Download Structure of Tetrameric Human Phenylalanine Hydroxylase and Its

THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 1998 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 273, No. 27, Issue of July 3, pp. 16962–16967, 1998
Printed in U.S.A.
Structure of Tetrameric Human Phenylalanine Hydroxylase and Its
Implications for Phenylketonuria*
(Received for publication, March 6, 1998)
Fabrizia Fusetti‡§, Heidi Erlandsen¶, Torgeir Flatmarki, and Raymond C. Stevens‡**
From the ‡Department of Chemistry, University of California and Physical Biosciences Division, Lawrence Berkeley
National Laboratory, Berkeley, California 94720, ¶Protein Crystallography Group, Chemistry Department, University of
Tromsø, N-9037 Tromsø, Norway, and iDepartment of Biochemistry and Molecular Biology, University of Bergen,
Årstadveien 19, N-5009 Bergen, Norway
Mammalian phenylalanine hydroxylase (PheOH,1 phenylalanine 4-monooxygenase, EC 1.14.16.1) is an iron- and tetrahydropterin-dependent enzyme that catalyzes the hydroxylation of L-phenylalanine (L-Phe) to L-tyrosine. The reaction is the
rate-limiting step in the catabolic pathway of phenylalanine
resulting in the complete degradation of the amino acid.
* This work received support from the Physical Biosciences Division
of Lawrence Berkeley National Laboratory (to R. C. S.), the Research
Council of Norway (to H. E. and T. F.), Rebergs Legat (to T. F.), the Novo
Nordisk Foundation (to T. F.), and the European Commission (to T. F.).
The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. Section 1734 solely to
indicate this fact.
The atomic coordinates and structure factors (code 2PAH) have been
deposited in the Protein Data Bank, Brookhaven National Laboratory,
Upton, NY.
§ Present address: Laboratory of Biophysical Chemistry, Dept. of
Chemistry, University of Groningen, Nijenborgh 4, 9747 AG Groningen,
The Netherlands.
** To whom correspondence should be addressed. Tel.: 510-643-8285;
Fax: 510-643-9290; E-mail: [email protected].
1
The abbreviations used are: PheOH, phenylalanine hydroxylase;
PKU, phenylketonuria; TyrOH, tyrosine hydroxylase; TrpOH, tryptophan hydroxylase; MES, (2-(N-morpholino)ethanesulfonic acid); BisTris-propane,2-[bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol.
PheOH activity is tightly regulated by reversible phosphorylation and substrate activation. Mutations in the human PheOH
gene result in the metabolic disorder known as phenylketonuria (PKU), a relatively common autosomal recessive disease
that, if not diagnosed and properly treated from birth, ultimately leads to severe mental retardation. Approximately 280
mutant alleles of the human PheOH gene have been identified
in patients with PKU and non-PKU hyperphenylalaninemia
(1). The mutations lead to a variety of clinical and biochemical
phenotypes with different degrees of severity (2, 3).
PheOH shares many biochemical and physical properties
with tyrosine hydroxylase (TyrOH) and tryptophan hydroxylase (TrpOH) suggesting that they evolved from a common
ancestor and might constitute a family of structurally related
proteins (for review, see Ref. 4). PheOH, TyrOH, and TrpOH
catalyze similar hydroxylation reactions using the same nonheme iron atom but differ in their substrate specificity. All
three hydroxylases are organized into three domains: a 12–19kDa N-terminal domain involved in regulation of the activity
followed by a segment of about 38 kDa, which consist of a
catalytic domain and a 5-kDa tetramerization domain. The
crystal structure of a dimeric form of PheOH (5) containing
only the catalytic domain shows a remarkable similarity to the
previously reported structure of TyrOH (6). The tetramerization domain, located at the C terminus of TyrOH, is characterized by the presence of a long helix that promotes oligomerization through a coiled-coil motif (6). The sequence homology
among the aromatic amino acid hydroxylases is very high,
particularly in the catalytic and tetramerization domains (6 –
8). Proteolysis experiments on rat PheOH and TyrOH have
shown that deletion of the N-terminal domain results in active
truncated proteins with reduced substrate binding specificity.
The attachment of either N-terminal regulatory domain (as
chimeric proteins) enhances the substrate specificity displayed
by the catalytic domain (9).
Since the discovery of the relationship between PKU and
mutations in PheOH, the enzyme has been the subject of intense biochemical and genetic studies but for many years it has
been difficult to gain detailed structural information on
PheOH. Crystallization and diffraction to 2.7 Å resolution of a
phosphorylated tetrameric form of full-length PheOH was reported several years ago (10), but no structure determination
was completed from those studies. More recent crystallization
experiments have shown that highly diffracting crystals can be
obtained only for dimeric forms of PheOH missing the C-terminal tetramerization domain (11, 12).
Here we report the first crystallization and structure determination of a truncated form of human tetrameric PheOH
(residues 118 – 452) to which we will refer as PheOHCterm. The
crystal structure of PheOH reveals that within one monomer
the catalytic and tetramerization domains can adopt two dif-
16962
This paper is available on line at http://www.jbc.org
Downloaded from http://www.jbc.org/ at Medical College of Wisconsin on July 27, 2015
Phenylalanine hydroxylase (PheOH) catalyzes the
conversion of L-phenylalanine to L-tyrosine, the ratelimiting step in the oxidative degradation of phenylalanine. Mutations in the human PheOH gene cause phenylketonuria, a common autosomal recessive metabolic
disorder that in untreated patients often results in varying degrees of mental retardation. We have determined
the crystal structure of human PheOH (residues 118 –
452). The enzyme crystallizes as a tetramer with each
monomer consisting of a catalytic and a tetramerization
domain. The tetramerization domain is characterized by
the presence of a domain swapping arm that interacts
with the other monomers forming an antiparallel coiledcoil. The structure is the first report of a tetrameric
PheOH and displays an overall architecture similar to
that of the functionally related tyrosine hydroxylase. In
contrast to the tyrosine hydroxylase tetramer structure,
a very pronounced asymmetry is observed in the phenylalanine hydroxylase, caused by the occurrence of two
alternate conformations in the hinge region that leads
to the coiled-coil helix. Examination of the mutations
causing PKU shows that some of the most frequent mutations are located at the interface of the catalytic and
tetramerization domains. Their effects on the structural
and cellular stability of the enzyme are discussed.
Structure of Phenylalanine Hydroxylase
16963
ferent orientations, thus forming a tetramer similar to that in
TyrOH (6) but with an unexpected asymmetry in the packing of
the four subunits. Because some of the most common PKU
mutations map at the junction between catalytic and tetramerization domain, this structure of PheOH can provide a template
for analyzing the molecular basis of PheOH deficiency.
EXPERIMENTAL PROCEDURES
FIG. 1. Ribbon diagram representations of the PheOHCterm and
TyrOHCterm tetramers. The four subunits are shown in different
colors. In the PheOH tetramer the 2-fold crystallographic axis relates
subunits A (green) and B (blue) to C (yellow) and D (orange), the
noncrystallographic 2-fold relates the dimer AC to BD. In the TyrOH
tetramer the four subunits are related by a crystallographic 222 symmetry. The coiled-coil motif is visible in the center of each tetramer. The
iron bound in the active sites are shown as red spheres. The figure was
generated using the programs MOLSCRIPT (48) and RASTER3D (49).
gram “O” (21). Strict noncrystallographic symmetry was initially
imposed during the refinement. After addition of the tetramerization
domain the restrain option was used because the relative positions of
the core and the tetramerization domains within the tetramer are
described by different noncrystallographic two-fold operators. The
refinement was performed imposing tight noncrystallographic symmetry restraints (weight of 300) for the core of the catalytic domain
(residues 118 –393) but relaxed restraints for the tetramerization
domain (weight of 10). Alternating refinement cycles with X-PLOR
(18) and model building produced the final structure. In the final
model all residues are defined in the electron density with the exception of residues 138 –144. The final model of the two subunits in the
asymmetric unit, comprising 654 residues, has an R-factor of 22.7%
for all measured data between 20 and 3.1 Å and an R-free of 33.7%
(22). The final structure has a root mean square deviation in bond
length and angles of 0.006 Å and 1.7°, respectively.
RESULTS
A truncated form of PheOH including the catalytic and tetramerization domains, residues 118 – 452, was expressed in E.
coli, purified and crystallized. This fragment of PheOH (PheOHCterm) retains full catalytic properties but is more soluble
than the wild-type enzyme. Dynamic light scattering analysis
and size-exclusion chromatography revealed that the recombinant purified protein is 99% monodisperse in solution with a
calculated molecular mass around 160 kDa, corresponding to a
tetramer (data not shown). PheOHCterm crystallizes in space
group P3112 with two molecules in the asymmetric unit forming a tetramer (dimer of dimers) through the crystallographic
two-fold symmetry (Fig. 1). The structure was solved by molecular replacement using a model derived from the crystal structure of TyrOH. The PheOHCterm catalytic domain contains 13
a-helices and 8 b-strands and is identical, within coordinate
error, to the structure of a dimeric form of PheOH recently
determined at 2.0 Å resolution, which we will refer to as PheOHCat (5). The PheOHCterm catalytic domain in the tetramer
structure can be superimposed on the corresponding domain of
the TyrOHCterm structure with a root mean square deviation of
0.64 Å for all the Ca atoms. The active site of human PheOHCterm is located in a deep cleft in the core of each monomer
and is very similar to that of TyrOHCterm. Although iron was
not added during the purification and crystallization procedures, electron density compatible with that of a metal ion is
present in the catalytic center in the same site as seen in the
structure of TyrOHCterm (6) and PheOHCat (5). His-285, His290, and Glu-330 coordinate the iron in agreement with site-
Downloaded from http://www.jbc.org/ at Medical College of Wisconsin on July 27, 2015
Protein Purification and Crystallization—The gene encoding the catalytic and the tetramerization domain of human PheOH (residues 118 –
452 of human phenylalanine hydroxylase, PheOHCterm), cloned into
vector pET-3a-d (Novagen), was overexpressed in Escherichia coli
BL21. For protein purification, frozen cell paste was resuspended in 50
mM Tris-HCl buffer, pH 8.0, containing 20 mM NaCl, 5% glycerol, and 3
mM methionine, 0.5 mM phenylmethylsulfonyl fluoride. Cell lysis was
started by adding lysozyme and completed by French press treatment.
All purification steps were performed at 4 °C. The soluble fraction of the
protein extract was diluted with one volume of 50 mM Tris-HCl buffer,
pH 8.0, 5% glycerol, 3 mM methionine and applied onto a DEAE Fast
Flow column (Amersham Pharmacia Biotech). The fractions containing
PheOHCterm were pooled, and the protein was further purified by ammonium sulfate precipitation at 30 –50% saturation. The precipitate
was resuspended in 50 mM MES buffer, pH 5.8, 2.5% glycerol, 3 mM
methionine and loaded onto a Perseptive Biosystem HS column and
eluted with a NaCl gradient. PheOHCterm fractions were pooled, concentrated and further purified by size-exclusion chromatography on a
Superdex 200 Pharmacia column eluted with 50 mM Bis-Tris-propane
buffer, pH 7.0, 200 mM NaCl, 2.5% glycerol, 3 mM methionine. After
adding 2 mM dithiothreitol, the protein was concentrated to about 10
mg/ml in an Amicon pressure cell. The purified PheOHCterm appeared
homogeneous (tetramer) as assessed by size-exclusion chromatography
and silver-stained SDS-polyacrylamide gel electrophoresis.
Initial crystallization conditions were identified by using the sparse
matrix approach (14). PheOHCterm was crystallized at room temperature by hanging drop vapor diffusion from 0.6 M MgSO4, 0.2 M sodium,
potassium tartrate, 10% polyethylene glycol 4000, 100 mM Bis-Trispropane buffer, pH 7.0. Crystals grew within 1–2 days as triangular
plates with dimension 0.6 3 0.6 3 0.2-mm3 in space group P3112 (a 5
b 5 119.5 Å, c 5 126.0 Å, a 5 b 5 90°, g 5 120°). The asymmetric unit
contains two PheOHCterm monomers, each consisting of a catalytic and
a tetramerization domain. Two asymmetric units form one tetramer by
a crystallographic two-fold symmetry axis. The solvent content is about
60%. To obtain good diffraction it was essential to harvest and flash
freeze the crystals immediately after growth was completed (1–2 days).
Data Collection and Structure Determination—X-ray diffraction data
to 3.1 Å were collected at Beam Line 7–1 of the Stanford Synchrotron
Radiation Laboratory (SSRL) on flash frozen crystals. A solution of 30%
glycerol in 25% mother liquor was used as a cryoprotectant. The programs DENZO and SCALEPACK were used for data processing and
reduction (15). A total of 120,922 measurements were recorded for
18,446 unique reflections. The data set is 99% complete between 20 and
3.1 Å with an Rmerge value of 7.5%.
The PheOHCterm structure was solved by molecular replacement.
The search model was based on the atomic coordinates of the crystal
structure of tyrosine hydroxylase (6), because the crystal structure of
PheOH (5) was not yet available when the present study was started.
The amino acid sequence identity between the catalytic domains of
PheOH and TyrOH is approximately 80%. The best search model
included only the catalytic domain of TyrOH in which all the residues
that were not identical in the two proteins were replaced with alanine. Molecular replacement was performed using the program
AMoRe (16, 17). In the cross-rotation function all data between 8 and
4.0 Å were used and model Patterson vectors were selected within a
radius of 30 Å. The translation function and a subsequent step of rigid
body refinement were carried out with all data between 8 and 4.0 Å.
This procedure allowed a clear identification of the two molecules in
the asymmetric unit. The search for monomer A gave a correlation
coefficient of 33% and an R-factor of 50%, whereas the search for
monomer B gave a correlation coefficient of 22% and an R-factor of
53%. Orientation and position of the dimer obtained by molecular
replacement were improved by rigid body refinement using the program X-PLOR (18). This lowered the R-factor to 41% and provided a
more accurate definition of the noncrystallographic two-fold operator.
The sA weighted uFou 2 uFcu maps (19) calculated at this stage showed
electron density for the helices forming the coiled-coil domain that
were not included in the search model. Simulated annealing omit
maps (20) were calculated for manual model building with the pro-
16964
Structure of Phenylalanine Hydroxylase
directed mutagenesis studies (23–25) and with the structure of
PheOHCat. Glu-286, proposed to be involved in pterin binding
in PheOH but not in TyrOH (26), occupies nearly identical
positions in both structures. The absence of significant structural differences in the active site of PheOHCterm and TyrOHCterm is not surprising because a similar low substrate binding specificity is observed for the catalytic domains in steadystate kinetic analyses (9).
The architecture of the PheOHCterm matches the overall fold
and organization found in the structure of TyrOHCterm (6). The
assembly of PheOH and TyrOH might be described by a domain swapping mechanism in which secondary structural elements mutually switch their position to promote oligomerization (27). The tetramerization domain is formed by a
C-terminal “arm” consisting of two b-strands, forming a b-ribbon, and a 40 Å long a-helix. The C-terminal arm extends over
a neighboring monomer bringing the four helices (one from
each monomer) into a tightly packed anti-parallel coiled-coil
motif in the center of the structure (Fig. 1). In the PheOH
structure the tetramer is formed by two conformationally different dimers resulting in a distortion of the 222 symmetry. A
superposition of the two monomers shows that the catalytic
domains and tetramerization domains, if taken separately, are
identical but they adopt different relative orientations (Fig.
2A). The distortion of the tetramer symmetry is also evident
from a surface calculation: in the PheOHCterm the subunits
interact with a buried surface area of about 1200 Å2 and 1700
Å2, across the crystallographic 2-fold dimerization interfaces
AC and BD, respectively (Fig. 1). The buried surface calculated
for the TyrOHCterm structure is about 1600 Å2 for both dimerization interfaces AC and BD. The discrepancy results from a
change in backbone conformation at the hinge region that
connects the b-ribbon to the coiled-coil helix accompanied by
another rotation in the coiled-coil helix.
Using the program DYNDOM (28) it was possible to identify
two rotation points in the tetramerization domain. The first
hinge axis is found at residue Thr 427 leading to a 22° movement of the catalytic domain with respect to the tetramerization domain. All the residues in the hinge (425– 429) are nicely
defined in electron density, except for Thr-427, which is very
disordered indicating flexibility in this part of the polypeptide
chain. The second change in backbone conformation is found at
residue Gly-442 that generates a kink in the coiled-coil helix
and a rotation of about 20° (Fig. 2B). A conformational change
Downloaded from http://www.jbc.org/ at Medical College of Wisconsin on July 27, 2015
FIG. 2. A, stereo view of the superposition of the PheOHCterm showing that in
PheOHCterm the relative orientation of the
coiled-coil helix in subunit A (green) is
different than in B (blue). B, stereo view
of the PheOHCterm tetramerization domain. The superposition of the coiled-coil
helices shows that the conformational
change in the tetramerization domain
causes a different orientation of the b-ribbon that leads to the catalytic domain.
The positions at which the conformational
changes are detected are highlighted in
red. Color code is as shown in A. For clarity, only the catalytic domain of B is
shown in gray. C, sequence alignment of
the PheOH and TyrOH tetramerization
domains. Proline residues located at the
hinge regions are shown in red. Secondary structure elements are represented as
boxes (a-helices) and arrows (b-strands).
An arrow indicates the positions at which
the conformational changes are detected.
Regions of identity are indicated with a
vertical bar. The figure was generated using the program MOLSCRIPT (48).
Structure of Phenylalanine Hydroxylase
16965
FIG. 3. A stereo view of the tetramerization domain of PheOHCterm,
(subunit A). PKU mutations sites are
shown in white as ball-and-stick representations (see text for details). The hydrogen-bond interactions of Arg-252 and
Arg-408 with the main chain carbonyl
groups of Ala-313, Leu-308, Leu-309, and
Leu-311, are shown as dashed lines. The
proline residues in the tetramerization
domain are shown in green as ball-andstick representations. The figure was generated using the program MOLSCRIPT
(48).
DISCUSSION
Full-length PheOH has been described to exist as both a
tetramer and dimer in a pH-dependent equilibrium and possibly as some monomer (4, 30). Lowering the pH, as well as
binding of L-phenylalanine, shifts the equilibrium toward the
tetrameric form of PheOH (31, 32). By contrast, TyrOH and
TrpOH are always reported to be a tetramer in solution (4).
Truncated forms of PheOH and TyrOH where the C-terminal
domain is left intact have so far only been isolated as tetramers. Removal of the last 34 residues at the C terminus results
in dimeric (PheOH) or monomeric (TyrOH) active hydroxlases.
The three-dimensional structure of the PheOHCterm tetramer
matches the overall fold and organization found in the structure of tetrameric TyrOHCterm (6). Despite the high degree of
sequence and structural identity, superposition of TyrOHCterm
and PheOHCterm shows that the relative orientation of the
subunits forming the tetramer is different in the two enzymes.
The difference is caused by a distortion of the 222-fold symmetry in the PheOH tetramer, whereas in the TyrOH tetramer
the four subunits are related by crystallographic symmetry and
form a dimer of dimers with only two unique subunit-subunit
interfaces (6). With the evidence of a conformational change at
hand it is tempting to hypothesize whether this structural
difference is physiologically meaningful. PheOH and TyrOH
share a common catalytic mechanism in which reduced pterin
cofactor, molecular oxygen, and iron are required to hydroxyl-
ate the aromatic amino acid, but they are regulated through
different mechanisms. PheOH activation is believed to be
achieved by specific phosphorylation on the regulatory domain
and/or by substrate activation. During the past few years the
mechanism of activation has been the subject of intense biochemical and spectroscopic studies, which have showed that
activation of PheOH by the substrate is highly cooperative and
always accompanied by alterations in the tertiary structure,
but the nature of the conformational change has not been
ascertained yet. Substrate activation has been proposed to
involve cooperativity among all four subunits of the tetramer
(33, 34) and radiation target analysis has shown that activation
of PheOH results in modification of the monomer structure and
promotes stronger association at the dimer interface (35).
These results are further supported by evidence of increase in
volume of the tetrameric protein (30), and by a shift of the
dimer 7 tetramer equilibrium toward the tetrameric form (31)
upon binding of L-phenylalanine. Recent infrared spectroscopic
studies (36) led to the same conclusion and have also shown
that the alterations measured upon PheOH activation are not
accompanied by any measurable changes in the secondary
structure. The asymmetric structure of the PheOHCterm tetramer could represent an intermediate state and could suggest
that high flexibility in the tetramerization domain is required
to optimally orient the domains for catalysis. Further experiments are required to understand if and how the observed
asymmetry of the tetramer is related to cooperative substrate
activation.
A possible contribution to the conformational heterogeneity
in the crystal structure of PheOHCterm could be a chemical
modification. PheOHCterm crystals lose their diffraction power
after a few days and mass spectroscopy analysis shows a timedependent heterogeneity of the protein in solution. Two-dimensional electrophoresis of full-length recombinant wild-type human PheOH (as isolated from E. coli) has shown a
microheterogeneity in terms of isoelectric point (five components), which has been proposed (37) and recently proven2 to be
a result of progressive deamidation of rather labile amide
groups. Studies on short peptides have demonstrated that the
rate of nonenzymatic deamidation of asparagine residues primarily depends on the amino acid at positions n 2 1 and n 1 1,
with half-lives from a few hours, when the residue in the n 1 1
position is a glycine, to several days (38). Polar residues preceding Asn and neighboring Ser and Thr increase deamidation
rates, whereas bulky, nonhydrophobic residues preceding Asn
2
T. Solstad and T. Flatmark, unpublished data.
Downloaded from http://www.jbc.org/ at Medical College of Wisconsin on July 27, 2015
in the helix could be necessary to preserve the coiled-coil interactions in response to the change in relative orientation of the
subunits. In the TyrOH sequence Thr-427 and Gly-442 are
replaced by a proline and a histidine residue, repectively (Fig.
2C). It has recently been proposed that the presence of a proline
in the exchanging arm induces conformational constraints on
the peptide backbone that favor the optimal conformation for
oligomerization (29). As shown in Fig. 2C and Fig. 3, both
PheOH and TyrOH sequences contain a total of three proline
residues in the tetramerization domain. In PheOHCterm two
prolines (Pro-407 and Pro-409) define the sharp turn that links
the tetramerization domain to the catalytic domain. Another
proline is also found at the C-terminal end of b-strand 7 (residue 416), but no proline residue is located at the last hinge
region connecting to the coiled-coil where the conformational
change is detected. The lack of a proline residue at this position
may allow more flexibility at this linker region and modulate
the arrangement of the quaternary structure in PheOH. In
TyrOH, a proline residue is present in each of the hinge regions
of the tetramerization domain, probably stabilizing its domainswapped conformation.
16966
Structure of Phenylalanine Hydroxylase
with high residual activity.3 Mutations R252W, A259L/T,
F299C, and L311P (43– 45) also result in extremely reduced
levels of PheOH and are associated with severe PKU phenotype. As described above, the carbonyl atom of Leu-311 forms a
hydrogen bond with Arg-408. Residue Ala-259 is buried in a
hydrophobic pocket about 4 Å from Leu-311 and Leu-308; a
larger or polar residue cannot be accommodated without disturbing the surrounding environment. Phe-299 is also involved
in hydrophobic interactions that stabilize the helix where Leu311 is located and Arg-252 forms a hydrogen bond with the
carbonyl atom of Ala-313. As shown in Fig. 3, these mutations
map in the same region as Arg-408, and they could dramatically alter the hydrogen bond network or produce steric hindrance, interfering with the correct positioning of the b-ribbon.
It is likely that their effect will have consequences for the
arrangement of the catalytic and tetramerization domains and
thus for the normal oligomerization process. Mutation Y414C
has also been reported, and the expressed enzyme retains
approximately 50% of normal activity (46). The mild effect of
this mutation can be observed also at structural level: a substitution Tyr 3 Cys in this position should not have drastic
consequences on the interactions with the neighboring residues
(Fig. 3). One of the most frequent alleles in the oriental population is the missense mutation R413P in exon 12, which causes
a severe metabolic form of PKU. This mutation results in
protein instability with almost undetectable levels of PheOH
when expressed in eukaryotic cells (47). Arg-413 is located in
the middle of the first strand (b7) of the b-ribbon in the tetramerization domain. A proline residue in this position would
disrupt the main chain to main chain hydrogen bond network
that stabilizes the b-ribbon. The PKU mutation R413P most
certainly leads to extreme disorder of the tetramerization arm
thus compromising the normal oligomeric state and stability of
PheOH.
The analysis of these mutations suggest that changes in the
b-ribbon have major effects on the stability of the enzyme. The
possibility that the b-ribbon represents a mutation “hot spot”
points to the structural importance of this element.
Acknowledgments—We thank Christelle Sabatier and Andy-Mark
Thunnissen for fruitful discussions and critical reading of the manuscript and also Steven Hayward for making the program DYNDOM
available prior to distribution. We also thank Randi S. Svebak and Ali
S. Muñoz for expert technical assistance.
REFERENCES
1. Hoang, L., Byck, S., Prevost, L., and Scriver, C. R. (1996) Nucleic Acids Res. 24,
127–131
2. Okano, Y., Eisensmith, R. C, Güttler, F., Lichter-Konecki, U., Konecki, D. S.,
Trefz, F. K., Dasovich, M., Wang, T., Henriksen, K., Lou, H., and Woo,
S. L. C. (1991) N. Engl. J. Med. 324, 1232–1238
3. Svensson, E., Iselius, L., and Hagenfeldt, L. (1994) J. Inherited Metab. Dis. 17,
215–222
4. Hufton, S. E., Jennings, I. G., and Cotton, R. G. H. (1995) Biochem. J. 311,
353–366
5. Erlandsen, H., Fusetti, F., Martinez, A., Hough, E., Flatmark, T., and Stevens,
R. C. (1997) Nat. Struct. Biol. 4, 995–1000
6. Goodwill, K. E., Sabatier, C., Marks, C., Raag, R., Fitzpatrick, P. F., and
Stevens, R. C. (1997) Nat. Struct. Biol. 4, 578 –585
7. Ledley, F. D., DiLella, A. G., Kwok, S. C. M., and Woo, S. L. C. (1985)
Biochemistry 24, 3389 –3394
8. Daubner, S. C., Lohse, D. L., and Fitzpatrick, P. F. (1993) Protein Sci. 2,
1452–1460
9. Daubner, S. C., Hillas, P. J., and Fitzpatrick, P. F. (1997) Biochemistry 36,
11574 –11582
10. Celikel, R., Davis, M. D., Dai, X., Kaufman, S., and Xuong, N.-H. (1991) J. Mol.
Biol. 218, 495– 498
11. Erlandsen, H., Martinez, A. Knappskog, P. M. Haavik, J., Hough, E., and
Flatmark, T. (1997) FEBS Lett. 406, 171–174
12. Kobe, B., Jennings, I. G., House, C. M., Feil, S. C., Michell, B. J., Tiganis, T.,
Parker, M. W., Cotton, R. G. H., and Kemp, B. E. (1997) Protein Sci. 6,
1352–1357
13. Daubner, S. C., Hillas, P. J., and Fitzpatrick, P. F. (1997) Arch. Biochem.
Biophys. 348, 295–302
3
P. M. Knappskog, E. Bjørgo, and T. Flatmark, unpublished data.
Downloaded from http://www.jbc.org/ at Medical College of Wisconsin on July 27, 2015
correlate with low deamidation rates (38). Deamidation may
result in marked conformational changes as first shown for
bovine heart cytochrome c (39). The sequence positions of the
labile amide groups in PheOH are, however, not yet identified.
Two putative targets for nonenzymatic deamidation are both
highly conserved within the mammalian PheOH, and map on
the tetramerization domain, Asn-426 (Asp-Asn-Thr), in the
hinge loop that seems to confer high flexibility to the tetramerization domain and Asn-438 (Ile-Asn-Ser) in the middle of the
coiled coil. Asn-426 and Asn-438 are well defined by the electron density, but Thr-427 is very weakly outlined. Mutation of
these residues may confirm if a nonenzymatic deamidation
mechanism is responsible for the heterogeneity of PheOHCterm
and for the difficulties encountered in the crystallization of this
enzyme in its tetrameric form.
In discussing the implications of the asymmetric organization of PheOHCterm we should also consider that the observed
arrangement could be the result of crystal packing interactions.
This possibility can be excluded because the catalytic domains
are nearly identical in both the dimeric and tetrameric form of
PheOH as well as in the TyrOH structure, even though they
crystallize in different space groups.
Phenylketonuria—PKU is a common recessive genetic disease caused by mutations in the human PheOH gene. PKU was
the first inborn error of metabolism shown to affect the central
nervous system and the first form of mental retardation related
to a specific enzymatic defect (40). Approximately 280 mutant
alleles have been identified so far and they have been associated with different degrees of PKU severity (for review see
Refs. 2 and 3 and PKU data base: http://www.mcgill.ca/pahdb).
The majority of PKU mutations map onto the catalytic domain,
but 12 mutations occur in residues at the interface of the
catalytic and tetramerization domain or within the tetramerization domain. Of particular interest in this study are the
mutations that have been shown by expression studies to severely affect the stability of the enzyme.
The splicing mutation IVS12ntg3a in intron 12 of the human PheOH gene is the most prevalent PKU allele among
Caucasians (41). The clinical and metabolic phenotype of homozygotes with the mutation is a severe (“classical”) form of
PKU. The mutation leads to the synthesis of a truncated form
lacking the last 52 C-terminal amino acids, which include the
tetramerization domain. Expression studies in eukaryotic systems indicated that the deletion abolished PheOH activity in
the cell as a result of protein instability. Another frequent
allele is caused by a CGG-to-TGG transition in exon 12, resulting in an amino acid substitution Arg/Trp at position 408. This
mutation also results in undetectable levels of PheOH in vivo
and in vitro and a severe metabolic PKU phenotype (42). Analysis of the TyrOH and PheOH structures shows that the highly
conserved Arg-408 is located on the hinge loop that connects
the tetramerization arm to the core of the PheOH monomer,
flanked by the above mentioned Pro-407 and Pro-409 (Fig. 3).
The arginine side chain is buried and forms a hydrogen bond
with the main chain carbonyl of Leu-308, Leu-309, and Leu-311
at the C-terminal end of helix 8. Arg-408 could be involved in
stabilizing domain swapping and promoting oligomerization by
anchoring the tetramerization arm onto the catalytic domain in
a proper orientation. Correct alignment of arm exchange in
domain swapping is believed to be necessary for the association
process (27). This suggestion is consistent with the analysis of
mutation R408Q that causes a non-PKU hyperphenylalaninemia in homozygotes (43). Expression of R408Q mutants in E.
coli results in the synthesis of a tetrameric form of the enzyme
Structure of Phenylalanine Hydroxylase
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
32.
33. Martinez, A., Olafsdottir, S., and Flatmark, T. (1993) Eur. J. Biochem. 211,
259 –266
34. Shiman, R., Jones, S. H., and Gray, D. W. (1990), J. Biol. Chem. 265,
11633–11642
35. Davis M. D., Parniak, M. A., Kaufman, S., and Kempner, E. (1997) Proc. Natl.
Acad. Sci. U. S. A. 94, 491– 495
36. Chehin, R., Thorolfsson, M., Knappskog, P. M., Martinez, A., Flatmark, T.,
Arrondo, J. L. R., and Muga, A. (1998) FEBS Lett. 422, 225–230
37. Døskeland, A. P., and Flatmark, T.(1996) Biochem. J. 319, 941–945
38. Tyler-Cross, R., and Schirch, V. (1991) J. Biol. Chem. 266, 22549 –22556
39. Flatmark, T., and Sletten, K. (1968) J. Biol. Chem. 243, 1623–1629
40. Følling, I., (1994) Acta Paediatr. 407, (suppl.) 4 –10
41. DiLella, A. G., Marvit, J., Lidsky, A. S., Güttler, F., and Woo, S. L. C. (1986)
Nature 322, 799 – 803
42. DiLella, A. G., Marvit, J., Brayton, K., and Woo, S. L. C. (1987) Nature 327,
333–336
43. Eiken, H. G., Stangeland, K., Skjelkvåle, L., Knappskog, P. M., Boman, H., and
Apold. J. (1992) Hum. Genet. 88, 608 – 612
44. Okano, Y., Wang, T., Eisensmith, R. C., Longhi, R., Riva, E., Giovannini, M.,
Cerone, R., Romano, C., and Woo, S. L. C. (1991) Genomics 9, 96 –103
45. Lichter-Konecki, U., Konecki, D. S., DiLella, A. G., Brayton, K., Marvit, J.,
Hahn, T. M., Trefz, F. K., and Woo, S. L. C. (1988) Biochemistry 27,
2881–2885
46. Okano, Y., Eisensmith, R. C., Dasovich, M. Wang, T. Guttler, F., and Woo,
S. L. C. (1991) Eur. J. Pediatr. 150, 347–352
47. Wang, T., Okano, Y., Eisensmith, R. C., Harvey, M. L., Lo, W. H. Y., Huang,
S. Z., Zeng, Y. T., Yuan, L. F., Furuyama, J. I., Oura, T., Sommer, S. S., Woo,
S. L. C. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 2146 –2150
48. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946 –950
49. Merritt, E. A., and Murphy, M. E. P. (1994) Acta Crystallogr. Sec. D 50,
869 – 873
Downloaded from http://www.jbc.org/ at Medical College of Wisconsin on July 27, 2015
31.
Jancarik, J., and Kim, S. H. (1991) J. Appl. Crystallogr. 24, 409 – 411
Otwinowski, Z., and Minor W. (1997) Methods Enzymol. 276, 307–326
Navaza, J. (1994) Acta Crystallogr. Sec. A 50, 157–163
CCP4 (1994) Acta Crystallogr. Sec. D 50, 760 –763
Brünger, A. T. (1992) X-PLOR. Version 3.1, Yale University Press, New
Haven, CT
Read, R. J. (1986) Acta Crystallogr. Sec. A 42, 140 –149
Hodel, A., Kim, S.-H., and Brünger, A. T. (1992) Acta Crystallogr. Sec. A 48,
851– 858
Jones, T. A., Zou, J. Y., Cowans, S. W., and Kjeldgaard, M. (1991) Acta
Crystallogr. Sec. A 47, 110 –119
Brünger, A. T. (1992) Nature 355, 472– 475
Gibbs, B. S., Wojchowski, D., and Benkovic, S. J. (1993) J. Biol. Chem. 268,
8046 – 8052
Balasubramanian, S., Carr, R. T., Bender, C. J., Peisach, J., and Benkovic,
S. J. (1994) Biochemistry 33, 8532– 8537
Ramsey, A. J., Daubner, S. C., Ehrlich, J. I., and Fitzpatrick, P. F. (1995)
Protein Sci. 4, 2082–2086
Dickson, P. W., Jennings, I. G., and Cotton, R. G. H. (1994) J. Biol. Chem. 269,
20369 –20375
Bennett, M. J., Schlunegger, M. P., and Eisenberg, D. (1995) Protein Sci. 4,
2455–2468
Hayward, S., and Berendsen, H. J. C. (1998) Proteins 30, 144 –154
Bergdoll, M., Remy, M.-H., Cagnon, C., Masson, J.-M., and Dumas, P. (1997)
Structure (Lond.) 5, 391– 401
Kappock, T. J., Harkins, P. C., Friedenberg, S., and Caradonna, J. P. (1995)
J. Biol. Chem. 270, 30532–30544
Martinez, A., Knappskog, P. M., Olafsdottir, S., Døskeland, A. P., Eiken, H. G.,
Svebak, R. M., Bozzini, M. L., Apold, J., and Flatmark, T. (1995) Biochem.
J. 306, 589 –597
Knappskog, P. M., Flatmark, T., Aarden, J. M., Haavik, J., and Martinez, A.
(1996) Eur. J. Biochem. 242, 813– 821
16967
PROTEIN CHEMISTRY AND
STRUCTURE:
Structure of Tetrameric Human
Phenylalanine Hydroxylase and Its
Implications for Phenylketonuria
Fabrizia Fusetti, Heidi Erlandsen, Torgeir
Flatmark and Raymond C. Stevens
Access the most updated version of this article at http://www.jbc.org/content/273/27/16962
Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity Sites.
Alerts:
• When this article is cited
• When a correction for this article is posted
Click here to choose from all of JBC's e-mail alerts
This article cites 43 references, 9 of which can be accessed free at
http://www.jbc.org/content/273/27/16962.full.html#ref-list-1
Downloaded from http://www.jbc.org/ at Medical College of Wisconsin on July 27, 2015
J. Biol. Chem. 1998, 273:16962-16967.
doi: 10.1074/jbc.273.27.16962