Loss of wild-type catalytic efficiency for reaction with phosphate kcat⁄ Km= 21 000 m1Æs1 was dramatic ‡107-fold in purified Arg507fi Ala R507A and Lys512 fi Ala K512A enzymes, reflecting a c
Trang 1Schizophyllum commune probed by site-directed
mutagenesis and chemical rescue studies
Christiane Goedl and Bernd Nidetzky
Institute of Biotechnology and Biochemical Engineering, Graz University of Technology, Austria
Glycosyltransferases (GTs) constitute a diverse class of
enzymes that catalyze the synthesis of glycosidic bonds
in oligosaccharides and glycoconjugates A nucleotide-,
phospho- or lipid-phospho-activated sugar is typically
utilized as the donor substrate, and transfer of the
gly-cosyl moiety to the acceptor molecule occurs with
either inversion or retention of configuration at the
reactive anomeric carbon [1] After detailed studies of glycogen phosphorylase spanning many decades [2–6], there has been recent rekindled interest in the mecha-nistic characterization of retaining glycosyltransferases, particularly in relation to glycoside hydrolases, the physiological counterpart enzymes that catalyze the breakdown of glycosidic linkages [7] The canonical
Keywords
catalytic mechanism; chemical rescue;
family GT-4; glycosyltransferase; a-retaining
glucosyl transfer
Correspondence
B Nidetzky, Institute of Biotechnology and
Biochemical Engineering, Graz University of
Technology, Petersgasse 12 ⁄ I, 8010 Graz,
Austria
Fax: +43 316 873 8434
Tel: +43 316 873 8400
E-mail: bernd.nidetzky@tugraz.at
(Received 5 October 2007, revised 10
December 2007, accepted 19 December
2007)
doi:10.1111/j.1742-4658.2007.06254.x
Schizophyllum commune a,a-trehalose phosphorylase utilizes a glycosyl-transferase-like catalytic mechanism to convert its disaccharide substrate into a-d-glucose 1-phosphate and a-d-glucose Recruitment of phosphate
by the free enzyme induces a,a-trehalose binding recognition and promotes the catalytic steps Like the structurally related glycogen phosphorylase and other retaining glycosyltransferases of fold family GT-B, the trehalose phosphorylase contains an Arg507-XXXX-Lys512 consensus motif (where
X is any amino acid) comprising key residues of its putative phosphate-binding sub-site Loss of wild-type catalytic efficiency for reaction with phosphate (kcat⁄ Km= 21 000 m)1Æs)1) was dramatic (‡107-fold) in purified Arg507fi Ala (R507A) and Lys512 fi Ala (K512A) enzymes, reflecting a corresponding change of comparable magnitude in kcat (Arg507) and Km (Lys512) External amine and guanidine derivatives selectively enhanced the activity of the K512A mutant and the R507A mutant respectively Analysis of the pH dependence of chemical rescue of the K512A mutant
by propargylamine suggested that unprotonated amine in combination with
H2PO4 ), the protonic form of phosphate presumably utilized in enzymatic catalysis, caused restoration of activity Transition state-like inhibition of the wild-type enzyme A by vanadate in combination with a,a-trehalose (Ki= 0.4 lm) was completely disrupted in the R507A mutant but only weakened in the K512A mutant (Ki= 300 lm) Phosphate (50 mm) enhan-ced the basal hydrolase activity of the K512A mutant toward a,a-trehalose
by 60% but caused its total suppression in wild-type and R507A enzymes The results portray differential roles for the side chains of Lys512 and Arg507 in trehalose phosphorylase catalysis, reactant state binding of phosphate and selective stabilization of the transition state respectively
Abbreviations
G1P, a- D -glucose 1-phosphate; GTs, glycosyltransferases; K512A, Lys512 fi Ala mutant; R507A, Arg507 fi Ala mutant; ScTPase,
Schizophyllum commune trehalose phosphorylase; SNi-like, internal return-like mechanism.
Trang 2mechanism of a retaining glycoside hydrolase is that of
a double-displacement reaction involving a covalent
glycosyl-enzyme intermediate [4,8–10] Because
evi-dence from structural and mutagenesis studies has so
far failed to support a similar type of intermediate for
retaining glycosyltransferases [6,7,11–14], an alternative
mechanistic scenario termed internal return-like (SN
i-like) was considered in which hypothetically, direct
front-side displacement of the leaving group by the
incoming nucleophile would result in retention of the
anomeric configuration (Fig 1) Reaction is proposed
to occur via a single transition state featuring a highly
developed oxocarbenium-ion character A strict
requirement for the tentative SNi-like mechanism is
that donor and acceptor substrates are precisely
posi-tioned in close proximity to each other in the enzyme
active site Indeed, reaction through a ternary complex
where both substrates must bind to the enzyme before
the first product is released appears to be a defining
catalytic feature of retaining glycosyltransferases;
structural insights into enzymatic glycosyl transfer via
a ternary complex are provided elsewhere [3,5,12,15]
However, little is known about catalytic factors that
could facilitate enzymatic glycosyl transfer via the
pro-posed SNi-like process, and the stabilization of the
oxocarbenium ion-like species in the transition state
The present study was concerned with the quantitative
analysis of the role of noncovalent interactions
between active-site residues of the enzyme and the
phosphate nucleophile⁄ leaving group in a reaction
cat-alyzed by a sugar 1-phosphate dependant transferase
Schizophyllum commune trehalose phosphorylase
(ScTPase; EC 2.4.1.231) utilizes a
glycosyltransferase-like catalytic mechanism to convert a,a-trehalose and
phosphate into a-d-glucose 1-phosphate (G1P) and
a-d-glucose in a freely reversible reaction [16,17] In
the direction of phosphorolysis, recruitment of
phos-phate by the free enzyme induces binding recognition
for a,a-trehalose and promotes the catalytic steps of glucosyl transfer d-glucose is released from the ternary enzyme–product complex, and dissociation of G1P regenerates the free enzyme [17] The unreactive phos-phate-analogue vanadate is a transition state-like inhibitor of ScTPase [18,19] whereby partial mimicry
of the transition state was proposed to derive from a hydrogen bond between vanadate and the a-anomeric hydroxyl of the glucose leaving group⁄ nucleophile (Fig 1) Unlike glycogen phosphorylase, which utilizes pyridoxal 5¢-phosphate to promote the attack of the phosphate [5] and other nucleotide sugar-dependent glycosyltransferases, which often require a metal ion for activation of the leaving group [20–22], ScTPase does not employ a cofactor in catalysis
Based on sequence similarity, ScTPase has been clas-sified into family (GT)-4 of the glycosyltransferase fam-ilies Recent crystal structures of three representatives
of family GT-4 [23,24] revealed a common protein structural organization typical of transferases of fold family GT-B where the catalytic centre, which features
a highly conserved architecture, is situated in a deep cleft formed by two Rossman-fold domains The struc-ture of Mycobacterium smegmatis phosphatidylinositol-mannosyltransferase (PimA) in complex with the natural sugar-donor substrate GDP-mannose showed that the distal phosphate moiety of the GDP leaving group was tightly coordinated by strong hydrogen bonds with Gly16, Arg196 and Lys202 [23] The suggestions from
a mutational analysis of ScTPase that the homologous Gly292, Arg507 and Lys512 (see supplementary Fig S1) serve a key role in the phosphorylase reaction
as phosphate-binding residues are thus strongly sup-ported [16] Numbering of ScTPase starts with the initiator methionine as 1 and does not consider the N-terminal 11 amino acid-long fusion peptide that is used for recombinant protein production In the pres-ent study, we have extended significantly the previous
Fig 1 Reaction of trehalose phosphorylase via an S N i-like mechanism proposed for retaining glycosyltransferases where direct front-side displacement results in retention
of anomeric configuration A predicted gen-eral feature of the mechanism is the devel-opment of a strong hydrogen bond between the incoming nucleophile and the leaving group in the transition state of the reaction.
Trang 3scanning mutagenesis of the active site of trehalose
phosphorylase [16] and report on results of
steady-state kinetic analysis and chemical rescue studies for
site-directed ScTPase mutants Arg507fi Ala (R507A)
and Lys512fi Ala (K512A) A detailed portrait of the
catalytic function of the two basic side chains is
pro-vided These new insights are of general interest
con-sidering the presence of an active-site consensus motif,
Arg507-XXXX-Lys512 (where X is any amino acid) as
in ScTPase, in glycogen phosphorylase [3,6] and other
retaining clan IV glycosyltransferases of fold family
GT-B [12,23,25] Interestingly, mammalian and yeast
glycogen synthases of family GT-3 appear to have
replaced the Arg-XXXX-Lys motif of their bacterial
and plant counterpart enzymes in family GT-5 by two
conserved clusters of Arg residues [26,27] Likewise,
GT-A fold glycosyltransferases also utilize a conserved
diad of Arg and Lys for binding of the donor substrate
pyrophosphate group and in catalysis The two basic
residues occur in a motif displaying the inverted GT-B
pattern, Lys359-XXXXX-Arg365 as in bovine
a-1,3-galactosyltransferase of family GT-6 [28] Despite the
evidence provided by the high-resolution crystal
struc-tures for many of these glycosyltransferases, the role
of the conserved residues for promoting a-retaining
glycosyl transfer has not been well defined using
muta-genesis and detailed kinetic analysis
Results
Analysis of kinetic consequences in R507A and
K512A mutants of ScTPase
CD spectra of purified wild-type and mutant trehalose
phosphorylases were almost superimposable on each
other (see supplementary Fig S2), indicating that the
relative proportion of secondary structural elements in
the folded structure of the wild-type enzyme was not
altered significantly in R507A and K512A mutants
Therefore, this strongly suggests that kinetic
conse-quences resulting from the replacement Arg507fi Ala
or Lys512fi Ala are not due to partial misfolding of the mutant enzymes Figure 2 displays results of the steady-state kinetic characterization of R507A and K512A, and kinetic parameters of wild-type and mutant phosphorylases for phosphorolysis of a,a-tre-halose are summarized in Table 1 Both R507A and K512A exhibited a dramatic (‡107-fold) loss of cata-lytic efficiency for reaction with phosphate (kcat⁄ Km)
in comparison with the wild-type enzyme In R507A, the effect on kcat⁄ Kmwas distributed between a major, 4.1· 105-fold decrease in apparent catalytic centre activity (kcat) and a comparably minor, 130-fold increase in the value of Km for phosphate In K512A,
by contrast, the initial phosphorolysis rate was linearly dependent on the concentration of phosphate up to 1.5 m (Fig 2), precluding determination of kcatand Km
Fig 2 Steady-state kinetic characterization of R507A (s) and K512A (d) mutant trehalose phosphorylases Reaction rates (V) were recorded in 50 m M Mes buffer, pH 6.6, using a constant con-centration of 400 m M a,a-trehalose [E] is the molar enzyme con-centration Reaction mixtures were incubated at 30 C for up to
40 h, and the release of G1P was determined enzymatically A plot
of concentration of G1P released against the incubation time was linear in all cases, allowing determination of V and showing that both enzymes were reasonably stable under the incubation condi-tions Error bars indicate the SD of four independent determina-tions.
Table 1 Comparison of kinetic parameters for wild-type, R507A and K512A mutant trehalose phosphorylases in a,a-trehalose phosphoroly-sis direction at 30 C and pH 6.6.
k cat ⁄ K m phosphate
[10)3Æ M )1Æs)1]
Fold decrease
K m phosphate
[m M ]
Fold increase
K ic vanadate
[l M ]
Fold increase
Vhydrolysis⁄ [E] [10)4Æs)1]
0 m M
phosphate a
50 m M
phosphate
a Data from [16].
Trang 4as independent kinetic parameters The value of
kcat⁄ Km was obtained from the slope of the straight
line fitted to the data in Fig 2 It was 1.5· 107-fold
lower than the corresponding catalytic efficiency of the
wild-type enzyme The decrease in kcat⁄ Km for K512A
must therefore reflect a very large (>2000-fold)
increase in Km, compared with the wild-type values,
suggesting that the site-directed replacement of Lys512
caused a more substantial disruption of binding
affin-ity for phosphate than that of Arg507
Table 1 also summarizes values of Kic vanadate for
wild-type and mutant trehalose phosphorylases As in
the wild-type enzyme, vanadate acted as a competitive
inhibitor against phosphate in K512A However,
Kic vanadate was increased 745-fold as result of the
site-directed replacement Lys512fi Ala By contrast,
R507A was not at all inhibited by the used
concentra-tions of vanadate (0.5–5.0 mm) The ratio of
Km⁄ Kic vanadate was therefore changed as result of the
site-directed substitution of Arg507 from a value of
2000 in the wild-type enzyme to an infinitesimally
small value (=103⁄ ¥) in the mutant It appears to
have been increased to a value significantly >2000
(=1600⁄ 0.3) in K512A
The hydrolase activities of wild-type and mutant
tre-halose phosphorylases towards a,a-tretre-halose were
com-pared under hydrolysis-only conditions [16] and under
conditions where, in the presence of 50 mm of
phos-phate, phosphorolysis competed with hydrolysis of the
disaccharide The results are summarized in Table 1
Inhibition by vanadate of the hydrolysis of
a,a-treha-lose and G1P catalyzed by wild-type ScTPase was also
measured With the methods used, it was not possible
to quantify, in the wild-type, a small proportion of
a,a-trehalose conversion by ‘error hydrolysis’ next to
an overwhelmingly predominant phosphorolysis
reac-tion, which also produces d-glucose However, within
limits of detection of the experimental procedures
(£1%), no hydrolysis of a,a-trehalose by wild-type
enzyme took place when 50 mm of phosphate was
present Because replacement of Arg507 caused
selec-tive slowing down of the phosphorolysis reaction
com-pared with the hydrolysis of a,a-trehalose [16], the
complete suppression of the hydrolase activity of
R507A towards a,a-trehalose upon addition of 50 mm
of phosphate could be established unambiguously By
marked contrast, the basal rate of hydrolysis of
a,a-trehalose by K512A was enhanced significantly
(approximately 1.6-fold) in the presence of 50 mm of
phosphate By contrast to the clear inhibitory effect of
vanadate on the hydrolysis of a,a-trehalose by
wild-type ScTPase (3.5-fold), vanadate did not inhibit the
hydrolysis of G1P by the same enzyme Values of
Vhydrolysis⁄ [E] were 4.0 · 10)4Æs)1 and 3.9· 10)4Æs)1 in the absence and presence of vanadate, respectively
Noncovalent complementation of trehalose phosphorylase activity in R507A and K512A Inclusion of 200 mm of guanidine into the assay for phosphorolysis of a,a-trehalose at pH 6.6 caused 45-fold enhancement of the activity of R507A seen
in the absence of guanidine (k0= 4.2· 10)5Æs)1) Likewise, a 23-fold stimulation of the basal activity
of K512A (k0 = 6.1· 10)5Æs)1) was observed in the presence of 200 mm of propargylamine Functional complementation of R507A and K512A, expressed as
krescue⁄ k0, displayed a hyperbolic dependence on the concentration of the respective rescue reagent (Fig 3) Values of kmax (R507A: 2.8· 10)3Æs)1; K512A: 1.6· 10)3Æs)1) and KR (guanidine, 100 mm; propargyl-amine, 58 mm) were obtained with a relative SD of approximately 6% and 10%, respectively, using non-linear fits of Eqn (1) to initial-rate data recorded in the absence and presence of rescue reagent concentra-tions in the range 10–200 mm Guanidine and propar-gylamine did not exhibit a significant effect on the activity of the wild-type enzyme, except for a weak inhibition (<50% reduction in rate) by concentrations
of guanidine higher than 100 mm No cross-reactiva-tion of R507A by propargylamine (10–200 mm) and K512A by guanidine (10–200 mm) was observed Addi-tion of 200 mm of NaCl did not alter the activity of either one of the site-directed mutants These results
Fig 3 Functional complementation of mutant trehalose phosphory-lases R507A was reactivated by guanidine (s) and K512A by prop-argylamine (d) No cross-reaction was observed, and the rescue agents did not significantly alter the wild-type activity (guanidine ,, propargylamine ) Lines show the fit of Eqn (1) to the data.
Trang 5provide strong evidence against the possibility of a
non-specific activation of ScTPase mutants and suggest that
external guanidine and propargylamine can partly
com-pensate for the loss of the original side chain in R507A
and K512A, respectively
A series of primary amines and derivatives of
guani-dine were therefore examined for their ability to restore
phosphorylase activity in K512A and R507A,
respec-tively The results obtained are summarized in Table 2,
along with relevant structural and electronic parameters
of the compounds used for chemical rescue Because
krescue for R507A and K512A appeared to exhibit a
complex dependence on steric factors of the rescue
reagents and the set of amine and guanidine compounds
tested was rather small (five each), we did not pursue
construction of the respective Brønsted plot using
quan-titative structure–activity relationship analysis
How-ever, even in the absence of correction for the influence
of molecular volume and hydrophobicity of the rescue
reagent, Table 2 shows clearly that the effect of the pKa
of the amine and guanidine derivatives on specific
resto-ration of activity in K512A and R507A, respectively,
was very small and probably not significant
Analysis of the pH dependence of functional
complementation of K512A
kmax⁄ KR for chemical rescue of K512A by
propargyl-amine and R507A by guanidine was pH-dependent
Its value decreased in K512A from 0.035 m)1Æs)1 at
pH 6.6 to 0.003 m)1Æs)1 at pH 8.2, and in R507A
from 0.028 m)1Æs)1 at pH 6.6 to 0.005 m)1Æs)1 at
pH 8.0 These pH effects are explicable on account
of changes in the ionization states of the enzyme
and the substrate phosphate (pKa,2= 7.2) and, in
the case of K512A, deprotonation of propargylamine
at high pH (pKa= 8.2) Analysis of pH-rate profiles
for wild-type ScTPase suggested that H2PO4) is the protonic form of phosphate utilized in the enzymatic reaction [17] The pH-dependence of functional com-plementation of K512A was therefore examined in more detail
Figure 4 compares pH-rate profiles of K512A assayed in the absence and presence of 200 mm of propargylamine with the corresponding pH-rate pro-file of the wild-type enzyme (Fig 4A) and summarizes the results of chemical rescue experiments with K512A carried out at four different pH values (Fig 4B) Figure 4A shows that pH-rate profiles of wild-type enzyme and K512A were similar, both showing maximum enzyme activity at an approximate
pH of 6.5 Addition of propargylamine caused an up-shift of the optimum pH of K512A by approximately
1 pH unit Restoration of trehalose phosphorylase activity in K512A by propargylamine was best at
pH 7.5 where a value of 140 was observed for
krescue⁄ k0 when the concentration of rescue reagent was saturating (Fig 4B) The presence of 100 mm of propargylamine caused an approximately 10-fold enhancement of the catalytic efficiency of K512A for reaction with phosphate at pH 6.6, in reasonable agreement with the results obtained in activity assays
at a single phosphate concentration of 50 mm (Table 2) Likewise, under conditions of chemical res-cue of K512A by 200 mm of propargylamine, kcat⁄ Km
for phosphate increased from a value of 0.014 m)1Æs)1
at pH 6.6 to 0.023 m)1Æs)1 at pH 7.5, suggesting that the corresponding pH-rate profile in Fig 4A reflects the pH dependence of kcat⁄ Km These results indicate that, for optimum restoration of activity in K512A, the protonation states of propargylamine and phosphate must be matched We found that there was
a good linear correlation between log(krescue⁄ k0) and the limiting concentration of either one of the
Table 2 Chemical rescue analysis for R507A and K512A The concentration of external reagent was 200 m M Values of V ⁄ [E] were recorded at 30 C in 50 m M Mes buffer, pH 6.6, using 400 m M a,a-trehalose and 50 m M potassium phosphate as substrates Molecular volume (Mol volume) and hydrophobicity (logP) were calculated using the programs SPARTAN 06, version 1.1.0 and KOWWIN , respectively.
pKavalues are from the literature [32,39].
Molecular volume [A˚3] logP
V ⁄ [E]
[10)4Æs)1]a
Fold increasea K512A pK a
Molecular volume [A˚3] logP
V ⁄ [E]
[10)4Æs)1]a
Fold increasea
Aminoguanidine 11.0 71.3 )1.99 1.6 3.8 2,2,2-Trifluoro
ethylamine
a V ⁄ [E] measured in the absence and presence of rescue reagent are referred to as k 0 and krescuein text, respectively Likewise, krescue⁄ k 0
in text corresponds to fold increase b Values in parentheses were corrected for the fraction of protonated amine.
Trang 6compounds, H2PO4) and unprotonated amine, in this
combination of protonic forms (Fig 4B, inset)
Discussion
Evidence from high-resolution X-ray structures
[3,12,25,29] and site-directed mutagenesis [13,30,31]
of glycosyltransferases of fold family GT-B supports
an important role for the consensus motif,
Arg507-XXXX-Lys512 as in ScTPase, in binding recognition
of the phospho leaving group of the glycosyl donor
substrate However, little is known about the
individ-ual contribution of each side chain to catalytic
efficiency The present study of ScTPase used
site-directed replacement by Ala and detailed kinetic
comparison of wild-type and mutant enzymes to
portray the tasks fulfilled by Arg507 and Lys512 in
the interaction network of active site residues during
binding of phosphate and in catalysis Although we
consider a-retaining glucosyl transfer via the SNi-like
mechanism plausible for ScTPase (Fig 1), the results
presented here do not provide evidence that would
settle the mechanistic debate surrounding this and
other retaining glycosyltransferases
Proposed roles for Arg507 and Lys512 in the
mechanism of ScTPase deduced from analysis
of kinetic consequences of their individual
replacements by Ala
The steady-state ordered kinetic mechanism of ScTPase
where phosphate binds before a,a-trehalose [17] implies
that kcat⁄ Kmfor phosphate is a second-order rate
con-stant for the association between the free
phosphory-lase and the nucleophile of the reaction kcatis thought
to measure the rate-determining conversion of the
ter-nary enzyme–substrate complex [17] Individual replacements of Arg507 and Lys512 caused disruption
of the phosphate binding rate by more than seven orders of magnitude, which is equivalent to an ener-getic destabilization of ‡41 kJÆmol)1 (=RT· ln107 where R is the gas constant and T is a temperature of 303.15 K), compared with the wild-type enzyme Equi-librium binding in terms of Kmfor phosphate appeared
to be completely destroyed in K512A whereas it was weakened in a comparatively moderate way (130-fold)
in R507A Interestingly, relevant single-site mutants
of family GT-35 maltodextrin phosphorylase (Arg535fi Gln; Lys540 fi Arg) [31] and family GT-5 glycogen synthase (Arg300 fi Ala; Lys305 fi Ala) [13], both from Escherichia coli, showed closely similar
Kmvalues to their wild-type forms Their catalytic cen-tre activities, however, were between three to four orders of magnitude below the corresponding wild-type levels In kcatterms, the dimension of loss of catalytic activity was significantly higher in R507A than the comparable Arg mutants of the two other transferases Noteworthy, a Lys211fi Ala mutant of Acetobacter xylinum a-mannosyltransferase, which shares with ScTPase the membership to family GT-4, was reported
to be devoid of any enzyme activity [30] The crystallo-graphically determined hydrogen bond distance between oxygens of the distal phospho group of UDP-glucose and the side chains of Arg196 and Lys202 of family GT-4 mannosyltransferase PimA was only 2.44 and 2.77 A˚, respectively [23] Removal of either one of the two strong bonds by mutagenesis would therefore
be expected to result in marked loss of the binding energy used by the wild-type enzyme to promote the reaction Kinetic consequences for R507A and K512A mutants of ScTPase are consistent with this structure-derived suggestion
Fig 4 pH-dependence of functional complementation of K512A by propargylamine Reaction rates were recorded using 50 m M potassium phosphate and 400 m M a,a-trehalose (A) pH profiles for wild-type enzyme (d), K512A (s) and K512A determined in the presence of
200 m M propargylamine (.) (B) Chemical rescue of K512A at pH 6.6 (d), 7.5 (s), 8.2 (.) and 8.5 (n) The inset displays the logarithmic dependence of k rescue ⁄ k 0 on the relative content of active compound (H 2 PO 4 )and NH
2 ) present.
Trang 7The comparison of apparent affinities for reversible
binding of phosphate (Km) and the transition state
mimic vanadate (Ki) by wild-type and mutant forms of
trehalose phosphorylase (Table 1) delineates
differen-tial roles for Lys512 and Arg507 in the enzymatic
mechanism The particular change in the ratio of
Km⁄ Ki resulting from site-directed substitution of
Lys512 (increase) and Arg507 (decrease) compared
with the wild-type value suggests that, whereas Lys512
appears to be primarily required for phosphate binding
in the reactant state, Arg507 promotes the catalytic
step of glucosyl transfer through a selective
stabiliza-tion of the transistabiliza-tion state of the reacstabiliza-tion Occupancy
of the phosphate binding site in wild-type and R507A
phosphorylases caused complete shut-down of their
hydrolytic activity towards a,a-trehalose in the absence
of phosphate whereas, in K512A, addition of
phos-phate stimulated weakly the breakdown of the
disac-charide via hydrolysis Steps involved in phosphate
binding by the wild-type enzyme arguably include an
obligatory exclusion of water from the catalytic site
Their drastic impairment resulting from the
site-direc-ted substitution of Lys512 is likely to be responsible
for this unusual property of the K512A mutant
Interpretation of results of functional
complementation studies
Chemical rescue experiments, in which a small
mole-cule compensates for the missing side chain of a
rele-vant site-directed mutant, often provide valuable
insights into the role of active-site arginine [32–36] and
lysine residues [37–40] for the catalytic function of
dif-ferent enzymes In the present study, we show that
activity lost in Arg507fi Ala and Lys512 fi Ala
vari-ants of ScTPase could be selectively restored by
deriv-atives of guanidine and primary amines, respectively
The failure of amines to rescue R507A and, likewise,
guanidine derivatives to rescue K512A is consistent
with observations made with relevant mutants of
sev-eral other enzymes [32–36], and it also supports the
notion that Arg507 and Lys512 fulfill different tasks in
trehalose phosphorylase catalysis (see above)
Partial functional complementation of the catalytic
defect in R507A by guanidine displayed saturation
behavior with respect to both the rescue agent and
the substrate Therefore, this suggests that guanidine
binds to the cleft vacated by the replacement of the
side chain of Arg507 in the mutant and both the
rescue agent and the substrate phosphate form a
ter-nary complex prior to catalysis Considering a pKa
for guanidine of approximately 13.6, we conclude
from analysis of the pH dependence of the
second-order rate constant for the chemical rescue process that the protonated guanidinium ion is most likely required for noncovalent restoration of phosphory-lase activity in R507A The observed 5.6-fold dec-rease in kmax⁄ KR in response to an increase in pH from 6.6 to 8.0 would be readily explained by depro-tonation of the phosphate monoanion, which is the form of the substrate presumably utilized in the enzymatic reaction [17], and parallels the effect of the same pH change on kcat⁄ Km for phosphate in wild-type ScTPase
Chemical rescue of K512A by propargylamine exhibited a complex pH dependence, likely explicable
on account of the similar pKavalues for the external reagent (pKa= 8.2) and the substrate phosphate (pKa,2= 7.2) However, the observed pH effects on
kmax⁄ KR for propargylamine and kcat⁄ Km for phos-phate determined in the absence and presence of a sat-urating concentration of propargylamine (200 mm;
4· KRat pH 6.6) would be best explained if unproto-nated propargylamine and H2PO4 ) were involved in the catalytic reaction of an optimally rescued K512A mutant The scenario proposed for propargylamine need not be the same for methylamine and ethylamine, which, in spite of their high pKa of 10.6, exhibit comparable efficiency to propargylamine as a rescue reagent of K512A at pH 6.6 (Table 2) Binding of propargylamine to K512A failed to restore, in terms of the Kmvalue, some of the affinity of wild-type treha-lose phosphorylase for phosphate Therefore, to what extent the function of the original side chain of Lys512 can be gauged by the results of our chemical rescue studies remains elusive
We plotted log(krescue⁄ k0) of R507A and K512A against the pKa of the rescue agent taking data from Table 2, assuming that all of the listed derivatives of guanidine and primary amines fit the respective cavity resulting from the replacement of the side chain of Arg507 (98 A˚3) and Lys512 (80 A˚3) by the side chain
of Ala (19 A˚3) These limited Brønsted plots did not detect a significant correlation between rescue efficacy and reagent pKa and therefore do not support a role for Arg507 and Lys512 in catalytic proton transfer by ScTPase However, in the phosphorolysis direction of the enzymatic reaction (Fig 1), partial protonation of the glycosidic oxygen of a,a-trehalose will be needed
to facilitate the departure of the leaving group In the hypothetical SNi-like catalytic mechanism of the treha-lose phosphorylase, enzyme-bound H2PO4 )is a strong candidate to fulfill the role of the proton donor The side chains of Arg507 and Lys512 could provide assis-tance in this process via electrostatic stabilization and positioning of the phosphate ligand
Trang 8Because of the similarity in the reactions catalyzed,
it is interesting to compare trehalose phosphorylase
with a-1,4-glucan phosphorylase Crystal structures of
E coli maltodextrin phosphorylase bound with
phos-phate and oligosaccharide show that hydrogen bonds
from the NE and NH2 atoms of the guanidine side
chain of Arg535 to two oxygen atoms of phosphate
dominate the contacts between the enzyme and the
phosphate group [3,6] The side chain of Lys540 and
the main chain N of Gly115 (corresponding to Gly292
in ScTPase) also interact with the phosphate ligand,
which is brought into a plausible catalytic position
within hydrogen-bonding distance of the reactive
gly-cosidic oxygen of the oligosaccharide Assuming that a
similar network of protein contacts positions vanadate
at the active site of ScTPase, mutation of the key
Arg507 into Ala would be expected to disrupt the
pro-posed hydrogen bond between vanadate and the
glyco-sidic oxygen of a,a-trehalose, consistent with the
observed complete loss of transition state-like
inhibi-tion by vanadate in R507A In maltodextrin
phosphor-ylase, the side chain of Lys540 also hydrogen bonds
with the 5¢-phosphate moiety of the pyridoxal
phosphate cofactor These contacts stabilize the
cata-lytic 5¢-phosphate group in a position within hydrogen
bonding distance of the substrate phosphate from
which the attack of inorganic phosphate on the
glyco-sidic oxygen is promoted Furthermore, the highly
con-served Glu638 (Glu606 in ScTPase), which is also
located at the sugar–phosphate contact region, forms a
salt bridge with Lys540 Results from 31P-NMR
stud-ies revealed an indirect interaction of Glu638 with the
5¢-phosphate group of the cofactor [41] and suggested
participation of the glutamate in establishing a
catalyt-ically relevant network of charged groups in the active
site [42] However, the absence of pyridoxal phosphate
in ScTPase implies that the role of the conserved lysine
in promoting the enzymatic reaction need not be
identical for the two phosphorylases
In summary, based on the evidence obtained in
the present study, we propose differential roles for
the side chains of Lys512 and Arg507 in trehalose
phosphorylase catalysis Although Lys512 is required
for binding of the phosphate nucleophile in the
reac-tant state, Arg507 facilitates the reaction through a
selective stabilization of the transition state In the
proposed SNi-like mechanism of ScTPase (Fig 1),
electrostatic ‘front-side’ stabilization of the
oxocarbe-nium ion-like transition state by the incoming
phosphate nucleophile could be a decisive catalytic
factor Arg507 might contribute indirectly to this
sta-bilization by bringing the phosphate into a suitable
position
Experimental procedures
Materials and enzymes
Unless otherwise noted, all materials used have been described elsewhere [17,43] Purified preparations of wild-type ScTPase as well as R507A and K512A mutants thereof were obtained using previously reported procedures [16] Enzyme stock solutions containing approximately
5 mg proteinÆmL)1 were stored in 50 mm potassium–phos-phate buffer, pH 7.0, and kept at)21 C until use
Protein characterization
Thawed protein samples were checked by SDS⁄ PAGE to ensure that partial N-terminal truncation of the phosphor-ylase preparations [16] had not occurred during storage Far-UV CD spectra of wild-type and mutant phosphory-lases were acquired at 30C with a J-715 spectropolari-meter (Jasco Inc., Easton, MD, USA) using a 0.1-cm path length cylindrical cell and instrument settings: step resolution = 0.2 nm; scan speed = 50 nmÆmin)1; response time = 1 s; bandwidth = 1 nm Triplicate spectra were recorded in the wavelength range 260–190 nm using enzymes (approximately 1.6 mgÆmL)1) dissolved in 50 mm potassium–phosphate buffer, pH 7.0 They were subse-quently averaged and corrected by a blank spectrum lack-ing enzyme Smoothlack-ing and normalizlack-ing was performed using a molecular mass of 82.8 kDa for full-length ScTPase Protein concentration was determined using the Bio-Rad dye-binding method (Bio-Rad, Vienna, Austria) referenced against BSA as the standard We are unaware
of a method for the titration of active sites in prepara-tions of trehalose phosphorylase Therefore, the molar enzyme concentration [E] was calculated, in a commonly used procedure, from the concentration of purified pro-tein Because all enzyme preparations were obtained and treated in exactly the same way and displayed similar sta-bilities of their activities during storage (data not shown), values of [E] for wild-type and mutant enzymes are without internal bias
Steady-state kinetic characterization
Buffer exchange to 50 mm Mes, pH 6.6–7.5, and
50 mm Tes, pH 7.5–8.5, was achieved through repeated ultrafiltration of protein stock solutions using 10-kDa cut-off Vivaspin 500 microconcentrator tubes (Sartorius, Gottingen, Germany) Initial rates of phosphorolysis of a,a-trehalose were recorded using a reported discontinu-ous assay [17,43] where the formation of G1P was mea-sured The concentration of G1P was determined as NADH produced in a second coupled enzymatic reaction catalyzed by phosphoglucomutase and glucose 6-phos-phate dehydrogenase The reaction mixtures for
Trang 9phospho-rolysis had a total volume of 200 lL and were incubated
in 1.5-mL tubes at 30C, using an Eppendorf
Thermo-mixer (Vienna, Austria) for temperature control and
gen-tle agitation using instrument settings of 300 r.p.m
Typical enzyme concentrations used were 40 lgÆmL)1 of
wild-type and 500 lgÆmL)1 of R507A and K512A The
reaction times varied between 0.1 h for wild-type enzyme
and up to 40 h for mutant phosphorylases A plot of
concentration of G1P released against the incubation time
was linear in all cases, indicating that enzyme inactivation
did not interfere with determination of the initial rate
under the conditions used Enzymatic rates (V) were
measured for conditions in which the concentration of
phosphate was varied in the range 5–300 mm whereas the
concentration of a,a-trehalose was 400 mm and constant
Vanadate added in concentrations of 0.5, 2.5, or 5.0 mm
was tested as reversible inhibitor of phosphorolysis of
a,a-trehalose catalyzed by wild-type and mutant
phos-phorylases at pH 6.6 Inhibition constants (Kic vanadate)
were calculated using initial-rate data acquired under
conditions in which the concentration of phosphate was
varied at a constant concentration of a,a-trehalose
(400 mm) in the absence or presence of different constant
concentrations of vanadate
Enzymatic rates of hydrolysis of a,a-trehalose or G1P
(Vhydrolysis) were determined at 30C in 50 mm Mes buffer,
pH 6.6, using 400 mm of disaccharide or 50 mm of sugar
1-phosphate substrate and measuring the concentration of
d-glucose released in samples taken at different times, up to
48 h A hexokinase-based spectrophotometric assay was
used for the determination of d-glucose Hydrolytic
reac-tions for wild-type phosphorylase were performed in the
absence and presence of 20 lm of vanadate
Functional complementation studies for R507A
and K512A mutants
Initial rate assays in the direction of phosphorolysis of
a,a-trehalose were used to analyze restoration of activity
in K512A or R507A caused by the addition of an external
primary amine or a derivative of guanidine Experiments
were carried out at 30C in 50 mm Mes buffer, pH 6.6,
containing 50 mm of potassium–phosphate and 400 mm of
a,a-trehalose The concentration of the amine or guanidine
derivative was typically 200 mM and constant, with the
exception of chemical rescue of R507A by guanidine and
K512A by propargylamine, which was analyzed at different
concentrations of external reagent in the range 10–200 mm
Suitable controls showed that none of the added amines or
guanidines had a significant effect on the activity of the
wild-type enzyme incubated under otherwise exactly
identi-cal conditions to the wild-type enzyme alone The increase
in ionic strength resulting from the addition of amine or
guanidine derivative was not corrected However, the
com-parison of initial rates measured in the absence and presence
of NaCl in concentrations in the range 10–200 mm at
pH 6.6 and 7.5 clearly indicated that the activities of R507A and K512A were not influenced by the relevant ionic strength changes The ratio krescue⁄ k0, where k0and krescue are V⁄ [E] values determined in the absence and presence of chemical rescue agent, respectively, is used to express the degree of activation of the mutant The pH dependence of functional complementation of K512A by propargylamine was determined in the pH range 6.6–8.5 using different reagent concentrations in the range 10–200 mm
Data processing
Processing of initial-rate data for the calculation of kinetic parameters and inhibitor binding constants used reported procedures [17] Equation (1) was fitted to data from activ-ity restoration experiments where kmax is the maximum initial rate, divided by [E], obtained at a saturating concen-tration of the rescue agent, and KR is the half-saturation constant for the reagent
krescue¼ kmax ½rescue agent=ðKRþ ½rescue agentÞ þ k0 ð1Þ
Acknowledgements
Financial support from the FWF Austrian Science Fund (project DK Molecular Enzymology W901-B05)
is gratefully acknowledged We thank Professor Walter Keller (Department of Chemistry, University of Graz) for help with CD spectroscopic analysis
References
1 Coutinho PM, Deleury E, Davies GJ & Henrissat B (2003) An evolving hierarchical family classification for glycosyltransferases J Mol Biol 328, 307–317
2 Madsen NB & Withers SG (1986) Glycogen phosphory-lase In Coenzymes and Cofactors: Pyridoxal Phosphate and Derivatives(Dolphin D, Paulson R & Avramovic
O, eds), pp 355–389 Wiley, New York, NY
3 Watson KA, McCleverty C, Geremia S, Cottaz S, Dri-guez H & Johnson LN (1999) Phosphorylase recogni-tion and phosphorolysis of its oligosaccharide substrate: answers to a long outstanding question EMBO J 18, 4619–4632
4 Davies GJ, Sinnott ML & Withers SG (1998) Glycosyl transfer In Comprehensive Biological Catalysis (Sinnott
ML, ed), pp.119–208 Academic Press, San Diego, CA
5 Palm D, Klein HW, Schinzel R, Buehner M & Helmr-eich EJ (1990) The role of pyridoxal 5¢-phosphate in glycogen phosphorylase catalysis Biochemistry 29, 1099–1107
6 Geremia S, Campagnolo M, Schinzel R & Johnson
LN (2002) Enzymatic catalysis in crystals of
Trang 10Escheri-chia colimaltodextrin phosphorylase J Mol Biol 322,
413–423
7 Lairson LL & Withers SG (2004) Mechanistic analogies
amongst carbohydrate modifying enzymes Chem
Com-mun 20, 2243–2248
8 Heightman TD & Vasella AT (1999) Recent insights into
inhibition, structure and mechanism of
configuration-retaining glycosidases Angew Chem Int Ed 38, 750–770
9 Ly HD & Withers SG (1999) Mutagenesis of
glycosid-ases Annu Rev Biochem 68, 487–522
10 Zechel DL & Withers SG (2000) Glycosidase
mecha-nisms: anatomy of a finely tuned catalyst Acc Chem
Res 33, 11–18
11 Schwarz A, Pierfederici FM & Nidetzky B (2005)
Catalytic mechanism of a-retaining glucosyl transfer
by Corynebacterium callunae starch phosphorylase: the
role of histidine-334 examined through kinetic
charac-terization of site-directed mutants Biochem J 387,
437–445
12 Gibson RP, Turkenburg JP, Charnock SJ, Lloyd R &
Davies GJ (2002) Insights into trehalose synthesis
pro-vided by the structure of the retaining
glucosyltransfer-ase OtsA Chem Biol 9, 1337–1346
13 Yep A, Ballicora MA & Preiss J (2004) The active site
of the Escherichia coli glycogen synthase is similar to
the active site of retaining GT-B glycosyltransferases
Biochem Biophys Res Commun 316, 960–966
14 Mitchell EP, Withers SG, Ermert P, Vasella AT,
Garman EF, Oikonomakos NG & Johnson LN (1996)
Ternary complex crystal structures of glycogen
phos-phorylase with the transition state analogue nojirimycin
tetrazole and phosphate in the T and R states
Biochem-istry 35, 7341–7355
15 Boix E, Zhang Y, Swaminathan GJ, Brew K &
Ach-arya KR (2002) Structural basis of ordered binding of
donor and acceptor substrates to the retaining
glycosyl-transferase, a-1,3-galactosyltransferase J Biol Chem
277, 28310–28318
16 Goedl C, Griessler R, Schwarz A & Nidetzky B (2006)
Structure-function relationships for Schizophyllum
com-munetrehalose phosphorylase and their implications for
the catalytic mechanism of family GT-4
glycosyltrans-ferases Biochem J 397, 491–500
17 Eis C, Watkins M, Prohaska T & Nidetzky B (2001)
Fungal trehalose phosphorylase: kinetic mechanism,
pH-dependence of the reaction and some structural
properties of the enzyme from Schizophyllum commune
Biochem J 356, 757–767
18 Nidetzky B & Eis C (2001) a-Retaining glucosyl
trans-fer catalysed by trehalose phosphorylase from
Schizo-phyllum commune: mechanistic evidence obtained from
steady-state kinetic studies with substrate analogues and
inhibitors Biochem J 360, 727–736
19 Eis C & Nidetzky B (2002) Substrate-binding
recogni-tion and specificity of trehalose phosphorylase from
Schizophyllum communeexamined in steady-state kinetic studies with deoxy and deoxyfluoro substrate analogues and inhibitors Biochem J 363, 335–340
20 Flint J, Taylor E, Yang M, Bolam DN, Tailford LE, Martinez-Fleites C, Dodson EJ, Davis BG, Gilbert HJ
& Davies GJ (2005) Structural dissection and high-throughput screening of mannosylglycerate synthase Nat Struct Mol Biol 12, 608–614
21 Boix E, Swaminathan GJ, Zhang Y, Natesh R, Brew K
& Acharya KR (2001) Structure of UDP complex of UDP-galactose: b-galactoside-a-1,3-galactosyltransferase
at 1.53-A˚ resolution reveals a conformational change in the catalytically important C terminus J Biol Chem
276, 48608–48614
22 Persson K, Ly HD, Dieckelmann M, Wakarchuk WW, Withers SG & Strynadka NC (2001) Crystal structure
of the retaining galactosyltransferase LgtC from Neisse-ria meningitidisin complex with donor and acceptor sugar analogs Nat Struct Biol 8, 166–175
23 Guerin ME, Kordulakova J, Schaeffer F, Svetlikova Z, Buschiazzo A, Giganti D, Gicquel B, Mikusova K, Jackson M & Alzari PM (2007) Molecular recognition and interfacial catalysis by the essential phosphatidyl-inositol mannosyltransferase PimA from mycobacteria
J Biol Chem 282, 20705–20714
24 Martinez-Fleites C, Proctor M, Roberts S, Bolam DN, Gilbert HJ & Davies GJ (2006) Insights into the synthe-sis of lipopolysaccharides and antibiotics through struc-tures of two retaining glycosyltransferases from family GT-4 Chem Biol 13, 1143–1152
25 Lariviere L, Sommer N & Morera S (2005) Structural evidence of a passive base-flipping mechanism for AGT,
an unusual GT-B glycosyltransferase J Mol Biol 352, 139–150
26 Buschiazzo A, Ugalde JE, Guerin ME, Shepard W, Ugalde RA & Alzari PM (2004) Crystal structure of glycogen synthase: homologous enzymes catalyze glyco-gen synthesis and degradation EMBO J 23, 3196–3205
27 Pederson BA, Cheng C, Wilson WA & Roach PJ (2000) Regulation of glycogen synthase Identification of resi-dues involved in regulation by the allosteric ligand glu-cose-6-P and by phosphorylation J Biol Chem 275, 27753–27761
28 Jamaluddin H, Tumbale P, Withers SG, Acharya KR & Brew K (2007) Conformational changes induced by binding UDP-2F-galactose to a-1,3 galactosyltransfer-ase – implications for catalysis J Mol Biol 369, 1270– 1281
29 Horcajada C, Guinovart JJ, Fita I & Ferrer JC (2006) Crystal structure of an archaeal glycogen synthase: insights into oligomerization and substrate binding of eukaryotic glycogen synthases J Biol Chem 281, 2923– 2931
30 Abdian PL, Lellouch AC, Gautier C, Ielpi L & Gere-mia RA (2000) Identification of essential amino acids in