Structure and reactivity of dinuclear cobalt(III) complexes with peroxide and phosphate diester analogues bridging the metal ions [2]

Full text of DOI:10.1021/ja9811905

J. Am. Chem. Soc. 1998, 120, 9943-9944
Structure and Reactivity of Dinuclear Cobalt(III)
9943
Complexes with Peroxide and Phosphate Diester
Analogues Bridging the Metal Ions
Jin Seog Seo, Rosemary C. Hynes, Daniel Williams, and
Jik Chin*
Department of Chemistry, McGill UniVersity
Montreal, Canada H3A 2K6
Nack-Do Sung*
Department of Agricultural Chemistry
Chung-Nam National UniVersity, Taejeon 305-764 Korea
ReceiVed April 8, 1998
There is much current interest both in chemistry1 and in
biology2 in understanding how two metal ions cooperatively
hydrolyze phosphate esters. Metal ion-catalyzed hydrolysis of
phosphate diesters with poor leaving groups is of particular interest
since some of the most important molecules of life (DNA, RNA,
phospholipids) contain such linkages. Hydrolyzing phosphate
diesters with poor leaving groups is an elusive goal as they are
enormously more stable compared to phosphate diesters with good
leaving groups.3,4 It has been suggested that metal-bound
peroxides can be effective nucleophiles for cleaving amides5 and
phosphates.6 Here we examine the conversion of 1a to 2. The
Figure 1. X-ray structure of 1b (ORTEP representation; ellipsoids at
50% probability level). Selected distances (Å) and angles (deg): Co1-
Co2 3.2319(25), Co1-N1 2.024(7), Co1-N2 1.924(7), Co1-N3 1.918-
(7), Co1-O1 1.863(6), Co1-O3 1.946(5), Co1-O5 1.915(6), Co2-N4
1.979(4), Co2-N5 2.017(7), Co2-N6 1.934(7), Co2-O2 1.862(6), Co2-
O4 1.928(5), Co2-O5 1.947(6), O1-O2 1.410(8); O3-P-O4 116.9-
(3), Co1-O3-P 119.7(3), Co2-O4-P 132.2(4), O1-Co1-O3 90.28-
(25), N1-Co1-N2 85.7(3), O2-Co2-O4 93.1(3), N4-Co2-N5 83.8(3).
Complex 1b was prepared by the same method used for making
2.7 The 31P NMR spectrum of the dinuclear Co(III) complex
shows a sharp singlet (1b: DMSO (trimethyl phosphate) δ 51
ppm) indicating that there is only one diastereomer in the sample.
The structure of 1b (Figure 1) closely resembles that of 2 except
that the bridging peroxide in 1b is replaced with solvent molecules
in 2.8 It appears that when phosphate monoester analogues are
used instead of phosphate diester analogues in the synthesis of
the dinuclear Co(III) complex, the bridging peroxide is more
readily exchanged with solvent water molecules. Replacing
diphenyl phosphinate in the above synthesis with methyl phenyl-
phosphonate gave 1a.9 Unfortunately, we were unable to prepare
the corresponding dinuclear Co(III) complex with a bridging
phosphate diester by the same synthetic procedure. However,
phosphonate monoesters are excellent models of phosphate
diesters in that they have comparable reactivities as discussed
below. Unlike in 1b, the phosphorus center in 1a is chiral and
two 31P NMR signals are observed for the two diastereomers
(1a: DMSO (trimethyl phosphate) δ 43 and 40 ppm in a ratio of
∼3:1).
structure of 1b was determined in this study as a model for the
structure of 1a while the structure of 2 was determined previ-
ously.7
(1) (a) Go¨bel, M. W. Angew. Chem., Int. Ed. Engl. 1994, 33, 1141. (b)
Chapman, W. H.; Breslow, R. J. Am. Chem. Soc. 1995, 117, 5462. (c) Wall,
M.; Hynes, R. C.; Chin, J. Angew. Chem., Int. Ed. Engl. 1993, 32, 1633. (d)
Young, M. J.; Chin, J. J. Am. Chem. Soc. 1995, 117, 10577. (e) Molenveld,
P.; Kapsabelis, S.; Engbersen, J. F. J.; Reinhoudt, D. N. J. Am. Chem. Soc.
1997, 119, 2948. (f) Ragunathan, K. G.; Schneider, H.-J. Angew. Chem., Int.
Ed. Engl. 1996, 35, 1219. (g) Yashiro, M.; Ishikubo, A.; Komiyama, M. J.
Chem. Soc., Chem. Commun. 1995, 1793. (h) Liu, S.; Luo, Z.; Hamilton, A.
D. Angew. Chem., Int. Ed. Engl. 1997, 36, 2678.
Hydrolysis of the phosphonate ester bond in 1a was monitored
1
by H and 31P NMR. In a typical kinetic experiment, 1a was
dissolved in DMSO-d6 (2 mM) and the hydrolysis reaction was
started by adding 0.5 equiv volume of a buffered solution of D2O
(10 mM N-ethylmorpholine, pD 7, 25 °C). Production of
methanol from the hydrolysis of 1a was monitored by 1H NMR.
The 1H NMR of the product of hydrolysis of 1a matches that of
a genuine sample of 2 synthesized from phenylphosphonate.
Conversion of 1a to 2 was also monitored by 31P NMR (Figure
S1).
(2) (a) Stra¨ter, N.; Lipscomb, W. N.; Klabunde, T.; Krebs, B. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 2024. (b) Wilcox, D. E. Chem. ReV. 1996, 96, 2435.
(3) Successful hydrolysis of unactivated phosphonate or phosphate esters
with lanthanide ions have been reported although the hydrolytic mechanism
is not clearly understood. (a) Takasaki, B. K.; Chin, J. J. Am. Chem. Soc.
1994, 116, 1121. (b) Sumaoka, J.; Miyama, S.; Komiyama, M. J. Chem. Soc.,
Chem. Commun. 1994, 1755. (c) Tsubouchi, A.; Bruice, T. C. J. Am. Chem.
Soc. 1994, 116, 11614.
(4) Chin, J.; Banaszczyk, B.; Jubian, V.; Zou, X. J. Am. Chem. Soc. 1989,
111, 186.
(5) (a) Rana, T. M.; Meares, C. F. J. Am. Chem. Soc. 1991, 113, 1859. (b)
Narasimha, N. M.; Mahroof-Tahir, M.; Karlin, K. D. J. Am. Chem. Soc. 1993,
115, 10404.
(6) (a) Takasaki, B. K.; Chin, J. J. Am. Chem. Soc. 1995, 117, 8582. (b)
Kimball, A. S.; Lee, J.; Jayaram, M.; Tullius, T. D. Biochemistry 1993, 32,
4698. (c) Breslow, R.; Zhang, B. J. Am. Chem. Soc. 1994, 116, 7893. (d)
Schnaith, L. M. T.; Hanson, R. S.; Que, L., Jr. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 569.
(7) Seo, J. S.; Sung, N.-D., Hynes, R. C.; Chin, J. Inorg. Chem. 1996, 35,
7472.
(8) Crystal structural data for 1b(ClO4)2: Co2C45H43N6O13PCl2, monoclinic,
space group P21/n, a ) 11.6526(11) Å, b ) 11.1944(9) Å, c ) 36.026(3) Å,
b ) 94.449(7)°, V ) 4685.2(7) Å3, Z ) 4; 5886 measured reflections, 3411
with I > 2.5s(I), 623 parameters, R ) 0.062 and Rw ) 0.062, GOF ) 1.66.
Anal. Calcd for C45H43N6O13PCl2Co2 (1b‚(ClO4)2): C, 49.33; H, 3.96; N, 7.67;
Cl, 6.47; Co, 11.76. Found: C, 49.49; H, 4.20; N, 7.93; Cl, 6.35; Co, 10.71.
(9) Anal. Calcd for C40H41N6O14PCl2Co2 (1a‚(ClO4)2): C, 45.78; H, 3.94;
N, 8.01; Cl, 6.76; Co, 11.23. Found: C, 45.30; H, 3.98; N, 7.90; Cl, 7.01;
Co, 10.94.
S0002-7863(98)01190-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/10/1998

9944 J. Am. Chem. Soc., Vol. 120, No. 38, 1998
Communications to the Editor
The pseudo-first-order rate constant for the hydrolysis of the
ester bond in 1a is 5 × 10-7 s-1 under the above experimental
conditions. This represents a substantial rate acceleration for the
hydrolysis of the metal-coordinated phosphonate ester over that
of the metal-free substrate. In general, phosphonate monoesters
are about as stable as phosphate diesters. The second-order rate
contant for hydroxide-catalyzed hydrolysis of bis(p-nitrophenyl)
phosphate,3 p-nitrophenyl phenylphosphonate,10 and p-nitrophenyl
methyl phosphonate11 at 25 °C are 2.0 × 10-5, 8.3 × 10-5, and
2.8 × 10-5 M-1 s-1, respectively.12 Assuming that the second-
order rate constant for hydroxide-catalyzed hydrolysis of methyl
phenyl phosphonate is comparable to that for the hydrolysis of
dimethyl phosphate (6.8 × 10-12 M-1 s-1),4 the pseudo-first-order
rate constant for hydroxide-catalyzed hydrolysis of the monoester
at pH 7 should be ∼10-18 s-1. This gives a crude estimate for
the rate acceleration of ∼1011-1012-fold.
the diester.17 In another recent study,18 it was shown that the
phosphate diester bond in 3a with a good leaving group is
hydrolyzed ∼1011 times more rapidly than the metal-free diester
while the phosphate diester in 3b with poor leaving groups
It is interesting to compare the reactivity of the dinuclear
complex to that of previously reported mononuclear metal
complexes. Mononuclear Co(III) complexes provide ∼107-fold
rate acceleration for hydrolyzing phosphate diesters4,13 when the
N-Co-N bond angle opposite the O-Co-O bond angle is
locked into a five-membered ring as it is in our dinuclear system
(1b). Hence, there appears to be considerable gain in reactivity
on going from the mononuclear to the dinuclear complex.
There are two reasonable mechanisms for the cleavage of the
phosphonate ester bond in 1a. The bridging peroxide may act
as an intramolecular nucleophilic catalyst14 or it may first be
replaced by solvent molecules followed by intramolecular nu-
cleophilic attack of the bridging phosphonate by the metal
hydroxide. Metal-bound peroxides have been implicated in the
cleavage of amides5 and phosphates.6 To distinguish the mecha-
nistic possibilities, 18O labeling experiments were performed. The
incorporation of 18O into the product phosphonate 2 can be
detected by 31P NMR spectroscopy since it results in an upfield
shift of the phosphonate signal by ∼0.02 ppm.15 When 1a was
allowed to react with 50% 18O-labeled D2O, the product phos-
phonate peak appeared as a doublet, clearly demonstrating the
incorporation of O-18 from the solvent to the product (Figure
S2). In contrast, if 1a is first synthesized with 50% 18O-labeled
O2 and then hydrolyzed with unlabeled D2O, there is no
incorporation of O-18 into the product.16 We therefore conclude
that the cleavage of the phosphonate ester in 1a takes place by
intramolecular nucleophilic attack of metal hydroxide on the
bridging phosphonate ester. The bridging peroxide in 1a may
still be involved in cleaving phosphonate esters with good leaving
groups. Consistent with this interpretation, we were unable to
synthesize the dinuclear Co(III) complex with p-nitrophenyl
phenylphosphonate bridging the two metal centers because
p-nitrophenol is released during the synthesis.
dissociates from the dinuclear Co(III) complex without any
observable hydrolysis of the diester bond.19 Hence, a catalyst
that provides large rate acceleration for hydrolyzing activated
phosphates do not necessarily provide comparable rate accelera-
tion for hydrolyzing unactivated phosphates. In sharp contrast
to the result for 3b, 1a is cleanly hydrolyzed to 2. This represents
the first hydrolysis of an unactivated phosphate diester analogue
coordinated to Co(III).
It is interesting that while the dinuclear Co(III) centers in 1-3
all appear to be active for cleaving phosphates with good leaving
groups, the dinuclear center in 2 is the only one that provides
large rate accelerations for hydrolyzing phosphates with poor
leaving groups. Unlike the bridging peroxide 1 or the bridging
oxide 3, the metal hydroxide 2 can begin to deprotonate during
expulsion of the leaving group which may be important for
hydrolyzing phosphates with poor leaving groups.20 Consistent
with this interpretation, metal alkoxides are more reactive than
metal hydroxides for cleaving phosphates with good leaving
groups but the order is reversed for hydrolyzing phosphates with
poorer leaving groups.21 Metal alkoxide, metal-bridging perox-
ides, and metal-bridging oxides should all be able cleave
phosphate diesters with poor leaving groups if the leaving group
oxygen is activated by protonation or by coodination to a metal.
In nature, there are dinuclear metallophosphoesterases whose
active sites resemble that of 2 (fructose-1,6-biphosphatase)22 and
3 (purple acid phosphatase).23 Undertanding the mechanism of
action of 2 and 3 in detail may give valuable insights into how a
variety of dinuclear metallophosphoesterases function.
Acknowledgment. We thank NSERC, Pioneer Hi-Bred International
Inc., and the U.S. Army research office for support of this work. We
also thank Dr. Kimin Park (at Pohang Institute of Technology, Korea)
for his kind help with Figure 1.
We recently reported that phenyl phosphate coordinated to the
dinuclear metal complex in 2 is hydrolyzed ∼1011 times more
rapidly than uncoordinated phenyl phosphate.7 While metal-
promoted hydrolysis of activated phosphates with good leaving
groups is quite common, hydrolyzing unactivated phosphate
diesters with poor leaving groups remains a major challenge.
Attempt at hydrolysis of dimethyl phosphate coordinated to a
mononuclear Co(III) complex resulted in simple dissociation of
Supporting Information Available: X-ray structural information for
1b including crystal packing diagram, positional parameter, interatomic
distances and angles and Figures S1 and S2 (13 pages, print/PDF). See
any current masthead page for ordering information and Web access
instructions.
JA9811905
(10) Seo, J. S. Hydrolyzing Phosphates with Metal Ions: Di and Trinuclear
Approaches. Ph.D. Dissertation, McGill University, March 1997.
(11) Rahil, J.; Pratt, R. F. J. Chem. Soc., Perkin Trans. 2 1991, 947.
(12) The reactivity trend is not expected to change much for esters with
poorer leaving groups. There is excellent correlation between the reactivity
of phosphate diesters and the basicity of the leaving groups (see ref 4). Similar
correlation can be obtained for phosphonate esters from the reactivity of phenyl
methyl phosphonate (5 × 10-7 M-1 s-1). See ref 10.
(17) Schmidt, W.; Taube, H. Inorg. Chem. 1963, 2, 698.
(18) Wahnon, D.; Lebuis, A.-M.; Chin, J. Angew. Chem., Int. Ed. Engl.
1995, 34, 2412.
(19) The coordinated phosphate diester in 3b is estimated to hydrolyze
∼105 times more rapidly than the metal-free phosphate. Williams, N. H.;
Cheung, W.; Chin, J. J. Am. Chem. Soc. 1998, 120, 8079.
(20) The metal-hydroxide in [Co(cyclen)(OH2)(OH)]2+ should also depro-
tonate during expulsion of poor leaving groups13 in a step that is not rate
determining.
(13) Kim, J. H.; Chin, J. J. Am. Chem. Soc. 1992, 114, 9792.
(14) The peroxyphosphate intermediate could then be reduced by hydrogen
peroxide that may be present in solution. See ref 6a.
(21) Young, M. J.; Wahnon, D.; Hynes, R. C.; Chin, J. J. Am. Chem. Soc.
1995, 117, 9441.
(15) (a) Cohn, M.; Hu, A. Proc. Natl. Acad. Sci. U.S.A. 1978, 75, 200. (b)
Lowe, G.; Sproat, B. S. J. Chem. Soc., Chem. Commun. 1978, 565.
(16) If we start with 2, there is no incorporation of O-18 from solvent to
the phosphonate.
(22) Zhang, Y.; Liang, J.-Y.; Huang, S.; Ke, H.; Lipscomb, W. N.
Biochemistry 1993, 32, 1844.
(23) Stra¨ter, N.; Klabunde, T.; Tucker, P.; Witzel, H.; Krebs, B. Science
1995, 268, 1489.