Crystal Structure of the Ternary Complex of theCatalytic Domain of Human PhenylalanineHydroxylase with Tetrahydrobiopterin and3-(2-Thienyl)-L-alanine, and its Implications for theMechanism of Catalysis and Substrate Activation

Ole Andreas Andersen1, Torgeir Flatmark2 and Edward Hough1*

1Department of ChemistryUniversity of Tromsø, N-9037Tromso, Norway

2Department of Biochemistryand Molecular BiologyUniversity of BergenÅrstadveien 19, N-5009 BergenNorway

Phenylalanine hydroxylase catalyzes the stereospecific hydroxylation ofL-phenylalanine, the committed step in the degradation of this aminoacid. We have solved the crystal structure of the ternary complex(hPheOH–Fe(II)·BH4·THA) of the catalytically active Fe(II) form of a trun-cated form (DN1–102/DC428–452) of human phenylalanine hydroxylase(hPheOH), using the catalytically active reduced cofactor 6(R)-L-erythro-5,6,7,8-tetrahydrobiopterin (BH4) and 3-(2-thienyl)-L-alanine (THA) as asubstrate analogue. The analogue is bound in the second coordinationsphere of the catalytic iron atom with the thiophene ring stacking againstthe imidazole group of His285 (average interplanar distance 3.8 Å) andwith a network of hydrogen bonds and hydrophobic contacts. Binding ofthe analogue to the binary complex hPheOH–Fe(II)·BH4 triggers structuralchanges throughout the entire molecule, which adopts a slightly morecompact structure. The largest change occurs in the loop region compri-sing residues 131–155, where the maximum r.m.s. displacement (9.6 Å) isat Tyr138. This loop is refolded, bringing the hydroxyl oxygen atom ofTyr138 18.5 Å closer to the iron atom and into the active site. The iron geo-metry is highly distorted square pyramidal, and Glu330 adopts a confor-mation different from that observed in the hPheOH–Fe(II)·BH4 structure,with bidentate iron coordination. BH4 binds in the second coordinationsphere of the catalytic iron atom, and is displaced 2.6 Å in the directionof Glu286 and the iron atom, relative to the hPheOH–Fe(II)·BH4 structure,thus changing its hydrogen bonding network. The active-site structure ofthe ternary complex gives new insight into the substrate specificity of theenzyme, notably the low affinity for L-tyrosine. Furthermore, the structurehas implications both for the catalytic mechanism and the molecular basisfor the activation of the full-length tetrameric enzyme by its substrate. Thelarge conformational change, moving Tyr138 from a surface position intothe active site, may reflect a possible functional role for this residue.

q 2002 Elsevier Science Ltd. All rights reserved

Keywords: phenylalanine hydroxylase; tetrahydrobiopterin; thienylalanine;conformational change; protein crystallography*Corresponding author

Introduction

The non-heme iron enzyme phenylalaninehydroxylase (PheOH, phenylalanine 4-monooxy-genase, EC 1.14.16.1) catalyzes the hydroxylationof the essential aromatic amino acid L-phenyl-alanine (L-Phe) to L-tyrosine (L-Tyr) in the presenceof the specific pterin cofactor 6(R)-L-erythro-5,6,7,8-tetrahydrobiopterin (BH4) and dioxygen. The reac-tion is the rate-limiting step in the degradation of

0022-2836/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved

E-mail address of the corresponding author:edward@james.chem.uit.no

Abbreviations used: 4a-OH-BH4, 4a-hydroxy-tetra-hydrobiopterin; BH2, L-erythro-7,8-dihydrobiopterin;BH4, 6(R)-L-erythro-5,6,7,8-tetrahydrobiopterin; HPA,hyperphenylalaninemia; hPheOH, human phenylalaninehydroxylase; hTyrOH, human tyrosine hydroxylase;L-Tyr, L-tyrosine; L-Phe, L-phenylalanine; PheOH,phenylalanine hydroxylase; PKU, phenylketonuria;rPheOH, rat phenylalanine hydroxylase; rTyrOH, rattyrosine hydroxylase; THA, 3-(2-thienyl)-L-alanine;TyrOH, tyrosine hydroxylase.

doi:10.1016/S0022-2836(02)00560-0 available online at http://www.idealibrary.com onBw

J. Mol. Biol. (2002) 320, 1095–1108

L-Phe to carbon dioxide and water.1 Inborn errorsthat reduce or destroy the activity of the enzymeare responsible for the human autosomal recessivedisease phenylketonuria (PKU)/hyperphenylalani-nemia (HPA). The disease causes elevated concen-trations of L-Phe in the blood, which can impairthe normal development of the brain and causesevere mental retardation. In most of Europe,approximately 1 in 10,000 live births reportedlyhas the disorder2 and more than 400 differentmutations are associated with PKU/HPA.3† Mostof the mutations are found in the catalytic domain3

and they demonstrate different clinical, metabolicand enzymatic phenotypes.4,5 Recent crystallo-graphic studies on human phenylalaninehydroxylase (hPheOH)6 – 9 and rat phenylalaninehydroxylase (rPheOH)10 have made it possible todefine the structural phenotypes of the differentgenotypes.11 A limitation in the assignment of thestructural phenotypes has been that they have

been based on the structures of catalytically inac-tive Fe(III) forms of the enzyme, which also lackstructural information on substrate binding.Following our recently solved crystal structureof the catalytically active Fe(II) form of the trun-cated form DN1–102/DC428–452-hPheOH and itsbinary complex with the reduced pterin cofactor(BH4),

12 we now present the crystal structure ofa ternary complex with the substrate analogue3-(2-thienyl)-L-alanine (THA). THA is a substratefor rPheOH13 and hPheOH14 and binds competi-tively to L-Phe at the active site.15 Binding of thesubstrate analogue also triggers a conformationalchange similar to that observed upon binding ofL-Phe.14,16,17

The structure reveals the binding sites of thepterin cofactor and the substrate under near-turn-over conditions, i.e. in the absence of dioxygen,and provides new insights into the substrate speci-ficity and catalytic mechanism of the enzyme. Itshows that substrate binding triggers a substantialstructural change in the catalytic domain, particu-larly in the active-site region. This change may

Figure 1. Stereo picture of the electron density at (a) the THA-binding site and (b) the BH4-binding site in the ternarycomplex. Blue electron density is from s-weighted 2Fo 2 Fc maps at 1.2s while red omit electron density is froms-weighted Fo 2 Fc maps at 2.7s (a) omitting THA and the side-chain of Tyr138 and (b) omitting BH4, the side-chainof Glu330 and Wat2. The Figure was produced using BOBSCRIPT.57

† http://data.mch.mcgill.ca/pahdb_new/

1096 Ternary Complex of Phenylalanine Hydroxylase

http://data.mch.mcgill.ca/pahdb_new/

represent the “epicenter” of the global confor-mational transition and catalytic activation thatoccurs in the full-length tetrameric enzyme uponsubstrate binding.18

Results

Well defined crystals of the binary hPheOH–Fe(II)·BH4 complex were treated with the substrateanalogue THA by adding solid THA to the crystal-lization drops. The whole procedure of crystalli-zation, post-crystallization diffusion soaking inTHA, flash-cooling in liquid nitrogen and mount-ing of crystals was carried out anaerobically asdescribed.12 When observed in the microscope, thecrystals appeared to be unaffected by the THAsoaking, but the diffraction pattern revealed amosaicity that was two to three times higher thanthat of the “binary” crystals. All three axes of theunit cell were 1–2 Å shorter than for the binarycomplex, and useable data were obtained to 2.5 Å.It should be noted, however, that the binary crys-tals of the enzyme were found to deform/disinte-grate when exposed to L-Phe. Data collected forcofactor-free crystals soaked in L-Phe revealed thesame high level of mosaicity as that observed forhPheOH–Fe(II)·BH4·THA but processing of thesedata was unsuccessful. The reason why we havebeen unsuccessful using L-Phe as substrate is not

clear, but does suggest that possible lattice or struc-tural changes may simply have exceeded thattolerable within the crystals. Co-crystallization ofhPheOH–substrate/substrate-analogue complexesusing previously known crystallization con-ditions19 failed, as did experiments to find newcrystallization conditions for such a complex.

The structure was refined to a final Rwork and Rfreeof 22.0% and 26.7%, respectively. The mean error ofthe atomic positions was determined to 0.29 Åusing the sA method.

20 The final model contains307 residues, a ferrous iron, 39 water molecules,one THA molecule and the reduced cofactor(BH4). The electron density for both THA and BH4is very good (Figure 1(a) and (b)), as is the electrondensity for most of the amino acid side-chains withthe exception of the loop residues 130–134 (seebelow) and a few surface-located side-chains.

Binding of the substrate analogue 3-(2-thienyl)-L-alanine (THA)

All atoms of the substrate analogue THA arewell defined, consistent with the low B-factors(below 22 Å2) estimated during refinement. THAbinds in the second coordination sphere of thecatalytic iron atom, with the five-membered thio-phene ring packing against the imidazole group ofthe iron ligand His285 (Figures 1(a) and 2) with anaverage interplanar distance of 3.8 Å. This distance

Figure 2. An illustration of the BH4 and THA-binding sites in the ternary complex. The Figure was produced usingLigplot,58 and edited using CorelDRAW 9.0.

Ternary Complex of Phenylalanine Hydroxylase 1097

is similar to that observed for the phenyl ring ofL-Phe in the ternary complex of hPheOH as deter-mined by a combined nuclear magnetic reso-nance (NMR) and molecular docking analysis.15

However, the hydrogen bonding pattern of themain-chain THA is somewhat different from thatfound for L-Phe in that structure. Thus, the aminoN forms a water-mediated hydrogen bond toTyr277 (3.2 and 3.1 Å), and hydrogen bonds toThr278 O (2.8 Å) and a water molecule (2.8 Å),which in turn is hydrogen bonded to Gly346 O(2.8 Å), Glu353 O12 (2.7 Å) and possibly Ser350 Og

(3.4 Å). THA OT1 is hydrogen bonded to Arg270Nh1 (3.3 Å), Thr278N (3.0 Å) and possibly Thr278O (3.1 Å) while THA OT2 is hydrogen bonded toArg270 Nh2 (2.9 Å), Ser349 Og (2.4 Å) and possiblySer349 O. Hydrophobic contacts are formed fromGly346 Ca (3.8 Å) and Phe331 Cz (3.7 Å) to THACd, from Phe331 Cz (3.8 Å) to THA C12 and fromSer350Cb to THA C (3.5 Å) (Figure 2).

Binding of the substrate analogue 3-(2-thienyl)-L-alanine (THA) triggers large-scalestructural changes

All previous crystal structures of the binary com-plexes of the double truncated form (“catalyticdomain”) of hPheOH8,9,12 can be superimposedonto the non-liganded structure6 (PDB entry1PAH) with r.m.s. deviations for main-chain atomsof between 0.21 and 0.31 Å. By contrast, a super-position of the ternary complex hPheOH–Fe(II)·BH4·THA onto the non-liganded structurereveals an r.m.s. deviation for main-chain atoms of2.2 Å, and superposition onto other hPheOHstructures8,9,12 gives similar r.m.s. values. Thus, theternary complex structure is significantly differentfrom the substrate-free structures and demon-strates that substrate-analogue binding triggerslarge-scale structural changes. The ternary enzymeis slightly smaller compared to the binary and

ligand-free structures of hPheOH due to a morecompact packing. The average atomic distance tothe centre of mass is 17.9 Å for the THA-boundstructure compared to 18.1 Å for the binary andligand-free structures. The superposition of theternary hPheOH–Fe(II)·BH4·THA complex ontothe binary hPheOH–Fe(II)·BH4 complex

12 (PDBentry 1J8U) is shown in Figure 3. In addition togeneral adjustments throughout the wholestructure, part of the chain comprising residues131–155 (mostly loop residues) is substantiallyrefolded. The largest displacement is observed forTyr138 (r.m.s. displacement of 9.6 Å). The hydroxylin this residue is displaced by 20.7 Å to a partiallyburied position in the active site (Figure 3),with its Oh only 6.5 Å away from the iron atom,5.7 Å away from BH4 C4a and 3.7 Å away fromTHA C11. The phenol ring is packed betweenLeu248 (closest distance 3.6 Å) and Val379 (closestdistance 3.4 Å) forming a hydrophobic cluster(Figure 4), and Tyr138 Oh forms an intramolecularhydrogen bond to a water molecule, which is alsohydrogen bonded to O20 in BH4 (Figure 2). In thesubstrate-free crystal structures of hPheOH,6 – 9,12

Tyr138 is located on the surface of the proteinwith a solvent-exposed side-chain and appears tohave no specific importance, except for a possiblecontribution to protein stability.21 Tyr277 adopts aconformation slightly different from that in thesubstrate-free structures of hPheOH and itshydroxyl oxygen atom is displaced 6.3 Å comparedwith the binary hPheOH–Fe(II)·BH4 complex toform a possible water-mediated hydrogen bondwith THA N.

Displacement of the pterin cofactor uponsubstrate binding

In addition to the large motion of the 131–155region, including the reorientation of Tyr138 froma surface position to a location in the active site, a

Figure 3. Stereo picture of the ternary hPheOH–Fe(II)·BH4·THA complex (blue/cyan) superimposed on the binaryhPheOH–Fe(II)·BH4 complex

12 (red/orange) (PDB entry 1J8U). Black ball-and-stick models of BH4, THA and iron areshown for the ternary structure. BH4 in the binary structure is omitted for clarity. The highest r.m.s. displacement(see the text) occurs in loop residues 129–146 coloured cyan in the ternary complex and orange in the binary complex.The r.m.s. maximum is located at Tyr138. The Figure was produced using MOLSCRIPT.59

1098 Ternary Complex of Phenylalanine Hydroxylase

significant displacement was observed for thepterin cofactor BH4 (Figures 2 and 5). Iron to pterin(C4a, O4 and N5) distances for all crystal and NMRstructures of PheOH and tyrosine hydroxylase(TyrOH) are compared in Table 1. The pterin-bind-ing site of the binary hPheOH–Fe(II)·BH4 is similarto that found in the crystal structure of the binarycomplex (hPheOH–Fe(III)·BH2) of hPheOH withthe oxidized L-erythro-7,8-dihydrobiopterin (BH2)cofactor (PDB entries 1LRM and 1DMW). Theorientation of the pterin cofactor in these two struc-tures is similar to that of the hPheOH–Fe(III)·BH2·-L-Phe NMR and molecular docking structure andthe present hPheOH–Fe(II)·BH4·THA structure.However, the pterin cofactor is displaced signifi-cantly compared to the binary complexes. Super-imposing the binary hPheOH–Fe(II)·BH4 complexon the ternary hPheOH–Fe(II)·BH4·THA complexbased on conserved active-site residues (His285,

His290 and Fe) revealed a mean displacement of2.6 Å for BH4 in the direction of Glu286 and ironupon THA binding (Figure 5). Iron distances areshortened from 5.9, 3.8 and 5.7 Å for C4a, O4 andN5, respectively, in the crystal structure of binaryhPheOH–Fe(II)·BH4 complex to 4.5, 3.4 and 3.7 Åin the current THA ternary structure (Table 1).However, the pterin cofactor is still not coordi-nated directly to the iron atom. This contrasts withthe combined NMR and molecular modelingstudies on hPheOH, in which the distance betweenBH2 O4 and the iron atom was estimated to be2.6 Å, and thus compatible with direct coordi-nation to the iron atom.15 However, this NMR/molecular modeling structure of cofactor and sub-strate bound at the active site was modeled intothe rigid structure of the non-liganded doubletruncated form of hPheOH and thus did nottake into account any bound water molecules or

Figure 5. Stereo picture of the binding site of BH4 in the ternary hPheOH–Fe(II)·BH4·THA structure. Side-chains forLeu248 and Leu249 are omitted for clarity. All potential hydrogen bonds to the pterin moiety are shown as dotted lines.The green model of BH4 illustrates its position in the binary hPheOH–Fe(II)·BH4 complex

12 when superimposed on theternary structure using conserved active-site residues (His285, His290 and Fe). The Figure was produced usingMOLSCRIPT.59

Figure 4. Stereo picture of the packing of Tyr138 in the hydrophobic core at the active site. The green model illus-trates the position of residues in the binary hPheOH–Fe(II)·BH4 complex

12 when superimposed on the ternary struc-ture based on conserved active-site residues (His285, His290 and Fe). The Figure was produced using MOLSCRIPT.59

Ternary Complex of Phenylalanine Hydroxylase 1099

possible conformational changes in the active siteassociated with substrate binding. The p-stackinginteractions of the cofactor with Phe254 are,however, similar for all three crystal structures ofhPheOH pterin complexes. The side-chain ofPhe254 in hPheOH–Fe(II)·BH4·THA is displacedabout 1.9 Å in the same direction as the pterindisplacement compared with the binary hPheOH–Fe(II)·BH4 and hPheOH–Fe(III)·BH2 structures,and the average interplanar phenyl-cofactor dis-tance is 3.7 Å. The loop residues 247–251 form thesame pattern of direct hydrogen bonds to thepterin as in the binary hPheOH–Fe(II)·BH4 andhPheOH–Fe(III)·BH2 complexes (except the pterinO20–Ser251 Og bond, which is not present in thepublished hPheOH–Fe(III)·BH2 crystal structure).

8

These residues are thus displaced about 2.6 Å(relative to the active site) in the same direction asthe pterin displacement. Glu286 is not displacedsignificantly and its carbonyl group forms hydro-gen bonds directly to N3 and O4 of BH4; thesetwo connections are water-mediated hydrogenbonds in the binary complexes hPheOH–Fe(II)·BH4and hPheOH–Fe(III)·BH2. BH4 C3

0 forms hydro-phobic contacts with Leu255 Cd1 (3.6 Å), Ser 251Ca (3.8 Å) and Phe254 Cd2 (3.8 Å), while BH4 C8aand C7 forms contacts with Leu248 Cd1 (3.7 and3.8 Å, respectively).

The torsion angle between the hydroxyl groupsin the dihydroxypropyl side-chain of BH4 in theternary complex is similar (2598) to that in thebinary hPheOH–Fe(II)·BH4 complex (2658) andthe ternary hPheOH–Fe(III)·BH2·L-Phe NMRstructure (2608), enabling the BH4 O2

0 to make ahydrogen bond with the side-chain oxygen atomof Ser251. A crystal structure of the binaryhPheOH–Fe(III)·BH2 complex to 2.1 Å resolution(data not shown) was obtained. The electrondensity maps showed unambiguous positions forall cofactor atoms, including the dihydroxypropylside-chain, and revealed that the angle betweenthe hydroxyl groups of the side-chain is 2628.Thus, the same hydrogen bond is formed betweenO20 and Ser251 Og (2.7 Å) as for the hPheOH–Fe(II)·BH4 complex.

The ternary complex has a distorted squarepyramidal, five-coordinated iron atom

Previously determined crystal structures ofamino acid hydroxylases6 – 10,12,22,23 have shown thatthe iron ligands are consistently two histidineresidues, a monodentate glutamic acid residue

and a varying number of water molecules. Theiron atom in the present ternary complex is five-coordinated by two histidine residues, a bidentateglutamic acid residue, and a single water molecule(Wat2) in a highly distorted square pyramidalgeometry with Glu330 O11 as the axial ligand(Figures 1(b) and 2). The ligand–iron distances are2.3 Å (His285), 2.3 Å (His290), 2.6 Å (Glu330 O11),2.4 Å (Glu330 O12) and 2.4 Å (Wat2). Both histidineresidues and the glutamic acid residue have goodelectron densities with low B-factors (below 32Å2). The water ligand has a slightly higher B-factor(44 Å2) and is displaced about 0.9 Å compared toWat2 in the binary hPheOH–Fe(II)·BH4 complex.No density appeared for either Wat1 or Wat3,which is in conformity with the 1H NMR studieson full-length hPheOH, suggesting that at leastone of the coordinating water molecules is dis-placed from coordination upon the bindingof L-Phe at the active site.24 Magnetic circulardichroism (MCD) studies on rPheOH further sup-ports a five-coordinate square pyramidal Fe(II) siteupon addition of pterin in the presence of L-Phe.25

Discussion

The present crystal structure of the ternary com-plex hPheOH–Fe(II)·BH4·THA has given valuablenew information related to the question of the sub-strate-binding site, the substrate specificity and theconformational transition (hysteresis) that occursin the enzyme upon substrate binding. Further-more, the structure has important implications forthe catalytic mechanism and defines clearly theamino acid residues of the active-site crevice struc-ture that are involved in the binding of pterincofactor and substrate under near-turnover con-ditions (in the absence of dioxygen) as well asproviding a structural explanation for the disease-associated PKU/HPA mutations related to thesebinding sites.

The substrate specificity and substrate-binding site

The substrate specificity of PheOH has beenstudied extensively by Kaufman.16 For maximumactivity, a substrate including an unmodified ala-nine residue must be attached to an aromatic ringthat may contain a number of subsitutions andstill be hydroxylated as long as the alanine part isintact. Of particular interest was the finding that

Table 1. Comparison of metal to pterin distances (Å)

Phenylalanine hydroxylase Tyrosine hydroxylase

X-ray hPheOH–Fe(II)·BH4·THA

X-ray hPheOH–Fe(II)·BH4

12

X-ray hPheOH–Fe(III)·BH2

8

NMR hPheOH–Fe(III)·BH2·L-Phe

15

X-ray rTyrOH–Fe(III)·BH2

22

NMR hTyrOH–Fe(III)·BH2·L-Phe

60

Fe–C4a 4.5 5.9 6.1 4.3 5.6 3.6–4.1Fe–O4 3.4 3.8 3.8 2.6 ^ 0.3 3.6 3.3–4.1Fe–N5 3.7 5.7 6.1 4.4 ^ 0.4 5.4 3.3–3.8

1100 Ternary Complex of Phenylalanine Hydroxylase

3-(2-thienyl)-L-alanine (THA) is hydroxylated byPheOH14,16 and that this analogue induces acomprehensive global conformational transition(and activation of the enzyme) similar toL-Phe,14,16,17 although it has a slightly lower affinityof binding than L-Phe.14 Interestingly, recent NMRstudies on the double truncated form DN1–102/DC428–452-hPheOH have demonstrated thatL-Phe bound at the active site is displaced byTHA.15 Since THA binds competitively toL-Phe,16,17 the present structure allows us to modelthe physiological substrate L-Phe into the activesite (Figure 6(a)) assuming the position of themain chain of the substrates and the orientation ofthe ring structure (x1 and x2 angles) to be con-served. In this model, the phenyl group ispositioned appropriately in the hydrophobiccluster, 3.6 Å from the phenyl group of Phe331,3.7 Å from the side-chain of Trp326 as well as3.7 Å from the Ca atoms of Gly346 and Pro281.The side-chains of Phe331 and Trp326 are bothdisplaced (about 2 Å and 3 Å, respectively) uponbinding of substrate, resulting in hydrophobiccontacts with its phenyl group. Thus, the modeledL-Phe was found to interact with four main resi-dues at the active site, i.e. Arg270, Thr278, His285and Ser349, as well as with Tyr277, Pro281,Trp326, Phe331, Gly346, Ser350 and Glu353, andinterestingly, human missense single-pointmutations related to PKU/HPA have been

reported for eight of these residues (Table 2). Itshould be noted that one of the interactions withArg270, the interaction with the side-chain ofSer349 and the interactions with Pro281 andHis285 are present also in the NMR/moleculardocking structure,15 and the displacement ofTrp326 to accommodate hydrophobic contactswith L-Phe was predicted in that study, whereasthe other observed interactions are unique for thecrystal structure.

Furthermore, studies on chimeric forms ofpterin-dependent aromatic amino acid hydroxyl-ases have revealed that their substrate specificityis determined by the catalytic domain and thatnone of the chimeric enzymes, containing the cata-lytic domain of PheOH, were able to hydroxylateL-Tyr.26 L-meta-tyrosine (L-mTyr), which is a sub-strate for PheOH,16 was modeled into the bindingsite in the same manner as L-Phe (see above)(Figure 6(a)). Its hydroxyl oxygen atom couldform a hydrogen bond (3.0 Å) with Tyr138 Oh andits nearest carbon atom is Tyr138C11 (3.2 Å). Onthe other hand, when L-Tyr (L-pTyr) was similarlymodeled into the active site, its oxygen atom wasonly 2.5 Å away from the nearest carbon atom inthe side-chain of Trp326. This steric hindrancecould be minimized slightly by rotating the side-chain x-values (while conserving the position ofthe main-chain). The best manual fit was found byrotating x1 , 98, bringing the hydroxyl oxygen

Figure 6. Stereo picture ofthe substrate-binding site with(a) THA (green), the modelledL-Phe (red phenyl group) andL-mTyr (red phenyl groupand blue hydroxyl oxygen atom)and (b) L-Tyr (blue; see the text).The Figure was produced usingMOLSCRIPT.59

Ternary Complex of Phenylalanine Hydroxylase 1101

atom 3.0 Å away from carbon atoms in both Trp326and Glu330 (Figure 6(b)). Whereas L-Phe and, to acertain degree, L-mTyr, are well accommodated inthe present active-site structure, and with specificinteractions similar to that of THA, the hydroxylgroup of L-Tyr is not accepted, due to sterichindrance. The resulting strained binding site forL-Tyr may well explain why this amino acid is nota substrate, but is an appropriate leaving product.Thus, the crystal structure is in complete agree-ment with the substrate specificity observed insteady-state kinetics.16

Site-directed mutagenesis has been performedon TyrOH to identify residues responsible forsubstrate binding.27 Arg316 (Arg270 in PheOH)was shown to be critical for substrate bindingwith a 400-fold higher Km value for the Arg316Lysmutation, whereas the Asp328Ser (Asp282 in

PheOH) mutation showed a 26-fold higher Kmvalue. In the present crystal structure, the Arg270side-chain forms a salt-bridge with the carboxylgroup of THA (and L-Phe), explaining the criticalrole of this residue in substrate binding. Asp282does not bind to the substrate, but its carboxylgroup forms a salt-bridge with Arg270 and is thusimportant for substrate binding by providingstability and correct positioning of Arg270.

Conformational changes at the active siteupon pterin cofactor and substrate binding

A comparison of the non-liganded and thebinary complexes of the double truncated formDN1–102/DC428–452-hPheOH with oxidized(BH2) and reduced (BH4) cofactor have revealedsome important conformational changes of theactive-site structure upon cofactor binding. Thus,in the hPheOH–Fe(III)·BH2 complex,

8 the loopbetween residues 245 and 250 shows the largestdisplacement (the Ca atom of Gly247 moves,1.3 Å toward the pterin ring), in the direction ofthe iron atom, and thus is able to form severalhydrogen bonds to the pterin ring. Furthermore,the Leu248 side-chain changes its conformation ascompared to the non-liganded structure, and nowfaces the active site. Leu255 also shifts its confor-mation to accommodate the dihydropropyl side-chain of the pterin molecule. In the hPheOH–Fe(II)·BH4 structure,

12 the overall fold is verysimilar to that reported for hPheOH–Fe(III)·BH2.However, superposition of the two structures hasrevealed that the reduced cofactor is displacedabout 0.5 Å away from Ser251, and that the pterinring is rotated about 108 (along the C4a–C8abond) with the pyrimidine ring rotated towardsPhe254. The angle between the hydroxyl groups inthe dihydroxypropyl group is 2658, which enablesthe BH4 O2

0 to make a strong hydrogen bond(2.4 Å) with the side-chain oxygen atom of Ser251,while O10 forms water-mediated hydrogen bondsbetween residues Ala322 and Glu330. In addition,Glu330 adopts a completely different conformationin the hPheOH–Fe(II)·BH4 structure. On the basisof these crystallographic data, it is evident that inboth oxidation states of the enzyme and cofactorthe pterin cofactor binds to the enzyme active siteby an induced-fit mechanism involving a confor-mational change in the active-site crevice structureof the protein.

In the present study, it is further demonstratedthat the binding of the substrate analogue THA tothe binary complex hPheOH–Fe(II)·BH4 triggerslarge additional conformational changes at theactive-site crevice structure. The largest deviationoccurs in the region comprising the residues 131–155 with a maximum main-chain r.m.s. displace-ment (9.6 Å) at Tyr138. The hydroxyl oxygen atomof this residue is indeed displaced 18.5 Å closerto the catalytic iron atom. Furthermore, Glu330adopts a bidentate coordination and conformationnot observed previously in any crystal structure of

Table 2. Amino acid residues in the active-site crevicestructure involved in the binding of pterin cofactor andsubstrate and missense single-point mutations that havebeen reported in these residues in human hyperphenyl-alaninemias

Residue Mutations MutNoa Comment Reference

BH4 Tyr138 NoGly247 G247V 229 4% r.a.b 44Leu248 L248R 230 – –

L248P 231 – –Leu249 L249F 232 – –

L249H 233 – –Ser251 NoPhe254 F254I 237 – –Leu255 L255V 238 11% r.a. 44, 45

L255S 239 1% r.a. 45His264c H264L 250 – –Glu286 NoAla322c A322T 314 – –

A322G 315 75% r.a. 46Tyr325c Y325C 324 PKUd –Glu330c E330D 328 PKUe –

L-Phe Arg270 R270K 257 – –R270S 258 2% r.a. 45

Tyr277 Y277D 268 – –Y277C 269 – –

Thr278 T278A 270 – –T278N 271 – –T278I 272 – –

Pro281 P281L 275 ,1% r.a. 47His285 NoTrp326 NoPhe331 F331L 329 – –

F331C 330 – –Gly346 G346R 350 – –Ser349 S349P 355 ,1% r.a. 48, 49

S349L 356 – –Ser350 S350T 358 – –Glu353 No

a Mutation number in the data base (http:/data.mch.mcgill.ca/pahdb_new/

b The abbreviation % r.a. represents the residual activity ofthe recombinant enzyme as a percentage of the wild-typeexpressed enzyme.

c Binds BH4 only in the binary complex via a water bridge.d Classic PKU, with genotype Y325C/L348V.e Classic PKU, with genotype E330D/R408W.

1102 Ternary Complex of Phenylalanine Hydroxylase

http:/data.mch.mcgill.ca/pahdb_new/http:/data.mch.mcgill.ca/pahdb_new/

PheOH. Thus, the binding of both the pterincofactor and the substrate induces conformationalchanges at the active site that have not beendetected by any alternative biophysical method.

Functional implications

The PheOH-catalyzed hydroxylation of L-Phe isa three-substrate reaction with specific bindingsites for L-Phe, BH4 and dioxygen, and there isgeneral agreement that the Fe(II) centre partici-pates directly in oxygen incorporation.21,28,29 In thepresent structure of the unproductive (anaerobic)ternary complex, the binding sites for L-Phe andBH4 and the iron coordination are clearly defined.The structure has revealed a five-coordinated ironatom with a distorted square-pyramidal coordi-nation, as proposed previously on the basis ofMCD spectral analysis of rPheOH.25 This findingis consistent with the ordered reaction mechanismproposed for the enzyme, wherein cofactor andsubstrate must be present before any product isreleased.28,30 The position of dioxygen binding tothe iron is not established unequivocally, andattempts to bind NO and CO (by diffusion intocrystals of the ternary complex) have not been suc-cessful so far. Dioxygen may bind either into theposition occupied by Wat2 (as shown in Figure7(b)) or into the open coordination position, asfavoured in a recent model.30 In the catalytic reac-tion, dioxygen is cleaved, incorporating one of theoxygen atoms into BH4 to form 4a-hydroxy-tetra-hydrobiopterin (4a-OH-BH4),

31,32 while the otheroxygen atom is incorporated into the substrate togenerate L-Tyr. L-Phe/L-Tyr were modeled manu-ally into the THA-binding site by conserving themain-chain and the x1 and x2 angles (Figure 7(a))and 4a-OH-BH4 (calculated using the perturbativeBecke–Perdew model (pBP/DNp p )33,34 of the PCSPARTAN PRO programme package) was super-imposed on BH4 (Figure 7(a)) to simulate thepositions of the oxygen atoms of dioxygen afterthe reaction but prior to product and cofactorrelease. On the basis of their relative positions,dioxygen was modeled manually into the activesite, positioning one of the atoms (proximal oxy-gen) at Wat2 and the other atom (distal oxygen) inthe direction of the hydroxyl oxygen atom in4a-OH-BH4 (Figure 7(b)). The proximal oxygenatom is then 3.3 Å from L-Tyr Oh and 3.0 Å fromthe closest carbon atom in THA/L-Phe, while thedistal oxygen atom is 0.9 Å from the 4a-OH-BH4hydroxyl oxygen atom and 2.2 Å from BH4 C4a.However, modeling dioxygen in the same mannerin the open coordination position (at the positionof Wat3 in the binary hPheOH–Fe(II)·BH4complex)12 leaves the proximal oxygen atom 5.6 Åfrom L-Tyr Oh and 5.1 Å from the closest carbonatom in THA/L-Phe, while the distal oxygen atomis 1.8 Å from the 4a-OH-BH4 hydroxyl oxygenatom and 1.9 Å from BH4 C4a. Binding of dioxygenat the open coordination position seems unlikely,since BH4 O4 is close (,1.0 Å). However, the

proposed dioxygen-binding site at the Wat2position is in a hydrophobic microenvironmentconsisting of His285, BH4, THA/L-Phe and Pro281with the side-chain of Pro281 3.4 Å from the distaloxygen atom. It is important to note here that theP281L mutation is associated with severe PKU.35,47

A displacement of the BH4 molecule upon sub-strate binding has been proposed to accommodatea possible Fe(II)-peroxo-BH4 intermediate,

12,36 andis indeed confirmed in the present structure(Figure 5). The crystallographic data are entirelyconsistent with the occurrence of a large-scale con-formational transition that brings the two sub-strates into the close proximity required forreaction. Thus, the pterin C4a–Fe(II) distance of4.5 Å in the ternary complex is far more appropri-ate for the formation of a bridging dioxygenmolecule between the iron atom and BH4 (as aputative Fe(II)–O–O–BH4 intermediate) than theC4a–Fe(II) distance of 5.9 Å in the binary BH4structure, implying that the bridging dioxygenintermediate can be formed only after substratebinding. On this basis, the following ordered reac-tion mechanism can be proposed (Figure 8),wherein cofactor and substrate are bound beforeany product is released. Although little is knownabout the mechanism of reduction,28 prereductionof the active-site iron (step 1) is an obligate eventprior to catalysis.37 Significant changes occur inthe active site upon reduction, including a reducedaffinity for two of the coordinated water molecules(Wat1 and Wat2), a displacement of Glu330 andpossible disorder of its side-chain.12 The reversiblebinding of BH4 (step 2) changes the overall coordi-nation geometry and causes the Glu330 ligand tochange its coordination to the iron atom.12 Thereversible binding of L-Phe (step 3) triggers afurther conformational change altering the ironcoordination to a highly distorted square pyrami-dal geometry where Glu330 adopts yet anotherconformation with bidentate iron coordination(Figure 2). Furthermore, the position of BH4 in theactive site is altered (Figure 5), favouring dioxygenbinding at the position occupied by Wat2 (Figure7(b)) and the formation of a putative Fe(II)–O–O–BH4 (4a-peroxy-BH4) intermediate

25,38 (step 4).Although the molecular mechanism of dioxygenactivation is still an unsolved question, a hetero-lytic cleavage of the oxygen–oxygen bond30,38,39

produces a molecule of 4a-OH-BH4 and an oxidi-zing species, the so-called activated oxygen inter-mediate (most likely an oxyferryl species) (step 5),leading finally to release of the products (step 6).Experimental evidence has been presented thatthe formation of the hydroxylating intermediate isthe rate-limiting step in the tyrosine hydroxylase-catalyzed reaction.39,40,63

In the non-heme iron enzyme extradiol dioxy-genase, a tyrosine residue at the active site hasbeen proposed to stabilize a radical intermediatein the catalytic cycle.41,42 In hPheOH, Tyr325 hasbeen considered to have a similar function,6 butsite-directed mutagenesis revealed that it plays no

Ternary Complex of Phenylalanine Hydroxylase 1103

direct role in the catalytic reaction.8 It remains tobe seen if the substrate-induced reorientation ofTyr138 gives this residue a similar function. Alter-natively, Tyr138 may contribute to determine thesubstrate specificity, since this is a phenylalanineresidue (Phe214) in TyrOH. It is interesting to notethat a tyrosine residue is present at the equivalentposition of tryptophan hydroxylase. Another possi-bility is that Tyr138 plays an important role in theregulation of the enzyme.

The relation between the substrate-inducedconformational change in the catalytic domainand that observed in the full-length tetramericwild-type enzyme

Although there is general agreement about theimportance of L-Phe (and some substrate ana-logues) for the activation of the tetrameric full-length wild-type enzyme,4,18 determination of themechanism by which the substrate induces therelated global conformational change has remainedelusive.4,18 In the present study, it has been shownthat binding of the substrate at the active site trig-gers large-scale structural changes in the catalytic

domain, including the active-site crevice structure,which are likely to represent the epicenter of theglobal conformational change observed in the full--length tetrameric enzyme, and thus delineate amolecular mechanism for substrate-activation ofthe tetramer. The conformational changes allowthe cofactor and substrate to access the active siteduring enzyme turnover more freely.10,43 It isnotable that, so far, hPheOH has not been crystal-lized in the full-length form, and that furtherstudies on this enzyme form are required todemonstrate how the conformational change inthe catalytic domain may be transmitted to otherparts of the enzyme.

Structural insight into the effect of single-pointmutations of active-site residues

Mutations in the human gene encoding thePheOH enzyme result in the autosomal recessivelyinherited disease PKU/HPA, and more than400 mutations have so far been identified†. The

Figure 7. (a) Stereo picture of the active site of the ternary structure. THA is replaced by the modelled L-Phe andL-Tyr (green), and BH4 is replaced by the superpositioned hydroxylated cofactor (4a-OH-BH4; blue). (b) Stereo pictureof the active site of the ternary structure with a proposed binding site of dioxygen (red). One of the oxygen atoms ispositioned at Wat2, while the other atom is positioned in the direction of the C4a attached hydroxyl oxygen atom in4a-OH-BH4 when superimposed on BH4. The Figure was produced using MOLSCRIPT.

59

† http://data.mch.mcgill.ca/pahdb_new/

1104 Ternary Complex of Phenylalanine Hydroxylase

http://data.mch.mcgill.ca/pahdb_new/

resulting metabolic and clinical phenotypes rangein severity from mild forms of non-PKU HPA tothe severe forms of PKU. The structural basis ofsome of the metabolic and enzymatic phenotypeshas been discussed recently11 in the light of thefour crystal structures of inactive Fe(III) forms ofthe enzyme, but without information on the modeof substrate binding and associated conformationalchanges in the catalytic domain. Table 2 presentsa summary of the current information on thereported missense single-point mutations ofactive-site residues, involved in the binding of thereduced pterin cofactor and the substrate. Of thesemutations, 13 are found in nine residues involvedin cofactor binding, and 14 mutations in are foundin eight residues involved in substrate binding.Only a few of these mutations have been expressedas recombinant enzymes.5,44 – 49 As expected, muta-tions of residues at the pterin cofactor andsubstrate-binding sites result in reduced affinitiesfor BH4 and L-Phe, respectively, and thus reducedcatalytic efficiency in steady-state kinetic analyses.

In addition, some of these mutations affect theoverall structure and stability of the enzyme.5,45

Materials and Methods

Crystallization and data collection

Expression and purification of the double truncatedmutant (DN1–102/DC428–452) of hPheOH were carriedout as described.50,51 Anaerobic co-crystallization ofhPheOH–Fe(II) in complex with BH4 was undertakenessentially as described12 but with some modifications.The drops contained initial concentrations of BH4 of5 mM, and 15 mM sodium dithionite was used as thereducing agent. After four days of growth, solid THAwas added in excess to the drop (anaerobic) and left for24 hours to allow diffusion. A crystal of approximatesize 0.6 mm £ 0.2 mm £ 0.05 mm was flash-frozen inliquid nitrogen and data were collected at 100 K on theSwiss–Norwegian Beamline (BM01) at the EuropeanSynchrotron Radiation Facility (ESRF) in Grenoble(France). A wavelength of 0.800 Å and a MAR345Research imaging plate system were used. Processing of

Figure 8. Reaction pathway of PheOH. The catalytic cycle consists of a reduction of the iron centre [1] which con-verts the six-coordinate, high-spin Fe(III) to a “four-coordinate”, high-spin Fe(II); reversible binding of BH4, whichresults in a six-coordinate Fe(II) and a change in coordination of Glu330 [2]; reversible binding of L-Phe to give a highlydistorted square pyramidal five-coordinated Fe(II), with a bidentate coordination of Glu330 [3]; reversible binding ofdioxygen to give a five-coordinated Fe(II)-O2 intermediate followed by the formation of the putative Fe(II)–O–O–BH4 intermediate [4]; heterolytic cleavage of the oxygen–oxygen bond and the formation of a 4a-OH-BH4 and anoxyferryl species [5], and products are finally released [6].

Ternary Complex of Phenylalanine Hydroxylase 1105

the data was done using DENZO,52 while scaling andmerging were carried out using SCALA in the CCP4program suite.53

Model building and refinement

The refinement was initiated by a rigid-body refine-ment of the catalytic domain of hPheOH–Fe(III)6 (PDBentry 1PAH) using CNS version 1.0:54 10% of the reflec-tions were excluded in the refinement for cross-validation.55 The initial R-factor was 50.7%, suggestinglarge deviations from the starting model, but threerounds of simulated annealing (CNS) dropped theR-factor to 33.6%. At this point, 88 residues (132–149,333–342 and 351–410) were outside electron densityand subsequently removed from the model. The omittedresidues were gradually re-added, in between rounds ofenergy minimization and refinement, in their correctpositions at the ends of the remaining four peptidechains. The model building was done using the graphi-cal programme O56 and sA-weighted

20 2Fo 2 Fc andFo 2 Fc maps calculated in CNS. The electron densitiesfor both BH4 and THA were good even after one roundof simulated annealing, but they were added to themodel after all three rounds of simulated annealing.Water oxygen atoms were added throughout the pro-ceeding refinement as clear 2Fo 2 Fc and Fo 2 Fc den-sities appeared in positions compatible with hydrogenbonds to the protein or other water molecules. Luzatti64

and sA coordinate errors and r.m.s. deviations concerningbond lengths, bond angles, dihedral angles and improperangles were calculated using CNS. Statistics of theprocessing and of the final model is presented in Table 3.

A crystal structure to 2.1 Å of the binary hPheOH–Fe(III)·BH2 complex was obtained (data not shown) byusing known crystallization conditions,8 modified witha high concentration (5 mM) of BH2. This structure (PDBentry 1LRM) was refined by CNS to a final Rwork andRfree of 21.1% and 23.8%, respectively.

Protein Data Bank accession numbers

Atomic coordinates and structure factors have beendeposited at the Protein Data Bank (PDB), ResearchCollaboratory for Structural Bioinformatics (RCSB), withaccession number 1KW0. The coordinates will remainprivileged until 28th January 2003.

Acknowledgments

This work has been supported by grants from theNorwegian Research Council (NFR), the NorwegianCouncil on Cardiovascular Diseases, Rebergs legat, L.Meltzer Høyskolefond, the Novo Nordisk Foundationand the European Commission. We thank Ali SepulvedaMuñoz for expert technical assistance in preparing thebacterial extracts and fusion protein, and the staff of theSwiss–Norwegian Beamlines in Grenoble (France).

References

1. Kaufman, S. (1993). The phenylalanine hydroxyl-ating system. Advan. Enzymol. Relat. Areas. Mol. Biol.67, 77–264.

2. Bickel, H., Bachmann, C., Beckers, R., Brandt, N. J.,Clayton, B. E., Corrado, G. et al. (1981). Neonatalmass screening for metabolic disorders. Eur. J.Pediatr. 137, 133–139.

3. Scriver, C. R., Waters, P. J., Sarkissian, C., Ryan, S.,Prevost, L., Côté, D. et al. (2000). PAHdb: A locus-specific knowledgebase. Hum. Mutat. 15, 99–104.

4. Flatmark, T. & Stevens, R. C. (1999). Structuralinsight into the aromatic amino acid hydroxylasesand their disease-related mutant forms. Chem. Rev.99, 2137–2160.

Table 3. Summary of data-collection and refinement statistics for the hPheOH–Fe(II)·BH4·THA complex

Resolution (Å) 10–2.5 (2.64–2.50)a

Cell dimensions a, b, c (Å) 65.17, 106.74, 123.44Space group C2221No. observations 59,237 (8578)a

No. unique reflections 14,178 (2071)a

Multiplicity 4.2 (4.1)a

Rmerge (%) 9.2 (37.7)a

I/s(I) 7.1 (2.1)a

Completeness (%) 94.4 (95.2)a

Wilson B-factor (Å2) 41.3No. atoms in refinement 2586No. solvent molecules 39Rwork (%) 22.0Rfree (%)

55 (10% of data) 26.7Average B-factor all atoms (Å2) 34.3Average THA B-factor (Å2) 18.6Average BH4 B-factor (Å

2) 31.7r.m.s. deviationsBond lengths (Å) 0.007Bond angles (deg.) 1.12Dihedral angles (deg.) 21.9Improper angles (deg.) 0.87Luzatti r.m.s. coordinate error (Å)64 0.35sA r.m.s. coordinate error (Å)

20 0.29f and c angles in most-favoured regions (%)b 87.3f and c angles in additional allowed regions (%)b 12.7

a Statistics for highest-resolution shell are given in parentheses.b Statistics from Ramachandran plots61 are calculated by PROCHECK.62

1106 Ternary Complex of Phenylalanine Hydroxylase

5. Waters, P. J., Parniak, M. A., Nowacki, P. & Scriver, C.R. (1998). In vitro expression analysis of mutations inphenylalanine hydroxylase: linking genotype tophenotype and structure to function. Hum. Mutat.11, 4–17.

6. Erlandsen, H., Fusetti, F., Martı́nez, A., Hough, E.,Flatmark, T. & Stevens, R. C. (1997). Crystal structureof the catalytic domain of human phenylalaninehydroxylase reveals the structural basis for phenyl-ketonuria. Nature Struct. Biol. 4, 995–1000.

7. Fusetti, F., Erlandsen, H., Flatmark, T. & Stevens, R.C. (1998). Structure of tetrameric human phenyl-alanine hydroxylase and its implications forphenylketonuria. J. Biol. Chem. 273, 16962–16967.

8. Erlandsen, H., Bjørgo, E., Flatmark, T. & Stevens, R.C. (2000). Crystal structure and site-specific muta-genesis of pterin-bound human phenylalaninehydroxylase. Biochemisty, 39, 2208–2217.

9. Erlandsen, H., Flatmark, T., Stevens, R. C. & Hough,E. (1998). Crystallographic analysis of the humanphenylalanine hydroxylase catalytic domain withbound catechol inhibitors at 2.0 Å resolution. Bio-chemistry, 37, 15638–15646.

10. Kobe, B., Jennings, I. G., House, C. M., Michell, B. J.,Goodwill, K. E., Santarsiero, B. D., Stevens, R. C.,Cotton, R. G. H. & Kemp, B. E. (1999). Structuralbasis of autoregulation of phenylalanine hydroxyl-ase. Nature Struct. Biol. 6, 442–448.

11. Erlandsen, H. & Stevens, R. C. (1999). The structuralbasis of phenylketonuria. Mol. Genet. Metab. 68,103–125.

12. Andersen, O. A., Flatmark, T. & Hough, E. (2001).High resolution crystal structures of the catalyticdomain of human phenylalanine hydroxylase in itscalalytically active Fe(II) form and binary complexwith tetrahydrobiopterin. J. Mol. Biol. 314, 279–291.

13. Kaufman, S. & Mason, K. (1982). Specificity of aminoacids as activators and substrates for phenylalaninehydroxylase. J. Biol. Chem. 257, 14667–14678.

14. Stokka, A. J. & Flatmark, T. (2002). 3-(2-Thienyl)-L-alanine as a competitive substrate analogue andactivator of human phenylalanine hydroxylase.In Chemistry and Biology of Pteridines and Folates(Milstien, S., Kapatos, G., Levine, R. A. & Shane, B.,eds.), pp. 109–113, Kluwer Academic Publishers.

15. Teigen, K., Frøystein, N. A. & Martı́nez, A. (1999).The structural basis of the recognition of phenyl-alanine and pterin cofactors by phenylalaninehydroxylase: implications for the catalytic mecha-nism. J. Mol. Biol. 294, 807–823.

16. Kaufman, S. (1987). Phenylalanine 4-monooxygenasefrom rat liver. Methods Enzymol. 142, 3–17.

17. Døskeland, A. P., Døskeland, S. O., Øgreid, D. &Flatmark, T. (1984). The effect of ligands of phenyl-alanine 4-monooxygenase on the cAMP-dependentphosphorylation of the enzyme. J. Biol. Chem. 259,11242–11248.

18. Flatmark, T., Stokka, A. J. & Berge, S. V. (2001). Use ofsurface plasmon resonance for real-time measure-ments of the global conformational transition inhuman phenylalanine hydroxylase in response tosubstrate binding and catalytic activation. Anal. Bio-chem. 294, 95–101.

19. Erlandsen, H., Martı́nez, A., Knappskog, P. M.,Haavik, J., Hough, E. & Flatmark, T. (1997). Crystalli-zation and preliminary diffraction analysis of atruncated homodimer of human phenylalaninehydroxylase. FEBS Letters, 406, 171–174.

20. Read, R. J. (1986). Improved Fourier coefficients formaps using phases from partial structures witherrors. Acta Crystallog. sect. A, 42, 140–149.

21. Pace, C. N., Horn, G., Hebert, E. J., Bechert, J., Shaw,K., Urbanikova, L. et al. (2001). Tyrosine hydrogenbonds make a large contribution to protein stability.J. Mol. Biol. 312, 393–404.

22. Goodwill, K. E., Sabatier, C. & Stevens, R. C. (1998).Crystal structure of tyrosine hydroxylase withbound cofactor analogue and iron at 2.3 Å resolution:self-hydroxylation of Phe300 and the pterin-bindingsite. Biochemistry, 37, 13437–13445.

23. Goodwill, K. E., Sabatier, C., Marks, C., Raag, R.,Fitzpatrick, P. F. & Stevens, R. C. (1997). Crystalstructure of tyrosine hydroxylase at 2.3 Å resolutionand its implications for inherited neurodegenerativediseases. Nature Struct. Biol. 4, 578–585.

24. Olafsdottir, S. & Martı́nez, A. (1999). The accessibilityof iron at the active site of recombinant humanphenylalanine hydroxylase to water as studied by1H NMR paramagnetic relaxation. Effect of L-Pheand comparison with the rat enzyme. J. Biol. Chem.274, 6280–6284.

25. Kemsley, J. N., Mitic, N., Zaleski, K. L., Caradonna, J.P. & Solomon, E. I. (1999). Circular dichroism andmagnetic circular dichroism spectroscopy of thecatalytically competent ferrous active site of phenyl-alanine hydroxylase and its interaction with pterincofactor. J. Am. Chem. Soc. 121, 1528–1536.

26. Daubner, S. C., Hillas, P. J. & Fitzpatrick, P. F. (1997).Characterization of chimeric pterin-dependenthydroxylases: contributions of the regulatorydomains of tyrosine and phenylalanine hydroxylaseto substrate specificity. Biochemistry, 36, 11574–11582.

27. Daubner, S. C. & Fitzpatrick, P. F. (1999). Site-directedmutants of charged residues in the active site oftyrosine hydroxylase. Biochemistry, 38, 4448–4454.

28. Kappock, T. J. & Caradonna, J. P. (1996). Pterin-dependent amino acid hydroxylases. Chem. Rev. 96,2659–2756.

29. Chen, D. & Frey, P. A. (1998). Phenylalanine hydroxyl-ase from Chromobacterium violaceum. Uncoupled oxi-dation of tetrahydropterin and the role of iron inhydroxylation. J. Biol. Chem. 273, 25594–25601.

30. Solomon, E. I., Brunold, T. C., Davis, M. I., Kemsley,J. N., Lee, S.-K., Lehnert, N. et al. (2000). Geometricand electronic structure/function correlations innon-heme iron enzymes. Chem. Rev. 100, 235–350.

31. Dix, T. A. & Benkovic, S. J. (1988). Mechanism ofoxygen activation by pteridine-dependent mono-oxygenases. Accts Chem. Res. 21, 101–107.

32. Haavik, J. & Flatmark, T. (1987). Isolation and charac-terization of tetrahydropterin oxidation productsgenerated in the tyrosine 3-monooxygenase (tyrosinehydroxylase) reaction. Eur. J. Biochem. 168, 21–26.

33. Becke, A. D. (1988). Density-functional exchange-energy approximation with correct asymptoticbehaviour. Phys. Rev. A, 38, 3098–3100.

34. Perdew, J. P. (1986). Density-functional approxi-mation for the correlation energy of the inhomo-geneous electron gas. Phys. Rev. ser. B, 33, 8822–8824.

35. Dworniczak, B., Grudda, K., Stumper, J., Bartholome,K., Aulehla-Scholz, C. & Horst, J. (1991). Phenyl-alanine hydroxylase gene: Novel missense mutationin exon 7 causing severe phenylketonuria. Genomics,9, 193–199.

36. Loeb, K. E., Westre, T. E., Kappock, T. J., Mitic, N.,Glasfeld, E., Caradonna, J. P. et al. (1997). Spectro-scopic characterization of the catalytically competent

Ternary Complex of Phenylalanine Hydroxylase 1107

ferrous site of the resting, activated, and substrate-bound forms of phenylalanine hydroxylase. J. Am.Chem. Soc. 119, 1901–1915.

37. Marota, J. J. A. & Shiman, R. (1984). Stoichiometricreduction of phenylalanine hydroxylase by itscofactor: a requirement for enzymatic activity. Bio-chemistry, 23, 1303–1311.

38. Dix, T. A., Bollag, G. E., Domanico, P. L. & Benkovic,S. J. (1985). Phenylalanine hydroxylase: absolute con-figuration and source of oxygen of the 4a-hydroxy-tetrahydropterin species. Biochemistry, 24, 2955–2958.

39. Benkovic, S., Wallick, D., Bloom, L., Gaffney, B. J.,Domanico, P., Dix, T. & Pember, S. (1985). On themechanism of action of phenylalanine hydroxylase.Biochem. Soc. Trans. 13, 436–438.

40. Fitzpatrick, P. F. (1991). Studies of the rate-limitingstep in the tyrosine hydroxylase reaction: alternatesubstrates, solvent isotope effects, and transition-state analogues. Biochemistry, 30, 6386–6391.

41. Han, S., Eltis, L. D., Timmis, K. N., Muchmore, S. W.& Bolin, J. T. (1996). Crystal structure of thebiphenyl-cleaving extradiol dioxygenase from aPCB-degrading pseudomonad. Science, 270, 976–980.

42. Senda, T., Sugiyama, K., Narita, H., Yamamoto, T.,Kimbara, K., Fukuda, M. et al. (1996). Three-dimen-sional structures of free form and two substratecomplexes of an extradiol ring-cleavage type dioxy-genase, the BphC enzyme from Pseudomonas sp.strain KKS102. J. Mol. Biol. 255, 735–752.

43. Jennings, I. G., Teh, T. & Kobe, B. (2001). Essentialrole of the N-terminal autoreguatory sequence inthe regulation of phenylalanine hydroxylase. FEBSLetters, 488, 196–200.

44. Li, J., Eisensmith, R. C., Wang, T., Lo, W. H., Huang,S. Z., Zeng, Y. T. et al. (1992). Identification of threenovel missense PKU mutations among Chinese.Genomics, 13, 894–895.

45. Bjørgo, E., Knappskog, P. M., Martı́nez, A., Stevens,R. C. & Flatmark, T. (1998). Partial characterizationand three-dimensional-structural localization ofeight mutations in exon 7 of the human phenyl-alanine hydroxylase gene associated with phenyl-ketonuria. Eur. J. Biochem. 257, 1–10.

46. Svensson, E., Eisensmith, R. C., Dworniczak, B., vonDobeln, U., Hagenfeldt, L., Horst, J. & Woo, S. L. C.(1992). Two missense mutations causing mild hyper-phenylalaninemia associated with DNA haplotype12. Hum. Mutat. 1, 129–137.

47. Okano, Y., Wang, T., Eisensmith, R. C., Longhi, R.,Riva, E., Giovannini, M. et al. (1991). Phenylketonuriamissense mutations in the Mediterranean. Genomics,9, 96–103.

48. Knappskog, P. M., Eiken, H. G., Martı́nez, A.,Flatmark, T. & Apold, J. (1995). The PKU mutationS349P causes complete loss of catalytic activity inthe recombinant phenylalanine hydroxylase enzyme.Hum. Genet. 95, 171–173.

49. Weinstein, M., Eisensmith, R. C., Abadie, V., Avigad,S., Lyonnet, S., Schwartz, G. et al. (1993). A missensemutation, S349P, completely inactivates phenyl-alanine hydroxylase in north African Jews withphenylketonuria. Hum. Genet. 90, 645–649.

50. Martı́nez, A., Knappskog, P. M., Olafsdottir, S.,Døskeland, A. P., Eiken, H. G., Svebak, R. M. et al.(1995). Expression of recombinant human phenyl-alanine-hydroxylase as fusion protein in Escherichiacoli circumvents proteolytic degradation by host cellproteases. Isolation and characterization of the wild-type enzyme. Biochem. J. 306, 589–597.

51. Knappskog, P. M., Flatmark, T., Aarden, J. M.,Haavik, J. & Martı́nez, A. (1996). Structure/functionrelationships in human phenylalanine hydroxylase.Effect of terminal deletions on the oligomerization,activation and cooperativity of substrate binding tothe enzyme. Eur. J. Biochem. 242, 813–821.

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

53. Collaborative Computational Project, No. 4 (1994).The CCP4 suite: programs for protein crystal-lography. Acta Crystallog. sect. D, 50, 760–763.

54. Brünger, A. T., Adams, P. D., Clore, G. M., DeLano,W. L., Gros, P., Grosse-Kunstleve, R. W. et al. (1998).Crystallography & NMR system: a new softwaresuite for macromolecular structure determination.Acta Crystallog. sect. D, 54, 905–921.

55. Kleywegt, G. J. & Brünger, A. T. (1996). Checkingyour imagination: applications of the free R value.Structure, 4, 897–904.

56. Jones, T. A., Zou, J.-Y., Cowan, S. W. & Kjeldgaard,M. (1991). Improved methods for building proteinmodels in electron density maps and the location oferrors in these models. Acta Crystallog. sect. A, 47,110–119.

57. Esnouf, R. M. (1997). An extensively modified ver-sion of MolScript that includes greatly enhancedcoloring capabilities. J. Mol. Graph. Model, 15,132–134.

58. Wallace, A. C., Laskowski, R. A. & Thornton, J. M.(1995). LIGPLOT: a program to generate schematicdiagrams of protein–ligand interactions. ProteinEng. 8, 127–134.

59. Kraulis, P. J. (1991). MOLSCRIPT: a program to pro-duce both detailed and schematic plots of proteinstructures. J. Appl. Crystallog. 24, 946–950.

60. Martı́nez, A., Vageli, O., Pfleiderer, W. & Flatmark, T.(1998). Proton NMR studies on the conformationof the pterin cofactor bound at the active site ofrecombinant human tyrosine hydroxylase. Pteridines,9, 44–52.

61. Ramachandran, G. N. & Sasisekharan, V. (1968).Conformation of polypeptides and proteins. Advan.Protein Chem. 23, 283–438.

62. Laskowski, R. A., MacArthur, M. W., Moss, D. S. &Thornton, J. M. (1993). Procheck—a program tocheck the stereochemical quality of protein struc-tures. J. Appl. Crystallog. 26, 283–291.

63. Fitzpatrick, P. F. (1991). Steady-state kinetic mecha-nism of rat tyrosine hydroxylase. Biochemistry, 30,3658–3662.

64. Luzzati, V. (1952). Traitement statistique des errorsdans la determination des structures cristallines.Acta Crystallog. 5, 802–810.

Edited by R. Huber

(Received 31 January 2002; received in revised form 30 May 2002; accepted 5 June 2002)

1108 Ternary Complex of Phenylalanine Hydroxylase

Crystal Structure of the Ternary Complex of the Catalytic Domain of Human Phenylalanine Hydroxylase with Tetrahydrobiopterin anIntroductionResultsBinding of the substrate analogue 3-(2-thienyl)-l-alanine (THA)Binding of the substrate analogue 3-(2-thienyl)-l-alanine (THA) triggers large-scale structural changesDisplacement of the pterin cofactor upon substrate bindingThe ternary complex has a distorted square pyramidal, five-coordinated iron atom

DiscussionThe substrate specificity and substrate-binding siteConformational changes at the active site upon pterin cofactor and substrate bindingFunctional implicationsThe relation between the substrate-induced conformational change in the catalytic domain and that observed in the full-length tStructural insight into the effect of single-point mutations of active-site residues

Materials and MethodsCrystallization and data collectionModel building and refinementProtein Data Bank accession numbers

AcknowledgmentsReferences