Cooperativity between Metal Ions in the Cleavage of Phosphate Diesters and RNA by Dinuclear Zn(II) Catalysts

Article abstract of DOI:10.1021/ic030131b

A series of ligands containing linked 1,4,7-triazacyclononane macrocycles are studied for the preparation of dinuclear Zn(II) complexes including 1,3-bis(1,4,7-triazacyclonon-1-yl)-2-hydroxypropane (L2OH), 1,5-bis(1,4,7-triazacyclonon-1-yl)pentane (L3), 2,9-bis(1-methyl-1,4,7-triazacyclonon-1-yl)-1,10-phenanthroline (L4), and α,α′-bis(1,4,7-triazacyclonon-1-yl)-m-xylene (L5). The titration of these ligands with Zn(NO₃)₂ was monitored by ¹H NMR. Each ligand was found to bind two Zn(II) ions with a very high affinity at near neutral pH under conditions of millimolar ligand and 2 equiv of Zn(NO₃)₂. In contrast, a stable mononuclear complex was formed in solutions containing 5.0 mM L20H and 1 equiv of Zn(NO ₃)₂. ¹H and ¹³C NMR spectral data are consistent with formation of a highly symmetric mononuclear complex Zn(L2OH) in which a Zn(II) ion is sandwiched between two triazacyclononane units. The second-order rate constant k_Zn for the cleavage of 2-hydroxypropyl-4-nitrophenyl phosphate (HPNP) at pH 7.6 and 25 °C catalyzed by Zn2(L2O) is 120-fold larger than that for the reaction catalyzed by the closely related mononuclear complex Zn(L1) (L1 = 1,4,7-triazacyclononane). By comparison, the observation that the values of k_Zn determined under similar reaction conditions for cleavage of HPNP catalyzed by the other Zn(II) dinuclear complexes are only 3-5-fold larger than values of k_Zn for catalysis by Zn(L1) provides strong evidence that the two Zn(II) cations in Zn₂(L2O) act cooperatively in the stabilization of the transition state for cleavage of HPNP. The extent of cleavage of an oligoribonucleotide by Zn(L1), Zn2(L5), and Zn2(L2O) at pH 7.5 and 37 °C after 24 h incubation is 4,10, and 90%. The rationale for the observed differences in catalytic activity of these dinuclear Zn(II) complexes is discussed in terms of the mechanism of RNA cleavage and the structure and speciation of these complexes in solution.

Full text of DOI:10.1021/ic030131b

Inorg. Chem. 2003, 42, 7737−7746
Cooperativity between Metal Ions in the Cleavage of Phosphate Diesters
and RNA by Dinuclear Zn(II) Catalysts
Olga Iranzo, Terry Elmer, John P. Richard, and Janet R. Morrow*
Department of Chemistry, UniVersity at Buffalo, The State UniVersity of New York,
Amherst, New York 14260
Received April 14, 2003
A series of ligands containing linked 1,4,7-triazacyclononane macrocycles are studied for the preparation of dinuclear
Zn(II) complexes including 1,3-bis(1,4,7-triazacyclonon-1-yl)-2-hydroxypropane (L2OH), 1,5-bis(1,4,7-triazacyclonon-
1
-yl)pentane (L3), 2,9-bis(1-methyl-1,4,7-triazacyclonon-1-yl)-1,10-phenanthroline (L4), and R,R′-bis(1,4,7-triazacy-
1
clonon-1-yl)-m-xylene (L5). The titration of these ligands with Zn(NO
was found to bind two Zn(II) ions with a very high affinity at near neutral pH under conditions of millimolar ligand
and 2 equiv of Zn(NO . In contrast, a stable mononuclear complex was formed in solutions containing 5.0 mM
L2OH and 1 equiv of Zn(NO
3 2
) was monitored by H NMR. Each ligand
3 2
)
1
13
3 2
) . H and C NMR spectral data are consistent with formation of a highly symmetric
mononuclear complex Zn(L2OH) in which a Zn(II) ion is sandwiched between two triazacyclononane units. The
second-order rate constant kZn for the cleavage of 2-hydroxypropyl-4-nitrophenyl phosphate (HPNP) at pH 7.6 and
2
25 °C catalyzed by Zn (L2O) is 120-fold larger than that for the reaction catalyzed by the closely related mononuclear
complex Zn(L1) (L1 ) 1,4,7-triazacyclononane). By comparison, the observation that the values of kZn determined
under similar reaction conditions for cleavage of HPNP catalyzed by the other Zn(II) dinuclear complexes are only
3
−5-fold larger than values of kZn for catalysis by Zn(L1) provides strong evidence that the two Zn(II) cations in
Zn (L2O) act cooperatively in the stabilization of the transition state for cleavage of HPNP. The extent of cleavage
of an oligoribonucleotide by Zn(L1), Zn (L5), and Zn (L2O) at pH 7.5 and 37 °C after 24 h incubation is 4,10, and
0%. The rationale for the observed differences in catalytic activity of these dinuclear Zn(II) complexes is discussed
in terms of the mechanism of RNA cleavage and the structure and speciation of these complexes in solution.
2
2
2
9
Introduction
Many metalloenzymes that catalyze the hydrolysis of
two metal ions might interact cooperatively at a catalyst
binding site to produce a catalytic rate acceleration far in
excess of that obtained from recruitment of only a single
metal ion.
1-3
phosphate diesters contain two metal ions in the active site,
and certain ribozymes that catalyze hydrolysis or transes-
terification of phosphate diesters also require two metal ions
for full catalytic activity.4,5 The catalytic rate acceleration
of protein or RNA catalysts that utilize one metal ion will
be at least doubled by recruitment of a second metal ion
which acts independently of the first, for example, at a second
identical subunit. It is much more difficult to understand how
An important challenge to inorganic and bioinorganic
chemists is the design of dinuclear metal ion catalysts, where
the metal ions act cooperatively to stabilize the transition
state for the catalyzed reaction in a manner that models
catalysis by proteins and RNA. Extremely impressive and
efficient “catalysis” has been observed by exchange inert
Co(III) complexes that contain a bridging phosphate diester
that shows extraordinary reactivity toward hydrolysis.6
These Co(III) complexes provide important insight into the
potential for transition state stabilization by the cooperatively
-8
*
Author to whom correspondence should be addressed. E-mail: jmor-
row@acsu.buffalo.edu.
(
(
1) Lipscomb, W. N.; Str a¨ ter, N. Chem. ReV. 1996, 96, 2375-2433.
2) Str a¨ ter, N.; Lipscomb, W. N.; Klabunde, T.; Krebs, B. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 2024-2055.
(
(
3) Wilcox, D. E. Chem. ReV. 1996, 96, 2435-2458.
4) Pontius, B. W.; Lott, W. B.; Von Hippel, P. H. Proc. Natl. Acad. Sci.
U. S. A. 1997, 94, 2290-2294.
(6) Hendry, P.; Sargeson, A. M. Inorg. Chem. 1990, 29, 92-97.
(7) Williams, N. H.; Cheung, W.; Chin, J. J. Am. Chem. Soc. 1998, 120,
8079-8087.
(8) Williams, N. H.; Takasaki, B.; Wall, M.; Chin, J. Acc. Chem. Res.
1999, 32, 485-493.
(5) Takagi, Y.; Warashina, M.; Stec, W. J.; Yoshinari, K.; Taira, K. Nucleic
Acids Res. 2001, 29, 1815-1834.
1
0.1021/ic030131b CCC: $25.00 © 2003 American Chemical Society
Inorganic Chemistry, Vol. 42, No. 24, 2003 7737
Published on Web 09/26/2003

Iranzo et al.
Scheme 1
sometimes observed for dinuclear compared to mononuclear
complexes are consistent with catalysis at noninteracting
catalyst “subunits”, while the larger, up to 500, rate enhance-
ments observed with other catalysts suggest that in some
cases it is possible to achieve cooperative stabilization of
the transition state for RNA cleavage by two metal ions.
The studies of these metal ion complexes often do not
proceed far beyond the determination of kinetic parameters
for catalysis of phosphodiester cleavage, so that the underly-
ing explanation for the differences in catalytic activity for
different metal ion complexes is not well understood. This
is partly because dinuclear metal ion complexes reported to
date often show low solubility in water. This limits their
applications in the study of biological molecules and requires
that their mechanism be characterized in mixed aqueous/
organic solvents, whose properties are less well-defined than
in water. In addition, the problem of determining the role of
the metal ion in catalysis takes on added layers of complexity
for dinuclear complexes because of the many roles possible
for each metal ion in providing stabilization of the reaction
interacting metal ions. However, they do not undergo
multiple turnover.
Efficient catalysis can only be observed by metal ion
complexes that undergo rapid exchange between free and
bound substrate. Many mononuclear Zn complexes have been
9
-11
shown to be effective catalysts for RNA cleavage,
and a
simple strategy for the design of dinuclear catalysts is to
connect two mononuclear zinc ligands by a short tether.
However, it is not clear that cooperative interaction between
metal ions should be obtained for such dinuclear ligands,
given the inherent electrostatic repulsion between metal ions
in these complexes. Cooperative interaction between metal
ions can be promoted by incorporation of functionality within
the tether, which will allow it to serve as a scaffold for
chelation of the two metal ions. This presents obvious
challenges in the design and synthesis of the appropriate
tether.
3
0-32
transition state.
We are interested in designing efficient catalysts for the
hydrolysis of RNA (Scheme 1). Varying degrees of success
We recently reported kinetic parameters and structural data
for Zn (L2O), an extremely effective dinuclear Zn(II) catalyst
for the cleavage of the model RNA substrate HPNP (HPNP
12-21
have been achieved with dinuclear catalysts using Zn(II),
2
Cu(II),2
2-28
and La(III). The 2-fold rate enhancements
29
3
3
)
2-hydroxypropyl-4-nitrophenyl phosphate). This Zn(II)
(
9) Breslow, R.; Berger, D.; Huang, D.-L. J. Am. Chem. Soc. 1990, 112,
complex is highly water soluble, binds two Zn(II) ions very
strongly, and shows relatively clear and simple speciation
in water. In addition, this complex is one of the best dinuclear
complex catalysts for HPNP cleavage with the highest
reported second-order rate constant in 100% aqueous solution
3686-3687.
(
(
10) Shelton, V. M.; Morrow, J. R. Inorg. Chem. 1991, 30, 4295-4299.
11) A ¨ı t-Haddou, H.; Sumaoka, J.; Wiskur, S. L.; Folmer-Andersen, J. F.;
Anslyn, E. V. Angew. Chem., Int. Ed. 2002, 41, 4014-4016.
(
(
(
12) Chapman, W. H., Jr.; Breslow, R. J. Am. Chem. Soc. 1995, 117, 5462-
5469.
13) Yashiro, M.; Ishikubo, A.; Komiyama, M. J. Chem. Soc., Chem.
Commun. 1995, 1793-1794.
33
for a Zn(II) catalyst. In this paper we report the synthesis,
1
13
H and C NMR spectral data, and speciation for several
14) Molenveld, P.; Stikvoort, W. M. G.; Kooijman, H.; Spek, A. L.;
Engbersen, J. F. J.; Reinhoudt, D. N. J. Org. Chem. 1999, 64, 3896-
additional Zn(II) complexes, including those of 1,5-bis(1,4,7-
triazacyclonon-1-yl)pentane (L3), 2,9-bis(1-methyl-1,4,7-
triazacyclonon-1-yl)-1,10-phenanthroline (L4), and R,R′-
bis(1,4,7-triazacyclonon-1-yl)-m-xylene (L5), that are struc-
3906.
(
(
(
15) Rossi, P.; Felluga, F.; Tecilla, P.; Formaggio, F.; Crisma, M.; Toniolo,
C.; Scrimin, P. Biopolymers 2000, 55, 496-501.
16) Yamada, K.; Takahashi, Y.; Yamamura, H.; Araki, S.; Saito, K.;
Kawai, M. Chem. Commun. 2000, 1315-1316.
2
turally related to Zn (L2O). A comparison of the second-
17) Gajda, T.; Kr a¨ mer, R.; Jancs o´ , A. Eur. J. Inorg. Chem. 2000, 1635-
1644.
order rate constants kZn for catalysis of HPNP cleavage by
this series of structurally related mono- and dinuclear
complexes provides strong evidence that the metal ions in
(
(
(
18) He, C.; Lippard, S. J. J. Am. Chem. Soc. 2000, 122, 184-185.
19) Kawahara, S.; Uchimaru, T. Eur. J. Inorg. Chem. 2001, 2437-2442.
20) Albedyhl, S.; Schnieders, D.; Jancs o´ , A.; Gadja, T.; Krebs, B. Eur. J.
Inorg. Chem. 2002, 1400-1409.
2 2 2
Zn (L3), Zn (L4), and Zn (L5) act nearly independently of
(21) Worm, K.; Chu, F.; Matsumoto, K.; Best, M. D.; Lynch, V.; Anslyn,
E. V. Chem.-Eur. J. 2003, 9, 741-747.
(22) Wall, M.; Hynes, R. C.; Chin, J. Angew. Chem., Int. Ed. Engl. 1993,
(28) Gadja, T.; Jancs o´ , A.; Mikkola, S.; L o¨ nnberg, H.; Sirges, H. J. Chem.
Soc., Dalton Trans. 2002, 1757-1763.
(29) Yashiro, M.; Ishikubo, A.; Komiyama, M. J. Biochemistry 1996, 120,
1067-1069.
(30) Chin, J. Curr. Opin. Chem. Biol. 1997, 1, 514-521.
(31) Mikkola, S.; Stenman, E.; Nurmi, K.; Yousefi-Salakdeh, E.; Str o¨ mberg,
R.; L o¨ nnberg, H. J. Chem. Soc., Perkin Trans. 2 1999, 1619-1625.
(32) Molenveld, P.; Engbersen, J. F. J.; Reinhoudt, D. N. Chem. Soc. ReV.
2000, 29, 75-86.
(33) Iranzo, O.; Kovalevsky, A. Y.; Morrow, J. R.; Richard, J. P. J. Am.
Chem. Soc. 2003, 125, 1988-1993.
3
2, 1633-1635.
(
(
23) Young, M. J.; Chin, J. J. Am. Chem. Soc. 1995, 117, 10577-10578.
24) Liu, S.; Hamilton, A. D. Bioorg. Med. Chem. Lett. 1997, 7, 1779-
1784.
(
25) Molenveld, P.; Engbersen, J. F. J.; Kooijman, H.; Spek, A. L.;
Reinhoudt, D. N. J. Am. Chem. Soc. 1998, 120, 6726-6737.
(26) Liu, S.; Hamilton, A. D. Chem. Commun. 1999, 587-588.
27) (a) Fritsky, I. O.; Ott, R.; Pritzkow, H.; Kr a¨ mer, R. Chem.-Eur. J.
(
2
001, 7, 1221-1231. (b) Fritsky, I. O.; Ott, R.; Pritzkow, H.; Kr a¨ mer,
R. Inorg. Chim. Acta 2003, 346, 111-118.
7738 Inorganic Chemistry, Vol. 42, No. 24, 2003

Metal Ion CooperatiWity
one another in catalysis of the cleavage of HPNP but that
addition of an alkoxide to the tether is sufficient to provide
a scaffold for the cooperative transition state stabilization
NaOH. The free amine base was extracted with chloroform (6 ×
30 mL), and the combined organic fractions were dried over Na
SO . The solvent was removed on a rotary evaporator, and the oily
residue was dried at room temperature under high vacuum to give
.25 g (9.67 mmol) of 1,4,7-triazacyclononane in 89% yield. 1,4,7-
1
2
-
4
2
by the Zn(II) cations in Zn (L2O).
1
4,10
Triazatricyclo-[5.2.1.0 ]decane was prepared from L1 by a
Experimental Section
39,40
published method.
1,5-Bis(1,4,7-triazacyclonon-1-yl)pentane
(L3)41 and R,R′-bis(1,4,7-triazacyclonon-1-yl)-m-xylene (L5)42 were
prepared using the orthoamide derivative of L1 (1,4,7- triazatricyclo-
[5.2.1.04 ]decane) according to literature procedures. The protect-
ing group was removed by refluxing for 4 h in water followed by
10-12 h reflux in a basic solution of 4 M NaOH. 2,9-Bis-
(bromomethyl)-1,10-phenanthroline was synthesized from neocu-
Materials. All reagents and solvents were of analytical reagent
grade and were used without further purification, unless otherwise
noted. Acetonitrile, dimethylformamide, and methanol were dried
over 4-Å molecular sieves. All aqueous solutions were prepared
using Millipore MILLI-Q purified water. All reactions were carried
,10
2
out under a N (g) atmosphere. Thin-layer chromatography (TLC)
4
3
plates were silica gel 60F254, 0.2-mm thickness (Aldrich). Merck
grade 9385, 230-400 mesh silica gel, 60 Å (Aldrich), was used in
all column flash chromatography. 1,3-Dibromo-2-propanol, 1,5-
dibromopentane, and R,R′-dibromo-m-xylene were purchased from
Aldrich. N-(2-Hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid)
proine hydrate (Aldrich) by following a literature procedure.
,4,7-Tris(p-toluenesulfonyl)-1,4,7-triazacyclononane (Ts
A modification of a reported procedure was used to form this
1
3
L1).
44
macrocycle. N,N′,N′′-Tris(p-toluenesulfonyl)-diethylenetriamine
28.3 g, 50.0 mmol) was dissolved in 270 mL of dimethylforma-
(
(HEPES) and N-(2-hydroxyethyl)piperazine-N′-(3-propanesulfonic
mide. Cesium carbonate (34.34 g, 105.4 mmol) was added to this
solution, and the resulting suspension was stirred at room temper-
ature for 1 h. 1,2-Bis(p-toluenesulfonyl)-1,2-ethanediol (18.65 g,
acid) (EPPS) were purchased from Sigma and/or Aldrich. The
barium salt of 2-hydroxypropyl-4-nitrophenyl phosphate (HPNP)
was prepared according to the literature procedure.34 The oligori-
bonucleotide A was synthesized on an Applied Biosystems DNA
6
synthesizer using phosphoramidites from Glen research and was
deprotected following established protocols.
50.3 mmol) in 125 mL of dimethylformamide was added dropwise
to this suspension. The reaction mixture was stirred under N
2
overnight at room temperature, and the resulting turbid-yellow
solution was added slowly to 1.5 L of water with rapid stirring to
give the product as a light-yellow precipitate. This solid was
collected by filtration and washed in 2:1 (v:v) dimethylformamide-
water by stirring for several hours. The resulting white solid was
collected by filtration, thoroughly washed with water, and dried
under vacuum. The product was recrystallized from 80:20 (v:v)
Aqueous stock solutions (40.0 mM) of the ligands were prepared
from their respective salts, and concentrations were determined by
1
H NMR using p-toluenesulfonic acid as internal standard. Solutions
3 2
of Zn(NO ) (Aldrich) were standardized with EDTA using
35
Eriochrome Black T. Solutions of oligoribonucleotides were
prepared using autoclaved triply distilled water, and all standard
precautions were taken to avoid ribonuclease contamination.
An Orion Research Digital ion analyzer/501 and an Orion
Research Ross Combination pH Electrode 8115BN were used for
all pH measurements. An UVIKON-XL spectrophotometer by Bio-
Tek instruments equipped with a thermostatic multicell transfer
1
dimethylformamide-water. Yield: 85-90%. H NMR (400 MHz,
3
3
CDCl
3
): δ 7.73 (d, J ) 8.4 Hz, 6H, Ar), 7.36 (d, J ) 8.4 Hz, 6H,
Ar), 3.46 (s, 12H, ring-CH ), 2.47 (s, 9H, CH ).
2
3
,9-Bis(1-methyl-1-azonia-4,7-diazatricyclo[5.2.14,10]decane)-
,10-phenanthroline Dibromide. 1,4,7-Triazatricyclo-[5.2.1.0 ]-
2
4,10
1
decane (0.571 g, 4.10 mmol) was dissolved in 9 mL of dry
acetonitrile, and 2,9-bis(bromomethyl)-1,10-phenanthroline (0.684
g, 1.89 mmol) in 60 mL of dry acetonitrile was added dropwise,
with stirring, over 2 h to give a beige precipitate. The slurry was
heated to reflux for 2 h and stirred overnight at room temperature.
The solid was collected by filtration, washed with cold acetonitrile
1
compartment was used for all kinetic measurements. H NMR
spectra were recorded on a Varian Inova 500, Varian Inova 400,
or Varian Gemini 300 spectrometer. 13C NMR spectra were
recorded on a Varian Inova 500 or Varian Gemini 300 spectrometer
and 31P NMR spectra on a Varian Inova 400 spectrometer. Chemical
shifts are reported as parts per million (ppm) downfield from
tetramethylsilane (TMS). A VG 70-SE mass spectrometer with fast
atom bombardment (FAB) and chemical ionization was utilized
for FAB m/z low-resolution analysis. A ThermoFinnigan Mat 95
XL spectrometer with a Cs ion gun at 20 kV was utilized for FAB
m/z high-resolution analysis. For analysis by FAB-MS, m-nitro
benzyl alcohol (NBA) was used as a matrix.
and dry ether, and dried at room temperature under vacuum to afford
1
0
.796 g of beige powder. Yield: 66%. H NMR (500 MHz, D
2
O,
see Supporting Information (SI) Figure S1 for labeling): δ 8.51
3
(
d, J ) 8.0 Hz, 2H, H(4) and H(7)), 7.94 (s, 2H, H(5) and H(6)),
3
7
.94 (d, J ) 8.0 Hz, 2H, H(8) and H(3)), 6.36 (s, 2H, methine H),
.96 (s, 4H, CH (15)), 4.11-4.09 (m, 4H, ring-CH ), 3.48-3.37
), 3.17-3.15 (m, 4H, ring-CH ), 3.07-3.05 (m,
4
2
2
(
m, 12H, ring-CH
H, ring-CH ).
HCl Salt of 2,9-Bis(1-methyl-1,4,7-triazacyclonon-1-yl)-1,10-
2
2
3
6
Syntheses. N,N′,N′′-Tris(p-toluenesulfonyl)-diethylenetriamine
4
2
and 1,2- bis(p-toluenesulfonyl)-1,2-ethanediol37 were prepared
according to literature procedures. 1,4,7-Tris(p-toluenesulfonyl)-
phenanthroline (L4‚6HCl). 2,9-Bis(1-methyl-1-azonia-4,7-
1
,4,7-triazacyclononane (Ts
3
L1) was deprotected according to
4,10
diazatricyclo[5.2.1 ]decane)-1,10-phenanthroline dibromide (0.796
38
established procedures. The free base form of 1,4,7-triazacy-
clononane (L1) was generated by adjusting the pH of a solution of
the trihydrobromide salt (10.9 mmol) to 12.5-13.0 with solid
g, 1.24 mmol) was added to 60 mL of aqueous 6 M HCl. The
(
(
39) Atkins, T. J. J. Am. Chem. Soc. 1980, 102, 6364-6365.
40) Weisman, G. R.; Johnson, V.; Fiala, R. E. Tetrahedron Lett. 1980,
21, 3635-3638.
(
(
34) Brown, D. M.; Usher, D. A. J. Chem. Soc. 1965, 6558-6564.
35) Bassett, J.; Denney, R. C.; Jeffery, G. H.; Mendham, J. Vogel’s
Textbook of QuantitatiVe Inorganic Analysis, 4th ed.; John Wiley &
Sons: New York, 1978; Chapter 10.
(41) Haidar, R.; Ipek, M.; DasGupta, B.; Yousaf, M.; Zompa, L. J. Inorg.
Chem. 1997, 36, 3125-3132.
(42) Graham, B.; Fallon, G. D.; Hearn, M. T. W.; Hockless, D. C. R.;
Lazarev, G.; Spiccia, L. Inorg. Chem. 1997, 36, 6366-6373.
(43) Chandler, C. J.; Deady, L. W.; Reiss, J. A. J. Heterocycl. Chem. 1981,
18, 599-601.
(44) Vriesema, B. K.; Buter, J.; Kellog, R. M. J. Org. Chem. 1984, 49,
110-113.
(
(
(
36) Atkins, T. J.; Richman, J. E.; Oettle, W. F. Org. Synth. 1978, 58,
8
6-98.
37) McAuley, A.; Norman, P. R.; Olubuyide, O. Inorg. Chem. 1984, 23,
938-1943.
38) Koyama, H.; Yoshino, T. Bull. Chem. Soc. Jpn. 1972, 45, 481-484.
1
Inorganic Chemistry, Vol. 42, No. 24, 2003 7739

Iranzo et al.
reaction mixture was heated to reflux for 24 h, the solvent was
removed on a rotary evaporator, and the yellow residue was
coevaporated with water (2 × 25 mL). This solid was washed with
30 mL of 1:2 (v:v) 12 M aqueous HCl-ethanol by stirring the
slurry for 4 h at room temperature. The solid was collected by
filtration, washed with ethanol, and dried at room temperature under
high vacuum to afford 0.811 mg of a beige powder. Yield: 96%.
1
H NMR (500 MHz, D
2
O, see SI Figure S1 for labeling): δ 8.75
3
(d, J ) 8.4 Hz, 2H, H(4) and H(7)), 8.08 (s, 2H, H(5) and H(6)),
3
8
3
.05 (d, J ) 8.4 Hz, 2H, H(8) and H(3)), 4.43 (s, 4H, CH
2
(15)),
3
.63 (s, 8H, ring-CH
2
(18)), 3.24 (t, J ) 5.5 Hz, 8H, ring-CH
2
-
3
13
(
D
1
17)), 3.13 (t, J ) 5.5 Hz, 8H, ring-CH
O, see SI Figure S1 for labeling): δ 156.11 (C(2) and C(9)),
42.89 (C(11) and C(14)), 137.91 (C(4) and C(7)), 129.67 (C(12)
2
(16)). C NMR (500 MHz,
2
and C(13)), 127.62 (C(5) and C(6)), 126.27 (C(4) and C(8)), 58.214
+
(
C(15)), 48.52, 44.17, 42.65 (ring carbons). FAB m/z: 463.3 (MH ).
HR FAB m/z: 463.3297 (calculated for C26
39 8
H N : 463.3292).
N,N′-Bis(tert-butoxycarbonyl)-1,4,7-triazacyclononane (Boc
2
-
L1). L1 (2.16 g, 16.72 mmol) and triethylamine (3.38 g, 33.42
mmol) were dissolved in 80 mL of dry chloroform, and the solution
was cooled in a 0-3 °C ice bath. Di-tert-butyl dicarbonate (6.78
g, 31.06 mmol) in 155 mL of dry chloroform was added slowly
with stirring. After the addition was complete, the reaction mixture
was allowed to warm to room temperature and then stirred for ca.
Figure 1. 1H NMR spectra of L2OH as a function of r (r ) [Zn(NO3)2]/
[
L2OH]) at 5.0 mM L2OH, pD ) 6.93, 22 °C and I ) 0.10 M, NaNO3.
20 h. The solvent was removed by rotary evaporator to afford an
oily residue. This was dissolved in a minimum amount of hexanes
to give a slurry, which was applied to a silica gel column. The
product was eluted under flash chromatography conditions with a
gradient of 100% hexanes to 30:70 (v:v) hexanes-ethyl acetate.
Fractions were collected and analyzed by TLC (30:70 (v:v)
1
,3-Bis(1,4,7-triazacyclonon-1-yl)-2-hydroxypropane HCl Salt
(
L2OH‚6HCl). Boc -L2OH (1.23 g, 1.72 mmol) was added to 18
4
mL of a solution containing 1:1 (v:v) 12 M aqueous HCl-ethanol,
and the solution was stirred for 4 h at room temperature. The white
solid that formed during this time was collected by filtration, washed
with cold ethanol, and washed by stirring in a solution containing
a 1:2 (v:v) 12 M aqueous HCl-ethanol. The final white solid was
collected by filtration, washed with cold ethanol, and dried at room
temperature under vacuum to afford 0.736 g of the pure hydro-
chloride salt. Yield: 81%. H NMR (500 MHz, D
S1 for labeling): δ 4.16 (tt, J12 ) 9.9 Hz, J12′ ) 2.4 Hz, 1H,
CH(1)-OH), 3.59-3.53 (m, 8H, ring-CH
Hz, 8H, ring- CH (4)), 3.08-3.03 (m, 4H, ring-CH
.93 (m, 4H, ring-CH
Hz, 2H, H(2′)), 2.59 (dd, J22′ ) 14.3 Hz, 3J21 ) 9.8 Hz, 2H, H(2)).
2
C NMR (300 MHz, D O, see SI Figure S1 for labeling): δ 65.96
C(1)), 58.71 (C(2)), 48.34, 43.94, 42.30 (ring carbons).
Synthesis of the Metal Complexes. The Zn(II) complexes of
L1, L2OH, L3, L4, and L5 were prepared in aqueous solution by
mixing Zn(NO and the corresponding HBr or HCl salt of the
hexanes-ethyl acetate, R
L1). Fractions containing Boc
f
) 0.80 for Boc
3
-L1, R
f
) 0.28 for Boc
2
-
2
-L1 were combined, the solvent was
removed by rotary evaporation, and the final residue was dried at
room temperature under high vacuum to give 3.43 g of a white
1
powder. Yield: 62%. The H NMR of this compound is complex
1
2
O, see SI Figure
due to the presence of two conformers. The two conformers are
more readily resolved in the 13C NMR spectrum. C NMR (300
3
3
13
3
2
(5)), 3.25 (t, J ) 5.6
MHz, CDCl
3
): δ 156.49, 156.24 (CO), 80.21, 80.11 (tBu: C(CH
3 3
) ),
2
2
(3)), 2.98-
(3)), 2.70 (dd, J2′2 ) 14.3 Hz, J2′1 ) 2.4
5
4
3.54, 52.96, 52.84, 52.10, 50.89, 50.26, 49.99, 48.74, 48.57, 48.17,
2
3
2
2
7.78 (ring carbons), 28.97 (tBu: C(CH
3
)
3
+
). FAB m/z (relative
2
+
intensity): 330.2 (MH , 100%), 274.3 (MH - 56, 24%).
13
1
,3-Bis(4,7-di(tert-butoxycarbonyl)-1,4,7-triazacyclonon-1-yl)-
-hydroxypropane (Boc -L2OH). Boc -L1 (3.14 g, 9.53 mmol),
,3-dibromo-2-propanol (0.98 g, 4.50 mmol), and triethylamine
(
2
1
4
2
(1.15 g, 11.36 mmol) were added to 110 mL of dry acetonitrile.
3 2
)
The solution was heated to reflux for 41 h, and the progress of the
reaction was monitored by TLC (30:70 (v:v) hexanes-ethyl acetate,
ligand in 1:1.1 (mononuclear system) and 2:1.1 (dinuclear system)
molar ratios and adjusting the pH to 6.5-7.0. Concentrations of
ligand and Zn(II) were determined as described above.
R
f
2
) 0.28 for Boc -L1, R
f
4
) 0.72 for Boc -L2OH). The solvent
was removed by rotary evaporation, and the oily residue was
dissolved in 110 mL of chloroform. The resulting solution was
washed with 10% aqueous NaOH (3 × 100 mL) and water (3 ×
1
H NMR Spectroscopic Measurements. The formation and
stoichiometry of the Zn(II) dinuclear complexes in D O were
2
1
1
studied by H NMR. H NMR spectra were recorded for solutions
that contained a constant concentration of ligand and increasing
2 4
100 mL), and the solution was dried over Na SO . The solvent
was removed under vacuum to afford an oil that was purified by
flash chromatography using the same conditions as for Boc -L1.
The fractions containing the product Boc -L2OH (R ) 0.72) were
concentrations of Zn(NO
concentrations of 5.0, 4.0, 1.0, and 5.0 mM, respectively, for L2OH,
L3, L5, and L4 and increasing Zn(NO from 0 to 3.0 equiv with
3
)
2
(I ) 0.10 M, NaNO
3
), using ligand
2
4
f
3 2
)
combined, the solvents were removed by rotary evaporation, and
the final residue was dried at room temperature under high vacuum
to give 1.63 g of a light-yellow powder. Yield: 50%. 13C NMR
respect to the corresponding ligand. The solution pD was calculated
45
by adding 0.4 to the reading from the pH meter. pDs were adjusted
1
to the desired value using NaOD or DCl. H NMR spectra for these
(
500 MHz, CDCl
3
, only the chemical shifts of the major isomer
), 67.62
13
compounds are given in Figure 1 and SI Figures S2-5. C NMR
were assigned): δ 155.78 (CO), 79.74 (tBu: C(CH )
3 3
spectra of the two Zn(II) complexes obtained with the L2OH ligand
are given below.
(
CHOH), 63.15 (CH
carbons), 28.73 (tBu: C(CH
m/z: 715.4953 (calculated for C35
2
-CHOH-CH
2
), 54.70, 50.54, 49.21 (ring
+
3 3
) ). FAB m/z: 715.6 (MH ). HR FAB
H
67
9 6
O N : 715.4964).
(45) Glasoe, P. K.; Long, F. A. J. Phys. Chem. 1960, 64, 188-190.
7740 Inorganic Chemistry, Vol. 42, No. 24, 2003

Metal Ion CooperatiWity
Zn
(L2O). 13C NMR (300 MHz, D
2
2
O, pD ) 6.93, see Figure 1
for labeling): δ 64.90 (C(1)), 62.44 (C(2)), 52.35, 49.09, 44.96,
Chart 1
4
3.86, 42.92, 42.48 (ring carbons).
Zn(L2OH). 13C NMR (300 MHz, D
2
O, pD ) 6.93, see Figure 1
for labeling): δ 64.72 (C(1)), 63.35 (C(2)), 54.40, 53.44, 43.63,
4
3.56, 43.00, 42.78 (ring carbons).
Kinetics of Transesterification of HPNP (2-Hydroxy-
propyl 4-nitrophenyl phosphate). The concentration of the Zn-
II) complexes examined in these experiments ranged from 0.20 to
.0 mM. In a typical experiment, a solution of the metal complex
in 20.0 mM buffer at I ) 0.10 M (NaNO ) was adjusted to the
(
2
3
desired pH, transferred to a cuvette, and equilibrated at 25 °C in a
thermostated spectrophotometer. The reaction was initiated by
injection of 10 µL of a solution of HPNP to give a final HPNP
concentration of 0.038 mM, and transesterification was monitored
by following the increase in absorbance at 400 nm due to the release
of 4-nitrophenolate. The pH values of these solutions were
determined at the end of each spectrophotometric experiment and
were found to be within 0.02 units of the initial value.
2
The reactions catalyzed by Zn (L2O) were monitored for three
half-lives, and pseudo-first-order rate constants kobsd were deter-
mined as the slopes of semilogarithmic plots of reaction progress
against time. In most other cases, the reaction of HPNP was
monitored for the disappearance of the first 5-10% of the substrate,
and the reaction endpoint was obtained by heating the reaction to
-1
6
0 °C. Values of kobsd (s ) were determined as kobsd ) V
is the initial reaction velocity and [S] is the initial substrate
concentration. Second-order rate constants kZn (M
i o
/[S] , where
V
i
o
-
1
-1
s
) for the
reactions catalyzed by the different Zn(II) complexes were deter-
mined as the slope of linear plots of kobsd against Zn(II) complex
concentration (correlation coefficients >0.997). Rate constants were
reproducible to (6%. The pseudo-first-order rate constants for
transesterification of HPNP in the absence of the catalyst kuncat (s )
were determined by the method of initial rates (5% conversion).
here were prepared from the orthoamide derivative of L1
4
,10
(1,4,7-triazatricyclo[5.2.1.0 ]decane) including L3, L4,
and L5 (Chart 1, Scheme 2). The synthetic procedure using
this protecting group leads to high yields of ligand over a
relatively short reaction time if linkers contain aromatic
groups adjacent to the methylene group. Such aromatic
linkers activate the methylene carbon to nucleophilic attack,
facilitating the synthesis of dinuclear ligands. In contrast,
the preparation of ligands from the orthoamide derivative
of L1 that contain alkyl linkers (L3) requires reaction times
-1
-
8
-1
At pH 7.6, the background rate constant was 5.2 × 10
s
and
this value was subtracted from the pseudo-first-order rate constant
-
1
k
obsd (s ) in the presence of Zn(NO
Cleavage of A . The oligonucleotide was purified by TLC on
silica gel plates by using a solvent system containing 1-propanol-
3 2
) .
6
3
2
NH
4
OH-H
2
O in a 55:35:10 ratio. A
6
was 3′ labeled with [R- P]-
41
of several days. Similarly, the preparation of the L2OH
pCp using RNA T4-ligase (Amersham Life Science). The oligo-
nucleotide was purified by polyacrylamide gel electrophoresis on
a gel containing 20% acrylamide and 1% bisacrylamide, eluted with
water, and desalted by size-exclusion chromatography (PD-10
column containing G-25M Sephadex, Pharmacia). All reaction
ligand is problematic because the alkyl linker is not activated.
In addition, previous work suggests that the alcohol group
in the linker needs protection if the orthoamide derivative
51,52
of L1 is used as a precursor.
For these reasons, the L2OH
ligand was prepared by using Boc protected L1 as a
precursor. This route was chosen over use of tosyl protecting
groups as reported in the original synthesis of this compound
in order to avoid the use of mercury amalgam for the
mixtures contained 1.00 mM complex, 2.5 µM A
6
, and 10.0 mM
and then
HEPES buffer. Solutions were spiked with 32P-labeled A
6
incubated for 24 h at 37 °C, pH 7.5. Following incubation, 10 µL
of loading buffer (9 M urea in tris-borate buffer) was added.
Samples were loaded on a 20% polyacrylamide gel, and reactions
were run alongside RNA ladders obtained from base hydrolysis of
the RNA. The gel was run at 10-20 W for 3-4 h. Radioactivity
was quantitated by phosphorimaging (Molecular Dynamics). Results
are the average of two or more experiments. A control reaction
run without catalyst had 2% cleavage after incubation for 24 h.
4
7
deprotection step. The Boc protecting group is readily
removed under mild conditions making this preparative route
(46) Wieghardt, K.; Tolksdorf, I.; Herrmann, W. Inorg. Chem. 1985, 24,
1
230-1235.
(
(
(
(
47) Sessler, J. T.; Sibert, J. W.; Burrell, A. K.; Lynch, V.; Markert, J. T.;
Wooten, C. L. Inorg. Chem. 1993, 32, 4277-4283.
48) McCue, K. P.; Voss, D. A.; Marks, C.; Morrow, J. R. J. Chem. Soc.,
Dalton Trans. 1998, 2961-2963.
Results and Discussion
49) Weisman, G. R.; Vachon, D. J.; Johnson, V. B.; Gronbeck, D. A. J.
Chem. Soc., Chem. Commun. 1987, 886-887.
Ligand Synthesis. There are several methods in the
literature for the preparation of linked 1,4,7-triazacy-
clononane macrocycles. Typically, these syntheses use a
50) Weller, H.; Siegfried, L.; Neuburger, M.; Zehnder, M.; Kaden, T. A.
HelV. Chim. Acta 1997, 80, 2315-2328.
(51) Blake, A. J.; Fallis, I. A.; Gould, R. O.; Parsonos, S.; Ross, S. A.;
Schroder, M. J. Chem. Soc., Dalton Trans. 1996, 4379-4387.
52) McGowan, P.; Podesta, T. J.; Thornton-Pett, M. Inorg. Chem. 2001,
4
6-48
deprotected derivative of L1
or, alternately, the ortho-
(
amide protecting group is used.4
1,42,49,50
Three of the ligands
40, 1445-1453.
Inorganic Chemistry, Vol. 42, No. 24, 2003 7741

Iranzo et al.
Scheme 2
in dinuclear complexes in solutions containing 10.0 mM
3 2
Zn(NO ) and 5.0 mM L3 or L2OH, respectively, at pH 6.53.
Here we use NMR spectroscopy to further study the structure
of the different metal complexes postulated in these titrations
and, for cases where the potentiometric data is not available,
to confirm that two Zn(II) ions bind to the ligand under
conditions similar to those in the kinetics experiments (vide
infra). The titration of dinuclear ligands L2OH, L3, L4, and
3 2
L5 with varying ratios of Zn(NO ) was studied by use of
1
H NMR spectroscopy as depicted in Figure 1 for L2OH
and Figures S2-5 in SI for L3, L4, L5, and L2OH.
1
For L2OH, L4, and L5, the H NMR spectrum of the free
ligand shows different conformers in solution at near neutral
1
pH values. In contrast, at acidic pH values the H NMR
spectra are consistent with a single conformer. In all sets of
1
experiments, the H NMR resonances of the free ligand
3 2
disappear upon addition of Zn(NO ) to give rise to new
resonances corresponding to the formation of new Zn(II)
complexes in slow exchange with free ligand. The protons
of each CH
lent to each other upon addition of Zn(NO )
2
group in the macrocycle ring become inequiva-
, and all ligand
3 2
resonances shift upon Zn(II) coordination. The resonances
corresponding to the protons in the different linkers generally
1
appear in a region of the H NMR spectrum relatively free
of other signals, facilitating assignments. In cases where they
overlap with the resonances of other groups, homonuclear
1
H correlation spectroscopy (COSY) NMR experiments were
used to make assignments.
Scheme 3
For the ligand L2OH, only two Zn(II)-containing species
3 2
are observed when the ratio of Zn(NO ) to ligand is
1
increased from 0 to 2.0, giving rise to relatively simple H
NMR spectra (Figure 1 and SI Figure S5). When r < 1.0,
signals of a new species A are observed in addition to signals
due to free L2OH (SI Figure S5). When r ) 1.0, only
resonances attributed to species A are present while the
resonances due to free L2OH completely disappear. When
3 2
more Zn(NO ) is added, 1.0 < r < 2.0, new resonances are
detected corresponding to species B (SI Figure S5) in
addition to the resonances of species A. At r ) 2.0, only
resonances attributed to species B are observed and further
3 2
addition of Zn(NO ) (r > 2.0) does not produce any change
in the spectrum, confirming that there is no further binding
of Zn(II). On the basis of these results, species A is assigned
to a mononuclear complex of Zn(II) and L2OH (Zn(L2OH))
2
and species B to a dinuclear complex (Zn (L2O)). These
dinuclear and mononuclear complexes of Zn(II) with L2OH
1
3
1
were further characterized by C NMR, homonuclear H
1
COSY NMR, and H nuclear Overhauser effect spectrometry
NOESY) 1D NMR spectroscopy. The C NMR spectra of
1
3
(
the method of choice for linkers containing alcohol groups
both the dinuclear complex Zn (L2O) and the mononuclear
2
(Scheme 3).
complex Zn(L2OH) show eight carbon resonances, two
corresponding to the linker and six corresponding to the
methylene groups of the L1 units. These results suggest that
Zn(II) Binding to Dinuclear Ligands. The macrocyclic
ligands L2OH and L3 strongly bind two Zn(II) ions in
solution as determined previously in studies using pH-
potentiometric titrations.3 The large binding constants of
these ligands lead to 97 and 98% of the Zn(II) being bound
1
both complexes are highly symmetric. H NOESY NMR
3,53
spectra gave further information about the conformation of
the linker protons (CH
complexes. For the dinuclear complex, irradiation of H
enhances the intensity of the resonance of H2′ but no effect
2 1 2
H2′-CH OH-CH H2′) in the two
1
(
53) DasGupta, B.; Haidar, R.; Hsieh, W.-Y.; Zompa, L. J. Inorg. Chim.
Acta 2000, 306, 78-86.
7742 Inorganic Chemistry, Vol. 42, No. 24, 2003

Metal Ion CooperatiWity
1
is observed for the resonance of H
2
. For the mononuclear
S2-4). For the ligands L3, L4, and L5, the H NMR spectra
complex, irradiation of H
1
notably enhances the resonance
of solutions containing equimolar amounts of Zn(NO ) and
3
2
of H2′ and also slightly enhances the intensity of H
2
.
ligand do not correspond to the exclusive formation of a
highly symmetric mononuclear Zn(II) complex as is observed
for the ligand L2OH. However, when the ratio is increased
to 2.0, the spectra are consistent with the presence of a single
type of complex, corresponding to the dinuclear complexes
These NMR spectral data are consistent with the following
solution structures for the mononuclear and dinuclear Zn(II)
complexes of L2OH. For the mononuclear complex, the
intramolecular cross-relaxation from H
1
to H2′ and from H
1
1
1
to H
2
observed by use of H NOESY NMR spectroscopy is
Zn (L3), Zn (L4), and Zn (L5). Notably, in all cases, the H
2
2
2
consistent with a relatively flexible linker, suggesting that
the alcohol group of the linker is not coordinated to the Zn-
NMR spectrum does not change further when the ratio is
greater than 2.0, suggesting that the dinuclear complex is
essentially completely formed with 2 equiv of Zn(II) under
the conditions of the NMR titration. Also consistent with
the formation of a dinuclear complex is the lack of any
observed precipitate of free Zn(II) ion when a second
equivalent of Zn(II) is added. This is especially significant
for the L4 ligand, which contains both L1 and phenanthroline
units that may bind Zn(II). That the complexed Zn(II) atoms
(
II) ion. This species is designated Zn(L2OH) because the
1
alcohol group is in protonated form. In addition, both the H
and 13C NMR spectra of this complex suggest a highly
symmetric structure with two equivalent L1 units. This NMR
solution data, together with structural data for complexes of
related bis(L1) ligands, suggests that the Zn(L2OH) complex
has a sandwich structure with both L1 units binding a single
Zn(II) ion to form a six-coordinate complex.5
3,54
Such a
in Zn
2
(L4) are bound to the macrocycle units is supported
structural motif, where the alcohol group is not coordinated
to the metal ion, has been found for other potentially
dinucleating systems under conditions of equimolar amounts
of metal ion and ligand.1 An alternate structure we cannot
rule out is a dimeric complex with each Zn(II) bound to two
L1 units from different ligands. Such a dimeric complex of
1
by the inequivalent H NMR resonances of each CH
2
group
in the L1 ring. Binding of the first two Zn(II) ions to the
macrocycle units is not surprising given the much larger
7,28
56a
propensity of L1 to bind Zn(II) (log K ) 11.6) compared
to substituted phenanthrolines (log K ) 4.1 for neocuproine).5
6b
What is more surprising is that a third Zn(II) does not appear
2 2
the form Zn (L2OH) could give rise to the NMR spectra
1
to bind as suggested by the H NMR spectra (compare r )
we observe. However, such dimeric complexes are generally
observed for alkyl linkers with four, five, or six methylene
units. Ligands containing two L1 units linked by a propyl
group form the intramolecular metal ion sandwich struc-
2
.0 and r ) 3.0) and the appearance of a precipitate when a
third equivalent of Zn(II) is added. However, the phenan-
throline unit may have a reduced affinity for Zn(II) in the
highly positively charged Zn (L4) complex. Kinetic data also
2
suggest that L4 binds two Zn(II) ions (see below).
4
1,53,55
ture.
2
For the dinuclear complex Zn (L2O), the linker
is relatively rigid as suggested by nuclear Overhauser effect
experiments. All the linker protons exhibit a marked upfield
shift compared to the free ligand (Figure 1). This is consistent
with coordination and deprotonation of the alcohol group to
Zompa and co-workers used pH-potentiometric titrations
to study the solution chemistry of a series of Zn(II)
complexes of ligands containing two L1 macrocycles linked
53
by alkyl groups with two to eight carbons. They found that
the ligand L3 is able to form stable dinuclear complexes with
Zn(II), consistent with our NMR studies. For the ligands L4
and L5, there are no previous studies on the stability of their
Zn(II) complexes. However, Zompa and co-workers studied
the solution chemistry of the Cu(II) complexes of L5 as well
as those of L3 and found that both ligands form stable
2
form Zn (L2O). On the basis of these observations, the
change in the chemical shift observed in our system suggests
binding of two Zn(II) to the alcohol group of the linker and
its deprotonation to form an alkoxo-bridged dinuclear Zn-
(II) complex. This result is supported by the X-ray crystal
structure obtained for the complex Zn (L2O) and the results
2
33
of pH-potentiometric titrations. These data indicate that the
ligand L2OH is able to form both stable mononuclear and
dinuclear Zn(II) complexes with L2OH, a result that differs
from that of the related ligand 1,3-bis(1,4,7-triaza-1-cy-
clononyl)-propane, which only forms a very stable mono-
nuclear sandwich complex with Zn(II).53 This distinct
behavior is likely due to the presence of the alcohol group
in the linker moiety that favors the opening of the stable
sandwich complex to form the dinuclear Zn(II) complex
4
1,57
dinuclear Cu(II)-ligand complexes.
This is consistent
with our NMR studies, which suggest that two units of L1
connected by an aromatic linker (L4 and L5) form dinuclear
complexes of Zn(II). In addition, our pH-potentiometric and
structural studies on L2OH complexation of Zn(II) are
3
3
consistent with the results obtained here. Our previous
studies suggested that an alkoxy-bridged dinuclear Zn(II)
complex was essentially completely formed in solutions
containing 1.0 mM L2OH and 2.0 mM Zn(NO ) at pH 6.0.
3 2
when the ratio of Zn(NO ) to ligand is increased to 2.0.
3
2
1
13
For the ligands L3, L4, and L5, a mixture of free ligand
and different Zn(II)-containing species, most likely mono-
nuclear as well as dinuclear species, is observed as the ratio
H and C NMR studies here confirm that there is a single
dinuclear Zn(II) complex formed under these conditions. In
summary, all the ligands studied here bind two Zn(II) ions
3 2
of Zn(NO ) to ligand is increased from 0 to 2.0 (SI Figures
(
56) (a) Martell, A. E.; Smith, R. M. Critical Stability Constants; Plenum
Press: New York, 1982; Vol. 5. (b) Martell, A. E.; Smith, R. M.
Critical Stability Constants; Plenum Press: New York, 1975; Vol. 2.
(
54) Fry, F. H.; Fallon, G. D.; Spiccia, L. Inorg. Chim. Acta 2003, 346,
7-66.
55) Moufarrej, T.; Bui, K.; Zompa, L. J. Inorg. Chim. Acta 2002, 340,
6-64.
5
(
(57) DasGupta, B.; Katz, C.; Israel, T.; Watson, M.; Zompa, L. J. Inorg.
Chim. Acta 1999, 292, 172-181.
5
Inorganic Chemistry, Vol. 42, No. 24, 2003 7743

Iranzo et al.
1
additional change in the H NMR spectrum upon addition
of a third equivalent of Zn(II), suggests that the primary
species present in solution under these conditions has two
Zn(II) ions, not three.
It was not possible to determine values for kobsd when [Zn-
(NO
3 2
) ] is g0.50 mM at pH 7.6 in solutions that contain no
ligand because of precipitation of Zn(II) hydroxide. However,
-
7
-1
the value of kobsd ) 7.2 × 10 for reaction in the
s
presence of 0.21 mM Zn(NO ) is sufficient to show that
3 2
the catalytic contribution of free Zn(II) ion toward the
cleavage of HPNP by the dinuclear Zn(II) complexes
examined in Table 1 is negligible. For example, in the case
of the reactions catalyzed by L3 and L2OH at pH 7.6 with
Figure 2. Dependence of the observed pseudo-first-order rate constant
on the concentration of Zn2(L3) at pH ) 7.6, 25 °C, and I ) 0.10 M NaNO3.
2
3 2
.0 mM ligand and 4.0 mM Zn(NO ) , only 1.14% (0.046
-
5
mM) and 0.0017% (6.8 × 10 mM), respectively, of the
total Zn(II) is present as free metal ion, so that the
contribution of this metal ion to kobsd should be much less
Table 1. Rate Constants for the Cleavage of HPNP by Zn(II)
Complexes at pH 7.6, 25 °C, and I ) 0.10 M NaNO3
kZn (M-1 s-1
relative ka
catalyst
)
-
7
-1
than the value of 7.2 × 10
s
determined for reaction in
. In the case of L4 and
-
-
-
-
-
2
Zn2(L2O)
Zn2(L3)
Zn2(L4)
Zn2(L5)
Zn(L1)
25 × 10
1.1 × 10
120
5
4
3
1
2
2
2
2
the presence of 0.21 mM Zn(NO )
3 2
0.89 × 10
0.58 × 10
0.21 × 10
L5, the failure to detect free ligand by NMR spectral analysis
of solutions that contain a 2-fold-excess concentration of zinc
over ligand shows that <5% of Zn(II) under these conditions
is present in the free form. We conclude that under the
conditions given in Table 1, all dinuclear Zn(II) catalysts
examined here are more effective than free Zn(II) in catalysis
of transesterification of HPNP.
a
kZn/kZn(L1).
to form a dinuclear complex under conditions similar to those
in our kinetics experiments described below.
Kinetics of Transesterification of HPNP and RNA. The
transesterification of HPNP catalyzed by the different Zn(II)
complexes was monitored by following the increase in
absorbance at 400 nm due to the release of 4-nitrophenolate.
The cyclic phosphate diester was identified by 31P NMR as
the sole phosphorus-containing product of each metal ion
catalyzed transesterification of HPNP examined in this work.
Figure 2 shows the effect of increasing concentrations of
The catalysis of the cleavage of an oligoribonucleotide
containing six adenosines (A
dinuclear Zn(II) catalysts was also studied. The extent of
cleavage of A after 24 h was 4, 10, and 90% for Zn(L1),
Zn (L5), and Zn (L2O), respectively. This represents all
6
) by Zn(L1) and by two
6
2
2
cleavage products, and no attempt was made to separately
quantify the oligoribonucleotides of different lengths. The
oligonucleotide products from the Zn(II) complex catalyzed
cleavage comigrate with those produced by hydroxide
catalyzed cleavage. The 3′-end-labeled oligonucleotide prod-
ucts of the hydroxide ladder contain 5′-OH as a terminal
group. Comigration of the Zn(II) catalyzed cleavage products
confirms that the Zn(II) complexes promote the hydrolytic
cleavage of RNA. Qualitatively, the extent of cleavage
correlates to the relative rate constants observed for the
HPNP substrate with the dinuclear complexes being the best
Zn
of HPNP. The second-order rate constant for the Zn
catalyzed reaction was determined as the slope of the linear
plot of kobsd against [Zn (L3)]. The reactions of other catalysts
were examined in the same way through [cat] ) 2.0 mM,
except for Zn
(L5), which was only soluble up to ≈1.0 mM.
Table 1 summarizes the second-order rate constants (kZn
2
(L3) on the observed first-order rate constant for cleavage
2
(L3)-
2
2
,
M- s ) for transesterification of HPNP catalyzed by several
different Zn(II) complexes at pH 7.6. The data for dinuclear
complexes are also reported as the relative rate acceleration
with respect to the mononuclear complex Zn(L1). The
catalytic activities of most of these dinuclear Zn(II) com-
plexes are only marginally larger than that for Zn(L1) (3- to
1 -1
2
catalysts and Zn (L2O) by far the more active of the two.
Comments on the Mechanism of Cleavage of RNA and
RNA Analogues by Dinuclear Catalysts. Table 1 shows
that the catalytic activity of the dinuclear complexes Zn
2
(L3),
2
5-fold), with the exception of Zn (L2O), which is a 120-
Zn (L4), and Zn (L5) is not much greater than the sum of
2
2
fold better catalyst than Zn(L1).
their parts, as determined from the activity of Zn(L1). These
results provide strong evidence that the subunits of these
dinuclear catalysts act nearly independently. The observation
that the activity of these dinuclear complexes is 3-5-fold
larger than that of the mononuclear Zn(L1) is consistent with
weak (<0.4 kcal/mol) stabilization of the transition state for
reaction at one tethered complex by long-range electrostatic
interactions with the second.
The L4 ligand contains a phenanthroline unit as a part of
the linker and may bind a third Zn(II) ion. Cleavage of HPNP
at pH 8.0 was studied in the presence of 1.0 mM L4 with 2
and 3 equiv of Zn(II). These reactions were run in parallel,
and although precipitation was observed with 3 equiv of
Zn(II), the reaction was allowed to proceed until completion.
-
1
The pseudo-first-order rate constants kobsd (s ), determined
-
5
by the method of three half-lives, are identical (1.4 × 10
The situation is more interesting for the reaction catalyzed
-1
s ) for experiments conducted with a 2:1 or a 3:1 ratio.
2
by Zn (L2O), for which the overall catalytic activity is 120-
1
This result, along with the H NMR data which shows no
fold greater than the mononuclear complex Zn(L1), or 60-
7744 Inorganic Chemistry, Vol. 42, No. 24, 2003

Metal Ion CooperatiWity
fold greater than the activity expected for a complex in which
the tethered complexes react independently. Here, the alkoxy
group shields the electrostatic interactions between the Zn(II)
ions and allows the cations to be drawn relatively close
together in a complex of greatly enhanced activity (I). This
catalysis or that two available coordination sites per metal
center are required to form a good dinuclear catalyst. For
Zn (L3), Zn (L4), and Zn (L5), a lack of available coordina-
2 2 2
tion sites is not the most likely reason that these complexes
are not good catalysts. These complexes are expected to have
two or three available coordination sites on each Zn(II)
center.
Another important feature of metal ion complex catalyzed
cleavage of RNA and RNA analogues is that pH-rate
profiles frequently track the formation of metal ion hydroxide
2
high density of positive charge at Zn (L2O) is ideal for
providing electrostatic stabilization of the transition state for
cleavage of HPNP relative to the reactant state because there
is a net unit increase in negative charge on proceeding from
the reactant to transition state.
complexes. For Zn
pH-rate profile is consistent with a single ionization to form
an active catalytic species and the kinetic pK of this species
matches the thermodynamic pK for formation of Zn (L2O)-
OH) within experimental error. Although there are alterna-
2
(L2O) catalyzed cleavage of HPNP, the
a
a
2
33
(
tive mechanisms possible, one interpretation of this obser-
vation is that the metal hydroxide ligand participates in RNA
cleavage by acting as general base catalyst (I). One reason
2
that Zn (L2O) is a good catalyst at near neutral pH values
is that the Zn(II) hydroxide complex forms at lower pH
values compared to the analogous mononuclear complex,
making catalysis more effective at neutral pH. Thus knowl-
2
The bridging alkoxide group in Zn (L2O) may play a
similar role to the bridging ligands in enzymes that utilize
two metal ions. Dinuclear functional motifs in the active
center of metalloenzymes have the metal ions bridged by
an endogenous anionic amino acid, usually Asp or Glu, or
edge of the pK of the Zn(II)-water ligands and a full pH
a
dependence of the catalyzed reaction is important in making
comparisons of the catalytic properties of dinuclear and
mononuclear Zn(II) complexes.
A second difference observed for dinuclear catalysts
compared to mononuclear catalysts is the strength of their
1
-3
by an exogenous ligand such as water or hydroxide.
Consequently, the electrostatic repulsions between metal ions
are reduced by the bridging ligands and they can come in
close proximity to function as a catalytic unit.
Mononuclear metal ion complexes that are effective
catalysts for cleavage of RNA have a least two coordination
sites available for catalysis.58 These coordination sites are
necessary for binding the substrate and catalyzing cleavage.
The tridentate ligand L1 was chosen because it binds Zn(II)
tightly (log K ) 11.6) while leaving two to three coordination
2
interaction with substrate. Zn (L2O) binds a phosphate diester
6
-fold more strongly than does an analogous mononuclear
3
3
Zn(II) complex. Strong substrate binding to the Zn(II)
complex has the effect of increasing the second-order rate
constant for cleavage because more of the Michaelis complex
is formed. Binding constants for the other dinuclear com-
plexes were not determined here. However, given that the
difference in binding to a dinuclear complex compared to
the mononuclear complex is only 6-fold, it is unlikely that
binding constants alone would be responsible for the dif-
ferences observed here in catalytic properties of the dinuclear
Zn(II) complexes.
2
sites open for catalysis. Zn (L2O) probably has fewer
available coordination sites than do the other dinuclear
complexes lacking a bridging group and, on this basis, might
be expected to be a poorer catalyst. The crystal structure of
[
2 2 2 4 2
Zn (L2O)(Cl)(H O) ](ClO ) has one five-coordinate and
The mechanism of cleavage of HPNP by the Zn(II)
complexes of L3, L4, and L1 is not easily interpreted because
the solution chemistry of these complexes is not well-defined.
one six-coordinate Zn(II) center with four coordination sites
on each Zn(II) occupied by L2O, including two tridentate
33
L1 units and a bridging alkoxide. This leaves one or two
coordination sites available for catalysis on each Zn(II)
center, and this is evidently a sufficient number to form a
good catalyst. For dinuclear catalysts, the minimum number
of available coordination sites for a good catalyst has not
been established and could be different from the requirements
for mononuclear catalysts if the two metal ion centers
cooperatively catalyze the reaction. Of note here is that the
dinuclear Cu(II) complex of L2OH, which has two five-
coordinate Cu(II) centers with a single available coordination
site on each Cu(II),59 is completely inactive in catalyzing
the cleavage of HPNP.60 This suggests either that the
coordination sites on Cu(II) are not suitable for binding and
a 2
For example, the pK of water ligands in Zn (L3) is not
readily measured. This is due to slow equilibria at pH values
where hydroxide complex formation is generally observed,
possibly due to the formation of hydroxide-bridged com-
plexes of higher order aggregates. Similarly, pH profiles for
2
the cleavage of HPNP by Zn (L4) complexes from pH 6 to
1
0 are not consistent with a single well-defined ionization
attributable to a Zn(II)-water ligand (data not shown). The
detailed mechanistic study of Zn(L1) as a catalyst is similarly
made difficult by its precipitation at pH values greater than
8
.0. Dinuclear or mononuclear complexes with complicated
(
59) Fry, F. H.; Moubaraki, B.; Murray, K. S.; Spiccia, L.; Warren, M.;
Skelton, B. W.; White, A. H. J. Chem. Soc., Dalton Trans. 2003, 866-
871.
(58) Amin, S. A.; Morrow, J. R.; Lake, C. H.; Churchill, M. R. Angew.
Chem., Int. Ed. Engl. 1994, 33, 773-775.
(60) Manuscript in preparation.
Inorganic Chemistry, Vol. 42, No. 24, 2003 7745

Iranzo et al.
speciation are not amenable to detailed mechanistic analysis.
In contrast, the bridging alkoxide in the L2OH ligand
facilitates the formation of a dinuclear Zn(II) complex with
relatively simple solution chemistry. This solution chemistry
can then be correlated to catalytic properties for the elucida-
tion of the mechanism of cleavage of RNA and RNA model
substrates. The high activity at neutral pH and straightforward
bridging the two Zn(II) centers, but we have no structural
or spectroscopic evidence for this binding mode. However,
Zn (L2O) binds phosphate diesters 6-fold more strongly than
2
does an analogous mononuclear complex, suggesting that
the second metal ion interacts with the phosphate diester
substrate. In addition, dinuclear Zn(II) and Cu(II) complexes
with similar bridging alkoxide groups have been shown to
form complexes containing bridging phosphate diesters.1
A Zn(II)-hydroxide complex is likely the active catalytic
form and may act as a general base catalyst although we
cannot rule out other possible mechanisms as discussed
7,22,28
solution chemistry of the Zn
study as a catalyst.
2
(L2O) complex facilitates its
It is interesting that Zn
2
(L2O) also effectively cleaves the
oligoribonucleotide A . One criticism of work with HPNP
6
3
3
is that this substrate has a much better leaving group than
does RNA and does not have the ribose ring. It has been
pointed out that free metal ions and Zn(II) mononuclear
complexes catalyze cleavage of HPNP more effectively than
previously. The Zn(II) complex also facilitates cleavage
through an electrophilic role, most likely through stabilization
of a negatively charged phosphorane transition state. Another
possible important role for these complexes is in facilitating
leaving group departure in the cleavage of RNA or RNA
3
1
RNA or substrates that more closely resemble RNA. Our
initial results here suggest that three of our Zn(II) catalysts
promote cleavage of RNA and that the extent of cleavage
correlates roughly to that for the HPNP substrate. Further
studies are underway using different RNA substrates to better
31
substrates that contain poor leaving groups. Further experi-
ments are underway in our laboratories with the goal of
delineating these possible roles for metal ion complexes in
the cleavage of RNA.
2
understand why dinuclear complexes in general and Zn (L2O)
in particular are so effective at promoting the cleavage of
RNA.
Acknowledgment. J.R.M. and J.P.R. thank the National
Science Foundation (CHE 9986332) for support of this work.
We thank Dr. Kevin P. McCue for preliminary studies. We
acknowledge the clip art from Corel Gallery 2 that was used
for the cover art for this article and Erik Farquhar, Olga
Iranzo, and Cliff Rossiter for designing the cover art.
To date, our scheme for the mechanism of cleavage of
RNA by dinuclear Zn(II) complexes has the following
features. For effective catalysis, two metal ions are main-
tained in close proximity. This is accomplished here and in
metalloenzymes that utilize two metal ions for phosphate
diester cleavage by use of a bridging ligand. The first step
in catalysis involves binding of the dinuclear complex to the
phosphate diester moiety of the RNA or RNA model
substrate. It is possible that the phosphate diester binds by
Supporting Information Available: Ligand labels for L4 and
1
L2OH and H NMR spectra of L3, L5, L4, and L2OH. This material
is available free of charge via the Internet at http://pubs.acs.org.
IC030131B
7746 Inorganic Chemistry, Vol. 42, No. 24, 2003