNA sequencing.

Expression and purification of rP4

The rP4 pET20b plasmid was transformed into E. coliBL21AI cells and plated on LB medium containing

ampicillin (50 μg/mL). A single colony of the transformantwas picked and used to inoculate 1 L of culture. The proteinwas expressed via autoinduction30 at 37 °C, with constantshaking at 300 rpm. The cells were harvested by centrifu-gation at 3500 rpm for 30 min at 4 °C and resuspended in20 mM phosphate, 20 mM imidazole, and 500 mM NaCl atpH7.0. The cell pelletwas flash frozen in liquid nitrogen andstored at −80 °C.Frozen cells were thawed at 4 °C and ruptured using a

French press at 1000 psi. Unbroken cells and cellulardebris were removed by centrifugation for 60 min at17,500 rpm and 4 °C. The supernatant was collected andsubjected to a second centrifugation step (30 min,17,500 rpm, 4 °C). The resulting supernatant was usedfor further purification by immobilized metal-ion affinitychromatography (Ni2+-charged HiTRAP; GE Health-care), followed by cation-exchange chromatography(HiTRAP SP; GE Healthcare). The purified enzyme wasdialyzed into 50 mM sodium acetate, 50 mM NaCl, and2.5 mM MgCl2 at pH 6.0. The sample was concentratedto 2–5 mg/mL (based on the bicinchoninic acid assay;Pierce) using a centrifugal ultrafiltration device (cutoff,10 kDa).

Expression and purification of D66N

The D66N mutant plasmid was transformed into E. coliBL21(DE3) cells and plated on LB medium containingampicillin (50 μg/mL). A single colony of the transformantwas picked and used to inoculate a 10-mL starter culturecontaining ampicillin (50 μg/mL). After overnight growthat 37 °C with shaking at 250 rpm, the starter culture wasused to inoculate 1 L of LB medium supplemented withampicillin (50 μg/mL). The culture was then grown at 37 °Cwith shaking at 250 rpm until the optical density at 600 nmhad reached 0.6. Protein expression was induced by addingisopropylβ-D-thiogalactoside (0.5mM), and the culturewasincubated for 8 h at 25 °C with shaking at 200 rpm. Theexpressed protein was purified as described above for rP4.After purification, the sample was dialyzed overnight into50 mM sodium acetate and 50 mM NaCl at pH 6.0 andconcentrated to 10 mg/mL.

Crystallization and preparation of enzyme–ligandcomplexes

Crystallization trials of rP4 and D66N were performed at20 °C using the sitting-drop method, with drops formedfrom 2 μL of enzyme solution and 2 μL of reservoir solution.Initial crystallization conditions were identified usingcommercially available screens (Hampton Research). Prom-ising results were obtained with reservoirs containingammonium citrate and polyethylene glycol (PEG) 3350.After optimization, diffraction-quality crystals with ahexagonal externalmorphologywere grownover reservoirscontaining 0.05–0.2 M ammonium citrate, 0.05–0.15 mMMgCl2, and 18–28% (wt/vol) PEG 3350 in the pH range 6.8–7.2. The best rP4 crystals typically grew in 18–23% (wt/vol)PEG 3350, whereas a higher concentration of 23–28% (wt/vol) PEG 3350 was used for crystallization of D66N. Thetypical protein concentrations used for optimal crystalgrowth were 1–3 mg/mL for rP4 and 8–10 mg/mL forD66N.

648 Structures of rP4 Complexes

Crystals of the D66N–substrate complexes wereobtained by soaking, as follows. Stock solutions of thesubstrates were prepared in water, and the pH wasadjusted to 6.0. D66N crystals were cryoprotected in 28–30% PEG 3350, 0.1 M ammonium citrate buffer (pH 7.0),and 20% PEG 200. The cryoprotected crystals weretransferred to a solution of the cryobuffer supplementedwith 5–20 mM substrate and 100–200 mM MgCl2. Thesoaking time was in the range of 5–45 min.Crystals of rP4 complexed with Pi were also obtained by

soaking. A stock solution of potassium phosphate(100 mM KH2PO4, pH 6.0) was first prepared. Next,crystals of rP4 were cryoprotected at room temperature in23–28% (wt/vol) PEG 3350, 0.1 M ammonium citratebuffer (pH 7.0), and 20% PEG 200. The cryoprotectedcrystals were transferred to a solution of the cryobuffersupplemented with 25 mM Pi and 200 mM MgCl2. After30 min, the crystals were picked up with Hampton loopsand plunged into liquid nitrogen.

Structure determination

X-ray diffraction data sets were collected at AdvancedLight Source beamline 4.2.2 and Advanced Photon Sourcebeamline 24-ID-C (Table 1). The data sets were processedwith HKL2000.31 The crystals belonged to space groupP6522 with unit cell lengths of a=98 Å and c=107 Å, onemolecule in the asymmetrical unit, 54% solvent, and a Vmof 2.65 Å3/Da.32,33 We note that this form is different fromthe tetragonal one used in our earlier work.14 Initialphases were estimated using molecular replacement asimplemented in PHASER,34 with the search modelderived from a previously determined rP4 structure(PDB code 3ET414). Coot35 was used for model building,and PHENIX36 was used for refinement. A common set oftest reflections (5%) was used for refinement calculations.For each structure, the B-factor model used during theinitial rounds of refinement consisted of an isotropicB-factor for each nonhydrogen atom and TLS refinement,with one TLS group corresponding to the protein chain.Anisotropic B-factors were used during the final fewrounds of refinement of the NMN and Pi complexes. Theintroduction of anisotropic B-factors decreased Rfree by0.08 for the NMN complex and by 0.06 for the Pi complex.

Kinetic characterization

Steady-state enzymatic activity was assessed at 25 °Cusing a discontinuous assay that measures the productionof Pi.

37,38 The assay buffer consisted of 100 mM sodiumacetate, and 1 mM MgCl2 at pH 5.5. For each substrateconcentration, the reaction was stopped using the mala-chite green reagent after reaction times of 15 s, 75 s, 135 s,and 195 s, and the citrate color development reagent (34%sodium citrate, wt/vol) was added 60 s after stoppingeach reaction. After 30 min, the Pi concentrations weredetermined spectrophotometrically at 625 nm by referenceto a standard curve constructed from solutions of knownPi concentration. The initial rate was estimated by fittingthe data from the four time points to a line. Apparentvalues of Km and Vmax were estimated by fitting the initialrate data to the Michaelis–Menten equation using Origin8 software.

PDB accession codes

Atomic coordinates and structure factor amplitudeshave been deposited in the PDB39 with accession codes3OCU (NMN), 3OCV (5′-AMP), 3OCW (3′-AMP), 3OCX(2′-AMP), and 3OCY (Pi).

Acknowledgements

We thank Dr. Jay Nix of Advanced Light Sourcebeamline 4.2.2 for help with data collection. Thisresearch was supported by National Institutes ofHealth grant U54 AI057160 to the Midwest RegionalCenter of Excellence for Biodefense and EmergingInfectious Disease Research and the University ofMissouri Research Board. H.S. was supported by apredoctoral fellowship from National Institutes ofHealth grant DK071510 and by a Chancellor'sDissertation Completion Fellowship from the Uni-versity of Missouri—Columbia. Part of this researchwas performed at the Advanced Light Source. TheAdvanced Light Source was supported by theDirector, Office of Science, Office of Basic EnergySciences, US Department of Energy, under contractno. DE-AC02-05CH11231. Part of this work wasbased on research conducted at the NortheasternCollaborative Access Team beamlines of theAdvanced Photon Source, supported by award RR-15301 from the National Center for ResearchResources at the National Institutes of Health. Useof theAdvanced Photon Sourcewas supported by theOffice of Basic Energy Sciences, US Department ofEnergy, under contract no. W-31-109-ENG-38.

References

1. Reilly, T. J., Chance, D. L. & Smith, A. L. (1999). Outermembrane lipoprotein e (P4) of Haemophilus influenzaeis a novel phosphomonoesterase. J. Bacteriol. 181,6797–6805.

2. Reilly, T. J. & Smith, A. L. (1999). Purification andcharacterization of a recombinant Haemophilus influ-enzae outer membrane phosphomonoesterase e (P4).Protein Expr. Purif. 17, 401–409.

3. Foxwell, A. R., Kyd, J. M. & Cripps, A. W. (1998).NontypeableHaemophilus influenzae: pathogenesis andprevention. Microbiol. Mol. Biol. Rev. 62, 294–308.

4. Murphy, T. F., Faden, H., Bakaletz, L. O., Kyd, J. M.,Forsgren, A., Campos, J. et al. (2009). NontypeableHaemophilus influenzae as a pathogen in children.Pediatr. Infect. Dis. J. 28, 43–48.

5. Green, B. A., Baranyi, E., Reilly, T. J., Smith, A. L. &Zlotnick, G. W. (2005). Certain site-directed, none-nzymatically active mutants of the Haemophilusinfluenzae P4 lipoprotein are able to elicit bactericidalantibodies. Infect. Immun. 73, 4454–4457.

6. Mason, K. W., Zhu, D., Scheuer, C. A., McMichael,J. C., Zlotnick, G. W. & Green, B. A. (2004). Reduction

649Structures of rP4 Complexes

of nasal colonization of nontypeable Haemophilusinfluenzae following intranasal immunization withrLP4/rLP6/UspA2 proteins combined with aqueousformulation of RC529. Vaccine, 22, 3449–3456.

7. Hotomi, M., Ikeda, Y., Suzumoto, M., Yamauchi, K.,Green, B. A., Zlotnick, G. et al. (2005). A recombinantP4 protein of Haemophilus influenzae induces specificimmune responses biologically active against naso-pharyngeal colonization in mice after intranasalimmunization. Vaccine, 23, 1294–1300.

8. Kemmer, G., Reilly, T. J., Schmidt-Brauns, J., Zlotnik,G.W., Green, B. A., Fiske,M. J. et al. (2001). NadNand e(P4) are essential for utilization of NAD and nicotin-amide mononucleotide but not nicotinamide ribosidein Haemophilus influenzae. J. Bacteriol. 183, 3974–3981.

9. Gerlach, G. & Reidl, J. (2006). NAD+ utilization inPasteurellaceae: simplification of a complex pathway.J. Bacteriol. 188, 6719–6727.

10. Singh, S. K., Kurnasov, O. V., Chen, B., Robinson, H.,Grishin, N. V., Osterman, A. L. & Zhang, H. (2002).Crystal structure of Haemophilus influenzae NadRprotein. A bifunctional enzyme endowed with NMNadenyltransferase and ribosylnicotinimide kinaseactivities. J. Biol. Chem. 277, 33291–33299.

11. Thaller, M. C., Schippa, S. & Rossolini, G. M. (1998).Conserved sequencemotifs amongbacterial, eukaryotic,and archaeal phosphatases that define a new phospho-hydrolase superfamily. Protein Sci. 7, 1647–1652.

12. Reilly, T. J., Green, B. A., Zlotnick, G.W. & Smith, A. L.(2001). Contribution of the DDDD motif of H.influenzae e (P4) to phosphomonoesterase activity andheme transport. FEBS Lett. 494, 19–23.

13. Ou, Z., Felts, R. L., Reilly, T. J., Nix, J. C. & Tanner, J. J.(2006). Crystallization of recombinant Haemophilusinfluenzae e (P4) acid phosphatase. Acta Crystallogr.Sect. F, 62, 464–466.

14. Felts, R. L., Ou, Z., Reilly, T. J. & Tanner, J. J. (2007).Structure of recombinant Haemophilus influenzae e (P4)acid phosphatase reveals a new member of thehaloacid dehalogenase superfamily. Biochemistry, 46,11110–11119.

15. Passariello, C., Schippa, S., Iori, P., Berlutti, F., Thaller,M. C. & Rossolini, G. M. (2003). The molecular class Cacid phosphatase of Chryseobacterium meningosepticum(OlpA) is a broad-spectrum nucleotidase with prefer-ential activity on 5′-nucleotides. Biochim. Biophys. Acta,1648, 203–209.

16. Malke, H. (1998). Cytoplasmic membrane lipoproteinLppC of Streptococcus equisimilis functions as an acidphosphatase. Appl. Environ. Microbiol. 64, 2439–2442.

17. Reilly, T. J. & Calcutt, M. J. (2004). The class C acidphosphatase of Helicobacter pylori is a 5′ nucleotidase.Protein Expr. Purif. 33, 48–56.

18. Reilly, T. J., Chance, D. L., Calcutt, M. J., Tanner, J. J.,Felts, R. L., Waller, S. C. et al. (2009). Characterizationof a unique class C acid phosphatase from Clostridiumperfringens. Appl. Environ. Microbiol. 75, 3745–3754.

19. Wang, R., Ohtani, K., Wang, Y., Yuan, Y., Hassan, S. &Shimizu, T. (2010). Genetic and biochemical analysisof a class C non-specific acid phosphatase (NSAP) ofClostridium perfringens. Microbiology, 156, 167–173.

20. Malke, H. & Steiner, K. (1999). XIV LancefieldInternational Symposium on Streptococci and Streptococ-cal Diseases, Aukland New Zealand.

21. DeLano, W. L. (2002). The PyMOL User's Manual.DeLano Scientific, Palo Alto, CA.

22. Allen, K. N. & Dunaway-Mariano, D. (2004). Phos-phoryl group transfer: evolution of a catalytic scaffold.Trends Biochem. Sci. 29, 495–503.

23. Engh, R. A. & Huber, R. (1991). Accurate bond andangle parameters for X-ray protein structure refine-ment. Acta Crystallogr. Sect. A, 47, 392–400.

24. Lovell, S. C., Davis, I. W., Arendall, W. B., III, deBakker, P. I., Word, J. M., Prisant, M. G. et al. (2003).Structure validation by Calpha geometry: phi,psi andCbeta deviation. Proteins, 50, 437–450.

25. Saenger, W. (1984). Principles of Nucleic Acid Structure.Springer-Verlag, New York, NY.

26. Cornish-Bowden, A. (1979). Fundamentals of EnzymeKinetics. Butterworth, London, UK.

27. Leone, R., Cappelletti, E., Benvenuti, M., Lentini, G.,Thaller, M. C. & Mangani, S. (2008). Structural insightsinto the catalytic mechanism of the bacterial class Bphosphatase AphA belonging to the DDDD superfam-ily of phosphohydrolases. J. Mol. Biol. 384, 478–488.

28. Flint, A. J., Tiganis, T., Barford, D.& Tonks, N. K. (1997).Development of “substrate-trapping” mutants to iden-tify physiological substrates of protein tyrosine phos-phatases. Proc. Natl Acad. Sci. USA, 94, 1680–1685.

29. Singh, H., Felts, R. L., Schuermann, J. P., Reilly, T. J. &Tanner, J. J. (2009). Crystal structures of the histidineacid phosphatase from Francisella tularensis provideinsight into substrate recognition. J. Mol. Biol. 394,893–904.

30. Studier, F. W. (2005). Protein production by auto-induction in high density shaking cultures. ProteinExpression Purif. 41, 207–234.

31. Otwinowski, Z. & Minor, W. (1997). Processing ofX-ray diffraction data collected in oscillation mode.Methods Enzymol. 276, 307–326.

32. Matthews, B. W. (1968). Solvent content of proteincrystals. J. Mol. Biol. 33, 491–497.

33. Kantardjieff, K. A. & Rupp, B. (2003). Matthewscoefficient probabilities: improved estimates for unitcell contents of proteins, DNA, and protein–nucleicacid complex crystals. Protein Sci. 12, 1865–1871.

34. 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.

35. Emsley, P. & Cowtan, K. (2004). Coot: model-buildingtools for molecular graphics. Acta Crystallogr. Sect. D,60, 2126–2132.

36. Adams, P. D., Afonine, P. V., Bunkoczi, G., Chen,V. B., Davis, I. W., Echols, N. et al. (2010). PHENIX:a comprehensive Python-based system for macromo-lecular structure solution. Acta Crystallogr. Sect. D, 66,213–221.

37. Lanzetta, P. A., Alvarez, L. J., Reinach, P. S. & Candia,O. A. (1979). An improved assay for nanomoleamounts of inorganic phosphate. Anal. Biochem. 100,95–97.

38. Carter, S. G. & Karl, D.W. (1982). Inorganic phosphateassay with malachite green: an improvement andevaluation. J. Biochem. Biophys. Methods, 7, 7–13.

39. Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G.,Bhat, T. N., Weissig, H. et al. (2000). The Protein DataBank. Nucleic Acids Res. 28, 235–242.