- Email: [email protected]
Zn2+, Mg2+, and H+ binding to d -fructose 1,6-bisphosphate studied by 31P and 1H NMR
OF BIOCHEMISTRY 219, No. 2, December,
ARCHIVES
Vol.
Zn2+, Mg2+,
GERARD *Department
AND BIOPHYSICS pp. 268-276, 1982
and H+ Binding to D-Fructose 1,6-Bisphosphate Studied by 31P and ‘H NMR B. VAN
DEN
BERG*,’
AND
AREND
HEERSCHAPT
of Biochemistry I, Medical Faculty, Erasmus University Rotterdam, Rotterdam, and tDepartment of Biophysical Chemistry, University Toerrwoiveld 1, 6525 ED Nijmegen, The Netherlands Received
March
2, 1982, and in revised
form
July
PO Box 1738, of Nijmegen,
3000 DR
2, 1982
The anomeric composition and mutarotation rates of fructose 1,6-bisphosphate were determined in the presence of 100 mM KC1 at pH 7.0 by 31P NMR. At 23 and 37°C the solution contains (15 + l)% of the (Y anomer. The anomeric rate constants at 37°C are (4.2 + 0.4) s1 for the /3 - (Y anomerization and (14.9 + 0.5) s-l for the reverse reaction. A DzO effect between 2.1 and 2.6 was found. From acid base titration curves it appeared that the pK values of the phosphate groups range from 5.8 to 6.0. MS+ and Zn2+ bind preferentially to the l-phosphate in the cY-anomeric position. Zn2+ has a higher afhnity for this phosphate group than MS+ has. At increasing pH the fraction LY anomer decreases slightly. At increasing M$+/fructose 1,6-bisphosphate ratios the fraction 01 anomer increases till 19% at a ratio of 20. Proton and probably Mgz+ binding decreases the anomerization rate. The time-averaged preferred orientation of the l-phosphate along the C&O1 bond of the LY conformer is strongly pH dependent, gauche rotamers being predominant at pH 9.4. In the presence of divalent cations the orientation is biased toward trans. A mechanistic model is proposed to explain the Zn2+, M$+, and pH-dependent behavior of the gluconeogenic enzyme fructose 1,6-bisphosphatase.
D-Fructose-1,6-bisphosphate has been shown by 13C NMR spectroscopy to be an equilibrated mixture in aqueous solution composed of approximately 20% cy, 80% /3, and 2-4% keto anomer (l-3) (see Fig. 1). The gluconeogenic enzyme fructose 1,6bisphosphatase (fructose-1,6-bisphosphate 1-phosphohydrolase, EC 3.1.3.11) specifically uses ~Fru 1,6-P; as a substrate (4). Based on in vitro data it was calculated that in rat liver the rate of spontaneous anomerization is about lo-15 times slower than the maximal activity of FBPase (5, 6). This led Koerner et al. (5) to propose a model for the regulation of the Fru 6-P,
Fru 1,6-P, cycle based on the anomeric specificities of phosphofructokinase and FBPase. Earlier it has been proposed that divalent cations might shift the equilibrium of the anomerization to the (Y form by the formation of a chelate between the cation and the c&oriented phosphate groups (7). This would mean that in viuo the formation of the (Y anomer is more rapid than assumed. Experiments, however, contradicted this proposal (7). It is also possible that divalent cations increase both anomeric rate constants. Therefore we reinvestigated the influence of cations on the anomerization of Fru l,6-P2. It appeared that binding of Mg2+ will have little effect in duo. In the course of this study it appeared that Zn2+ and M$+ bind preferentially to the l-phosphate in the a! position. This phenomenon might have implications for
1 To whom all correspondence should be addressed. ’ Abbreviations used: Fru 1,6-Pz, fructose 1,6-bisphosphate; UMP, uridine 5’.monophosphate; Fru 6P, fructose 6-phosphate; EDTA, ethylenediaminetetraacetic acid; FBPase, fructose l&bisphosphatase. 0003.9861/62/140268-09$02.00/O Copyright All rights
Q 1982 by Academic Press. Inc. of reproduction in any form reserved.
268
CATION
TO
BINDING
D-FRUCTOSE
H20P CH20P
i)H
-
bH
I C-OH L H20P
tures
of Fru
a
keto
B
FIG. 1. Schematic
representation
of anomeric
struc-
1,6-P2.
the inhibition of rat liver FBPase (8-10) and Mg+ (11). MATERIALS
AND
by Zn2+
METHODS
Chemicals. Fru 1,6-Pz was purchased from Boehringer, Mannheim. KC1 and Mg(NO& (suprapure) were from E. Merck, Darmstadt. ZnCl, (ultrapure) was obtained from Ventron, Karlsruhe. Fru 1,6-Pz was prepared in a 200 mM stock solution. If indicated this solution was chromatographed over a column loaded with Dowex 50W. The acidic fractions were collected and set to pH 7.0 with KOH. Thereafter the Fru 1,6-P* was lyophilized and stored at 4°C until use. The powder was redissolved in a concentration as indicated in the legends of the figures in the pres-
269
1,6-BISPHOSPHATE
ence of 100 mM KC1 in 50% D20 for field frequency locking and other additions as indicated. The pH was set to the desired value with KOH or HCl. NMR spectroscopy. The 3’P NMR spectra were recorded on a Varian XL-100 spectrometer operating in the Fourier transform mode at 40.5 MHz. Heteronuclear proton noise decoupling was used to remove the J coupling induced by fructose protons. A pulse width of 20 ps was employed corresponding to a flip angle of 45”. Usually 200-1000 scans were accumulated with an aquisition time of 5 s, a computer delay of 0.5 s, and a digital resolution of 0.06 Hz/pt. Temperature was set with an accuracy of fl”C by a Varian temperature controller. Chemical shifts are given relative to 20% H3P04 as an external reference with downfield shifts defined as positive. ‘H NMR spectra were recorded on a Bruker WM 90. Chemical shifts are referred to 4,4-dimethyl-4-silapentane-1-sulfonate. RESULTS
Figure 2 shows a 31P NMR spectrum of Fru 1,6-Pz at 23°C in the presence of EDTA. On the basis of their relative chemical shifts, intensities, and 31P-1H coupling constants the resonances at 5.2 and 4.0 ppm can be assigned to the l- and 6-phosphates of cr-Fru 1,6-Pz, respectively, while those at 4.4 and 4.1 ppm are the corre-
/:I 23’C
1 5
1 4 Chemical
FIG. mM Fru
2. ‘lP
NMR spectrum 1,6-P,, 100 mM KCl,
Shift
[PPm)
of Fru 1,6-P* at 23 and 37’C. and 0.5 mM EDTA at pH 7.0.
Spectrum
of a solution
containing
50
270
VAN
DEN
BERG
sponding resonances of the p conformer (12). No resonances could be detected for the keto conformation, even in spectra measured at -10°C in the presence of dimethyl sulfoxide. It appeared that the resonance of the analog of keto-Fru 1,6-PZ, the keto conformer of dihydroxyacetone phosphate, measured under identical conditions, has the same chemical shift as the l-phosphate of P-Fru 1,6-PZ. So most likely the resonance of the keto conformer of Fru 1,6-PZ coincides with those of the /3 conformation. In the absence of EDTA considerable line broadening is observed, most likely due to contaminating paramagnetic ions. It is also observed that in the absence of EDTA the resonances are shifted upfield. This could be due to binding of divalent cations present in the commercial preparation of Fru 1,6-Pa. Indeed, when the Fru 1,6-Pa solution was pretreated with a cation exchanger (see Materials and Methods) a spectrum was obtained comparable to the spectrum in the presence of EDTA (not shown). Kinetics of the Mutarotation When the temperature is increased, the resonance of the (Y and p conformers belonging to the l-phosphate as well as the 6-phosphate broaden and move to each other (Fig. 2). This is typical for an exchange situation. Kinetic information can be obtained from the excess linewidth due to this exchange. We shall restrict ourselves to the resonances of the l-phosphate, since these are best resolved. According to the reaction sequence given in Fig. 1, the exchange of this phosphate should be treated as a three-sites problem. From the reaction rates given by Midelfort et al. (3) and the position of the keto conformer as outlined above, it follows that the lifetime of the keto conformation is small compared to its chemical shift differences (in Hz) with the (Y and /3 conformations. Furthermore, the keto conformation constitutes only 2% of the total amount of Fru 1,6-PZ (3). With these considerations it can be shown that the threesites exchange situation can be simplified to a two-sites problem (13); in this case between the (Y and /3 conformations.
AND
HEERSCHAP TABLE LINEWIDTH
T (‘-2)
pH
23 23 23 37 23
5.1 7.0 9.8 7.0 7.0
OF UMP
I
3’P NMR
RESONANCE
Addition
300 mM
AU0
Mg(NO&
0.45 0.45 0.44 0.45 0.67
f * + f f
0.07 0.06 0.05 0.06 0.04
Note. Spectra were recorded from a solution containing 100 mM UMP and 100 mM KCl. EDTA was present in a concentration of 0.5-10 mM and the D20 concentration was 30-100%. The concentration of EDTA and DzO did not influence the linewidth. Values are given kS.D. (n = 3).
Because, in the present case, the limit of slow exchange applies, the rate constants of the (Y - /3 anomerization (K,,) and the 0 - (Y anomerization (k,,) can be determined from the linewidth of the two individual resonances by k,, = II(Av,
- A&
k,, = II(Au,, - A$), in which Au stands for the width at halfheight of the resonance (14). The value Au’ is the width in the absence of exchange. This value was determined by measuring the linewidth of UMP. This is a good approximation since UMP has almost the same molecular weight as Fru 1,6-Pp. Relevant values of Av’ are given in Table I. Only Mg2+ has a significant effect on Au’. It appeared that the linewidths of the resonances of Fru l,6-P2 depend on the D20 concentration (Fig. 3). Because such a D20 effect could not be measured on the linewidth of the 31P resonance of UMP and dimethylhydroxyacetone phosphate, which do not exhibit conversions (not shown), it is concluded that this is due to a change in mutarotation rate of Fru l,6-P2. The rate constants were evaluated from the extrapolated linewidth at zero D,O concentration (Fig. 3). At 23°C this gives a k,, of (5.1 -t 0.2) s-l and a kpa of (1.2 + 0.3) S-l. These values were also evaluated at 37°C resulting in a k,, of (14.9 * 0.5) s-l and a k,, of (4.2 ? 0.4) s-l. The D,O effect
CATION
BINDING
TO
D-FRUCTOSE
was determined by integration of the (Yand p resonances of the l-phosphate. At 23 and 37°C the fraction (Y anomer is (15 + 1)s. This value is not dependent on the D20 concentration. Because nuclear Overhauser effects, due to the decoupling of the sugar protons, could be different for (Y and /3 resonances and hence affect the calculated ratio a/P-anomer, spectra were also recorded without decoupling. In this way we found the fraction LYanomer to be 16% which is in good agreement with the above established value.
‘1
1__1;:_
01
271
1,6-BISPHOSPHATE
Effect of Mg2+, Zn’+,
;0
[D20]in
% loo
FIG. 3. The effect of D20 on the linewidth of Fru 1,6-P, resonances. Spectra were recorded from a solution containing 120 mM Fru 1,6-P*, 10 mM EDTA, 100 mM KC1 at pH 7.0 and 23°C. Linewidth of 01 anomeric (m) and p-anomeric (0) resonances of the l-phosphate group.
(defined as kCHz0)/kCD20j was found to be between 2.1 and 2.6. The anomeric composition of Fru 1,6-P*
and H+
The effect of divalent cations on the 31P resonances was studied with a Dowextreated Fru 1,6-P2 solution without EDTA. Figure 4 shows that in the presence of Mg(NO,), all resonances are shifted upfield. Up to a Mg(NO&/Fru 1,6-P2 ratio of about 1, however, the resonances corresponding to the 6-phosphates are not affected which indicates that Mg2+ binds preferentially to the l-phosphate. Nonlinear regression of the titration curves yields
.
a
210
FIG. 4. Chemical shifts of Fru 1,6-Pz resonances as a function of the MgCl, concentration. Spectra were recorded from a Dowex-treated Fru 1,6-P, sample. The final solution contained 60 mM Fru 1,6-P,, 100 mM KC1 at pH 7.0. Temperature was 23’C. Circles, resonances of l-phosphates; squares, resonances of 6-phosphates. 0 q , (Y Resonances; 0 n , /3 Resonances.
272
FIG.
details
5. Chemical see Fig. 4.
shifts
VAN
DEN
of Fru
1,6-P,
BERG
resonances
a maximum shift for the l-phosphate group of the OLanomer of 1.05 ppm and for the p anomer of 0.80 ppm. The titration points up to a Ms+/Fru 1,6-Pz ratio of 1.2 were used for a Scatchard analysis. If one M?+ cation binds to a phosphate, we calculate an apparent binding constant for the l-phosphate of the (Y anomer of 13 and 6 M-l for the fl anomer. This indicates that Mg2+ binds preferentially to the I-phosphate of the CYanomer. When we consider the binding of Zn2’ (Fig. 5) we may also conclude that Zn2+ has a relatively high affinity for the lphosphate group of the CYanomer. However, a quantitative evaluation is hampered by the fact that at high concentrations of ZnC12 a precipitate is formed of Zn(OH)2. If we assume that the maximal shift is the same as with Mg(N03)2, then we calculate that at a cation/Fru l,6-P2 ratio of 0.17, the affinity for Zn2+ is four times higher than for M2+. The effect of divalent cations on the linewidth and anomeric composition were also evaluated. To avoid linebroadening from contaminating paramagnetic ions, a small amount of EDTA was added. Because an effect on the T2 relaxation time of resonances of phosphate groups in the
AND
HEERSCHAP
as a function
of the
Z&l,
concentration.
For
presence of Mg2+ has been reported (15), linebroadening can be expected. As can be seen in Table I, linebroadening is observed upon addition of M$+ to UMP. Considering Fru l,6-P2 we observed a
1 5
6
7 PH
FIG. 6. Chemical a function of pH.
shifts of Fru 1,6-P, For details see Fig.
resonances 4.
as
CATION
BINDING
TO
D-FRUCTOSE
273
1,6-BISPHOSPHATE
8 gives the relevant nomenclature on the orientation of the l-phosphate. The fractional population of the transrotamer (P,) can be estimated from (16): p
I
24 6
a
10 PH
FIG. 7. Influence of pH on the rate constant of the anomerization and anomeric composition of Fru 1,6Pp. Conditions as in Fig. 4.
linebroadening of 0.38 Hz for the /3 resonance, but only 0.04 Hz for the LYresonance of the l-phosphate at a Mg’+/Fru 1,6-P2 ratio of 6. This might indicate that k,, is decreased upon Mg2+ binding. This is substantiated by the observation that at a Mg’+/Fru l,6-P2 ratio of 6 the fraction (Y anomer is 17% and at a ratio of 20 this fraction is 19%. However the effect is small. It is well known that the chemical shift of phosphate monoesters is very sensitive to the degree of ionization of these acids. Figure 6 represents an acid base titration curve of Fru l,6-Pz. The pK values for the (Y and p conformations of the l-phosphate are 5.8 and 5.9, respectively. These values are both 6.0 for the 6-phosphate. The effect of pH on the rate constant k,, and anomeric composition is shown in Fig. 7. It can be concluded that the degree of ionization influences this rate constant. Furthermore at high pH the anomeric equilibrium shifts slightly in the direction of the fl anomer. Conformation Fru 1,6-P,
of the Phosphate
Groups
t
= 24 -
(JIHP +
JI'HP)
18
Coupling constants and calculated Pt values are given in Table II. It can be seen that the l-phosphate of the (Yanomer prefers gauche orientations, while the timeaveraged preferred orientation of the /3 anomer is trans. The time-averaged preferred orientations are dependent on a number of conditions. An increase in temperature from 6 to 34°C slightly increases the contribution of gauche orientations. At low pH both phosphates tend to increase the population of their trans rotamer, which is also observed in the presence of divalent cations. These latter effects are more pronounced on the a-phosphate than on the P-phosphate. The effect of divalent cation binding and protonation was further studied by ‘H NMR. Figure 9a shows a 90-MHz ‘H NMR spectrum of a Fru l,6-P2 solution. Up till now none of the resonances has been assigned to particular protons of the sugar. After addition of ZnC12some minor changes around 3.85 ppm can be observed (Fig. 9b). However the resonance at 4.00 ppm clearly shifts downfield to 4.07 ppm (arrow). Under these conditions only the resonance of the l-phosphate of the (Y anomer shifts (see Fig. 5). Therefore the resonance at 4.00 ppm in the ‘H NMR spectrum is tentatively assigned to protons of the CYanomer. Upon addition of Mg(NO,), also a downfield shift of this resonance is observed (not shown).
of
‘H-“‘P coupling constants are sensitive to the time-averaged preferred conformation of phosphate esters along the C-O bonds (16). We measured the effect of temperature, pH, and divalent cation binding on the coupling constants (JHP) of the resonances of the l-phosphate group. Figure
gauche
FIG. 8. Newman projections entations of the l-phosphate viewed along the C,-0, bond.
+
gauche-
of the dominant origroup of Fru 1,6-P*
274
VAN
DEN
BERG
AND
TABLE ‘H-alp
HEERSCHAP II
COUPLINGCONSTANTSANDTIME-AVERAGEDPREFERREDORIENTATIONS ALONG
THE
C,-0,
BOND
OF
FRU
1.6-P,
cy Anomer T (“C) 6 23 34 23 23 23 23 23
Addition
PH
7.0 7.0 7.0 5.4 9.4 7.0 7.0 7.0
ZnCla (0.17)” MgWU (3.31” M&‘JW~ (5)”
0 Anomer
J in Hz
Pt in %
J in Hz
7.8 7.9 8.0 6.5 8.3 7.5 7.2 7.0
47 46 44 61 41 50 54 56
5.9
6.0 6.0 5.6 6.2 6.0 5.9 5.8
Pt in % 68 67 67 71 64 67 68 69
a b a a a b b b
Note. Samples contained 60 mM Fru 1,6-P2, 100 mM KC1 in the presence of 5-10 mM EDTA (a). In some cases a Dowex-treated Fru 1,6-P* sample was used in the absence of EDTA (b). H, and H,, are virtually isochronous in their coupling to the phosphates and are given as one value. Pt is the fraction transrotamer. a Value in parentheses represents the ratio of cation to Fru 1,6-P* concentration.
At pH 5.4 again some shifts around 3.85 ppm are visible but the resonance at 4.00 ppm virtually does not change its position (Fig. 9c). From this we conclude that divalent cations bind in a different way to a-Fru 1,6-Pz than protons do.
DISCUSSION
Kinetics and Mechanism of Mutarotation The anomeric composition of Fru 1,6-P* has been determined by several investigators (Table III). The composition of 15%
a
CHEMICAL
SHIFT(PPM)
FIG. 9. ‘H NMR spectrum of Fru 1,6-Pz. (a) Spectrum of a Dowex-treated Fru 1,6-P* solution (40 mM) in the presence of 100 mM KC1 at pH 7.0. (b) Idem after adding ZnClz to a ZnCls/Fru 1,6P, ratio of 0.15. (c) Spectrum at pH 5.4 in the presence of 10 mM EDTA. t, Extra resonance intensity due to EDTA protons.
CATION
BINDING
TO
D-FRUCTOSE
l$BISPHOSPHATE
275
can be seen that divalent cations that are bound to the cx anomer might coordinate DISTRIBUTION OFTHEFURANOSEFORMS with the 3-OH group. The distance of the OF FRU 1,6-Pz l-phosphate of the (Y anomer to this hydroxyl group and its conformation (see Method Reference X(3 Table I) makes this less likely to be so for 13C NMR 2 23:70 the p anomer. This could explain the rel3 16:84 =C NMR ative high affinity of the (Yanomer for di20:80 ‘+2 NMR I valent cations. Supporting evidence is ob10:90 ‘IP NMR 11 tained by ‘H NMR: Coordination with the =P NMR This study 15:85 3-OH group could result in a deshielding effect on protons located on C-3 and C-4. Indeed in the ‘H NMR spectrum a downCYand 85% p anomer reported here agrees field shift was observed of a resonance that well with the most accurate values ob- was tentatively attributed to cu-anomeric tained by Midelfort et al. (3) using 13C protons. From the upfield shift of the resonances NMR. The mutarotation rate constants observed in “crude” (i.e., not treated with derived by us are also in good agreement Dowex) preparations of Fru 1,6-Pz, we conwith those obtained by these authors. The D20 effect of 2.1-2.6 reported here clude that these preparations are contamhas been frequently observed in the mu- inated with divalent cations. One of these tarotation of nonphosphorylated sugars ions could be Zn’+, which is then responsible for the kinetic hysteresis of FBPase (17). A DzO effect greater than 1 indicates (see accompanying paper (25)). that proton transfer is a rate-controlling step in the mutarotation. In comparison In Viva Situation to nonphosphorylated sugars the mutarotation rates of phosphorylated sugars are In uivo the free Mg2+ concentration is very rapid (3, 17, 18). Baily et al. (18) have about 1 mM (19) and the concentration of proposed that this is due to intramolecular Fru 1,6-Pz in rat liver is only 20 pM (20) phosphate catalysis. In their model study so that the Mg’+/Fru 1,6-Pz ratio is about on the mutarotation of a-D-glucose they 50. At such a high ratio the anomeric comfound that this reaction can be accelerated position shifts in the direction of the CY by inorganic phosphate; the dibasic species anomer. The /3/a ratio will be about 4.0. being a better catalyst than the monobasic Our measurements on the linewidth of species. A similar behavior is reported here the phosphate resonance and the (Y/P ratio for Fru 1,6-Pz (see Fig. 7). We also found in the presence of Mg’+ indicate that the that the time-averaged preferred orientakinetics of Fru 1,6-Pz are affected to a tion of the l-phosphate group of the ol ansmall extent by this cation. This means omer is sensitive to the degree of ionization that in the in uivo situation, when Fru 1,6of this group. More gauche orientations Pz is completely saturated with Mg2+, the coincide with an increased mutarotation. mutarotaion rate constants could be It cannot be excluded that the orientation slightly different from the values obtained of the l-phosphate is also important in the by 13CNMR (3) and ‘lP NMR. From the phosphate catalysis. Considering this type considerations given above it is concluded of catalysis it can be expected that binding that cellular changes in Mgzf have no efof divalent cations has some influence on fect on the mutarotation rate and anothe mutarotation rates and anomeric commerit composition of Fru 1,6-P2. position. However, our measurements with Mg2+ indicate that these effects are small Relation with Enzyme Activity (see also Ref. (7)). From the titration curves with Mg2+ and FBPase shows rather complex kinetic Zn” (Figs. 4 and 5) we conclude that the properties, including activation by chelatl-phosphate has the highest affinity for ing agents accompanied by a shift of the divalent cations. From model building it pH optimum from more alkaline to neutral TABLE
III
276
VAN
DEN
BERG
AND
pH (19). The inactivation has been attributed to binding of Zn2+ to the enzyme (8). Apart from these inhibitory sites it was necessary to postulate the existence of activating sites (8) to account for the activity in the presence of Zn2’ alone. Binding studies revealed three classes of binding sites(9, 10). It might be speculated that, depending on the ionic conditions, the hydrolyses of Fru l,6-P2 occurs via a dissociative or an associative mechanism, each mechanism characterized by certain kinetic parameters. Metal ions can inhibit a dissociative mechanism or activate an associative mechanism or vice versa (21, 22), and so change the kinetics of rat liver FBPase. As pointed out by Mildvan (22), there are differences in space requirements at the catalytic site between an associative and dissociative mechanism. From Table II it appears that the time-averaged orientation of the l-phosphate of a-Fru 1,6P2 along the C,-0, bond is strongly dependent upon the pH. The sharp pH optimum in FBPase activity in the presence of EDTA (23, 24) might indicate that a particular conformation of the l-phosphate group is necessary for hydrolyses. As Zn2+ inhibits at extremely low concentrations (Ki = 0.3 PM, Ref. (8)), we assume that, after turnover, Zn2’ remains bound to the enzyme near the catalytic site. Evidence for this proposal is given in the accompanying paper (25).
HEERSCHAP
3. MIDELFORT, C. F., GUPTA, J. A. (1976) Biochemistry 4. FREY, W. A., FISHBEIN, R., AND
Haasnoot Koster, Hilbers
Dr.
J. A. L.
Walters
and
Dr.
C. A.
G.
for valuable suggestions. Prof. Dr. J. F. Dr. Th. J. C. van Berkel, and Prof. Dr. C. W. are thanked for reading the manuscript. Miss
A. C. Hanson is thanked for typing the manuscript. The Netherlands Foundation for Fundamental Medical Research (FUNGO) is acknowledged for partial financial support (Grant 13-39-18).
6. BLOXHAM,
Biophys.
Res. Commun.
51, 543-549.
D. P., ANDYORK,
Biochem.
D. A. (1976)
Sot.
Trans. 4, 989-993. 7. BENKOVIC, S. J. (1972) Biochem. Commun. 47, 852-858. 8. TEJWANI, G. O., PEDROSA, S., AND HORECKER, B.
Acad.
Sci. USA
9. PEDROSA, RECKER,
74, 10.
S.,
F. O., PONTREMOLI, L. (1976) P~oc. Nat.
2692-2695.
MELLONI,
B., AND
S.,
AND
Biochem.
E.,
DE
14.
CARRINGTON, in Introduction
SALAMINO,
M.,
AND
248,
BYRNE,
8567-8573.
10,4705-4711.
Biochemistry
BOER, J. W. M. (1977) Lanthanide Shifts Reagents Binding, Shift Mechanisms and Exchange, Ph.D. thesis, University of Nymegen, Krips Repro, Meppel. A., AND MCLACHLAN, A. D. (1967) to Magnetic Resonance, Har-
per & Row, Tokyo. COHEN,
New
S. M., AND
Acad.
Sci. USA
York,
and
BURT,
74,
John
Weatherhill,
P~oc. Nat.
C. T. (1977)
4271-4275.
16.
WOOD, D. J., HRUSKA, F. E., AND OGILVIE, (1974) Canad. J. Chem. 52, 3353-3366.
17.
ISBELL,
20.
F.,
B. L. (1978)
188, 90-97.
Biophys.
G. R. (1971)
GRAY,
13.
18.
Ho-
Acad. Sci. USA
HORECKER,
MARCUS, C. J., GELLER, A. W. L. (1973) J. Biol. Chem.
12.
15.
Res.
2742-2745.
PONTREMOLI,
Arch. 11.
73,
Biophys.
F. O., PONTREMOLI, B. L. (1977) Proc. Nat.
SPARATORE,
H.
S., AND
Carbohydr.
Chem.
BAILEY, J. M., P. G. (1970)
PIGMAN,
W.
Biochem.
FISHMAN,
Biochemistry
24,
K. K.
Aduan.
(1969) 14-63.
P. H., AND PERTCHEV, 9, 1189-1194.
VELOSO, D., GUYNN, R. W., OSKARSSON, M., AND VEECH, R. L. (1973) J. Biol. Chem. 248,48114819. FAUPEL, (1972)
R. P., SEITZ,
Arch.
H. J., AND
Biochem.
TARNOWSKI,
Biophys.
JENCKS, and New
22.
MILDVAN, (Opella, Dekker,
23.
HORECKER, B. L., MELLONI, E., AND PONTREMOLI, S. (1975) in Advances in Enzymology (Meister, A., ed.), Vol. 42, pp. 1933226, Wiley,
24.
New NIMMO,
J. 25.
W. P. (1969). in Catalysis Enzymology, pp. 112-115, York.
W.
148, 509-522.
21.
REFERENCES 1. BENKOVIC, S. J., ENGLE, J. L., AND MILDVAN, A. S. (1972) Biochem. Biophys. Res. Commun. 47, 852-858. 2. KOERNER, T. A. W., CARY, L. W., BHACCA, N. S., AND YOUNATHAN, E. S. (1973) Biochem.
16,
Biochemistry
5. KOERNER, A. W., VOLL, R. J., AND YOUNATHAN, E. S. (1977) FEBS Z&t. 84, 207-213.
19.
thank
DE
S. J. (1977).
BENKOVIC,
K., AND ROSE, 21782185. MAINE, M. M.,
2479-2484.
ACKNOWLEDGMENTS We
R.
15,
VAN H.,
A. S. (1979) S. J., and New York.
in NMR and Lu, P., eds.),
York. H. G., ANDTIPTON,
in Chemistry McGraw-Hill, Biochemistry pp. 3033304,
K. F. (1975)
Biochem.
145,323-334. DEN BERG, G. B., VAANDRAGER-VERDUIN, VAN BERKEL, T. J. C., AND KOSTER,
(1982)
Arch.
Biochem.
Biophys.
J. F.
219, 277-285.
ARCHIVES
Vol.
Zn2+, Mg2+,
GERARD *Department
AND BIOPHYSICS pp. 268-276, 1982
and H+ Binding to D-Fructose 1,6-Bisphosphate Studied by 31P and ‘H NMR B. VAN
DEN
BERG*,’
AND
AREND
HEERSCHAPT
of Biochemistry I, Medical Faculty, Erasmus University Rotterdam, Rotterdam, and tDepartment of Biophysical Chemistry, University Toerrwoiveld 1, 6525 ED Nijmegen, The Netherlands Received
March
2, 1982, and in revised
form
July
PO Box 1738, of Nijmegen,
3000 DR
2, 1982
The anomeric composition and mutarotation rates of fructose 1,6-bisphosphate were determined in the presence of 100 mM KC1 at pH 7.0 by 31P NMR. At 23 and 37°C the solution contains (15 + l)% of the (Y anomer. The anomeric rate constants at 37°C are (4.2 + 0.4) s1 for the /3 - (Y anomerization and (14.9 + 0.5) s-l for the reverse reaction. A DzO effect between 2.1 and 2.6 was found. From acid base titration curves it appeared that the pK values of the phosphate groups range from 5.8 to 6.0. MS+ and Zn2+ bind preferentially to the l-phosphate in the cY-anomeric position. Zn2+ has a higher afhnity for this phosphate group than MS+ has. At increasing pH the fraction LY anomer decreases slightly. At increasing M$+/fructose 1,6-bisphosphate ratios the fraction 01 anomer increases till 19% at a ratio of 20. Proton and probably Mgz+ binding decreases the anomerization rate. The time-averaged preferred orientation of the l-phosphate along the C&O1 bond of the LY conformer is strongly pH dependent, gauche rotamers being predominant at pH 9.4. In the presence of divalent cations the orientation is biased toward trans. A mechanistic model is proposed to explain the Zn2+, M$+, and pH-dependent behavior of the gluconeogenic enzyme fructose 1,6-bisphosphatase.
D-Fructose-1,6-bisphosphate has been shown by 13C NMR spectroscopy to be an equilibrated mixture in aqueous solution composed of approximately 20% cy, 80% /3, and 2-4% keto anomer (l-3) (see Fig. 1). The gluconeogenic enzyme fructose 1,6bisphosphatase (fructose-1,6-bisphosphate 1-phosphohydrolase, EC 3.1.3.11) specifically uses ~Fru 1,6-P; as a substrate (4). Based on in vitro data it was calculated that in rat liver the rate of spontaneous anomerization is about lo-15 times slower than the maximal activity of FBPase (5, 6). This led Koerner et al. (5) to propose a model for the regulation of the Fru 6-P,
Fru 1,6-P, cycle based on the anomeric specificities of phosphofructokinase and FBPase. Earlier it has been proposed that divalent cations might shift the equilibrium of the anomerization to the (Y form by the formation of a chelate between the cation and the c&oriented phosphate groups (7). This would mean that in viuo the formation of the (Y anomer is more rapid than assumed. Experiments, however, contradicted this proposal (7). It is also possible that divalent cations increase both anomeric rate constants. Therefore we reinvestigated the influence of cations on the anomerization of Fru l,6-P2. It appeared that binding of Mg2+ will have little effect in duo. In the course of this study it appeared that Zn2+ and M$+ bind preferentially to the l-phosphate in the a! position. This phenomenon might have implications for
1 To whom all correspondence should be addressed. ’ Abbreviations used: Fru 1,6-Pz, fructose 1,6-bisphosphate; UMP, uridine 5’.monophosphate; Fru 6P, fructose 6-phosphate; EDTA, ethylenediaminetetraacetic acid; FBPase, fructose l&bisphosphatase. 0003.9861/62/140268-09$02.00/O Copyright All rights
Q 1982 by Academic Press. Inc. of reproduction in any form reserved.
268
CATION
TO
BINDING
D-FRUCTOSE
H20P CH20P
i)H
-
bH
I C-OH L H20P
tures
of Fru
a
keto
B
FIG. 1. Schematic
representation
of anomeric
struc-
1,6-P2.
the inhibition of rat liver FBPase (8-10) and Mg+ (11). MATERIALS
AND
by Zn2+
METHODS
Chemicals. Fru 1,6-Pz was purchased from Boehringer, Mannheim. KC1 and Mg(NO& (suprapure) were from E. Merck, Darmstadt. ZnCl, (ultrapure) was obtained from Ventron, Karlsruhe. Fru 1,6-Pz was prepared in a 200 mM stock solution. If indicated this solution was chromatographed over a column loaded with Dowex 50W. The acidic fractions were collected and set to pH 7.0 with KOH. Thereafter the Fru 1,6-P* was lyophilized and stored at 4°C until use. The powder was redissolved in a concentration as indicated in the legends of the figures in the pres-
269
1,6-BISPHOSPHATE
ence of 100 mM KC1 in 50% D20 for field frequency locking and other additions as indicated. The pH was set to the desired value with KOH or HCl. NMR spectroscopy. The 3’P NMR spectra were recorded on a Varian XL-100 spectrometer operating in the Fourier transform mode at 40.5 MHz. Heteronuclear proton noise decoupling was used to remove the J coupling induced by fructose protons. A pulse width of 20 ps was employed corresponding to a flip angle of 45”. Usually 200-1000 scans were accumulated with an aquisition time of 5 s, a computer delay of 0.5 s, and a digital resolution of 0.06 Hz/pt. Temperature was set with an accuracy of fl”C by a Varian temperature controller. Chemical shifts are given relative to 20% H3P04 as an external reference with downfield shifts defined as positive. ‘H NMR spectra were recorded on a Bruker WM 90. Chemical shifts are referred to 4,4-dimethyl-4-silapentane-1-sulfonate. RESULTS
Figure 2 shows a 31P NMR spectrum of Fru 1,6-Pz at 23°C in the presence of EDTA. On the basis of their relative chemical shifts, intensities, and 31P-1H coupling constants the resonances at 5.2 and 4.0 ppm can be assigned to the l- and 6-phosphates of cr-Fru 1,6-Pz, respectively, while those at 4.4 and 4.1 ppm are the corre-
/:I 23’C
1 5
1 4 Chemical
FIG. mM Fru
2. ‘lP
NMR spectrum 1,6-P,, 100 mM KCl,
Shift
[PPm)
of Fru 1,6-P* at 23 and 37’C. and 0.5 mM EDTA at pH 7.0.
Spectrum
of a solution
containing
50
270
VAN
DEN
BERG
sponding resonances of the p conformer (12). No resonances could be detected for the keto conformation, even in spectra measured at -10°C in the presence of dimethyl sulfoxide. It appeared that the resonance of the analog of keto-Fru 1,6-PZ, the keto conformer of dihydroxyacetone phosphate, measured under identical conditions, has the same chemical shift as the l-phosphate of P-Fru 1,6-PZ. So most likely the resonance of the keto conformer of Fru 1,6-PZ coincides with those of the /3 conformation. In the absence of EDTA considerable line broadening is observed, most likely due to contaminating paramagnetic ions. It is also observed that in the absence of EDTA the resonances are shifted upfield. This could be due to binding of divalent cations present in the commercial preparation of Fru 1,6-Pa. Indeed, when the Fru 1,6-Pa solution was pretreated with a cation exchanger (see Materials and Methods) a spectrum was obtained comparable to the spectrum in the presence of EDTA (not shown). Kinetics of the Mutarotation When the temperature is increased, the resonance of the (Y and p conformers belonging to the l-phosphate as well as the 6-phosphate broaden and move to each other (Fig. 2). This is typical for an exchange situation. Kinetic information can be obtained from the excess linewidth due to this exchange. We shall restrict ourselves to the resonances of the l-phosphate, since these are best resolved. According to the reaction sequence given in Fig. 1, the exchange of this phosphate should be treated as a three-sites problem. From the reaction rates given by Midelfort et al. (3) and the position of the keto conformer as outlined above, it follows that the lifetime of the keto conformation is small compared to its chemical shift differences (in Hz) with the (Y and /3 conformations. Furthermore, the keto conformation constitutes only 2% of the total amount of Fru 1,6-PZ (3). With these considerations it can be shown that the threesites exchange situation can be simplified to a two-sites problem (13); in this case between the (Y and /3 conformations.
AND
HEERSCHAP TABLE LINEWIDTH
T (‘-2)
pH
23 23 23 37 23
5.1 7.0 9.8 7.0 7.0
OF UMP
I
3’P NMR
RESONANCE
Addition
300 mM
AU0
Mg(NO&
0.45 0.45 0.44 0.45 0.67
f * + f f
0.07 0.06 0.05 0.06 0.04
Note. Spectra were recorded from a solution containing 100 mM UMP and 100 mM KCl. EDTA was present in a concentration of 0.5-10 mM and the D20 concentration was 30-100%. The concentration of EDTA and DzO did not influence the linewidth. Values are given kS.D. (n = 3).
Because, in the present case, the limit of slow exchange applies, the rate constants of the (Y - /3 anomerization (K,,) and the 0 - (Y anomerization (k,,) can be determined from the linewidth of the two individual resonances by k,, = II(Av,
- A&
k,, = II(Au,, - A$), in which Au stands for the width at halfheight of the resonance (14). The value Au’ is the width in the absence of exchange. This value was determined by measuring the linewidth of UMP. This is a good approximation since UMP has almost the same molecular weight as Fru 1,6-Pp. Relevant values of Av’ are given in Table I. Only Mg2+ has a significant effect on Au’. It appeared that the linewidths of the resonances of Fru l,6-P2 depend on the D20 concentration (Fig. 3). Because such a D20 effect could not be measured on the linewidth of the 31P resonance of UMP and dimethylhydroxyacetone phosphate, which do not exhibit conversions (not shown), it is concluded that this is due to a change in mutarotation rate of Fru l,6-P2. The rate constants were evaluated from the extrapolated linewidth at zero D,O concentration (Fig. 3). At 23°C this gives a k,, of (5.1 -t 0.2) s-l and a kpa of (1.2 + 0.3) S-l. These values were also evaluated at 37°C resulting in a k,, of (14.9 * 0.5) s-l and a k,, of (4.2 ? 0.4) s-l. The D,O effect
CATION
BINDING
TO
D-FRUCTOSE
was determined by integration of the (Yand p resonances of the l-phosphate. At 23 and 37°C the fraction (Y anomer is (15 + 1)s. This value is not dependent on the D20 concentration. Because nuclear Overhauser effects, due to the decoupling of the sugar protons, could be different for (Y and /3 resonances and hence affect the calculated ratio a/P-anomer, spectra were also recorded without decoupling. In this way we found the fraction LYanomer to be 16% which is in good agreement with the above established value.
‘1
1__1;:_
01
271
1,6-BISPHOSPHATE
Effect of Mg2+, Zn’+,
;0
[D20]in
% loo
FIG. 3. The effect of D20 on the linewidth of Fru 1,6-P, resonances. Spectra were recorded from a solution containing 120 mM Fru 1,6-P*, 10 mM EDTA, 100 mM KC1 at pH 7.0 and 23°C. Linewidth of 01 anomeric (m) and p-anomeric (0) resonances of the l-phosphate group.
(defined as kCHz0)/kCD20j was found to be between 2.1 and 2.6. The anomeric composition of Fru 1,6-P*
and H+
The effect of divalent cations on the 31P resonances was studied with a Dowextreated Fru 1,6-P2 solution without EDTA. Figure 4 shows that in the presence of Mg(NO,), all resonances are shifted upfield. Up to a Mg(NO&/Fru 1,6-P2 ratio of about 1, however, the resonances corresponding to the 6-phosphates are not affected which indicates that Mg2+ binds preferentially to the l-phosphate. Nonlinear regression of the titration curves yields
.
a
210
FIG. 4. Chemical shifts of Fru 1,6-Pz resonances as a function of the MgCl, concentration. Spectra were recorded from a Dowex-treated Fru 1,6-P, sample. The final solution contained 60 mM Fru 1,6-P,, 100 mM KC1 at pH 7.0. Temperature was 23’C. Circles, resonances of l-phosphates; squares, resonances of 6-phosphates. 0 q , (Y Resonances; 0 n , /3 Resonances.
272
FIG.
details
5. Chemical see Fig. 4.
shifts
VAN
DEN
of Fru
1,6-P,
BERG
resonances
a maximum shift for the l-phosphate group of the OLanomer of 1.05 ppm and for the p anomer of 0.80 ppm. The titration points up to a Ms+/Fru 1,6-Pz ratio of 1.2 were used for a Scatchard analysis. If one M?+ cation binds to a phosphate, we calculate an apparent binding constant for the l-phosphate of the (Y anomer of 13 and 6 M-l for the fl anomer. This indicates that Mg2+ binds preferentially to the I-phosphate of the CYanomer. When we consider the binding of Zn2’ (Fig. 5) we may also conclude that Zn2+ has a relatively high affinity for the lphosphate group of the CYanomer. However, a quantitative evaluation is hampered by the fact that at high concentrations of ZnC12 a precipitate is formed of Zn(OH)2. If we assume that the maximal shift is the same as with Mg(N03)2, then we calculate that at a cation/Fru l,6-P2 ratio of 0.17, the affinity for Zn2+ is four times higher than for M2+. The effect of divalent cations on the linewidth and anomeric composition were also evaluated. To avoid linebroadening from contaminating paramagnetic ions, a small amount of EDTA was added. Because an effect on the T2 relaxation time of resonances of phosphate groups in the
AND
HEERSCHAP
as a function
of the
Z&l,
concentration.
For
presence of Mg2+ has been reported (15), linebroadening can be expected. As can be seen in Table I, linebroadening is observed upon addition of M$+ to UMP. Considering Fru l,6-P2 we observed a
1 5
6
7 PH
FIG. 6. Chemical a function of pH.
shifts of Fru 1,6-P, For details see Fig.
resonances 4.
as
CATION
BINDING
TO
D-FRUCTOSE
273
1,6-BISPHOSPHATE
8 gives the relevant nomenclature on the orientation of the l-phosphate. The fractional population of the transrotamer (P,) can be estimated from (16): p
I
24 6
a
10 PH
FIG. 7. Influence of pH on the rate constant of the anomerization and anomeric composition of Fru 1,6Pp. Conditions as in Fig. 4.
linebroadening of 0.38 Hz for the /3 resonance, but only 0.04 Hz for the LYresonance of the l-phosphate at a Mg’+/Fru 1,6-P2 ratio of 6. This might indicate that k,, is decreased upon Mg2+ binding. This is substantiated by the observation that at a Mg’+/Fru l,6-P2 ratio of 6 the fraction (Y anomer is 17% and at a ratio of 20 this fraction is 19%. However the effect is small. It is well known that the chemical shift of phosphate monoesters is very sensitive to the degree of ionization of these acids. Figure 6 represents an acid base titration curve of Fru l,6-Pz. The pK values for the (Y and p conformations of the l-phosphate are 5.8 and 5.9, respectively. These values are both 6.0 for the 6-phosphate. The effect of pH on the rate constant k,, and anomeric composition is shown in Fig. 7. It can be concluded that the degree of ionization influences this rate constant. Furthermore at high pH the anomeric equilibrium shifts slightly in the direction of the fl anomer. Conformation Fru 1,6-P,
of the Phosphate
Groups
t
= 24 -
(JIHP +
JI'HP)
18
Coupling constants and calculated Pt values are given in Table II. It can be seen that the l-phosphate of the (Yanomer prefers gauche orientations, while the timeaveraged preferred orientation of the /3 anomer is trans. The time-averaged preferred orientations are dependent on a number of conditions. An increase in temperature from 6 to 34°C slightly increases the contribution of gauche orientations. At low pH both phosphates tend to increase the population of their trans rotamer, which is also observed in the presence of divalent cations. These latter effects are more pronounced on the a-phosphate than on the P-phosphate. The effect of divalent cation binding and protonation was further studied by ‘H NMR. Figure 9a shows a 90-MHz ‘H NMR spectrum of a Fru l,6-P2 solution. Up till now none of the resonances has been assigned to particular protons of the sugar. After addition of ZnC12some minor changes around 3.85 ppm can be observed (Fig. 9b). However the resonance at 4.00 ppm clearly shifts downfield to 4.07 ppm (arrow). Under these conditions only the resonance of the l-phosphate of the (Y anomer shifts (see Fig. 5). Therefore the resonance at 4.00 ppm in the ‘H NMR spectrum is tentatively assigned to protons of the CYanomer. Upon addition of Mg(NO,), also a downfield shift of this resonance is observed (not shown).
of
‘H-“‘P coupling constants are sensitive to the time-averaged preferred conformation of phosphate esters along the C-O bonds (16). We measured the effect of temperature, pH, and divalent cation binding on the coupling constants (JHP) of the resonances of the l-phosphate group. Figure
gauche
FIG. 8. Newman projections entations of the l-phosphate viewed along the C,-0, bond.
+
gauche-
of the dominant origroup of Fru 1,6-P*
274
VAN
DEN
BERG
AND
TABLE ‘H-alp
HEERSCHAP II
COUPLINGCONSTANTSANDTIME-AVERAGEDPREFERREDORIENTATIONS ALONG
THE
C,-0,
BOND
OF
FRU
1.6-P,
cy Anomer T (“C) 6 23 34 23 23 23 23 23
Addition
PH
7.0 7.0 7.0 5.4 9.4 7.0 7.0 7.0
ZnCla (0.17)” MgWU (3.31” M&‘JW~ (5)”
0 Anomer
J in Hz
Pt in %
J in Hz
7.8 7.9 8.0 6.5 8.3 7.5 7.2 7.0
47 46 44 61 41 50 54 56
5.9
6.0 6.0 5.6 6.2 6.0 5.9 5.8
Pt in % 68 67 67 71 64 67 68 69
a b a a a b b b
Note. Samples contained 60 mM Fru 1,6-P2, 100 mM KC1 in the presence of 5-10 mM EDTA (a). In some cases a Dowex-treated Fru 1,6-P* sample was used in the absence of EDTA (b). H, and H,, are virtually isochronous in their coupling to the phosphates and are given as one value. Pt is the fraction transrotamer. a Value in parentheses represents the ratio of cation to Fru 1,6-P* concentration.
At pH 5.4 again some shifts around 3.85 ppm are visible but the resonance at 4.00 ppm virtually does not change its position (Fig. 9c). From this we conclude that divalent cations bind in a different way to a-Fru 1,6-Pz than protons do.
DISCUSSION
Kinetics and Mechanism of Mutarotation The anomeric composition of Fru 1,6-P* has been determined by several investigators (Table III). The composition of 15%
a
CHEMICAL
SHIFT(PPM)
FIG. 9. ‘H NMR spectrum of Fru 1,6-Pz. (a) Spectrum of a Dowex-treated Fru 1,6-P* solution (40 mM) in the presence of 100 mM KC1 at pH 7.0. (b) Idem after adding ZnClz to a ZnCls/Fru 1,6P, ratio of 0.15. (c) Spectrum at pH 5.4 in the presence of 10 mM EDTA. t, Extra resonance intensity due to EDTA protons.
CATION
BINDING
TO
D-FRUCTOSE
l$BISPHOSPHATE
275
can be seen that divalent cations that are bound to the cx anomer might coordinate DISTRIBUTION OFTHEFURANOSEFORMS with the 3-OH group. The distance of the OF FRU 1,6-Pz l-phosphate of the (Y anomer to this hydroxyl group and its conformation (see Method Reference X(3 Table I) makes this less likely to be so for 13C NMR 2 23:70 the p anomer. This could explain the rel3 16:84 =C NMR ative high affinity of the (Yanomer for di20:80 ‘+2 NMR I valent cations. Supporting evidence is ob10:90 ‘IP NMR 11 tained by ‘H NMR: Coordination with the =P NMR This study 15:85 3-OH group could result in a deshielding effect on protons located on C-3 and C-4. Indeed in the ‘H NMR spectrum a downCYand 85% p anomer reported here agrees field shift was observed of a resonance that well with the most accurate values ob- was tentatively attributed to cu-anomeric tained by Midelfort et al. (3) using 13C protons. From the upfield shift of the resonances NMR. The mutarotation rate constants observed in “crude” (i.e., not treated with derived by us are also in good agreement Dowex) preparations of Fru 1,6-Pz, we conwith those obtained by these authors. The D20 effect of 2.1-2.6 reported here clude that these preparations are contamhas been frequently observed in the mu- inated with divalent cations. One of these tarotation of nonphosphorylated sugars ions could be Zn’+, which is then responsible for the kinetic hysteresis of FBPase (17). A DzO effect greater than 1 indicates (see accompanying paper (25)). that proton transfer is a rate-controlling step in the mutarotation. In comparison In Viva Situation to nonphosphorylated sugars the mutarotation rates of phosphorylated sugars are In uivo the free Mg2+ concentration is very rapid (3, 17, 18). Baily et al. (18) have about 1 mM (19) and the concentration of proposed that this is due to intramolecular Fru 1,6-Pz in rat liver is only 20 pM (20) phosphate catalysis. In their model study so that the Mg’+/Fru 1,6-Pz ratio is about on the mutarotation of a-D-glucose they 50. At such a high ratio the anomeric comfound that this reaction can be accelerated position shifts in the direction of the CY by inorganic phosphate; the dibasic species anomer. The /3/a ratio will be about 4.0. being a better catalyst than the monobasic Our measurements on the linewidth of species. A similar behavior is reported here the phosphate resonance and the (Y/P ratio for Fru 1,6-Pz (see Fig. 7). We also found in the presence of Mg’+ indicate that the that the time-averaged preferred orientakinetics of Fru 1,6-Pz are affected to a tion of the l-phosphate group of the ol ansmall extent by this cation. This means omer is sensitive to the degree of ionization that in the in uivo situation, when Fru 1,6of this group. More gauche orientations Pz is completely saturated with Mg2+, the coincide with an increased mutarotation. mutarotaion rate constants could be It cannot be excluded that the orientation slightly different from the values obtained of the l-phosphate is also important in the by 13CNMR (3) and ‘lP NMR. From the phosphate catalysis. Considering this type considerations given above it is concluded of catalysis it can be expected that binding that cellular changes in Mgzf have no efof divalent cations has some influence on fect on the mutarotation rate and anothe mutarotation rates and anomeric commerit composition of Fru 1,6-P2. position. However, our measurements with Mg2+ indicate that these effects are small Relation with Enzyme Activity (see also Ref. (7)). From the titration curves with Mg2+ and FBPase shows rather complex kinetic Zn” (Figs. 4 and 5) we conclude that the properties, including activation by chelatl-phosphate has the highest affinity for ing agents accompanied by a shift of the divalent cations. From model building it pH optimum from more alkaline to neutral TABLE
III
276
VAN
DEN
BERG
AND
pH (19). The inactivation has been attributed to binding of Zn2+ to the enzyme (8). Apart from these inhibitory sites it was necessary to postulate the existence of activating sites (8) to account for the activity in the presence of Zn2’ alone. Binding studies revealed three classes of binding sites(9, 10). It might be speculated that, depending on the ionic conditions, the hydrolyses of Fru l,6-P2 occurs via a dissociative or an associative mechanism, each mechanism characterized by certain kinetic parameters. Metal ions can inhibit a dissociative mechanism or activate an associative mechanism or vice versa (21, 22), and so change the kinetics of rat liver FBPase. As pointed out by Mildvan (22), there are differences in space requirements at the catalytic site between an associative and dissociative mechanism. From Table II it appears that the time-averaged orientation of the l-phosphate of a-Fru 1,6P2 along the C,-0, bond is strongly dependent upon the pH. The sharp pH optimum in FBPase activity in the presence of EDTA (23, 24) might indicate that a particular conformation of the l-phosphate group is necessary for hydrolyses. As Zn2+ inhibits at extremely low concentrations (Ki = 0.3 PM, Ref. (8)), we assume that, after turnover, Zn2’ remains bound to the enzyme near the catalytic site. Evidence for this proposal is given in the accompanying paper (25).
HEERSCHAP
3. MIDELFORT, C. F., GUPTA, J. A. (1976) Biochemistry 4. FREY, W. A., FISHBEIN, R., AND
Haasnoot Koster, Hilbers
Dr.
J. A. L.
Walters
and
Dr.
C. A.
G.
for valuable suggestions. Prof. Dr. J. F. Dr. Th. J. C. van Berkel, and Prof. Dr. C. W. are thanked for reading the manuscript. Miss
A. C. Hanson is thanked for typing the manuscript. The Netherlands Foundation for Fundamental Medical Research (FUNGO) is acknowledged for partial financial support (Grant 13-39-18).
6. BLOXHAM,
Biophys.
Res. Commun.
51, 543-549.
D. P., ANDYORK,
Biochem.
D. A. (1976)
Sot.
Trans. 4, 989-993. 7. BENKOVIC, S. J. (1972) Biochem. Commun. 47, 852-858. 8. TEJWANI, G. O., PEDROSA, S., AND HORECKER, B.
Acad.
Sci. USA
9. PEDROSA, RECKER,
74, 10.
S.,
F. O., PONTREMOLI, L. (1976) P~oc. Nat.
2692-2695.
MELLONI,
B., AND
S.,
AND
Biochem.
E.,
DE
14.
CARRINGTON, in Introduction
SALAMINO,
M.,
AND
248,
BYRNE,
8567-8573.
10,4705-4711.
Biochemistry
BOER, J. W. M. (1977) Lanthanide Shifts Reagents Binding, Shift Mechanisms and Exchange, Ph.D. thesis, University of Nymegen, Krips Repro, Meppel. A., AND MCLACHLAN, A. D. (1967) to Magnetic Resonance, Har-
per & Row, Tokyo. COHEN,
New
S. M., AND
Acad.
Sci. USA
York,
and
BURT,
74,
John
Weatherhill,
P~oc. Nat.
C. T. (1977)
4271-4275.
16.
WOOD, D. J., HRUSKA, F. E., AND OGILVIE, (1974) Canad. J. Chem. 52, 3353-3366.
17.
ISBELL,
20.
F.,
B. L. (1978)
188, 90-97.
Biophys.
G. R. (1971)
GRAY,
13.
18.
Ho-
Acad. Sci. USA
HORECKER,
MARCUS, C. J., GELLER, A. W. L. (1973) J. Biol. Chem.
12.
15.
Res.
2742-2745.
PONTREMOLI,
Arch. 11.
73,
Biophys.
F. O., PONTREMOLI, B. L. (1977) Proc. Nat.
SPARATORE,
H.
S., AND
Carbohydr.
Chem.
BAILEY, J. M., P. G. (1970)
PIGMAN,
W.
Biochem.
FISHMAN,
Biochemistry
24,
K. K.
Aduan.
(1969) 14-63.
P. H., AND PERTCHEV, 9, 1189-1194.
VELOSO, D., GUYNN, R. W., OSKARSSON, M., AND VEECH, R. L. (1973) J. Biol. Chem. 248,48114819. FAUPEL, (1972)
R. P., SEITZ,
Arch.
H. J., AND
Biochem.
TARNOWSKI,
Biophys.
JENCKS, and New
22.
MILDVAN, (Opella, Dekker,
23.
HORECKER, B. L., MELLONI, E., AND PONTREMOLI, S. (1975) in Advances in Enzymology (Meister, A., ed.), Vol. 42, pp. 1933226, Wiley,
24.
New NIMMO,
J. 25.
W. P. (1969). in Catalysis Enzymology, pp. 112-115, York.
W.
148, 509-522.
21.
REFERENCES 1. BENKOVIC, S. J., ENGLE, J. L., AND MILDVAN, A. S. (1972) Biochem. Biophys. Res. Commun. 47, 852-858. 2. KOERNER, T. A. W., CARY, L. W., BHACCA, N. S., AND YOUNATHAN, E. S. (1973) Biochem.
16,
Biochemistry
5. KOERNER, A. W., VOLL, R. J., AND YOUNATHAN, E. S. (1977) FEBS Z&t. 84, 207-213.
19.
thank
DE
S. J. (1977).
BENKOVIC,
K., AND ROSE, 21782185. MAINE, M. M.,
2479-2484.
ACKNOWLEDGMENTS We
R.
15,
VAN H.,
A. S. (1979) S. J., and New York.
in NMR and Lu, P., eds.),
York. H. G., ANDTIPTON,
in Chemistry McGraw-Hill, Biochemistry pp. 3033304,
K. F. (1975)
Biochem.
145,323-334. DEN BERG, G. B., VAANDRAGER-VERDUIN, VAN BERKEL, T. J. C., AND KOSTER,
(1982)
Arch.
Biochem.
Biophys.
J. F.
219, 277-285.