|
Originally published In Press as doi:10.1074/jbc.M202107200 on April 17, 2002
J. Biol. Chem., Vol. 277, Issue 26, 23807-23814, June 28, 2002
Structure and Function of Human Erythrocyte Pyruvate
Kinase
MOLECULAR BASIS OF NONSPHEROCYTIC HEMOLYTIC ANEMIA*
Giovanna
Valentini §¶,
Laurent R.
Chiarelli§,
Riccardo
Fortin ,
Manuela
Dolzan ,
Alessandro
Galizzi ,
Donald J.
Abraham ,
Changqing
Wang ,
Paola
Bianchi**,
Alberto
Zanella**, and
Andrea
Mattevi 
From the Dipartimento di Genetica e Microbiologia,
Università di Pavia, via Abbiategrasso 207, 27100 Pavia, Italy,
the § Dipartimento di Biochimica, Università di Pavia,
via Taramelli 3b, 27100 Pavia, Italy, the Department of
Medicinal Chemistry, Virginia Commonwealth University, Richmond,
Virginia 23219, and the ** Divisione di Ematologia, IRCCS
Ospedale Maggiore di Milano, via Francesco Sforza 35, 20122 Milano, Italy
Received for publication, March 4, 2002, and in revised form, April 12, 2002
 |
ABSTRACT |
Deficiency of human erythrocyte isozyme (RPK) is,
together with glucose-6-phosphate dehydrogenase deficiency, the most
common cause of the nonspherocytic hemolytic anemia. To provide a
molecular framework to the disease, we have solved the 2.7 Å resolution crystal structure of human RPK in complex with fructose
1,6-bisphosphate, the allosteric activator, and phosphoglycolate, a
substrate analogue, and we have functionally and structurally
characterized eight mutants (G332S, G364D, T384M, D390N, R479H, R486W,
R504L, and R532W) found in RPK-deficient patients. The mutations target
distinct regions of RPK structure, including domain interfaces and
catalytic and allosteric sites. The mutations affect to a different
extent thermostability, catalytic efficiency, and regulatory
properties. These studies are the first to correlate the clinical
symptoms with the molecular properties of the mutant enzymes. Mutations greatly impairing thermostability and/or activity are associated with
severe anemia. Some mutant proteins exhibit moderate changes in the
kinetic parameters, which are sufficient to cause mild to severe
anemia, underlining the crucial role of RPK for erythrocyte metabolism.
Prediction of the effects of mutations is difficult because there is no
relation between the nature and location of the replaced amino acid and
the type of molecular perturbation. Characterization of mutant proteins
may serve as a valuable tool to assist with diagnosis and genetic counseling.
 |
INTRODUCTION |
Pyruvate kinase (PK)1
catalyzes the conversion of phosphoenolpyruvate (PEP) to pyruvate with
the synthesis of ATP. The enzyme requires K+ and
Mg2+ (or Mn2+) for activity (1-3). The
PK-catalyzed reaction represents the last step of glycolysis with the
reaction product, pyruvate, being involved in a number of energetic and
biosynthetic pathways. PK is activated homotropically by PEP and
heterotropically by monophosphorylated or bisphosphorylated sugars (2).
In addition, Mg2+, H+, and other cations
modulate enzymatic activity (4). The regulatory behavior of PK varies
depending on the enzyme source. Four PK isozymes have been identified
in mammals (5). The M1 (muscle) and M2 (fetal) proteins are products of
the alternative splicing of the same mRNA. M2 PK is allosterically
activated by fructose 1,6-bisphosphate (FBP) and PEP, whereas the M1
enzyme is exceptional in that it is the only known PK that displays
hyperbolic kinetics. The other two mammalian PK isozymes, liver and
erythrocyte, are coded by the same PKLR gene through
the use of tissue-specific alternate promoters. Both erythrocyte and
liver isozymes are activated by PEP and FBP (2).
The three-dimensional structures of several PKs from prokaryotic and
eukaryotic organisms have been elucidated (6-10). They reveal a
conserved architecture. PK is a 200-kDa tetramer with four identical
subunits, each consisting of four domains (Fig. 1): the small
N-terminal helical domain (absent in bacterial PKs); the A domain with
( / )8 barrel topology; the B domain, which is inserted
between strand 3 and helix 3 of the A domain
( / )8 barrel; and the C domain with an +
topology. This multidomain architecture is instrumental to the
regulation of PK activity. The enzyme activation is thought to involve
a combination of domain and subunit rotations coupled to alterations in
the active site geometry. In this mechanism, the residues located at
the domain and subunit interfaces are crucial in that they function in
the communication between the activator-binding site and the catalytic center (8-10).
Deficiency of human erythrocyte isozyme (RPK) is, together with
glucose-6-phosphate dehydrogenase deficiency, the most common cause of
nonspherocytic hemolytic anemia (11). RPK deficiency severely affects
the erythrocyte metabolism, causing ATP depletion, which ultimately
leads to hemolysis. Worldwide, more than 150 mutations in the gene
coding RPK have been found in RPK-deficient patients (12). The disease
is transmitted as a recessive trait, and the pathological symptoms
occur only in homozygotes or compound heterozygotes. The clinical
manifestations vary from mild to severe anemia, which can be life
threatening and require continuous transfusion therapy. Here, we
describe the first crystal structure of recombinant RPK and the
biochemical characterization of eight mutants found in patients
subjected to clinical follow-up. These studies allow a correlation
between the clinical symptoms and the molecular properties of the
mutant enzymes.
 |
EXPERIMENTAL PROCEDURES |
Expression Vectors--
The vectors used to express RPK and its
mutant and truncated forms were derivatives of pTrcHisB (Invitrogen).
RPK cDNA insert was obtained from pGG1 (13) after introduction of
NcoI-NdeI sites around the ATG initiation codon.
The mutagenic primer (14) used to introduce the two sites into pGG1 was
5'-CAAGGAGGCTGAAACCATGGCTAGCCAGGAGAACATATCATT. It altered the second and the third codon of the insert. The underlined letters indicate the mutated bases. The selection primer used to
abolish the vector unique restriction site AflIII was
5'-CGCAGGAAAGACCTTGGGAGCAAAAGGCC. The pTrcHisB with RPK cDNA inserted in
NcoI/EcoRI and designated pCW3 was
mutagenized to restore the codons previously changed. The mutagenic primer was
5'-TAAGGAGGAATAAACCATGTCGATCCA
GGAGAACATATCAT. The selection was performed by digesting the parental
pCW3 with NcoI. The new plasmid, which contains the correct
insert, was named pLC1. To obtain the desired RPK mutants, the pLC1 was
subjected to site-directed mutagenesis (14). The same selection primer (5'-CCCCCCTGAATTCGAACCTTGGCTG) was used to abolish the
unique HindIII site in all cases. The specific mutagenic
primers were: 5'-CTGGAGGTGAGCGACAGCATCATGGTGGCA for G332S;
5'-CTGCAACTTGGCGGACAAGCCTGTTGTCTG for G364D;
5'-CAAGCCCCGGCCAATGAGGGCAGAGACAAG for T384M;
5'-AGGGCAGAGACAAGCAATGTCGCCAATGCTG for D390N;
5'-CTGACCACAACTGGCCACTCAGCCCAGCTTCTG for R479H;
5'-AGCCCAGCTTCTGTCTTGGTACCGACCTCGG for R486W;
5'-CTGCCCAGGCTGCCCTCCAGGTCCACTTAT for R504L; and
5'-ATGATGTAAGATCGCTGGGTGCAATTTGGCA for R532W. To obtain a
truncated form of RPK lacking the first 49 residues, pCW3 was
mutagenized by using the primer
5'-TAAGGAGGAATAAACCATGGAGCTGGGCACTGCCTTCTTCC. This sequence
corresponds to that of the plasmid upstream of the initiation triplet
ATG and continues with that of the RPK insert starting from the GAG
codon of Glu at position 50. The selection was performed by digesting
pCW3 with NheI restriction enzyme. The plasmid with the
insert encoding the truncated RPK (residues 50-574) was
designated pLC3. All of the inserts were sequenced.
Protein Purification and Enzymatic
Analysis--
Escherichia coli DH5 transformed with the
specific expression vectors were grown at 37 °C in Luria-Bertani
medium containing 100 µg/ml ampicillin. When the culture optical
density at 600 nm reached a value of 0.5, the expression was induced by
addition of isopropyl- -D-thiogalactopyranoside at a
final concentration of 0.5 mM. The induction time was
12 h, whereas the induction temperature was 30 °C, with the
exception of the mutants G332S, G364D, R504L, and R532W, for which the
induction temperature was 21 °C. Wild-type and mutant enzymes were
purified by the procedure of Wang et al. (13). Enzyme
activities were measured at 37 °C by the assay (13) recommended by
the International Committee for Standardization in Hematology. The
kinetic parameters were determined with the Enzyme Kinetic
ModuleTM 1.1 (SPSS Science Software Gmb). Thermal stability
was measured by incubating the enzyme (100-200 µg/ml) at 53 °C in
a solution consisting of 20 mM potassium phosphate, pH 6.5, 100 mM KCl, and 1 mM EDTA. The samples were
removed at intervals and immediately assayed.
Crystallography--
Recombinant wild-type RPK was crystallized
using the vapor diffusion method at 22 °C. Well solutions consisted
of 50 mM Mes/KOH, pH 6.4, 10 mM
MnSO4, and 10-14% w/v PEG8000. Hanging drops were formed
by mixing equal volumes of 12 mg/ml protein in 50 mM KCl, 5 mM FBP, 5 mM phosphoglycolate, 20 mM potassium phosphate, pH 7.0, and well solutions. The
crystals were difficult to reproduce. The recombinant enzyme undergoes
partial proteolysis, and about 50% of the purified protein chains lack
the first 47 amino acids (13). On this basis, we produced a mutant
truncated enzyme lacking the first 49 residues (see above). Employment
of the truncated protein greatly improved the crystallization, which
was carried out using the above-described protocol. Crystals were
obtained for the T384M, R479H, and R486W truncated mutants by the same protocol used for the truncated wild-type RPK.
RPK crystals belong to space group P21 with one tetramer in
the asymmetric unit. The diffraction data were measured at 100 K on
beamline ID14-EH2 of the European Synchrotron Radiation Facility (Grenoble, France) using a MarCCD detector and beamline B7WB of DESY/EMBL (Hamburg, Germany) using a Mar Imaging Plate. Data processing and reduction were carried out using MOSFLM (15) and programs of the
CCP4 suite (16). The data collection statistics are reported in Table
I. The structure of the wild-type RPK was
solved by molecular replacement using the program Molrep (16). The
search model was the structure of rabbit muscle M1 PK in complex with pyruvate (Ref. 7; Protein Data Bank entry 1PKN). The Phases were
improved by 4-fold averaging (17), producing an electron density of
excellent quality. Model building was carried out with the program O
(18). The model was refined using Refmac (19). All of the measured data
(no cut-off) were employed, and 2.5% of unique reflections were
used to monitor the progress of the refinement by
Rfree validation. The refined wild-type
coordinates provided the starting model for the refinement of the
mutants. The set of reflections for calculation of
Rfree was identical to that of the wild-type
structure refinement. A summary of refinement statistics is presented
in Table I. Analysis and inspection of the structures were carried out
with the program O (18) and programs of the CCP4 package (16). The
figures were generated with Molscript (20).
 |
RESULTS |
The Three-dimensional Structure of RPK--
The crystallographic
studies were performed using a truncated RPK in which the 49 N-terminal
amino acids are absent. Use of the truncated protein resulted in
considerable improvement in the reproducibility of the crystallization
experiments. The truncated protein exhibits kinetic properties
virtually identical to those of wild-type RPK. A more detailed analysis
of this and other mutants targeting the N-terminal residues will be
published elsewhere.
The truncated recombinant RPK was crystallized in complex with
phosphoglycolate (a PEP analogue), FBP, Mn2+, and
K+. The presence of the allosteric activators implies that
the crystalline enzyme is in the active R state. The 2.7 Å resolution
structure of RPK reveals the typical four-domain subunit architecture
found in all PKs of known three-dimensional structure (Fig.
1A). The A (residues 85-159
and 263-431) and C domains (residues 432-574), together with the
small N-terminal domain (residues 57-84), form the main body of the
subunit. The B domain (residues 160-262) is loosely packed to the rest
of the molecule and adopts slightly different orientations (about 4°)
in the four crystallographically independent polypeptide chains. The
four subunits of the RPK tetramer are assembled to form a
D2 symmetric oligomer. The intersubunit interactions define
two large contact areas; the A/A' interface involves the A domains of
subunits related by the vertical 2-fold axis, as defined in Fig.
1B, whereas the C/C' interface involves the C domains of
subunits interacting along the horizontal axis.

View larger version (40K):
[in this window]
[in a new window]
|
Fig. 1.
Three-dimensional crystal structure of
RPK. The N-terminal domain is yellow, the A domain is
red, the B domain is cyan, and the C domain is
green. A, the RPK subunit. The gray
spheres indicate the C atoms of the residues subjected to
mutagenesis. B, the RPK tetramer. In this
orientation, a molecular 2-fold axis is perpendicular to the plane of
the paper, whereas the other two molecular 2-fold axes are vertical and
horizontal to the plane of the paper (indicated with
vertical and horizontal lines,
respectively).
|
|
The structure of the RPK subunit closely resembles that of rabbit
muscle M1 PK, as expected from the 59% sequence identity between the
two proteins. The similarity is highest with M1 PK in complex with
pyruvate (7), with a root-mean-square difference of 1.2 Å for 512 C
atom pairs. This M1 PK complex exhibits the same B domain orientation
found in RPK. In other M1 structures, the B domain is either more open,
as in the phospholactate complex (21), or more closed, as in the
complex with ATP (22, 23).
The Allosteric Site and the Catalytic Center--
RPK was
cocrystallized with FBP, phosphoglycolate, and the K+ and
Mn2+ ions. All ligands are clearly visible in the electron
density map. Phosphoglycolate, a potent PK inhibitor (9), is positioned in the PEP-binding site, which is located at the top of the A domain
( / )8 barrel, facing a cleft between the A and B
domains (Fig. 1A). It is at the heart of an intricate
network of hydrogen bonds, which involve protein residues and the
Mn2+ and K+ cations (Fig.
2A). The phosphate group is
bound to the K+ atom and the side chain of
Arg116, whereas the carboxylate moiety is anchored
through interactions with the Mn2+ ion, the side chain of
Thr371, and the main chain nitrogen atoms of
Gly338 and Asp339, which are located at the N
terminus of a short helical segment belonging to loop 6 of the A domain
( / )8 barrel. This binding geometry is identical to
that observed in the yeast PK-phosphoglycolate complex (9) and closely
resembles the binding of pyruvate and phospholactate to rabbit M1 PK
(for a discussion of the implications of this binding mode for
catalysis see Refs. 7 and 21). These similarities are in keeping with
the strict conservation among the PK sequences of all residues
surrounding the substrate-binding site.

View larger version (42K):
[in this window]
[in a new window]
|
Fig. 2.
The allosteric and catalytic sites of RPK.
A, stereo view of the active site with bound
phosphoglycolate (indicated by gray bonds),
Mn2+, and K+. With respect to Fig.
1A, the model has been rotated by ~30° around an axis
horizontal to the plane of the paper. B, stereo view of the
allosteric site with bound FBP (gray bonds). The orientation
is as in Fig. 1A.
|
|
The FBP activator is hosted in the allosteric site in the C domain
(Fig. 1A). The ligand is sandwiched between loops 475-479 and 557-566 (Fig. 2B) and extensively interacts with
protein. The 6'-phosphate group is engaged in a salt bridge with
Arg532, whereas the 1'-phosphate is hydrogen-bonded to the
side chains of Thr475 and Ser480 and the
backbone nitrogen atoms of Thr476 and Thr477.
Moreover, the dipole of -helix 480-486 points toward the
1'-phosphate, further compensating the ligand negative charge. This
geometry in FBP-binding is identical to that found in yeast PK
crystallized in the R state (9).
Rationale for the Mutagenesis Studies--
A survey of the
missense mutations associated with the nonspherocytic hemolytic anemia
shows that most of them cluster in specific regions of the protein
three-dimensional structure: the interface between the A and C domains,
the A/A' intersubunit interface, the hydrophobic core of the A domain,
and the FBP-binding site (24). We generated eight RPK mutants (Fig.
1A), targeting residues belonging to each of these regions
of the protein. Almost all selected mutations have been found in
homozygote patients. The kinetic, allosteric, and thermostability
parameters of mutant proteins were evaluated (Table
II), and the crystal structures of
three mutants (T384M, R479H and R486W) were solved. The mutations did
not induce significant conformational changes in the overall protein
conformation, and therefore we shall restrict the description of the
mutant structures mainly to the sites affected by the mutations.
View this table:
[in this window]
[in a new window]
|
Table II
Kinetic parameters of the wild-type and mutant RPKs
The results are the means ± S.E. for three determinations from
four different protein preparations.
|
|
G332S Mutation in the A Domain Hydrophobic Core--
Many RPK
mutations are localized in the hydrophobic core of the A domain. An
example is the G332S mutation (nucleotide G994 A),
which affects a residue that is strictly conserved among PK sequences.
Gly332 is located on strand 6, being buried inside the
domain core (Fig. 1A). The G332S protein exhibits a 9-fold
decrease in the catalytic efficiency (5-fold for the FBP activated
protein) and is considerably less thermostable than the wild-type
enzyme (Table II and Fig. 3). These
substantially altered molecular properties account for the clinical
data. In homozygous form, the G332S mutation leads to severe anemia
with the need of regular transfusions (25, 26).

View larger version (18K):
[in this window]
[in a new window]
|
Fig. 3.
Characterization of RPK mutants. ,
wild type; , G332S; , G364D; , T384M; , D390N; , R479H;
, R486W; , R532W. A, steady state kinetics of
wild-type and mutant RPKs as a function of PEP. B, thermal
stability of wild-type and mutant RPKs. The residual activity after
incubation at 53 °C is expressed as a percentage of the initial
activity.
|
|
Mutations at the A/C Interface: G364D,
R486W, and R504L--
The interface between the A and C domains is
characterized by many polar interactions that often involve charged
side chains. Many of these residues represent sites of pathological
mutations, which cause RPK deficiency at variable levels of severity.
The mutation R504L (nucleotide G1511 T) affects
Arg504, a C domain residue that is partly
solvent-accessible and engaged in an interdomain salt bridge with
Asp281 (Fig. 1A). The R504L mutation removes
this interdomain interaction and introduces a hydrophobic Leu side
chain in a solvent exposed site close to a negatively charged Asp. Such
amino acid replacement is clearly unfavorable, providing a reason for
the extreme instability of the protein, which prevented functional
analysis (Table II). This feature explains the severe anemia found in
RPK-deficient patients homozygous for this mutation (27).
The other two investigated mutants targeting the A/C interface affect
Gly364 and Arg486, which are part of a region
of close association between the A and C domains (Figs. 1A
and 4A). Arg486 is
hydrogen-bonded to the carbonyl oxygen of Leu362 at the C
terminus of the A domain helix 6, whereas the neighboring Gly364 allows a sharp turn of the polypeptide chain with a
backbone conformation ( = 85°, = 98°) that is
unfavorable for a nonglycine residue. The G364D (nucleotide
G1091 A) mutation has a drastic effect on the enzyme
stability, which is coupled to a 3-fold reduction of the catalytic
efficiency (Fig. 3B and Table II). Given the tightly packed
environment and the backbone conformation of Gly364, it is
conceivable that introduction of a charged Asp side chain at this site
of the A/C interface can greatly perturb the domain assembly, thus
being deleterious for stability. Fully consistent with these
observations is the severe anemia found in patients homozygous for
G364D (28). Together with R504L, the G364D mutant highlights the notion
that the interdomain interactions at the A/C interface are critical for
the stability of the protein.

View larger version (30K):
[in this window]
[in a new window]
|
Fig. 4.
The A/C interface in the region surrounding
Arg486. The orientation is as in Fig. 1A.
A, stereo diagram of the wild-type structure. B,
stereo diagram of the R486W mutant structure.
|
|
R486W (nucleotide C1456 T) is among the most frequent
mutations found in RPK-deficient patients (11). Characterization of this mutant reveals that such a drastic amino acid replacement results
in small effects on the molecular properties. The mutant three-dimensional structure shows that the Trp side chain is
accommodated without any structural perturbation. With respect to the
wild-type structure, no atomic movements larger than 0.25 Å can be
detected, whereas the indole nitrogen atom is able to establish a
hydrogen bond with the carbonyl oxygen atoms of Leu362 and
Asp361. Such structural conservation matches the limited
changes in biochemical parameters. The thermostability is even higher
than that of the wild-type protein (Fig. 3B), and the
allosteric properties are essentially unmodified (Table II and Fig.
3A). The only significant perturbation is in the catalytic
efficiency, which drops to 30% of the value measured for the wild-type
RPK (Table II). These moderate variations in the molecular parameters
correlate with the clinical symptoms because patients homozygous for
the R486W mutation generally exhibit a mild anemia (25).
The perturbed kinetics of the R486W protein is puzzling because
Arg486 is >20 Å away from the catalytic center (Fig.
1A), which is left unperturbed by the mutation as shown by
the mutant crystal structure. The "long range" effect of the R486W
mutation might reflect altered dynamic properties. It is known that the
B domain adopts different conformations depending on ligand binding
(21-23). The introduction of the Trp aromatic ring may restrict the
overall ability of the enzyme to undergo the conformational changes
occurring during catalysis, thereby perturbing the reaction kinetics.
The A/A' Interface: T384M and
D390N--
Asp390 is a solvent-inaccessible residue
located in the A/A' interface, at the heart of a hydrogen bond network
that involves Arg337 and Ser389' (the prime
symbol denotes a residue of a different subunit). Based on the
comparison between the structures of the T state E. coli PK and of the M1 isozyme, it was found that
Asp390 is crucial for the allosteric transition by coupling
changes in the quaternary structure with alterations in the active site geometry (8). A pathological mutation affecting this residue (D390N
corresponding to nucleotide G1168 A) has been detected
in a heterozygote patient (25). The molecular analysis shows that the
D390N amino acid replacement causes the almost complete inactivation of
the protein, which, however, is not less thermostable than the
wild-type RPK (Table I and Fig. 3B). These results are very
similar to those obtained with the E. coli enzyme for which
the same mutation was investigated (29). These observations support the
idea that Asp390 has a key role in the allosteric
regulation, suggesting that the D390N mutation may lock the protein in
an inactive conformation, impairing the transition to the R state.
The mutation T384M (nucleotide C1151 T) affects a
residue, which, although not directly involved in intersubunit
interactions, lays very close to the A/A' molecular 2-fold axis (Fig.
5). Thr384 is located at the
N terminus of helix 7 of the A domain ( / )8 barrel.
Its OG atom is hydrogen-bonded to the backbone nitrogen atoms of
Ala386 and Glu387, thus acting as helix-capping
element. The three-dimensional structure of the T384M mutant reveals
that the mutation does not cause atomic shifts larger than 0.3 Å (Fig.
5). The bulkier Met side chain is easily accommodated, the only change
being the removal of the helix-capping hydrogen bonds. Also the kinetic
characterization shows limited variations, the main difference between
T384M and the wild-type protein being a 3-fold reduction of the
catalytic efficiency (5-fold for the FBP activated form; Table II) that is mainly accounted for by a reduction in kcat.
Likewise, the mutation does not alter the thermostability parameters
(Fig. 3).

View larger version (32K):
[in this window]
[in a new window]
|
Fig. 5.
The A/A' interface close to
Thr384. The helices 7 of 2-fold related subunits
are shown. The residues of the opposite subunits are denoted by the
prime symbol. Attached to Thr384 is the Met side
chain (gray bonds) of the crystal structure of the T384M
mutant. The orientation is as in Fig. 1A.
|
|
Thr384 is not part of the binding sites for PEP and ADP,
and the crystal structure of the T384M protein shows that the active site geometry is not affected by the mutation. Thus, the altered kinetics displayed by the T384M mutant is difficult to rationalize. Modified enzymatic parameters were observed also for the equivalent mutation in the rabbit kidney isozyme (30). Thr384 is close
to the contact region between the A and B domains (Fig. 1A)
and therefore the T384M mutation may disturb the "closure" of the B
domain occurring upon ATP binding (22). An alteration of the
equilibrium between the "open" and "closed" B domain
conformations may affect the enzymatic activity. It is remarkable that
homozygosity for the T384M mutation is associated with anemia with mild
to severe symptoms (31, 32), implying that even moderate changes in the
enzyme catalytic power can have pathological effects.
The Allosteric Site: R532W and R479H--
The negative charges of
FBP are compensated by the N terminus of helix 479-486 for the
1'-phosphate and Arg532 side chain for the 6'-phosphate
(Fig. 2B). We have investigated two mutations that target
both of the elements involved in FBP binding. The first of these
mutations is R532W (nucleotide C1594 T), which has been
found in compound heterozygotes, in which the other allele had a
mutation causing the truncation of the protein. The clinical symptoms
in the patients carrying the mutation were severe (33). Molecular
analysis of the R532W protein indicates a complete loss in the
responsiveness to FBP, highlighting the essential
Arg532 role in activator binding (Table II). These
perturbed allosteric properties are associated with a decreased
thermostability (Fig. 3B), possibly reflecting the
energetically unfavorable exposure on the protein surface of the
hydrophobic Trp residue.
The R479H mutation has been found in RPK-deficient patients affected by
severe anemia (34, 35). The side chain of Arg479 is located
in the neighborhood of FBP, although it does not directly interact with
the activator (Fig. 2B). The crystal structure of R479H is
identical to that of the wild-type protein, with the His side chain
being fully solvent-exposed. Similarly, the kinetic parameters (Table
II) appear to be essentially unaffected by the mutation. These features
are in contrast with the severe clinical symptoms (34, 35). An
explanation for this riddle is given by the observation that the
mutation affects nucleotide 1436, which is located on a splicing site
at the 3'-end of exon 10. This fact, together with our biochemical
analysis, suggests that, rather than the amino acid replacement,
defects in mRNA splicing process are the actual cause of the RPK deficiency.
 |
DISCUSSION |
Implications for the Allosteric Regulation--
PK is a typical
allosteric enzyme of the K type. The allosteric signal is transmitted
across the long distance (>20 Å) separating the FBP-binding site from
the catalytic center. The exact mechanism of the allosteric transition
is not known in detail because no PK has been so far crystallized in
both the T and R states. The comparison between the structure of the T
state E. coli PK and that of the rabbit M1 enzyme (8)
suggested that the allosteric transition involves modifications in the
relative orientations of the domains and subunits coupled to
conformational changes in the PEP-binding site. The x-ray analysis of
the T state Leishmania mexicana PK (10) and the R state
yeast PK (9) led to a further refinement of this model, allowing a
better discrimination between the structural differences that are
consequences of the inherent divergence between eukaryotic and
prokaryotic proteins and the conformational changes that are genuinely
caused by the allosteric transition. The mechanism of PK regulation has
also been the subject of many mutagenesis experiments (Refs. 23 and 36
and references therein). The general picture emerging from the mutant
analysis is that the intersubunit interactions at the A/A' and C/C'
interfaces and the interdomain interactions at the A/B interface are
key to determining the allosteric responsiveness and to defining
the distribution of the conformations between active and inactive states. Moreover, the mutagenesis analyses combined with the
crystallographic data provide clear evidence for the idea that the T
and R forms correspond to ensembles of conformations characterized by
the rotational flexibility of the B domain (23).
Our study on human RPK is fully consistent with these features. The
crystal structure shows that the B domain is flexible in RPK also,
adopting different orientations in the crystallographically independent
subunits. The key functional role of the A/A' interface is highlighted
by the D390N mutation, which targets a residue located in the core of
the A/A' interface, producing an enzyme that retains a stable
tetrameric state but almost entirely lacks enzymatic activity.
Conversely, none of the mutations targeting residues at the A/C
interface alters the enzyme allosteric properties. Thus, in agreement
with the recent mutagenesis data on the yeast PK (36), this domain
interface appears to have little role in the transduction of the
allosteric signal; rather it is important for the stability of the
domain assembly within the enzyme subunit.
Molecular Basis of Nonspherocytic Hemolytic
Anemia--
Characterization of mutant proteins shows that amino acid
substitutions can affect thermostability, catalytic efficiency, and
response to the allosteric effector. Various regions of the RPK
structure, including domain interfaces and functional sites, are
affected by the pathological mutations (Fig. 1A). However, there appears to be no relation between the nature and location of the
replaced amino acid and the type of molecular perturbation. For
instance, both R504L and R486W mutations affect Arg residues involved
in interdomain polar interactions at the A/C interface, but their
effects are substantially different. The R504L protein is extremely
unstable, whereas the R486W mutant is even slightly more
thermoresistant than the wild type (Fig. 3B). These
observations emphasize the difficulty of predicting the consequences of
mutations simply from the location and the nature of the target
residues. They also warn against predictions of the effects of
mutations in human RPK based on the molecular analysis of other PK isozymes.
The clinical manifestations of a genetic disease reflect the
interactions of a variety of physiological and environmental factors
and do not solely depend on the molecular properties of the altered
molecule. Given this caution, it is evident that there is a general
correlation between the clinical manifestations and the biochemical
parameters of the mutant proteins. Mutants exhibiting strongly
perturbed kinetic and thermostability parameters (G332S, G364D, R504L,
and R532W) are associated with severe RPK deficiency. Conversely, in
the case of less abnormal molecular properties, the disease has milder
manifestations. It is remarkable that pathological conditions are
present in association with mutations such as T384M or R486W, which are
simply characterized by a moderate reduction of the catalytic
efficiency. The physiological concentrations of RPK substrates and
effectors are in the micromolar range (37). Therefore, in
vivo RPK operates in subsaturating conditions that may amplify the
effects of the different catalytic efficiencies between the wild-type
and mutant proteins (Fig. 3A).
In conclusion, our studies indicate that the functional parameters of
RPK are so finely tuned that even moderate molecular alterations may
significantly perturb cell metabolism. The correlation between
molecular and clinical parameters in PK deficiency suggests that
biochemical characterization of mutant proteins may serve as a valuable
tool to understand and assist with diagnosis and genetic counseling.
 |
ACKNOWLEDGEMENTS |
We thank the staffs of the synchrotron beam
lines of DESY/EMBL and ESRF for help during data collection.
 |
FOOTNOTES |
*
This work was supported by grants from University of Pavia
Progetto d'Ateneo "Nuove Tecnologie Molecolari e Cellulari,"
Ministero della Ricerca Scientifica e Tecnologica Progetto Genomica
Funzionale, and Allos Therapeutics Inc. (Westminster, CO).The costs of publication of this
article were defrayed in part by the
payment of page charges. The 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 the structure factors (code 1LIU, 1LIW, 1LIX, and 1LIY) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).
¶
To whom correspondence may be addressed. Fax: 39-0382-423108;
E-mail: giovale@unipv.it.

To whom correspondence may be addressed. Fax: 39-0382-528496;
E-mail: mattevi@ipvgen.unipv.it.
Published, JBC Papers in Press, April 17, 2002, DOI 10.1074/jbc.M202107200
 |
ABBREVIATIONS |
The abbreviations used are:
PK, pyruvate kinase;
RPK, human erythrocyte pyruvate kinase;
PEP, phosphoenolpyruvate;
FBP, fructose 1,6-bisphosphate;
Mes, 4-morpholineethanesulfonic
acid.
 |
REFERENCES |
| 1.
|
Kayne, F. J.
(1973)
The Enzymes
, 3rd Ed.
, pp. 353-382, Academic Press Inc., New York
|
| 2.
|
Fothergill-Gilmore, L. A.,
and Michels, P. A.
(1992)
Prog. Mol. Biol. Biophys.
59,
105-227
|
| 3.
|
Gupta, R. K.,
Oesterling, R. M.,
and Mildvan, A. S.
(1976)
Biochemistry
15,
2881-2887[CrossRef][Medline]
[Order article via Infotrieve]
|
| 4.
|
Mesecar, A. D.,
and Nowak, T.
(1997)
Biochemistry
36,
6792-6802[CrossRef][Medline]
[Order article via Infotrieve]
|
| 5.
|
Hall, E. R.,
and Cottam, G. L.
(1978)
Int. J. Biochem.
9,
785-793[CrossRef][Medline]
[Order article via Infotrieve]
|
| 6.
|
Allen, S. C.,
and Muirhead, H.
(1996)
Acta Crystallogr. Sect. D Biol. Crystallogr.
52,
499-504[CrossRef][Medline]
[Order article via Infotrieve]
|
| 7.
|
Larsen, T. M.,
Laughlin, L. T.,
Holden, H. M.,
Rayment, I.,
and Reed, G. H.
(1994)
Biochemistry
33,
6301-6309[CrossRef][Medline]
[Order article via Infotrieve]
|
| 8.
|
Mattevi, A.,
Valentini, G.,
Rizzi, M.,
Speranza, M. L.,
Bolognesi, M.,
and Coda, A.
(1995)
Structure
3,
729-741[Medline]
[Order article via Infotrieve]
|
| 9.
|
Jurica, M. S.,
Mesecar, A.,
Heath, P. J.,
Shi, W.,
Nowak, T.,
and Stoddard, B. L.
(1998)
Structure
6,
195-210[Medline]
[Order article via Infotrieve]
|
| 10.
|
Rigden, D. J.,
Phillips, S. E. V.,
Michels, P. A. M.,
and Fothergill-Gilmore, L. A.
(1999)
J. Mol. Biol.
291,
615-635[CrossRef][Medline]
[Order article via Infotrieve]
|
| 11.
|
Zanella, A.,
and Bianchi, P.
(2000)
Baillieres Best Pract. Res. Clin. Haematol.
13,
57-81[Medline]
[Order article via Infotrieve]
|
| 12.
|
Bianchi, P.,
and Zanella, A.
(2000)
Blood Cells Mol. Dis.
26,
47-53[CrossRef][Medline]
[Order article via Infotrieve]
|
| 13.
|
Wang, C. Q.,
Chiarelli, L. R.,
Bianchi, P.,
Abraham, D. J.,
Galizzi, A.,
Mattevi, A.,
Zanella, A.,
and Valentini, G.
(2001)
Blood
98,
3113-3120[Abstract/Free Full Text]
|
| 14.
|
Nickoloff, J. A.,
Deng, W. P.,
Miller, E. M.,
and Ray, F. A.
(1996)
Methods Mol. Biol.
58,
455-468[Medline]
[Order article via Infotrieve]
|
| 15.
|
Leslie, A. G.
(1999)
Acta Crystallogr. Sect. D Biol. Crystallogr.
55,
1696-1702[CrossRef][Medline]
[Order article via Infotrieve]
|
| 16.
|
Collaborative Computational Project Number 4.
(1994)
Acta Crystallogr. Sect. D Biol. Crystallogr.
50,
760-767[CrossRef][Medline]
[Order article via Infotrieve]
|
| 17.
|
Cowtan, K. D.,
and Main, P.
(1996)
Acta Crystallogr. Sect. D Biol. Crystallogr.
52,
43-48[CrossRef][Medline]
[Order article via Infotrieve]
|
| 18.
|
Jones, T. A.,
Zou, J. Y.,
Cowan, S. W.,
and Kjeldgaard, M.
(1991)
Acta Crystallogr. Sect. A
47,
110-119
|
| 19.
|
Murshudov, G. N.,
Vagin, A. A.,
and Dodson, E. J.
(1997)
Acta Crystallogr. Sect. D Biol. Crystallogr.
53,
240-255[CrossRef][Medline]
[Order article via Infotrieve]
|
| 20.
|
Kraulis, P. J.
(1991)
J. Appl. Crystallogr.
24,
946-950[CrossRef]
|
| 21.
|
Larsen, T. M.,
Benning, M. M.,
Wesenberg, G. E.,
Rayment, I.,
and Reed, G. H.
(1997)
Arch. Biochem. Biophys.
345,
199-206[CrossRef][Medline]
[Order article via Infotrieve]
|
| 22.
|
Larsen, T. M.,
Benning, M. M.,
Rayment, I.,
and Reed, G. H.
(1998)
Biochemistry
37,
6247-6255[CrossRef][Medline]
[Order article via Infotrieve]
|
| 23.
|
Wooll, J. O.,
Friesen, R. H.,
White, M. A.,
Watowich, S. J.,
Fox, R. O.,
Lee, J. C.,
and Czerwinski, E. W.
(2001)
J. Mol. Biol.
312,
525-540[CrossRef][Medline]
[Order article via Infotrieve]
|
| 24.
|
Mattevi, A.,
Bolognesi, M.,
and Valentini, G.
(1996)
FEBS Lett.
389,
15-19[CrossRef][Medline]
[Order article via Infotrieve]
|
| 25.
|
Zanella, A.,
Bianchi, P.,
Baronciani, L.,
Zappa, M.,
Bredi, E.,
Vercellati, C.,
Alfinito, F.,
Pelissero, G.,
and Sirchia, G.
(1997)
Blood
89,
3847-3852[Abstract/Free Full Text]
|
| 26.
|
Lenzner, C.,
Nurnberg, P.,
Thiele, B. J.,
Reis, A.,
Brabec, V.,
Sakalova, A.,
and Jacobasch, G.
(1994)
Blood
83,
2817-2822[Abstract/Free Full Text]
|
| 27.
|
Demina, A.,
Varughese, K. I.,
Barbot, J.,
Forman, L.,
and Beutler, E.
(1998)
Blood
92,
647-652[Abstract/Free Full Text]
|
| 28.
|
van Solinge, W. W.,
Kraaijenhagen, R. J.,
Rijksen, G.,
van Wijk, R.,
Stoffer, B. B.,
Gajhede, M.,
and Nielsen, F. C.
(1997)
Blood
90,
4987-4995[Abstract/Free Full Text]
|
| 29.
|
Valentini, G.,
Chiarelli, L.,
Fortin, R.,
Speranza, M. L.,
Galizzi, A.,
and Mattevi, A.
(2000)
J. Biol. Chem.
275,
18145-18152[Abstract/Free Full Text]
|
| 30.
|
Friesen, R. H.,
and Lee, J. C.
(1998)
J. Biol. Chem.
273,
14772-14779[Abstract/Free Full Text]
|
| 31.
|
Kanno, H.,
Fujii, H.,
Hirono, A.,
and Miwa, S.
(1991)
Proc. Natl. Acad. Sci. U. S. A.
88,
8218-8221[Abstract/Free Full Text]
|
| 32.
|
Neubauer, B.,
Lakomek, M.,
Winkler, H.,
Parke, M.,
Hofferbert, S.,
and Schröter, W.
(1991)
Blood
77,
1871-1875[Abstract/Free Full Text]
|
| 33.
|
Lakomek, M.,
Huppke, P.,
Neubauer, B.,
Pekrun, A.,
Winkler, H.,
and Schroter, W.
(1994)
Ann. Hematol.
69,
253-260[CrossRef][Medline]
[Order article via Infotrieve]
|
| 34.
|
Kanno, H.,
Ballas, S. K.,
Miwa, S.,
Fujii, H.,
and Bowman, H. S.
(1994)
Blood
83,
2311-2316[Abstract/Free Full Text]
|
| 35.
|
Kanno, H.,
Wei, D. C.,
Chan, L. C.,
Mizoguchi, H.,
Ando, M.,
Nakahata, T.,
Narisawa, K.,
Fujii, H.,
and Miwa, S.
(1994)
Blood
84,
3505-3509[Abstract/Free Full Text]
|
| 36.
|
Fenton, A. W.,
and Blair, J. B.
(2002)
Arch. Biochem. Biophys.
397,
28-39[CrossRef][Medline]
[Order article via Infotrieve]
|
| 37.
|
Beutler, E.
(1984)
Red Cell Metabolism: A Manual of Biochemical Methods
, Grune and Stratton, New York
|
| 38.
|
Laskowski, R. A.,
MacArthur, M. W.,
Moss, D. S.,
and Thornton, J. M.
(1993)
J. Appl. Crystallogr.
26,
283-291[CrossRef]
|
Copyright © 2002 by The American Society for Biochemistry and Molecular Biology, Inc.

CiteULike Complore Connotea Del.icio.us Digg Reddit Technorati What's this?
This article has been cited by other articles:

|
 |

|
 |
 
A. Zanella, P. Bianchi, and E. Fermo
Pyruvate kinase deficiency
Haematologica,
June 1, 2007;
92(6):
721 - 723.
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
R. van Wijk and W. W. van Solinge
The energy-less red blood cell is lost: erythrocyte enzyme abnormalities of glycolysis
Blood,
December 15, 2005;
106(13):
4034 - 4042.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
A. Diez, F. Gilsanz, J. Martinez, S. Perez-Benavente, N. W. Meza, and J. M. Bautista
Life-threatening nonspherocytic hemolytic anemia in a patient with a null mutation in the PKLR gene and no compensatory PKM gene expression
Blood,
September 1, 2005;
106(5):
1851 - 1856.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
L. R. Chiarelli, P. Bianchi, E. Fermo, A. Galizzi, P. Iadarola, A. Mattevi, A. Zanella, and G. Valentini
Functional analysis of pyrimidine 5'-nucleotidase mutants causing nonspherocytic hemolytic anemia
Blood,
April 15, 2005;
105(8):
3340 - 3345.
[Abstract]
[Full Text]
[PDF]
|
 |
|
Copyright © 2002 by the American Society for Biochemistry and Molecular Biology.
|
Advertisement
Advertisement
|