(Received for publication, August 16, 1996, and in revised form, December 12, 1996)
From the Department of Veterans Affairs Medical Center, To investigate the role in catalysis and/or
substrate binding of the Walker motif residues of rat testis fructose
6-phosphate,2-kinase:fructose-2,6-bisphosphatase (Fru
6-P,2-kinase:Fru-2,6-Pase), we have constructed and characterized mutant enzymes of Asp-128, Thr-52, Asn-73, Thr-130, and Tyr-197. Replacement of Asp-128 by Ala, Asn, and Ser resulted in a small decrease in Vmax and a significant increase in
Km values for both substrates. These mutants
exhibited similar pH activity profiles as that of the wild type enzyme.
Mutation of Thr-52 to Ala resulted in an enzyme with an infinitely high
Km for both substrates and an 800-fold decreased
Vmax. Substitution of Asn-73 with Ala or Asp
caused a 100- and 600-fold increase, respectively in
KFru 6-P with only a small increase in
KATP and small changes in
Vmax. Mutation of Thr-130 caused small changes in the kinetic properties. Replacement of Tyr-197 with Ser resulted in
an enzyme with severely decreased binding of Fru 6-P with 3-fold decreased Vmax. A fluorescent analog of ATP,
2 Fru 2,6-P21 is the most
potent activator of phosphofructokinase (PFK), and its synthesis and
degradation are catalyzed by a bifunctional enzyme, Fru
6-P,2-kinase:Fru-2,6-bisphosphatase (Fru 6-P + ATP The reaction catalyzed by Fru 6-P,2-kinase follows ternary complex
formation (8) with direct transfer of the The Fru 6-P,2-kinase reaction is very similar to that catalyzed by PFK
(Fru 6-P + ATP Recently, we have crystallized (19) and solved the three-dimensional
structure of rat testis Fru 6-P,2-kinase:Fru-2,6-Pase (RT2K) complexed
with Mg2+-ATP (20). These results demonstrated that the Fru
6-P,2-kinase domain of the enzyme does not resemble the structure of
PFK as had been presumed (21), but rather this domain is structurally similar to the family of nucleoside monophosphate kinases (NMP kinases), including adenylate kinase (22) and uridylate kinase (23).
The Mg2+-ATP binding regions of the NMP kinases and the Fru
6-P,2-kinase domain are remarkably similar, consisting of classical
Walker-A (GXXGXGKT) and -B (ZZZD;
where Z represents a hydrophobic amino acid) motifs (24),
which comprise a phosphate binding loop and a
Mg2+-coordinating aspartate residue, respectively. In the
Fru 6-P,2-kinase domain, the conserved Lys and Thr residues of the
Walker-A motif are Lys-51 and Thr-52, while Asp-128 is the conserved
residue of the Walker-B motif. No Fru 6-P was observed in the crystal structure, but the Fru 6-P binding site can be accurately predicted both by homology with the NMP binding sites of the NMP kinases and by
the location of residues previously identified to affect Fru 6-P
binding (12-14). By modeling Fru 6-P in this presumed substrate binding site, we have been able to search for a potential nucleophile for the activation of the 2-OH of Fru 6-P. With the exception of
Asp-128, there is no Asp or Glu in the active site that might fulfill
this role. We have identified Asn-73, Thr-130, and Tyr-197 as the only
amino acids in the vicinity of the binding site that might potentially
act as weak nucleophiles and/or assist in substrate binding.
To investigate the role in catalysis and/or substrate binding of the
conserved Walker motif residues and the weak nucleophiles in the
substrate binding site, we have prepared mutant enzymes at Asp-128,
Thr-52, Asn-73, Thr-130, and Tyr-197. The kinetic properties and the
nucleotide binding of these mutant enzymes have been determined and are
compared with the WT enzyme.
Rabbit muscle PFK was prepared as described (25). The cDNA
encoding RT2K was prepared as described (7). Restriction enzymes, T4
DNA ligase, and T4 polynucleotide kinase were purchased from New
England BioLabs (Beverly, MA). The Muta-Gene M13 in vitro mutagenesis kit was purchased from Bio-Rad. The pT7-7 RNA
polymerase/promoter plasmid (26) was a gift of Dr. Stan Tabor (Harvard
Medical School). The Sequenase version 2 sequencing kit was purchased
from U.S. Biochemical Corp. mant-ATP and mant-ADP were synthesized by
the method of Hiratsuka (27) and purified by HPLC on a Partisil 10 SAX
column (86 × 250 mm; Whatmann, Hillsboro, OR), eluting with 0.6 M ammonium phosphate (pH 4.0) (28). The product was an
equilibrium mixture of 70% 3 Plasmid RT2K/pT7-7 containing
the RT2K gene cloned in a pT7-7 vector (7) was digested with
XbaI and HindIII, and the isolated 1.7-kilobase
pair fragment was ligated into the XbaI-HindIII
site of M13mp18 (M13/RT2K). The ligation mixture was used to transform E. coli JM109 competent cells. The phage harboring M13/RT2K
was harvested and transfected into E. coli CJ236
(dut
Oligonucleotides used for mutagenesis
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
Acknowledgments
REFERENCES
(3
)-O-(N-methylanthraniloyl)ATP (mant-ATP)
served as a substrate with Km = 0.64 µM, and Vmax = 25 milliunits/mg
and was a competitive inhibitor with respect to ATP. When mant-ATP
bound to the enzyme, fluorescence intensity at 440 nm increased.
mant-ATP binding of the wild type and the mutant enzymes were compared
using the fluorometric method. The Kd values of the
T52A and D128N enzymes were infinitely high and could not be measured,
while those of the other mutant enzymes increased slightly. These
results provide evidence that those amino acids are involved in
substrate binding, and they are consistent with the crystallographic
data. The results also suggest that Asp-128 does not serve as a
nucleophile in catalysis, and since there are no other potential
nucleophiles in the active site, we hypothesize that the Fru
6-P,2-kinase reaction is mediated via a transition state stabilization
mechanism.
Fru
2,6-P2 + ADP and Fru 2,6-P2
Fru 6-P + Pi). Several isozymic forms of the enzyme from mammalian
tissues have been characterized (1). They are all homodimers with
Mr ranging from 108,000 to 120,000. The primary
structures of these enzymes revealed that the catalytic domains are
highly conserved, and the kinase and the phosphatase domains reside in
the N-terminal half and the C-terminal half, respectively (2-7).
-phosphate of ATP to the
2-OH of
-D-Fru 6-P (9). Little is known about the amino
acid residues involved in substrate binding and catalysis. The Fru 6-P
binding sites of Fru 6-P,2-kinase have been studied by chemical
modification and site-directed mutagenesis. The result of affinity
labeling experiments revealed that Cys-107 and Cys-196 of rat liver Fru
6-P,2-kinase:Fru-2,6-Pase and Cys-105 of the rat heart enzyme appear to
be near or at the Fru 6-P binding site (10, 11). Site-directed
mutagenesis of the rat testis enzyme demonstrated that Arg-102 (Arg-105
of the liver enzyme) is essential for Fru 6-P binding (12). Similarly,
the importance of this Arg residue in the liver and muscle enzymes was
shown by Rider et al. (13) and Kurland et al.
(14) by mutagenesis of the Arg residue to Ala, which resulted in a
200-fold increase in KFru 6-P. Arg-195 and
Gly-48, respectively, were shown to be essential for Fru 6-P and ATP
binding in the kinase reaction by site-directed mutagenesis in the
liver isozyme (15). More recently, Vertommen et al. (16)
showed that site-directed mutagenesis of Lys-54 and Thr-55 of the liver
isozyme (which correspond to Lys-51 and Thr-52 of the testis enzyme)
resulted in a 5000-fold decrease in the kinase activity. Since mutation
of Thr-55 to Cys resulted in loss of the kinase activity but the
mutated enzyme still binds mant-ATP and ATP, they suggested that the
Thr residue may be involved in catalysis.
fructose 1,6-bisphosphate + ADP). The crystal
structure of Escherichia coli PFK has been solved (17).
Based on the structures of complexes with Fru 6-P and imidoadenosine
5
-triphosphate (AMPPNP), Hellinga and Evans (18) concluded that
Asp-127 acts as a base, which increases nucleophilicity by abstracting
a H+ from the 1-OH of Fru 6-P. In support of this idea is
the observation that mutation of Asp-127 to Ser decreases the enzyme
activity by 18,000-fold compared with WT PFK (18). In the event that the mechanism of the Fru 6-P,2-kinase reaction is similar to that of
PFK, as has been widely presumed, then a nucleophile should be
identifiable that would activate the 2-OH of Fru 6-P for attack on the
ATP
-phosphate.
isomer and 30% 2
isomer (28). All
other chemicals were reagent grade and obtained from commercial sources.
ung
). The
purified recombinant M13/RT2K phage was used to prepare uracil-containing single-stranded template. Synthetic oligonucleotide primers used for constructing various mutants are shown in Table I. The oligonucleotide-directed in vitro
mutagenesis was performed as described by Kunkel (29) using the
Muta-Gene M13 in vitro mutagenesis kit. The double mutants
designated as T52A/D128A and T52S/D128A were constructed as follows.
The pT7-7 vector containing D128A mutant of RT2K/pT7-7 DNA was
digested with EcoRI, isolated, and introduced into
EcoRI-digested RT2K/pT7-7 containing the mutation of T52A.
The synthesized double-stranded DNA was used to transform E. coli MV1190 competent cells. Mutant derivatives were identified by
DNA sequencing (30), and the DNAs were digested with NdeI and HindIII. The DNA fragments containing the mutated RT2K
genes were subcloned into the NdeI-HindIII sites
of RT2K/pT7-7 and expressed in E. coli as before (7). The
WT and mutant Fru 6-P,2-kinase:Fru-2,6-Pase enzymes were purified as
described previously (31). However, some of the mutant enzymes required
slight modification of this procedure, which are described under
"Results."
Oligonucleotide
Sequence
T52A
5
-AGA AAT GTA GG
CTT GCC CCT-3
T52S
5
-AGA AAT GTA GG
CTT GCC CCT-3
N73A
5
-TA CTG ACC CAC G
GAA TTC CC-3
N73D
5
-TA CTG ACC CAC GT
GAA TTC CC-3
D128A
5
-ATT GGT AGC A
C AAA AAC CGC-3
D128S
5
-ATT GGT AGC A
AAA AAC CGC-3
D128K
5
-ATT GGT AGC
T
AAA AAC CGC-3
D128E
5
-ATT GGT AGC
TC AAA AAC CGC-3
D128N
5
-ATT GGT AGC AT
AAA AAC CGC-3
T130A
5
-GGT GGT ATT GG
AGC ATC-3
T130S
5
-GGT GGT ATT GG
AGC ATC AAA-3
T130V
5
-GGT GGT ATT G
AGC ATC AAA-3
Y197F
5
-GTT TTC A
A GCA TTC AAT-3
Y197S
5
-GTT TTC A
A GCA TTC AAT-3
This assay was based on the determination of Fru 2,6-P2 and is the same as described previously (32) with slight modification. The reaction mixture in a final volume of 50 µl contained 100 mM Tris/HCl (pH 7.5), 0.1 mM EDTA, 10 mM MgCl2, 2 mM ATP, and 2 mM Fru 6-P. The mixture was incubated at 30 °C for 10 min. At the end of the reaction, 0.1 N NaOH (50 µl) was added, and the mixture was heated for 90 s at 80 °C. Suitable aliquots were assayed for Fru 2,6-P2 as described by Uyeda et al. (33). One unit of enzyme activity is defined as the amount of enzyme that catalyzes the formation of 1 µmol of Fru 2,6-P2/min under these conditions.
Assay Method for Fru-2,6-PaseThis assay measures the formation of Fru 6-P fluorometrically coupled to NADPH formation and was described previously (31). The reaction mixture (in a final volume of 1.0 ml) contained 100 mM Tris/HCl (pH 7.5), 0.2 mM EDTA, 100 µM NADP, 0.4 unit of Glu 6-P dehydrogenase, 1 unit of phosphoglucose isomerase, and varying amounts of Fru 2,6-P2. The enzyme was desalted by column centrifugation (34) in 15 mM Tris sulfate (pH 7.5), 0.5 mM EDTA, and 5 mM dithiothreitol. The reaction was initiated with the addition of enzyme and followed at 25 °C by measuring the NADPH formation at 452-nm emission and excitation at 350 nm using an Aminco-Bowman Series 2 luminescence spectrometer (SLM Aminco/Urbana, IL).
Fluorescence MeasurementsFluorescence spectra were determined at 25 °C with an SLM Aminco Bowman Series 2 spectrofluorometer. Excitation and emission slits were set to 4 nm. Spectra were corrected for the buffer background but not for the instrument response functions. Binding of mant-ATP to enzymes was measured at 25 °C in a reaction mixture in a final volume of 0.2 ml, containing 50 mM Tris phosphate (pH 7.5), 0.1 mM EDTA, 5 mM dithiothreitol, and 1 µM Fru 6-P,2-kinase:Fru-2,6-Pase. Suitable aliquots (1 µl) of 10 µM mant-ATP were added and corrected for dilution. Fluorescence intensity due to free mant-ATP was measured in the absence of enzyme and used for correction. The excitation and emission were 280 nm (or 350 nm) and 450 nm, respectively. The advantage of using 280-nm excitation was that the fluorescence of mant-ATP was much lower than that at 340 nm.
Separation and Determination of mant-ADPmant-ADP and mant-ATP were separated on a Partisil 10 SAX column (4.6 × 250 mm) using a Dionex HPLC (Dionex/Sunnyvale, CA) and a Ratio-2 fluorometer (Optical Technology Devices, Elmsford, NY) equipped with a flow cell. mant-ADP was eluted from the column with a linear gradient of 0.25-0.5 M ammonium phosphate (pH 4.0), containing 25% ethanol with a flow rate of 1 ml/min. Under these conditions mant-ADP was usually eluted after 7.5 min, while mant-ATP was eluted with 0.5 M ammonium phosphate (after 15 min). To determine the concentration of mant-ADP, a standard curve of mant-ADP was generated by the HPLC chromatography of varying concentrations (0-10 pmol) of mant-ADP and by calculating the concentration from the peak heights of the mant-ADP fluorescence peaks. Samples (10 µl) containing mant-ADP were chromatographed under the identical conditions, and the concentrations were calculated using the standard curve.
Other MethodsSDS-polyacrylamide slab gel electrophoresis was performed with the Phast System (Pharmacia Biotech Inc.). Protein concentration was determined by the Bradford method (35) using bovine serum albumin as a standard.
The mutated enzymes described here were purified using the same procedure as that for the WT enzyme (31). However, D128N and D128S mutant enzymes were eluted from the Yellow-3 column (fractionation step 4) with 20 mM ATP, while T52A and Y197S were eluted from the same column with 0.2 or 0.3 M potassium phosphate included in buffer A (Tris/P, 50 µM, pH 7.5, 0.1 µM EDTA; 0.1 µM EGTA, 5% glycerol, 2 mM dithiothreitol, 1% polyethyleneglycol (M = 300)). All of the enzymes were purified to apparent homogeneity as judged by SDS-polyacrylamide gel electrophoresis.
Steady State Fru 6-P,2-kinase KineticsTable II summarizes the kinetic parameters of the WT RT2K and various mutant enzymes.
|
Thr-52 to Ala altered the Km
for both Fru 6-P and ATP to infinite value, i.e. not
saturable, and the Vmax was decreased to
1/1000 that of the WT enzyme. However, Thr-52 to Ser increased the Km and Vmax values about
2-fold. Similar changes in the kinetic parameters were observed with
T55C and T55S mutants of rat liver Fru 6-P,2-kinase (16). Thus, the
side chain hydroxyl group appears to be essential for binding of both
substrates and also for the catalysis. As depicted in Fig.
1, the crystal structure showed that the side chain OH
of Thr is hydrogen-bonded to Mg2+, which is chelated to the
PO4 of ATP (or 2-phosphate of Fru 2,6-P2;
Ref. 20). Thus, this interaction may stabilize the pentacoordinated transition state. Apparently, the extra methyl group present in Thr
does not significantly interfere with this interaction.
Asp-128 Mutants
Mutation of Asp-128 to Ala, Asn, and Ser resulted in an 8-24-fold decrease in Vmax and a 30-250-fold increase and 33-200-fold increase in KATP and KFru 6-P values, respectively. The changes in the kinetic properties of the Ala mutant were similar to those reported with the liver isozyme (13). Asp-128 to Glu increased Vmax 2-fold, and Km values of KATP and KFru 6-P increased 13- and 210-fold, respectively. Surprisingly, the Asp to Lys mutation increased the Km values comparable with the other mutants, but the Vmax was decreased by 1/100. Since the carboxyl group of Asp is chelated to Mg2+ (Fig. 1), one would expect the negative charge to be essential in the substrate binding and stabilization of the transition state. However, it appears that while the anionic interaction with Mg2+ may not be essential for ATP binding, it is required for catalysis.
T52A/D128A and T52S/D128ABoth of these double mutant enzymes showed extremely high Km values for both substrates as did the single mutant, T52A. The T52A/D128A mutant displayed only trace kinase activity, but surprisingly, the T52S/D128A mutant retained the same Fru 6-P,2-kinase activity as the WT enzyme although half of that of the single mutant T52S. These results may suggest a more important role for Thr-52 than for Asp-128 in efficient catalysis.
Thr-130 MutantsRelatively small changes in the kinetic properties, especially the Vmax values, were observed with Thr-130 to Ala or Ser mutation. Mutation to Val increased KFru 6-P 350-fold without affecting the Vmax. Thr-130 is located near the furanose ring of Fru 6-P and may not have strong interaction except to place the sugar moiety in proper orientation for efficient catalysis.
Tyr-197 MutantsMutation of Tyr to Phe increased KATP and KFru 6-P by 6- and 70-fold, respectively, without change in Vmax. However, mutation to Ser produced a large increase in KFru 6-P and a 19-fold increase in KATP. The Vmax decreased to 1/3 that of the WT enzyme. These results suggest that Tyr-197 provides a hydrophobic pocket for Fru 6-P.
Asn-73 MutantsThe N73A mutant showed a 130-fold increase in KFru 6-P and a 4-fold increase in KATP, but the Vmax value was only slightly decreased (50% of WT). The N73D mutant increased KFru 6-P to 600-fold and decreased Vmax to 9% of WT without affecting KATP. The Vmax/KFru 6-PEt value of the N73D mutant was 0.02% of WT, which was 1/21 that of N73A. Asn-73 appears to be essential for Fru 6-P binding, probably through hydrogen bonding, but not essential for ATP binding or catalysis. This interpretation is consistent with the crystal structure (Fig. 1), which shows that Asn-73 is located near the furanose ring of Fru 6-P and may orient the Fru 6-P to a proper position for catalysis through hydrogen bonding with the sugar moiety. Introduction of a negative charge at Asn-73 (N73D) may affect the interaction of Thr-52 and Asp-128, which are essential for catalysis with Fru 6-P, and causes a large decrease in Vmax/KFru 6-PEt.
Other Asp and Glu MutantsWe have altered all of the highly conserved aspartates among the various isozymes to Asn and some of the conserved glutamates to Gln including Asp-94, -112, -160, -189, -205, -209, and -231 and Glu-97, -120, -135, -155, and -195. These mutant enzymes were purified to homogeneity, and their kinetic parameters were determined. These mutations did not cause significant changes in the kinetic properties of the kinase activity compared with the WT enzyme.
Fru-2,6-Pase Activities of the Mutant EnzymesIn general, all of these mutant enzymes showed only small changes (2-4-fold increase) in KFru 2,6-P2 and Vmax values of Fru-2,6-Pase.
mant-ATP as a SubstrateThe binding of mant-ATP, a fluorescent derivative of ATP, to various mutant enzymes and the WT enzymes was compared. mant-ATP serves as a substrate for the WT Fru 6-P,2-kinase, and the apparent Kmant-ATP and Vmax were 0.64 µM and 25 milliunits/mg, respectively (data not shown). It is surprising that the Km value for mant-ATP is 156-fold smaller than that for ATP, since subsequent examination of the crystal structure reveals that there is space available to accommodate the methylanthraniloyl substituent with the potential for favorable packing interactions. To determine if mant-ATP binds at the same catalytic site as ATP, Fru 6-P,2-kinase activity was measured at varying ATP or Fru 6-P concentrations in the presence of constant concentration of 2 µM mant-ATP. The kinase activity with mant-ATP alone as a substrate in this reaction mixture (containing ATP as well) was determined by measuring the formation of mant-ADP. The mant-ADP and mant-ATP in the reaction mixture were first separated by HPLC chromatography and quantitated fluorometrically as described under "Experimental Procedures." The results demonstrated that mant-ATP was a competitive inhibitor with respect to ATP but noncompetitive with respect to Fru 6-P. The KiATP and KiFru 6-P values estimated from the plots were 0.82 µM and 2.8 µM, respectively.
Binding of mant-ATP to WT Fru 6-P,2-kinase:Fru-2,6-Pase was determined
fluorometrically. When mant-ATP (2 µM) was bound to the
enzyme, the fluorescence maximum decreased from approximately 450 to
440 nm when excited at 350 nm, and the fluorescence intensity increased
compared with free mant-ATP (Fig. 2, a versus
c). The mant-ATP binding to the enzyme required Mg2+
(Fig. 2, b versus c), since there was no change in the
fluorescence in the absence of Mg2+. The addition of 1 mM ATP decreased the fluorescence at 440 nm, and the
resulting spectrum was identical to that of free mant-ATP (Fig.
2d), suggesting that ATP displaced the enzyme-bound mant-ATP completely. The same qualitative results were obtained when mant-ATP was excited at 280 nm, but a much larger difference in the fluorescent increase of bound mant-ATP was observed compared with 350-nm
excitation.
This increase in the fluorescence intensity was used to study the binding of mant-ATP to the WT and some of the mutant enzymes, and the dissociation constants calculated from the double reciprocal plots are summarized in Table III. The dissociation constants for mant-ATP of the WT enzyme (0.80 µM) were comparable with the Kmant-ATP (0.64 µM) and considerably lower than KATP (100 µM) and confirmed the tighter binding than ATP. T52A and D128N did not bind mant-ATP, but other mutant enzymes bound the ATP derivative as well as the WT enzyme. Thus, the differences in the mant-ATP bindings to these enzymes were similar to the differences in the KATP values (Table I), although the former values were nearly 2 orders of magnitude lower than the latter values (Table II). These differences suggest that the binding of mant-ATP by the enzyme was determined partly by hydrophobic interaction of the mant- group.
|
These results, such as the large differences in the binding constants
of mant-ATP versus ATP, were in agreement with those reported recently with the rat liver Fru 6-P,2-kinase and its derivative containing six His residues (H6) at the C
terminus (16). However, there were several important differences. The rat liver enzyme (H6) contains high and low affinity
binding sites for Mg-mant-ATP and mant-ATP with Kd
values of 108 M and 10
5
M, respectively. Moreover, mant-ATP binding to the low
affinity site is unaffected by mutation of Thr-55 and Lys-54. We found that the rat testis enzyme had one binding site for Mg-mant-ATP with
the Kd values ranging from 10
7 to
10
6 M, and as shown in Fig. 2, its binding
requires Mg2+. The mutant enzymes showed different
Kd values, which were consistent with the changes in
the KATP values. These differences may be due to
the differences between these isozymes, which may be a reflection of
differences in the regulatory domains since their catalytic domains are
very similar.
If Asp-128 is the nucleophile involved
in catalysis, mutation of this residue may show an altered pH activity
profile. To investigate this possibility, the Fru 6-P,2-kinase activity
of the WT, D128A, and D128N enzymes at pH values between 6 and 9.5 were
determined. The results shown in Fig. 3 indicated that
the WT enzyme showed the pH optimum between 7 and 8.5, and the apparent pK values for the ascending and the descending rims were 6.6 and 9, respectively. The pH profiles of the mutant enzymes were similar to
that of the WT enzyme. The results suggested that Asp-128 does not
serve as a base in the catalysis, confirming the mutation results.
These pH activity curves are completely different from that of rat
muscle bifunctional enzyme in which the kinase activity increases
linearly from 6 to 10, and the optimum is pH 10 (13). Furthermore,
mutation of Asp-128 to Ala of the muscle enzymes results in loss of the
pH optimum at 10 and becomes independent of pH. Since the active sites
of all of these isozymes are highly conserved, it is unlikely that
these differences are due to tissue-specific isozymic differences.
All kinase reactions or phosphoryl transfer reactions require a
base (nucleophile) for catalysis, positive charges to
neutralize/stabilize the negative charges of the ATP phosphates, and
residues to position the substrates and stabilize the transition state.
The determination of the three-dimensional structure has demonstrated
that the active site of the Fru 6-P,2-kinase domain of the Fru
6-P,2-kinase:Fru 2,6-P2 appears to satisfy these
requirements with the exception of a catalytic nucleophile. It was the
goal of this study to probe by mutagenesis the residues in the active
site for their potential role as the apparently missing nucleophile.
The effect of mutagenesis of the residues located at or near the active
site presented herein is consistent with the three-dimensional model
determined by crystallography as shown in Fig. 1. Thr-52 and Asp-128
are hydrogen-bonded/salt-bridged to the Mg2+ ion of the
Mg2+-ATP complex. Accordingly, all but the most
conservative mutations of these residues have a very significant impact
on the KATP. Thr-52 is also close enough to the
-phosphate of the ATP to be directly hydrogen-bonded to it, and this
may explain the most extreme effect on KATP of
the T52A mutation. The profound effect on Vmax
of the T52A mutation and the comparably modest effects of the Asp-128
mutations were not predicted. We would have expected the charge
interaction of Asp-128 with the Mg2+ ion to dominate the
coordination of the Mg2+-ATP, while our results indicate
that Thr-52 is the more important Mg2+ ligand. The
Mg2+ ion may play a role in catalysis by stabilizing the
negative charge involved in transition state formation. Thr-52 may have an additional role in the ATP binding. According to the crystal structure of the bifunctional enzyme, the kinase active site is nearly
identical to the GTP binding site of p21ras, and Thr-52 in the
former is situated at the position analogous to Ser-17 in the latter
(36, 37). Muegge et al. (38) proposed that the interaction
of Ser-17 with the GTP
-phosphate maintains a conformation of the
P-loop main chain atoms, which optimize the orientation of the main
chain dipoles for interaction with the GTP phosphates. Thus, a
component of the decreased catalysis observed in our T52A mutants might
be due to a loss of an essential main chain conformation in the P-loop,
which promotes optimal electrostatic interactions with the ATP. Further
mutagenesis and crystallography of mutant enzymes or transition state
mimics will be necessary to resolve the structural basis for this
observation.
It should be noted that the observed changes in the kinetic properties of the mutated enzymes were not due to global structural changes for the following reasons. 1) All of these enzymes were purified to homogeneity with the same procedure as that used for the WT enzyme. The only exception was that some of the mutants bound to an affinity column (Yellow-3) more tightly than the WT enzyme, requiring slightly higher ATP concentration and/or higher phosphate for elution, although they showed increased Km values for the substrate(s). 2) Intrinsic tryptophan fluorescence spectra of Asp-128 and Thr-52 mutants were identical to that of the WT enzyme (data not shown).
The effect on the KFru 6-P of the mutants, which would apparently only affect Mg2+-ATP binding, can best be explained by assuming that ligand binding in the Fru 6-P,2-kinase active site is an ordered process in which ATP binds to the enzyme first followed by Fru 6-P. Earlier kinetic studies (8) suggested that the addition of the substrates to liver Fru 6-P,2-kinase is random, but it is possible that under these assay conditions the ordered substrate addition was favored. Thus, mutations affecting the ability to bind ATP will concomitantly affect Fru 6-P binding, while the inverse is not necessarily true.
While the crystal structure of the Fru 6-P,2-kinase:Fru-2,6-Pase did
not include Fru 6-P bound at its active site, the homology of this
enzyme with the NMP-kinases and the location of several residues known
to affect KFru 6-P have allowed for a confident prediction of the Fru 6-P binding site. As shown in Fig. 1, there are
several residues in the vicinity of the Fru 6-P binding site of the
enzyme that are potential (albeit weak) nucleophiles, including Thr-130, Tyr-197, and Asn-73. While mutations of each of these residues
(or Thr-52 and Asp-128) has an impact on KFru
6-P (or KATP), their low impact on
Vmax (1-150-fold reductions) would indicate
that none of them act as a nucleophile for activating substrate
(compared with the 13,000-fold reduction observed for the D127A
mutation of the PFK nucleophile (39)). Again, turning to the structure
in Fig. 1, this is reasonable, since their relation to a modeled Fru
6-P substrate places them in positions to bind the ribose moiety
(Asn-73, Thr-130, Tyr-197) or the Mg2+-ATP (Thr-52,
Asp-128) and not in the immediate vicinity of the gap between the
-phosphate and the Fru 6-P 2-OH, where an activating nucleophile
would be expected to reside. In addition, we have mutated all of the
highly conserved Asp and Glu residues in the Fru 6-P,2-kinase domain,
but none of these mutant enzymes were affected significantly.
The evidence presented here supports our contention that the catalytic
mechanism of the Fru 6-P,2-kinase reaction may not involve a strong
nucleophile but instead is mediated via transition state stabilization.
This may be the reason for the extremely low value for
kcat of 0.07 s1 of Fru
6-P,2-kinase. This is in contradiction to the widely held belief that
the Fru 6-P,2-kinase reaction would be "PFK-like" (21). This is not
surprising in light of the structural differences between the Fru
6-P,2-kinase domain and the PFK structures and the unanticipated
similarity between the Fru 6-P,2-kinase domain and the NMP-kinase and
G-protein structures. These latter enzymes are believed to catalyze
their reactions by a transition state stabilization mechanism.
We thank Drs. Paul A. Srere and Paul F. Cook for critical reading of this paper.