Simultaneous interaction with base and phosphate moieties modulates the phosphodiester cleavage of dinucleoside 3′,5′-monophosphates by dinuclear Zn2+complexes of Di(azacrown) ligands

Article abstract of DOI:10.1021/ja058806s

Five dinucleating ligands (1-5) and one trinucleating ligand (6) incorporating 1,5,9-triazacyclododecan-3-yloxy groups attached to an aromatic scaffold have been synthesized. The ability of the Zn²⁺ complexes of these ligands to promote the transesterification of dinucleoside 3′,5′-monophosphates to a 2′,3′-cyclic phosphate derived from the 3′-linked nucleoside by release of the 5′-linked nucleoside has been studied over a narrow pH range, from pH 5.8 to 7.2, at 90 °C. The dinuclear complexes show marked base moiety selectivity. Among the four dinucleotide 3′,5′-phosphates studied, viz. adenylyl-3′,5′-adenosine (ApA), adenylyl-3′,5′-uridine (ApU), uridylyl-3′,5′-adenosine (UpA), and uridylyl-3′, 5′-uridine (UpU), the dimers containing one uracil base (ApU and UpA) are cleaved up to 2 orders of magnitude more readily than those containing either two uracil bases (UpU) or two adenine bases (ApA). The trinuclear complex (6), however, cleaves UpU as readily as ApU and UpA, while the cleavage of ApA remains slow. UV spectrophotometric and ¹H NMR spectroscopic studies with one of the dinucleating ligands (3) verify binding to the bases of UpU and ApU at less than millimolar concentrations, while no interaction with the base moieties of ApA is observed. With ApU and UpA, one of the Zn²⁺- azacrown moieties in all likelihood anchors the cleaving agent to the uracil base of the substrate, while the other azacrown moiety serves as a catalyst for the phosphodiester transesterification. With UpU, two azacrown moieties are engaged in the base moiety binding. The catalytic activity is, hence, lost, but it can be restored by addition of a third azacrown group on the cleaving agent.

Full text of DOI:10.1021/ja058806s

Published on Web 07/29/2006
Simultaneous Interaction with Base and Phosphate Moieties
Modulates the Phosphodiester Cleavage of Dinucleoside
3′,5′-Monophosphates by Dinuclear Zn2+ Complexes of
Di(azacrown) Ligands
Qi Wang* and Harri Lo¨nnberg
Contribution from the Department of Chemistry, UniVersity of Turku, FIN-20014 Turku, Finland
Received December 29, 2005; E-mail: qiwang@utu.fi
Abstract: Five dinucleating ligands (1-5) and one trinucleating ligand (6) incorporating 1,5,9-triazacy-
clododecan-3-yloxy groups attached to an aromatic scaffold have been synthesized. The ability of the Zn2+
complexes of these ligands to promote the transesterification of dinucleoside 3′,5′-monophosphates to a
2′,3′-cyclic phosphate derived from the 3′-linked nucleoside by release of the 5′-linked nucleoside has
been studied over a narrow pH range, from pH 5.8 to 7.2, at 90 °C. The dinuclear complexes show marked
base moiety selectivity. Among the four dinucleotide 3′,5′-phosphates studied, viz. adenylyl-3′,5′-adenosine
(ApA), adenylyl-3′,5′-uridine (ApU), uridylyl-3′,5′-adenosine (UpA), and uridylyl-3′,5′-uridine (UpU), the dimers
containing one uracil base (ApU and UpA) are cleaved up to 2 orders of magnitude more readily than
those containing either two uracil bases (UpU) or two adenine bases (ApA). The trinuclear complex (6),
however, cleaves UpU as readily as ApU and UpA, while the cleavage of ApA remains slow. UV
spectrophotometric and 1H NMR spectroscopic studies with one of the dinucleating ligands (3) verify binding
to the bases of UpU and ApU at less than millimolar concentrations, while no interaction with the base
moieties of ApA is observed. With ApU and UpA, one of the Zn2+-azacrown moieties in all likelihood
anchors the cleaving agent to the uracil base of the substrate, while the other azacrown moiety serves as
a catalyst for the phosphodiester transesterification. With UpU, two azacrown moieties are engaged in the
base moiety binding. The catalytic activity is, hence, lost, but it can be restored by addition of a third azacrown
group on the cleaving agent.
with substitution-labile divalent metal ions. The dinuclear Zn2+
and Cu2+ complexes are typically 1 or 2 orders of magnitude
Introduction
Cleavage of RNA phosphodiester bonds by dinuclear 3d
transition metal ion complexes has received considerable
interest1 because many phosphoesterases contain two metal ions
in their catalytic center.2 Since the intermetallic distance in
phosphoesterases is of the order of 4 Å, it has been reasoned
that binding of two metal ions to a ligand that affords a 3-5 Å
distance between the metal centers is optimal for their coopera-
tive action. In fact, studies with substitution by inert Co3+
complexes have lent support to this reasoning. Two Co3+ ions
at a distance of 2.9 Å have been shown to result in a 4 × 105-
fold rate acceleration in the hydroxide-ion-catalyzed hydrolysis
of 2-hydroxypropyl 4-nitrophenyl phosphate (HPNP) by binding
simultaneously to both of the nonbridging phosphoryl oxygen
atoms.3 Equally high rate accelerations have not been obtained
more efficient catalysts than their mononuclear counterparts
when HPNP is used as a substrate. Dinuclear complexes
containing a bridging alkoxide group that keeps the two metal
ions in close proximity have been shown to exhibit up to 100-
fold rate enhancements over the corresponding mononuclear
complexes.4,5 The acceleration has been attributed to increased
acidity of the metal aquo ion, increased affinity to phosphate,
and additional transition-state stabilization.5 Additional examples
of synergic action of two Zn2+ ions are offered by a dinuclear
complex of a cyclic â-sheet peptide, gramicidin S,6 and a
dinuclear complex of calix[4]arene bearing two 2,6-bis(dialkyl-
aminomethyl) moieties as nucleating centers.7 Care should,
(4) (a) Wall, M.; Hynes, R. C.; Chin, J. Angew. Chem., Int. Ed. Engl. 1993,
32, 1633-1635. (b) Gajda, T.; Kra¨mer, R.; Jancso, A. Eur. J. Inorg. Chem.
2000, 1635-1644. (c) Albedyhl, S.; Schnieders, D.; Jancso, A.; Gajda, T.;
Krebs, B. Eur. J. Inorg. Chem. 2002, 1400-1409. (d) Gajda, T.; Jancso,
A.; Mikkola, S.; Lo¨nnberg, H.; Sirges, H. J. Chem. Soc., Dalton Trans.
2002, 1757-1763.
(5) (a) Iranzo, O.; Kovalevsky, A. Y.; Morrow, J. R.; Richard, J. P. J. Am.
Chem. Soc. 2003, 125, 1988-1993. (b) Iranzo, O.; Elmer, T.; Richard, J.
P.; Morrow, J. R. Inorg. Chem. 2003, 42, 7737-7746.
(6) Yamada, K.; Takahashi, Y.-i.; Yamamura, H.; Araki, S.; Saito, K.; Kawai,
M. Chem. Commun. 2000, 1315.
(1) Morrow, J. R.; Iranzo, O. Curr. Opin. Chem. Biol. 2004, 8, 192-200.
(2) (a) Weston, J. Chem. ReV. 2005, 105, 2151-2174. (b) Stra¨ter, N.; Lipscomb,
W. N.; Klabunde, T.; Krebs, B. Angew. Chem., Int. Ed. Engl. 1996, 35,
2024-2055. (c) Lipscomb, W. N.; Stra¨ter, N. Chem. ReV. 1996, 96, 2375-
2433. (d) Wilcox, D. E. Chem. ReV. 1996, 96, 2435-2458. (e) Gani, D.;
Wilkie, J. Chem. Soc. ReV. 1995, 24, 55-63. (f) Lahm, A.; Volbeda, S.;
Suck, D. J. Mol. Biol. 1990, 215, 207-210. (g) Davies, J. F.; Hostomska,
Z. Science 1991, 252, 88-95. (h) Beese, L. S.; Steitz, T. A. EMBO J.
1991, 10, 25-33.
(3) (a) Williams, N. H.; Takasaki, B.; Wall, M.; Chin, J. Acc. Chem. Res. 1999,
32, 485-493. (b) Williams, N. H.; Chin, J. Chem. Commun. 1996, 131-
132. (c) Williams, N. H.; Cheung, W.; Chin, J. J. Am. Chem. Soc. 1998,
120, 8079-8087.
(7) (a) Molenveld, P.; Kapsabelis, S.; Engbersen, J. F. J.; Reinhoudt, D. N. J.
Am. Chem. Soc. 1997, 119, 2948-2949. (b) Molenveld, P.; Stikvoort, W.
M. G.; Kooijman, H.; Spek, A. L.; Engbersen, J. F. J.; Reinhoudt, D. N. J.
Org. Chem. 1999, 64, 3896-3906.
9
10716
J. AM. CHEM. SOC. 2006, 128, 10716-10728
10.1021/ja058806s CCC: $33.50 © 2006 American Chemical Society

Base-Selective Cleavage of Dinucleoside Monophosphates
A R T I C L E S
however, be exercised on interpreting the rate accelerations by
bimetallic systems, since considerable rate enhancements may
also result from interaction between an aromatic ligand and the
departing 4-nitrophenoxide group.8
The only clear indication of base selectivity has been obtained
with a dinuclear Cu2+ complex containing terpyridine and
bipyridine as nucleating moieties.16 This complex cleaves ApA,
adenylyl-3′,5′-cytidine (ApC), and cytidylyl-3′,5′-adenosine
(CpA) approximately 1 order of magnitude more rapidly than
dimers containing no adenine base. The observed selectivity
has been attributed to stacking between the aromatic ligand and
the adenine base. Another example of marked base selectivity
is offered by the trinuclear Zn2+ (and Cu2+) complex of calix-
[4]arene, bearing three 2,6-bis(aminomethyl)pyridine coordina-
tion sites.17 This complex cleaves guanylyl-3′,5′-guanosine
(GpG) more than 1 order of magnitude faster than other dinu-
cleoside monophosphates. The trinuclear Zn2+ complex of 1,3,5-
tris[di(pyridin-2-ylmethyl)aminomethyl]benzene has been shown
to cleave UpU much more readily than its 2′,5′-isomer, while
with the corresponding adenosine diesters, the situation is the
opposite.18 However, no clear selectivity toward the base moiety
has been observed.
While the previous studies on cleavage of RNA phosphodi-
ester bonds by dinuclear transition metal complexes have mainly
concerned the interaction of metal ions with the phosphoryl
oxygens and/or the entering and departing nucleophile, the
present work is aimed at elucidating the effects that anchoring
of a dinuclear complex to the base moiety may have on its
catalytic activity. The Zn2+ chelates of small azacrowns have
been established to bind very tightly to the deprotonated N3
atom of uracil and thymine bases.19 In addition, ditopic
complexes containing two Zn2+-azacrown moieties attached
to an aromatic scaffold have been shown to exhibit simultaneous
interaction with the base and phosphate moieties of thymidine
5′-phosphate.19g,j Accordingly, such dinuclear complexes show
potential for development of cleaving agents that simultaneously
interact with the base and phosphate moiety, resulting in base-
selective cleavage. As an indication of the feasibility of this
approach, we have previously shown that the heterodinuclear
Ni2+,Zn2+ complex of a spiro di(azacrown) ligand, 2,6,10,14,-
18,22-hexaazaspiro[11.11]tricosane, cleaves UpU and adenylyl-
3′,5′-uridine (ApU) much faster than uridylyl-3′,5′-adeno-
sine (UpA) and exerts no cooperative acceleration on the
cleavage of ApA.20 Recently, the dinuclear Zn2+ complex
of 1,3-bis(1,4,7-triazacyclononan-1-yl)-2-propanol has been
shown to cleave UpU more efficiently than HPNP, but the more
efficient catalysis has been attributed to stabilization of
the transition state by interaction of the cleaving agent with the
C5′-oxyanion.21 To learn more about the selectivity of various
The results obtained with simple model esters containing no
nucleic acid base cannot be directly extrapolated to the cleavage
of RNA. First, the mechanism of transesterification of the
unactivated phosphodiester bonds in RNA is, in all likelihood,
different from that utilized by the highly labile 4-nitrophenyl
esters. With aryl phosphodiesters, the formation rather than
cleavage of the phosphorane intermediate obtained by the attack
of the neighboring hydroxyl group on the phosphorus atom is
rate-limiting, while with alkyl esters, such as RNA, the break-
down of the phosphorane constitutes the slow step.9 Second,
the plethora of donor atoms in RNA may result in formation of
nonproductive complexes that compete with the desired species,
benefiting from double Lewis acid activation. For example, a
dinuclear Zn2+ complex, where the intermetallic distance had
been adjusted to 4.5 Å by steric gearing, exhibited an 80-fold
catalytic advantage over the mononuclear complex when HPNP
was used as a substrate, but resulted in no additional acceleration
on the cleavage of uridylyl-3′,5′-uridine (UpU).10 Nevertheless,
clear evidence for double Lewis acid activation has been
obtained even with dinucleoside 3′,5′-monophosphates. The
dinuclear Cu2+ complex of 1,8-bis(1,4,7-triazacyclononan-1-
ylmethyl)naphthalene cleaves adenylyl-3′,5′-adenosine (ApA)
500 times more readily than the mononuclear Cu2+ complex of
1,4,7-triazacyclononane.11 More modest rate accelerations have
been observed with other bimetallic agents. The dinuclear Zn2+
complexes of N,N,N′,N′-tetrakis(pyridin-2-ylmethyl)-1,3-diami-
nopropan-2-ol12 and N,N,N′,N′-tetrakis(pyridin-2-ylmethyl)-m-
and -p-xylyldiamine12,13 cleave the internucleosidic phosphodi-
ester bond 2-15 times as fast as the mononuclear complex of
bis(pyridin-2-ylmethyl)amine, without any marked selectivity
toward the base moiety. Among complexes of macrocyclic
ligands, the dinuclear Zn2+ complex of bisdien (1,13-dioxa-
4,7,10,16,19,22-hexaazacyclotetracosane) cleaves UpU 1 order
of magnitude faster than the mononuclear complex of 1,5,9-
triazacyclododecane ([12]aneN3),14 and the dinuclear Zn2+
complex of 1,4,7,16,19,22-hexaaza-10,13,25,28-tetraoxacyclo-
triacontane cleaves ApA more rapidly than its mononuclear
counterparts.15 This kind of modest rate acceleration may,
however, reflect enhanced binding rather than real double Lewis
acid activation.
Binding of dinuclear cleaving agents to the base moiety of
ribonucleoside 3′-phosphodiesters has not received much at-
tention. Only in a few cases has the effect of the base moiety
structure on the catalytic activity been systematically studied.
(16) Liu, S.; Hamilton, A. D. Chem. Commun. 1999, 587-588.
(17) Molenveld, P.; Engbersen, J. F. J.; Reinhoudt, D. N. Angew. Chem., Int.
Ed. 1999, 38, 3189-3192.
(18) Komiyama, M.; Kina, S.; Matsumura, K.; Sumaoka, J.; Tobey, S.; Lynch,
V. M.; Anslyn, E. J. Am. Chem. Soc. 2002, 124, 13731-13736.
(19) (a) Shionoya, M.; Kimura, E.; Shiro, M. J. Am. Chem. Soc. 1993, 115,
6730-6737. (b) Shionoya, M.; Ikeda, T.; Kimura, E.; Shiro, M. J.
Am. Chem. Soc. 1994, 116, 3848-3859. (c) Kimura, E.; Ikeda, T.;
Aoki, S.; Shionoya, M. J. Biol. Inorg. Chem. 1998, 3, 259-267. (d)
Kikuta, E.; Murata, M.; Katsube, N.; Koike, T.; Kimura, E. J. Am. Chem.
Soc. 1999, 121, 5426-5436. (e) Kimura, E.; Kikuchi, M.; Kitamura,
H.; Koike, T. Chem. Eur. J. 1999, 5, 3113-31123. (f) Kimura, E.;
Kitamura, H.; Ohtani, K.; Koike, T. J. Am. Chem. Soc. 2000, 122, 4668-
4677. (g) Aoki, S.; Kimura, E. J. Am. Chem. Soc. 2000, 122, 4542-4548.
(h) Kimura, E.; Kikuta, E. J. Biol. Inorg. Chem. 2000, 5, 139-155. (i)
Kikuta, E.; Aoki, S.; Kimura, E. J. Biol. Inorg. Chem. 2002, 7, 473-482.
(j) Aoki, S.; Kimura, E. ReV. Mol. Biotechnol. 2002, 90, 129-155. (k)
Gao, J.; Reibenspies, J. H.; Martell, A. E. Org. Biomol. Chem. 2003, 1,
4242-4247. (l) Aoki, S.; Kimura, E. Chem. ReV. 2004, 104, 769-787.
(m) Rossiter, C. S.; Mathews, R. A.; Morrow, J. R. Inorg. Chem. 2005,
44, 9397-9404.
(8) Leivers, M.; Breslow, R. Bioorg. Chem. 2001, 29, 345-356.
(9) (a) Lo¨nnberg, H.; Stro¨mberg, R.; Williams, A. Org. Biomol. Chem. 2004,
2, 2165-2167. (b) Komiyama, M.; Matsumoto, Y.; Takahashi, H.; Shiiba,
T.; Tsuzuki, H.; Yajima, H.; Yashiro, M.; Sumaoka, J. J. Chem. Soc., Perkin
Trans. 2 1998, 691-695. (c) Mikkola, S.; Stenman, E.; Nurmi, K.; Yousefi-
Salakdeh, E.; Stro¨mberg, R.; Lo¨nnberg, H. J. Chem. Soc., Perkin Trans. 2
1999, 1619-1625.
(10) Worm, K.; Chu, F.; Matsumoto, K.; Best, M. D.; Lynch, V.; Anslyn, E. V.
Chem. Eur. J. 2003, 9, 741-747.
(11) Young, M. J.; Chin, J. J. Am. Chem. Soc. 1995, 117, 10577-10578.
(12) Yashiro, M.; Kaneiwa, H.; Onaka, K.; Komiyama, M. J. Chem. Soc., Dalton
Trans. 2004, 605-610.
(13) Yashiro, M.; Kawahara, R. J. Biol. Inorg. Chem. 2004, 9, 914-921.
(14) Mikkola, S.; Wang, Q.; Zoltan, J.; Helkearo, M.; Lo¨nnberg, H. Acta Chem.
Scand. 1999, 53, 453-456.
(15) Bencini, A.; Berni, E.; Bianchi, A.; Giorgi, C.; Valtancoli, B. Supramol.
Chem. 2001, 13, 489-497.
(20) Wang, Q.; Mikkola, S.; Lo¨nnberg, H. Chem. BiodiVersity 2004, 1, 1316-
1326.
9
J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006 10717

A R T I C L E S
Wang and Lo¨nnberg
Chart 1. The Ligands Studied
Scheme 1 a
a Reagents and conditions: (i) PhCHO, H2SO4; (ii) PhCH2Cl, KOH, benzene, reflux; (iii) HCl, MeOH(aq); (iv) TsCl, THF(aq), NaOH, 3-5 °C; (v) 1.
t
TBD, KOH, dioxane, reflux, 48 h, 2. NaBH4, dioxane, rt, 2 d; (vi) 4 mol L-1 HCl(aq), reflux, 4 d; (vii) Boc2O, BuOH(aq), NaOH, 40 °C, 2 d.
di(azacrown) complexes as cleaving agents, five dinucleating
ligands (1-5) and one trinucleating ligand (6) incorporating
two and three 1,5,9-triazacyclododecan-3-yloxy groups, re-
spectively, as coordination sites (Chart 1) have been synthe-
sized and studied in the presence of Zn2+ ion as catalysts for
the cleavage of the phosphodiester bond of ApA, ApU, UpA,
and UpU. The ligands employed allow the intermetallic dis-
tance to vary within rather wide limits and enable bridging
between the base and phosphate moiety. Efficient cleavage and
high base moiety selectivity have been obtained with these
complexes.
benzyl ethers (8) and hydrolyzed to 2-(benzyloxy)propane-1,3-
diol (9). The hydroxyl functions were tosylated (10), and the
tosyloxy groups were displaced with 1,5,7-triazabicyclo[4.4.0]-
dec-5-ene (TBD) to obtain a tricyclic orthoamide (11), as
described originally by Alder et al.23 and Kim et al.24 Prolonged
treatment (4 d) of 11 with 4 mol L-1 aqueous hydrogen chloride
under reflux gave 1,5,9-triazacyclododecan-3-ol trihydrochloride
(12), which was converted to the N-Boc-protected form (13)
by conventional methods.
Ligands 1-6 were obtained in N-Boc-protected form
(14-19) by treating an appropriate bis(bromomethyl)benzene,
4,4′-bis(chloromethyl)biphenyl, 2,6-(tosyloxy)pyridine,25 or 1,3,5-
tris(bromomethyl)benzene26 in dimethylformamide (DMF) with
the oxyanion of 13 (Scheme 2). The Boc groups were removed
by acid-catalyzed hydrolysis, which gave 1-6 as hydrochloric
salts. The free bases were obtained by passing the aqueous
solutions of the hydrochloric salts through a strong anion-
exchange resin (Dowex 1X2, OH- form). The products were
Results
Syntheses of Ligands 1-6. Scheme 1 depicts the synthesis
of N-tert-butoxycarbonyl (Boc)-protected 1,5,9-triazacyclodode-
can-3-ol (13) used to construct the di- and tri(azacrown) ligands
studied (1-6). Glycerol was first reacted with benzaldehyde to
obtain cis-5-hydroxy-2-phenyl-1,3-dioxane (7), as described
previously.22 This was converted to a mixture of cis and trans
(23) Alder, R. W.; Mowlam, R. W.; Vachon, D. J.; Weisman, G. R. J. Chem.
Soc., Chem. Commun. 1992, 507-508.
(24) Kim, R. D. H.; Wilson, M.; Haseltine, J. Synth. Commun. 1992, 24, 3109-
3114.
(25) Cabezon, B.; Cao, J.; Raymo, F. M.; Stoddart, J. F.; White, A. J. P.;
Williams, D. J. Chem. Eur. J. 2000, 6, 2262-2273.
(26) Langer, P.; Anders, J. T. Eur. J. Org. Chem. 2002, 686-691.
(21) O’Donoghue, A. M.; Pyun, S. Y.; Yang, M.-Y.; Morrow, J. R.; Richard, J.
P. J. Am. Chem. Soc. 2006, 128, 1615-1621.
(22) Crich, D.; Beckwith, A. L. J.; Chen, C.; Yao, Q.; Davison, I. G. E.;
Longmore, R. W.; Anaya de Parrodi, C.; Quintero-Cortes, L.; Sandoval-
Ramirezs, J. J. Am. Chem. Soc. 1995, 117, 8757-8768.
9
10718 J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006

Base-Selective Cleavage of Dinucleoside Monophosphates
A R T I C L E S
Scheme 2 a
a Reagents and conditions: (i) 13, NaH, DMF; (ii) 1. HCl(aq), 2. Dowex 1X2, OH- form.
Scheme 3
characterized by 1H and 13C NMR spectroscopy and high-
resolution mass spectroscopy.
then checked. In case any deviation from the initial pH was
observed, the pH was adjusted back to the original value. No
precipitation of zinc hydroxide was observed. The pH values
measured at 25 °C were extrapolated to the temperature of the
kinetic measurements (90 °C) with the aid of the known
temperature dependence of the pKa value of HEPES.27
Kinetic Measurements. The progress of the Zn2+-complex-
promoted cleavage of dinucleoside 3′,5′-monophosphates was
followed by withdrawing aliquots from the reaction solutions
at suitable intervals and analyzing their composition by RP-
HPLC. The compounds studied were UpU, UpA, ApU, and
ApA. The products were identified by spiking with authentic
compounds, viz. adenosine and uridine and their 2′,3′-cyclic
phosphates and 2′- and 3′-monophosphates. The rate constants
It is worth noting that the 1,5,9-triazacyclododecanyl groups
in ligands 1-6 are bonded to the aromatic scaffold through a
carbon atom (C3), not through a nitrogen atom as in many
previously reported constructs.5,11,19g,j All the ring nitrogen
atoms, hence, retain their ability to serve as hydrogen bond
donors, which has been suggested to play a role in binding of
azacrown chelates to nucleic acid bases.19
Preparation of the Reaction Solutions. The reaction solu-
tions were prepared as follows. A buffer solution (HEPES, 0.1
mol L-1, I ) 0.1 mol L-1 with NaNO3) having an exactly known
pH value in the pH range 6.5-8.0 (at 25 °C) was first prepared.
Appropriate amounts of the ligand and Zn(NO3)2 were then
added to give the desired final concentration for each. The
solutions were allowed to stand a few hours, and their pH was
(27) Good, N. E.; Winget, G. D.; Winter, W.; Connolly, T. N.; Izawa, S.; Singh,
R. M. M. Biochemistry 1966, 5, 467-477.
9
J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006 10719

A R T I C L E S
Wang and Lo¨nnberg
Figure 1. pH-rate profiles for the cleavage of ApU (b), UpA (2), ApA
(9), and UpU (1) (0.05 mmol L-1) in an excess of 1,2-bis[(1,5,9-
triazacyclododecan-3-yl)oxymethyl]benzene (1, 0.5 mmol L-1) and Zn2+
(1.0 mmol L-1) at 90 °C (I ) 0.1 mol L-1).
Figure 2. pH-rate profiles for the cleavage of ApU (b), UpA (2), ApA
(9), and UpU (1) (0.05 mmol L-1) in an excess of 1,3-bis[(1,5,9-
triazacyclododecan-3-yl)oxymethyl]benzene (2, 0.5 mmol L-1) and Zn2+
(1.0 mmol L-1) at 90 °C (I ) 0.1 mol L-1).
were calculated from the time-dependent diminution of the
concentration of the starting material. In each case, the 3′-linked
nucleoside was expectedly released as a 2′,3′-cyclic phosphate
and the 5′-linked nucleoside as a free alcohol (Scheme 3). The
cyclic phosphate was not, however, markedly accumulated, but
it was hydrolyzed to a 2:1 mixture of 2′- and 3′-phosphates.
The maximal mole fraction of the cyclic phosphate during a
kinetic run ranged from 0.01 to 0.05. The kinetics of the
hydrolysis of cyclic phosphate to phosphomonoesters was not
studied.
Pseudo-first-order rate constants for the cleavage of the four
dinucleoside 3′,5′-monophosphates were determined in an excess
of the cleaving agent over a narrow pH range, viz. pH 5.8-7.2
(T ) 90 °C, I ) 0.1 mol L-1 with NaNO3). The kinetic data
obtained with ligands 1-5, containing two azacrown moieties,
are presented in Figures 1-5. All these ligands exhibit a
common feature in the presence of 2 equiv of Zn2+: with ApU
and UpA, the cleavage is up to 2 orders of magnitude faster
than the cleavage promoted by the mononuclear Zn2+ chelate
of 1,5,9-triazacyclododecane ([12]aneN3]), whereas no such
acceleration is observed with ApA or UpU. Figure 6 shows
similar data for the cleavage in an excess of a trinucleating
ligand (6) and Zn2+. Now the cleavage of all uracil-containing
dimers (UpU, UpA, ApU) is markedly accelerated, while the
cleavage of ApA still remains slow. Table 1 summarizes the
rate constants at pH 6.84 (refers to 90 °C), where the cleavage
rate is maximal. Table 2 records the rate constants obtained at
two different concentrations of one of the dinucleating ligands
(3) and the trinucleating ligand (6). For comparative purposes,
the rate constants for the cleavage promoted by Zn2+[12]aneN3
and free Zn2+-aquo ion are also included.
Figure 3. pH-rate profiles for the cleavage of ApU (b), UpA (2), ApA
(9), and UpU (1) (0.05 mmol L-1) in an excess of 1,4-bis[(1,5,9-
triazacyclododecan-3-yl)oxymethyl]benzene (3, 0.5 mmol L-1) and Zn2+
(1.0 mmol L-1) at 90 °C (I ) 0.1 mol L-1), and the pH-rate profile
for the cleavage of ApU (O) at [3] ) 2.0 mmol L-1 and [Zn2+] ) 4.0
mmol L-1
.
mixture of ligand 3 and Zn2+. Comparative measurements were
carried out with ApU and ApA. The concentration of the
dinucleoside monophosphates studied was the same as that used
in kinetic experiments. The pH values at 25 °C correspond to
0.8 unit lower values at 90 °C, since the ionic product of water
at 90 °C is 1.6 logarithmic units lower than that at 25 °C.28
Figures 7 and 8 show the results obtained with UpU and ApA
UV Spectrophotometric Titrations. To elucidate the stability
of the mixed-ligand Zn2+ complex of UpU and a binucleating
azacrown ligand, a 50 µmol L-1 solution of UpU was titrated
at pH 7.5, 7.0, and 6.5 (T ) 25 °C, I ) 0.1 mol L-1) with a 1:2
(28) Marshall, W. L.; Franck, E. U. J. Phys. Chem. Ref. Data 1981, 10, 295-
304.
9
10720 J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006

Base-Selective Cleavage of Dinucleoside Monophosphates
A R T I C L E S
Figure 6. pH-rate profiles for the cleavage of ApU (b), UpA (2), ApA
(9), and UpU (1) (0.05 mmol L-1) in an excess of 1,3,5-tris[(1,5,9-
triazacyclododecan-3-yl)oxymethyl]benzene (6, 0.5 mmol L-1) and Zn2+
(1.5 mmol L-1) at 90 °C (I ) 0.1 mol L-1). In addition, the rate constant
for the cleavage of UpU at [6] ) 1.5 mmol L-1 and [Zn2+] ) 4.5 mmol
L-1 is indicated (×).
Figure 4. pH-rate profiles for the cleavage of ApU (b), UpA (2), ApA
(9), and UpU (1) (0.05 mmol L-1) in an excess of 4,4′-bis[(1,5,9-
triazacyclododecan-3-yl)oxymethyl]biphenyl (4, 0.5 mmol L-1) and Zn2+
(1.0 mmol L-1) at 90 °C (I ) 0.1 mol L-1).
Table 1. Pseudo-First-Order Rate Constants for the Cleavage of
Dinucleoside 3′,5′-Monophosphates (0.05 mmol L-1) in Excess of
Ligands 1-6 (0.5 mmol L-1) and Zn2+ (1.0 mmol L-1 with 1-5
and 1.5 mmol L-1 with 6) at pH 6.84 (90 °C, I ) 0.1 mol L-1
)
1
a
k
obs/10-3 s-
complex
ApU
UpA
ApA
UpU
(Zn2+)2-1 1.86 ( 0.03 1.59 ( 0.02 0.112 ( 0.003 0.043 ( 0.001
(Zn2+)2-2 1.38 ( 0.01 1.07 ( 0.01 0.078 ( 0.001 0.033 ( 0.001
(Zn2+)2-3 1.56 ( 0.02 0.66 ( 0.05 0.065 ( 0.002 0.049 ( 0.002
(Zn2+)2-4 0.42 ( 0.01 0.56 ( 0.01 0.063 ( 0.001 0.038 ( 0.001
(Zn2+)2-5 1.03 ( 0.01 1.13 ( 0.03 0.072 ( 0.001 0.022 ( 0.001
(Zn2+)3-6 2.03 ( 0.07 1.33 ( 0.03 0.143 ( 0.002 1.58 ( 0.02
a The errors indicated are standard deviations of the mean of single kinetic
runs.
of ApU resembles that of UpU, although the change in ab-
sorption is smaller, since only one of the base moieties (uracil)
participates in complex formation, and the plateau is reached
at [3]:[ApU] ≈ 1.6. ApA does not exhibit marked interaction
with (Zn2+)2-3.
NMR Spectroscopic Titrations. To obtain additional evi-
dence for the high stability of the mixed-ligand complex UpU-
(Zn2+)2-3, the changes in the 1H NMR spectrum of UpU upon
addition of 3 and Zn2+ (in a 1:2 molar ratio) were studied. The
H6 protons of UpU were observed to exhibit two doublets at
7.8 ppm, referring to the 3′- and 5′-linked nucleosides. When 3
and Zn2+ were added, the signals gradually disappeared, and
two new resonances at 7.5 and 7.6 ppm appeared. Upon addition
of 1 equiv of 3 and 2 equiv of Zn2+ to a 4 mmol L-1 solution
of UpU in D2O, this change was quantitative (Figure 10). For
comparison, the H2 and H8 signals of ApA did not respond on
a similar addition, suggesting that no interaction existed between
the adenine bases and the Zn2+ complex.
Figure 5. pH-rate profiles for the cleavage of ApU (b), UpA (2), ApA
(9), and UpU (1) (0.05 mmol L-1) in an excess of 2,6-bis[(1,5,9-
triazacyclododecan-3-yl)oxymethyl]pyridine (5, 0.5 mmol L-1) and Zn2+
(1.0 mmol L-1) at 90 °C (I ) 0.1 mol L-1).
at pH 7.5. The decrease in the absorbance at the wavelength of
the absorption maximum is plotted in Figure 9 as a function
of the composition of the reaction mixture. With UpU at pH
7.5, the absorption decreases linearly upon addition of ligand
3 and Zn2+ (in 1:2 molar ratio) until 1 equiv of 3 has been
added and remains constant thereafter. At pH 7.0, the break is
not as sharp, and a plateau is reached at [3]:[UpU] ≈ 1.6; at
pH 6.5, only a modest continuous decrease of absorption takes
place. The UV spectrum of ApA, in turn, remains almost
unchanged upon similar additions of 3 and Zn2+. The behavior
To estimate the stability of the binuclear Zn2+ complexes of
1
1-5, the changes in the H NMR spectrum of ligand 3 upon
addition of Zn2+ were studied in D2O at 35 °C (pD 7.1) and 90
°C (pD 6.5). The protons at C7 and C11 resonated at 1.88 ppm
(quintet). Upon addition of Zn2+, this resonance was decreased
with concomitant appearance of two new multiplets at 1.69 and
2.04 ppm. Figure 11 shows the mole fraction of the uncom-
9
J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006 10721

A R T I C L E S
Wang and Lo¨nnberg
Table 2. Comparison of the Cleavage Rate of Dinucleoside 3′,5′-Monophosphates at Two Concentrations of Dinucleating Ligand 3
and Trinucleating Ligand 6 to the Rate of the Cleavage Promoted by Zn2+[12]aneN3 and Free Zn2+-Aquo Ion at pH 6.84 (90 °C,
I ) 0.1 mol L-1
)
k
obs/10-3 s-1 a
1
complex
c/mmol L-
ApU
UpA
ApA
UpU
(Zn2+)2-3
(Zn2+)2-3
(Zn2+)3-6
(Zn2+)3-6
0.5
2.0
0.5
2.0
1.0
4.0
1.0
1.56 ( 0.02
1.35 ( 0.02
2.03 ( 0.07
2.10 ( 0.02
0.027 ( 0.001
0.067 ( 0.001
0.024b
0.66 ( 0.05
0.76 ( 0.01
1.33 ( 0.03
1.53 ( 0.02
0.060 ( 0.001
0.181 ( 0.001
0.034b
0.065 ( 0.002
0.127 ( 0.004
0.143 ( 0.002
0.30 ( 0.01
0.054 ( 0.001
0.155 ( 0.001
0.036b
0.049 ( 0.002
0.158 ( 0.002
1.58 ( 0.02
1.56 ( 0.01
0.026 ( 0.001
0.072 ( 0.01
0.030b
Zn2+[12]aneN3
Zn2+[12]aneN3
Zn2+
a The errors indicated are standard deviations of the mean of single kinetic runs. b Extrapolated from the data in ref 32 on the basis of first-order dependence
of the cleavage rate on [Zn2+] and [OH-].
Figure 7. UV spectra of UpU (50 µmol L-1 aqueous solution) at various
concentrations of 3 and Zn2+ at 25 °C (pH 7.5, I ) 0.1 mol L-1 with
NaNO3). Notation: the molar ratios of UpU, 3, and Zn2+ indicated by the
arrow refer to the curves from the top to bottom (at 260 nm).
Figure 9. Changes observed in the absorbance of 50 µmol L-1 solutions
of UpU (261 nm), ApU (260 nm), and ApA (258 nm) (T ) 25 °C, I ) 0.1
mol L-1) as a function of the molar ratio of NpN (N ) U or A), 3, and
Zn2+. The contribution of the absorbance of 3 has been subtracted.
Notation: UpU at pH 7.5 (b), UpU at pH 7.0 (2), UpU at pH 6.5 (1),
ApU at pH 7.5 (9), and ApA at pH 7.5 (O).
Figure 10. H6 NMR signals of the uracil bases of UpU (4 mmol L-1) in
the absence and in the presence of 1 equiv of 3 and 2 equiv of Zn2+ at pD
7.1 (T ) 25 °C).
mixing a dinucleating ligand (1-5, 2.0 mmol L-1, pD 7.1) with
2 equiv of Zn2+, the mole fraction of the uncomplexed azacrown
moiety at 35 °C was 0.36 with 1, 0.17 with 2, 0.45 with 3, 0.59
with 4, and 0.26 with 5. In a 1:3 mixture of ligand 6 (2.0 mmol
L-1) and Zn2+, the mole fraction of the uncomplexed azacrown
moiety was 0.24.
The data presented in Figures 9 and 10 suggest that the
complex UpU-(Zn2+)2-3 is more stable than the complex
(Zn2+)2-3. To obtain further evidence for this argument, binding
of Zn2+ to ligand 3 was studied by NMR spectroscopy in the
presence of UpU. Figure 12 shows the H7,H11 resonance of
the azacrown moieties of ligand 3 ([3] ) 2.0 mmol L-1). Upon
addition of 2 equiv of Zn2+, the H7,H11 resonance of the free
ligand was weakened and two new multiplets appeared at 1.69
and 2.04 ppm, referring to the Zn2+-complexed azacrown
Figure 8. UV spectra of ApA (50 µmol L-1 aqueous solution) at various
concentrations of 3 and Zn2+ at 25 °C (pH 7.5, I ) 0.1 mol L-1 with
NaNO3). Notation: the molar ratios of ApA, 3, and Zn2+ indicated by the
arrow refer to the curves from the bottom to top (at 258 nm).
plexed [12]aneN3 moieties as a function of the overall Zn2+
concentration at [3] ) 2.0 mmol L-1. Interestingly, the com-
plexing efficiency is independent of temperature. Evidently, the
temperature dependence of protonation of 3 is very similar to
that of complexation of Zn2+ with neutral ligand, and hence,
the Zn2+ ion competes for 3 with proton in a temperature-
independent manner. The other ligands gave similar results. On
9
10722 J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006

Base-Selective Cleavage of Dinucleoside Monophosphates
A R T I C L E S
Figure 11. Titration of 1,4-bis[(1,5,9-triazacyclododecan-3-yl)oxymethyl]-
benzene (3) with Zn2+ at 35 °C (b) and 90 °C (O). Mole fraction of
uncomplexed azacrown moieties (determined by 1H NMR) plotted against
the number of equivalents of Zn2+ compared to 3.
Figure 13. Assumed structure of the dinuclear mixed-ligand Zn2+
complexes of UpU and ligands 1-5.
mmol L-1 accelerates the cleavage of UpU 50 times as effi-
ciently as its difunctional counterpart (2) at [Zn2+] ) 1.0 mmol
L-1 (1000-fold acceleration compared to the metal-ion-inde-
pendent reaction). Since it has been shown that the dinuclear
Zn2+ complex of a closely related di(azacrown) ligand (20) binds
Figure 12. 1H NMR signal of H7,H11 protons of ligand 3 in D2O at pD
7.1 and 35 °C. (a) Ligand 3 (2.0 mmol L-1); (b) ligand 3 (2.0 mmol L-1
)
under neutral conditions at micromolar concentrations to thy-
midylyl-3′,5′-thymidine by displacement of the N3 proton of
both base moieties,19e it appears reasonable to assume that lig-
ands 1-6 in the presence of Zn2+ form a similar mixed-ligand
complex with UpU. The assumed structure of the complexes
obtained with ligands 1-5 is depicted in Figure 13. Formation
of this kind of complex does not accelerate the cleavage, since
both azacrown moieties are engaged in the base moiety binding.
The observed slow cleavage evidently is of the same origin as
the cleavage promoted by Zn2+[12]aneN3 or Zn2+-aquo ion,
involving binding of the catalytically active metal ion only to
the scissile monoanionic phosphodiester linkage.9b,c An aquo
ligand of the phosphate-bound Zn2+ then serves as an intramo-
lecular general acid, protonating the departing oxygen in concert
with the PO bond cleavage. Evidently, binding of another di-
nuclear complex to the base moieties does not prevent this reac-
tion. The trinuclear complex of ligand 6, in turn, undergoes a
high-affinity pre-equilibrium binding to the substrate. Two of
the azacrown moieties anchor as Zn2+ chelates the cleaving
agent to the base moieties of UpU, while the third participates
as an intracomplex catalyst in the cleavage of the phosphodiester
linkage. The mechanism of the actual cleavage step is, in all
likelihood, similar to that described above, viz. a pre-equilibrium
formation of a dianionic phosphorane intermediate followed by
rate-limiting proton transfer from an aquo ligand of the phos-
+ Zn2+ (4.0 mmol L-1); and (c) ligand 3 (2.0 mmol L-1) + Zn2+ (4.0
mmol L-1) + UpU (2.0 mmol L-1).
moieties. However, 45% of the azacrown moieties remained
uncomplexed. Addition of 1 equiv of UpU then resulted in com-
plete disappearance of the signal of uncomplexed [12]aneN3.
Discussion
Cleavage of UpU. As seen from Table 2, the rate constant
for the cleavage of UpU by the monomeric Zn2+ complex of
[12]aneN3 is 2.6 × 10-5 s-1 under the experimental conditions
of the present work ([Zn2+] ) 1.0 mmol L-1, T ) 90 °C, pH
6.84, I ) 0.1 mol L-1),20,29 representing a 20-fold rate
acceleration compared to the metal-ion-independent cleavage.30
The cleavage rate obtained in this work with bifunctional ligands
1-5 at the same Zn2+ concentration is only slightly higher
(Table 1). Accordingly, incorporation of two [12]aneN3 moieties
into a single ligand structure does not result in any marked
change in the cleavage rate. However, the situation is dramati-
cally changed upon addition of a third [12]aneN3 moiety. The
trifunctional ligand 6 at the overall Zn2+ concentration of 1.5
(29) Kuusela, S.; Lo¨nnberg, H. J. Phys. Org. Chem. 1993, 6, 347-356.
(30) Ja¨rvinen, P.; Oivanen, M.; Lo¨nnberg, H. J. Org. Chem. 1991, 56, 5396-
5401.
9
J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006 10723

A R T I C L E S
Wang and Lo¨nnberg
According to 1H NMR spectroscopic measurements, the
dinuclear complex (Zn2+)2-3 is considerably less stable than the
complex UpU-(Zn2+)2-3. While addition of 1 equiv of ligand 3
and 2 equiv of Zn2+ to a solution of UpU (4 mmol L-1, pD
7.1) is sufficient to convert UpU entirely to UpU-(Zn2+)2-3, as
evidenced by complete disappearance of the H6 resonances at
7.8 ppm and appearance of new signals at 7.4 and 7.5 ppm
(Figure 10), addition of 2 equiv of Zn2+ to a solution of 3 (2
mmol L-1, pD 7.1) still leaves 40% of the azacrown moieties
uncomplexed (Figure 11). Interestingly, the situation is very
similar at both 35 and 90 °C. Addition of 1 equiv of UpU to
this mixture, however, results in quantitative binding of Zn2+
to ligand 3 (Figure 12).
Since the kinetic experiments of UpU cleavage have been
carried out in excess of the ligand and Zn2+ ([6]:[Zn2+] 1:3),
the reaction solutions contain uncomplexed Zn2+ in addition to
UpU-(Zn2+)3-6. As mentioned above, the first-order rate constant
for the cleavage of UpU by Zn2+[12]aneN3 is 2.6 × 10-5 s-1
under the experimental conditions of the present work ([Zn2+
]
) 1 mmol L-1, pH 6.84). Catalysis by Zn2+-aquo ion has
previously been studied at pH < 6 and shown to be first-order
in hydroxide ion concentration.33 Extrapolation to pH 6.84 shows
the rate constant for the Zn2+-promoted cleavage to be 3 × 10-5
s-1 at [Zn2+] ) 1.0 mmol L-1. In the pH range used in the
present study, both reactions are approximately first-order in
hydroxide ion concentration. Since the rate constant for the
cleavage reaction promoted by the Zn2+ complex of 6 varies
from 3.5 × 10-4 s-1 at pH 5.8 to 1.6 × 10-3 s-1 at pH 6.8, the
contribution of uncomplexed Zn2+ ion to the observed rate
constants may be neglected. Consistent with this argument, the
rate constants obtained at pH 6.84 for the cleavage of UpU at
0.5 and 1.5 mmol L-1 concentrations of 6 ([Zn2+] ) 1.5 and
4.5 mmol L-1, respectively) are equal within the limits of
experimental errors: (1.58 ( 0.02) × 10-3 and (1.56 ( 0.03)
× 10-3 s-1 (Table 2). Evidently, at this pH, UpU is quantita-
tively engaged in the complex UpU-(Zn2+)3-6. The pH-rate
profile is, hence, quite similar to that reported previously for
the Zn2+[12]aneN3,28 leveling off to constant value on passing
the pKa value of the chelated Zn2+-aquo ion (with Zn2+[12]-
Figure 14. Assumed transition-state structure for the cleavage of the
trinuclear mixed-ligand Zn2+ complex of UpU and ligand 6.
phorane-bound metal ion to the departing 5′-oxyanion. Figure
14 shows the assumed transition-state structure for the cleavage
reaction.
1
The UV and H NMR spectroscopic studies on a ternary
system consisting of UpU, ligand 3, and Zn2+ verify that the
complex UpU-(Zn2+)2-3 (cf. Figures 10 and 12) really is stable
at millimolar concentrations. The 1H NMR spectroscopic
measurements show that both of the uracil bases are involved
in metal ion binding: upon addition of 3 and Zn2+ ([3]:[Zn2+
]
1:2), the H6 proton doublets at 7.8 ppm disappear and new
signals at 7.5 and 7.6 ppm appear. For comparison, the H2 and
H8 signals of ApA do not respond to addition of 3 and Zn2+
.
UV spectrophotometric titrations, in turn, allow estimation of
the stability of the UpU-(Zn2+)2-3 complex. The UV absorption
of a 50 µmol L-1 solution of UpU (pH 7.5, T ) 25 °C, I ) 0.1
mol L-1) decreases linearly upon addition of ligand 3 and Zn2+
([3]:[Zn2+] 1:2) until the molar ratio of UpU, 3, and Zn2+
reaches the value 1:1:2, and remains constant thereafter (Figure
9). At pH 7.0, 3 has to be present in 70% excess to reach the
plateau value (at a molar ratio of 1:1.6:3.2). On the basis of the
latter observation, the complex UpU-(Zn2+)2-3 may be estimated
to be half dissociated when 3 and Zn2+ are present in 15 and
30 µmol L-1 excess, respectively, at pH 7.0, referring to pH
6.2 at 90 °C. For comparison, the dissociation constant for the
complex TpT-(Zn2+)2-20 has been reported to be 0.6 µmol L-1
at pH 7.3.19e On going to lower pH, the stability of the complex
UpU-(Zn2+)2-3 is dramatically decreased: at pH 6.5 (5.84 at
90 °C), the absorption experiences only a modest continuous
decrease upon addition of 3 and Zn2+. This is expected since
the Zn2+ ions have to compete with protons for both the uracil
bases and the azacrown ligands. The pKa value of a uracil moiety
in a dinucleoside 3′,5′-monophosphate is 9.0 (25 °C, I ) 0.1
mol L-1),31 and the pKa values of ligand 3 expectedly are
somewhat lower than those reported for [12]aneN3 (12.6 and
7.6 at 25 °C, I ) 0.1 mol L-1).32
aneN3, 7.5 at 25 °C, I ) 0.1 mol L-1 31
) that participates in the
catalysis. By contrast, the rate constants obtained with ligands
1-5 undoubtedly contain contributions of uncomplexed Zn2+
ion and various Zn2+ complexes; hence, the pH-rate profiles
cannot be analyzed on the basis any single reaction.
Cleavage of ApU and UpA. In striking contrast to UpU,
dinucleoside 3′,5′-monophosphates containing only one uracil
base, viz. ApU and UpA, are cleaved by the Zn2+ complexes
of the dinucleating ligands (1-5) up to 2 orders of magnitude
more rapidly than by Zn2+[12]aneN3. Interestingly, a similar
100-fold acceleration has been obtained by tethering [12]aneN3
to a 2′-O-methylribo oligonucleotide complementary to the
target oligoribonucleotide.34 Evidently, a 100-fold acceleration
may be attributed to a proximity effect, i.e., to increased
concentration of the cleaving agent in the vicinity of the scissile
bond. On using an oligonucleotide conjugate, the proximity
effect is achieved by hybridization. On using (Zn2+)2-1, a similar
effect appears to be obtained by anchoring one of the Zn2+
chelate moieties of the cleaving agent to the uracil base. The
(31) Ogasawara, N.; Inoue, Y. J. Am. Chem. Soc. 1976, 98, 7048-7053.
(32) Zompa, L. J. Inorg. Chem. 1978, 2531-2536.
(33) Kuusela, S.; Lo¨nnberg, H. Nucleosides Nucleotides 1996, 15, 1669-1678.
(34) Niittyma¨ki, T.; Lo¨nnberg, H. Bioconjugate Chem. 2004, 15, 1275-1280.
9
10724 J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006

Base-Selective Cleavage of Dinucleoside Monophosphates
A R T I C L E S
Cleavage of ApA. ApA is cleaved by the Zn2+ complexes
of 1-6 only slightly more rapidly than by Zn2+[12]aneN3 (Table
2). With ApA, the only feasible interaction between the cleaving
agent and the base moieties is a stacking interaction, and this
interaction is too weak to result in a marked proximity effect.
While with ApU and UpA the cleavage rate is independent of
the concentration of (Zn2+)2-3, consistent with quantitative
formation of the catalytically active mixed-ligand complex,
the cleavage rate of ApA expectedly depends on the catalyst
concentration, analogously to Zn2+[12]aneN3-promoted cleav-
age. The somewhat higher catalytic activity of the Zn2+ complex
of 6 may be attributed to the fact that, with this ligand, the
overall Zn2+ concentration is 1.5 mmol L-1, compared to the
1.0 mmol L-1 concentration used with the other ligands. It is
also worth noting that, since a stable complex between ApA,
Zn2+, and the azacrown-derived ligand is not formed and the
ligands employed do not bind Zn2+ quantitatively over the entire
pH range studied, the rate constants most likely include
contributions of various metal-ion-containing species.
Dependence of the Cleavage Rate on pH and the Ligand
Structure. As discussed above, the pH-rate profiles for the
cleavage of ApU and UpA with 1-6 and UpU with 6 resemble
the profile reported previously for the Zn2+[12]aneN3-promoted
cleavage.28 In other words, the observed rate constant levels
off to a constant value on passing the pKa value of the chelated
Zn2+-aquo ion (with Zn2+[12]aneN3, 7.5 at 25 °C, I ) 0.1
Figure 15. Assumed transition-state structure for the cleavage of ApU by
dinuclear Zn2+ complexes of ligands 1-5.
remaining Zn2+ chelate moiety is then able to bind to the
phosphodiester linkage and facilitate the departure of the 5′-
linked nucleoside as an intracomplex general acid. A hypotheti-
cal structure of the transition state is depicted in Figure 15.
Evidently, binding to a single uracil moiety is sufficiently strong
to ensure quantitative complex formation in an excess of a
dinucleating ligand and Zn2+, since the pH-rate profiles (Figure
mol L-1 31
that participates in the catalysis. At pH less than
)
the pKa of the chelated Zn2+-aquo ion, one proton is lost on
going from the initial state to the transition state indicated in
Figure 13, since the attacking 2′-OH has to be deprotonated to
obtain the dianionic phosphorane. At pH greater than the pKa
of the chelated Zn2+-aquo ion, the initial state and transition
state are tautomeric structures: in the initial state, 2′-OH bears
a proton and one of the Zn2+ ligands is deprotonated, whereas
on going to the transition state, 2′-OH undergoes deprotonation
and the Zn2+-hydroxo ligand protonation. Accordingly, rate
equation (1) is obeyed. Here, Ka is the acidity constant of the
3) obtained with ApU at [3] ) 0.5 and 2.0 mmol L-1 ([Zn2+
]
) 1.0 and 4.0 mmol L-1, respectively) are almost identical
(Figure 3). Only at pH 5.8, a moderate rate acceleration takes
place upon increasing the concentration of 3. Evidently, the pre-
equilibrium anchoring is not any more quantitative at such a
low pH when the concentration of 3 is 0.5 mmol L-1. The
situation is similar with UpA: only a slight rate acceleration
with increasing catalyst concentration is observed (Table 2).
The cleavage by uncomplexed Zn2+ or mononuclear Zn2+
complexes is too slow to compete with the intracomplex reaction
around neutrality (Table 2). Accordingly, the pH-rate profiles
again resemble those obtained with Zn2+[12]aneN3. The tri-
nucleating ligand (6) cleaves ApU and UpA approximately as
efficiently as UpU. Most likely, one of the azacrown moieties
is engaged in base moiety binding, and only one of the
remaining coordination sites binds Zn2+. The latter assumption
is based on the observation that, at pH 7.5 (6.84 at 90 °C), the
Zn2+-azacrown complex is considerably less stable than the
mixed-ligand complex with the uracil base and azacrown.
Ka
kobs ) k
(1)
Ka + [H+]
catalytically active chelated Zn2+-aquo ion and k is the first-
order rate constant for the cleavage of the complexed dinucleo-
side monophosphate. Figures 1-6 show the curves obtained
by least-squares fitting for the cleavage of ApU, UpA, and UpU
(only on using 6 as a cleaving agent). While this equation in
many cases fits excellently to the experimental rate constants,
some of the curves tend to exhibit more marked curvature than
predicted by eq 1. Evidently, the complex formation is not
always quantitative over the whole pH range studied. The pKa
values obtained for the chelated Zn2+-aquo ion, serving as an
intracomplex catalyst, and the first-order rate constants, k, for
the cleavage of the fully complexed dinucleoside monophos-
phates are given in Table 3. The cleavage of UpU and ApA is
not markedly accelerated by the dinucleating ligands (1-5), and
the rate constants obtained undoubtedly contain contributions
of uncomplexed Zn2+ ion and various Zn2+ complexes; hence,
the pH-rate profiles cannot be analyzed on the basis any single
reaction.
The argument that one uracil base is sufficient to anchor the
cleaving agent to ApU receives support from the UV spectro-
photometric titrations. At pH 7.5, 2 equiv of ligand 3 and 4
equiv of Zn2+ are sufficient to convert ApU at a concentration
of 50 µmol L-1 to ApU-(Zn2+)n-3. The complex is half-
dissociated when 3 and Zn2+ are present in 15 and 30 µmol
L-1 excess, respectively. Whether the complex contains one or
two Zn2+ ions cannot be deduced from the spectrophotometric
data. Obviously, only the dinuclear complex can be catalytically
active.
As seen from Table 1, the cleaving activities of all five
dinuclear complexes (1-5) studied are surprisingly similar. Only
9
J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006 10725

A R T I C L E S
Wang and Lo¨nnberg
Table 3. Kinetic Parameters for the Cleavage of ApU, UpA, and
UpU When Engaged in a Mixed-Ligand Complex NpN-(Zn2+)2-L (L
) 1-5) or NpN-(Zn2+)3-6a
chloride (0.233 mol, 29.5 g), and powdered KOH (0.243 mol, 13.62
g) in dry benzene (600 mL) was heated in a Dean-Stark apparatus
until the separation of water was complete (12 h), as described
previously for the preparation of closely related compounds.36 The
solution was cooled to room temperature, washed with water, 5%
NaHCO3 aqueous solution, and water, and then dried with Na2SO4.
Evaporation of the solvent under reduced pressure gave a yellow solid,
which was further purified on a silica gel column, eluting first with
neat dichloromethane (DCM) and then with a 5:95 (v/v) mixture of
MeOH and DCM. Although the starting material (7) was a pure cis
isomer, compound 8 was obtained as a mixture of cis and trans isomers.
Facile isomerization of this compound in CDCl3 has been previously
reported.37 The overall yield was 99% (31.0 g). 1H NMR (CDCl3), trans
nucleoside
1
monophosphate
ligand
pKa
k/10-3 s-
ApU
1
2
3
6.7 ( 0.3
6.8 ( 0.1
6.8 ( 0.2
6.5 ( 0.1
6.8 ( 0.2
6.9 ( 0.1
6.1 ( 0.2
2.8 ( 0.9
2.7 ( 0.5
2.7 ( 1.1b
1.9 ( 0.1c
0.77 ( 0.02
2.1 ( 0.5
2.4 ( 0.2
4
5
6
UpA
UpU
1
2
3
4
5
6
6.6 ( 0.4
6.8 ( 0.1
6.6 ( 0.2
6.6 ( 0.2
6.9 ( 0.2
6.1 ( 0.2
2.3 ( 0.5
2.2 ( 0.3
1.0 ( 0.2
1.0 ( 0.4
2.6 ( 0.7
1.6 ( 0.2
isomer δ 7.41 (m, 10H), 5.43 (s, 1H), 4.64 (s, 2H), 4.39 (dd, J4eq5ax
J6eq5ax ) 5.0 Hz, J4eq4ax ) J6eq6ax ) 11.1 Hz, 2H), 3.83 (tt, J4eq5ax
)
)
J6eq5ax ) 5.0 Hz, J4ax5ax ) J6ax5ax ) 10.1 Hz, 1H), 3.69 (t, J ) 11.0 Hz,
2H); cis isomer δ 7.44 (m, 10H), 5.59 (s, 1H), 4.74 (s, 2H), 4.41 (dd,
J4eq5eq ) J6eq5eq ) 1.3 Hz, J4eq4ax ) J6eq6ax ) 12.6 Hz, 2H), 4.08 (dd,
J4ax5eq ) J6ax5eq ) 1.7 Hz, J4eq4ax ) J6eq6ax ) 12.6 Hz, 2H), 3.38 (quint.,
J ) 1.7 Hz, 1H). 13C NMR (CDCl3): trans isomer δ 137.89, 137.64,
129.01, 128.59, 128.31, 128.08, 127.79, 126.06, 101.31, 71.87, 70.23,
67.93; cis isomer δ 138.14, 138.11, 128.93, 128.47, 128.22, 127.78,
127.73, 126.22, 101.42, 70.32, 69.20, 69.04.
6
6.3 ( 0.2
1.7 ( 0.4
a pKa values refer to the chelated Zn2+-aquo ion serving as an
intracomplex catalyst and the first-order rate constants, k, to the cleavage
°
of a fully complexed dinucleoside monophosphate at 90 C (I ) 0.1 mol
L-1). b At [3] ) 0.5 mmol L-1 and [Zn2+] ) 1.0 mmol L-1 c At [3] ) 2.0
.
mmol L-1 and [Zn2+] ) 4.0 mmol L-1
.
2-(Benzyloxy)propane-1,3-diol (9). 5-(Benzyloxy)-2-phenyl-1,3-
dioxane (8) (55.6 mmol, 15.0 g) was dissolved in MeOH (300 mL),
and 1 mol L-1 HCl (50 mL) was added. The mixture was heated for 1
h under reflux. All volatiles were removed under reduced pressure,
and the residue was subjected to a similar treatment. Compound 9 was
obtained as a colorless oil in quantitative yield (10.1 g) and used without
the 4,4′-biphenyl-bridged complex (Zn2+)2-4 is somewhat less
efficient than the others, the cleaving activity toward ApU and
UpA being 22% and 35% of that of the most efficient cleaving
agent, the o-phenylene-bridged complex (Zn2+)2-1. The p-
phenylene-bridged complex (Zn2+)2-3 is, in turn, the only one
that exhibits marked selectivity between ApU and UpA, the
former being cleaved almost 3 times as fast as the latter. In
other words, the proximity effect is rather insensitive to the
structure of the scaffold that links the two nucleating moieties
to each other. Evidently, the freely rotating -CH2O- bridge
between the aromatic nucleus and the azacrown moieties allows
the cleaving agent to find a conformation where simultaneous
binding to the base and phosphate moieties is possible. Both of
the latter interactions are of electrostatic origin and, hence, as
long-range interactions are able to guide the cleaving agents to
a proper conformation.
1
purification for the next reaction. H NMR (CDCl3): δ 7.37 (m, 5H),
4.69 (s, 2H), 3.83 (dd, J ) 4.4, 11.7 Hz, 2H), 3.76 (dd, J ) 5.0, 11.7
Hz, 2H), 3.64 (quint., J ) 4.7 Hz, 1H), 2.02 (br s, 2H). 13C NMR
(CDCl3): δ 137.99, 128.64, 128.06, 127.86, 79.18, 72.02, 62.39.
2-(Benzyloxy)propane-1,3-diyl Bis(4-methylbenzenesulfonate) (10).
A mixture of 2-(benzyloxy)propane-1,3-diol (9) (55.6 mmol, 10.1 g)
in THF (250 mL) and aqueous sodium hydroxide (100 mL of a 2.2
mol L-1 solution) was placed in a flask equipped with a dropping funnel
and a thermometer. The mixture was cooled on an ice-salt bath with
vigorous stirring, and p-toluenesulfonyl chloride (122 mmol, 23.4 g)
in THF (150 mL) was added dropwise during 1 h, as described in the
literature for preparation of similar tosylates.38 The temperature was
kept between 3 and 5 °C during the addition. Stirring at 3-5 °C was
continued, and the progress of the reaction was monitored by TLC
every 10-15 min. Upon completion of the reaction, concentrated
aqueous HCl was added to quench the reaction. THF was removed
under reduced pressure, and the residue was extracted with DCM (5 ×
60 mL). All extracts were combined, dried with Na2SO4, and evaporated
to dryness under reduced pressure. Recrystallization from ethanol gave
the pure product in 85% yield (23.2 g). 1H NMR (CDCl3): δ 7.75 (d,
J ) 8.3 Hz, 4H), 7.33 (m, 7H), 7.20 (m, 2H), 4.51 (s, 2H), 4.06 (m, J
) 5.0, 10.7 Hz), 3.82 (quint., J ) 5.0 Hz, 1H), 2.47 (s, 6H). 13C NMR
(CDCl3): δ 145.17, 136.97, 132.38, 129.97, 128.46, 128.06, 127.98,
127.80, 73.83, 72.53, 67.71, 21.71.
Conclusions
Dinuclear complexes containing two Zn2+[12]aneN3 moieties
attached to an aromatic core show potential as base-selective
cleaving agents of RNA phosphodiester bonds. With dinucleo-
side 3′,5′-monophosphates containing only one uracil base, one
of the Zn2+[12]aneN3 moieties anchors the cleaving agent to
the substrate, while the other Zn2+ chelate moiety serves as an
intracomplex catalyst for the cleavage of the neighboring
phosphodiester bond. A 100-fold rate acceleration, compared
to that obtained with a monomeric Zn2+[12]aneN3, is achieved
at 0.5 mmol L-1 concentration of the cleaving agent. Substrates
2-(Benzyloxy)hexahydro-3a,6a,9a-triazaphenalene (11). A mixture
of ditosylate 10 (20.4 mmol, 10 g), 1,5,7-triazabicyclo[4.4.0]dec-5-
ene (TBD, 20.4 mmol, 2.84 g), and powdered KOH (82 mmol, 4.6 g)
in dry 1,4-dioxane (300 mL) was added to a flask. The mixture was
heated for 48 h under reflux. The completeness of the reaction was
verified by LC/ESI-MS (m/z 286 [M]+). The mixture was then cooled
containing two uracil bases are not cleaved, since both Zn2+
-
[12]aneN3 moieties are engaged in uracil binding. The catalytic
activity is, however, restored by addition of a third azacrown
group on the cleaving agent.
Experimental Section
General. All reagents employed were of reagent grade, and they
were used without further purification. All nucleotides and nucleosides
were products of Sigma. They were used as received after checking
the purity by HPLC.
cis- and trans-5-(Benzyloxy)-2-phenyl-1,3-dioxane (8). A mixture
of cis-2-phenyl-1,3-dioxan-5-ol (7)35 (0.113 mol, 20.95 g), benzyl
(35) Crich, D.; Beckwith, A. L. J.; Chen, C.; Yao, Q.; Davison, I. G. E.;
Longmore, R. W.; Anaya de Parrodi, C.; Quintero-Cortes, L.; Sandoval-
Ramirezs, J. J. Am. Chem. Soc. 1995, 117, 8757-8768.
(36) Weber, E. J. Org. Chem. 1982, 47, 3478-3486.
(37) Peng, S.; Wang, D.; Huang, H.; Cai, M. Chinese J. Magn. Reson. 1988, 5,
31-36.
(38) Ouchi, M.; Inoue, Y.; Liu, Y.; Naganune, S.; Nakamura, S.; Wada, K.;
Hakushi, T. Bull. Chem. Soc. Jpn. 1990, 63, 1260-1262.
9
10726 J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006

Base-Selective Cleavage of Dinucleoside Monophosphates
A R T I C L E S
to room temperature, and sodium borohydride (0.10 mol, 3.8 g) was
slowly added. Stirring was continued for 48 h at room temperature,
and the formation of 11 was checked by LC/ESI-MS (m/z 288 [M +
H]+ and 310 [M + Na]+). The solvent was removed under reduced
pressure. Water (100 mL) was added to the residue, and the aqueous
phase was extracted with DCM (5 × 50 mL). The combined DCM
extracts were dried with Na2SO4 and evaporated to dryness under
reduced pressure. The crude product was used for the next step without
purification.
4H), 4.60 (s, 4H), 3.86 (m, 2H), 3.38 (m, 24H), 2.01 (m, 4H), 1.76 (m,
4H), 1.46 (s, 36H), 1.44 (s, 18H). 13C NMR (CDCl3, 50 °C): δ 156.20,
155.94, 138.67, 128.28, 126.78, 126.55, 79.77, 79.53, 75.68, 71.78,
49.70, 46.89, 44.98, 29.40, 28.40. HR-MS for (MH - C5H8O2)+,
C56H96N6O14: requires 977.6538, found 977.6503.
Hexa-tert-butyl 3,3′-[1,4-Phenylenebis(methylene)bis(oxy)]bis-
(1,5,9-triazacyclododecane-1,5,9-tricarboxylate) (16). 1,4-Bis(bro-
momethyl)benzene (0.72 mmol, 0.19 g) was used. Compound 16 was
1
obtained in 85% yield (0.66 g). H NMR (CDCl3, 50 °C): δ 7.28 (s,
4H), 4.61 (s, 4H), 3.88 (m, 2H), 3.41 (m, 24H), 2.04 (m, 4H), 1.79 (m,
4H), 1.48 (s, 36H), 1.47 (s, 18H). 13C NMR (CDCl3, 50 °C): δ 156.25,
155.97, 137.93, 127.49, 79.82, 79.59, 75.67, 71.72, 49.69, 47.00, 45.01,
29.45, 28.43, 28.42. HR-MS for (MH - C5H8O2)+, C56H96N6O14:
requires 977.6538, found 977.6520.
1,5,9-Triazacyclododecan-3-ol (12). The crude product of 11 was
heated for 4 d in 4 mol L-1 aqueous HCl (600 mL) under reflux and
evaporated to dryness under reduced pressure. Addition of MeOH (80
mL) to the residue gave 12 as a pure trihydrochloride in 93% yield
1
(5.63 g) (calculated by comparison to 10). H NMR (D2O): δ 4.25
(quint., J ) 5.3 Hz, 1H), 3.37 (m, 12H), 2.16-2.11 (quint., J )
6.6 Hz, 4H). 13C NMR (D2O): δ 61.65, 47.93, 43.20, 41.00, 19.24.
HR-MS for C9H21N3O: requires 187.1685, found 187.1683.
Hexa-tert-butyl 3,3′-[Biphenyl-4,4′-diylbis(methylene)bis(oxy)]bis-
(1,5,9-triazacyclododecane-1,5,9-tricarboxylate) (17). 4,4′-Bis(chlo-
romethyl)biphenyl (0.72 mmol, 0.18 g) was used. Compound 17 was
obtained in 82% yield (0.68 g). 1H NMR (CDCl3): δ 7.51 (d, J ) 8.1
Hz, 4H), 7.37 (d, J ) 8.1 Hz, 4H), 4.64 (s, 4H), 3.87 (quint., J ) 5.9
Hz, 2H), 3.60 (dd, J ) 5.9, 14.9 Hz, 4H), 3.40 (m, 12H), 3.16 (m,
8H), 2.01 (m, 4H), 1.74 (m, 4H), 1.46 (s, 36H), 1.42 (s, 18H). 13C
NMR (CDCl3): δ 156.29, 156.01, 140.28, 137.50, 128.11, 127.01,
79.91, 79.66, 75.30, 71.55, 49.50, 47.05, 44.93, 29.46, 28.43. HR-MS
for (MH - C5H8O2)+, C62H100N6O14: requires 1053.6851, found
1053.6873.
Tri-tert-butyl 3-Hydroxy-1,5,9-triazacyclododecane-1,5,9-tricar-
boxylate (13). 1,5,9-Triazacyclododecan-3-ol trihydrochloride (19.0
mmol, 5.63 g) was dissolved in a mixture of tert-butyl alcohol (120
mL) and aqueous sodium hydroxide (80 mL of 2.9 mol L-1 solution).
Di-tert-butyl dicarbonate (Boc2O, 62.8 mmol, 13.7 g) in tert-butyl
alcohol (50 mL) was added to this solution with vigorous stirring. The
stirring was continued at room temperature overnight and then at
40-50 °C for 2 d. Formation of 13 was followed by LC/ESI-MS (m/z
488 [M + H]+, 510 [M + Na]+). Because the reactivity of the hydroxyl
group turned out to be only moderately lower than that of the ring
nitrogen atoms, Boc2O could not be used in excess, and still ester
formation could not be entirely avoided. The 2 days of incubation at
40-50 °C, however, resulted in isomerization to the desired product.
The solvent was removed under reduced pressure, water (50 mL) was
added, and the product was extracted to DCM (5 × 40 mL). The
combined extracts were dried with Na2SO4 and evaporated to dryness.
The residue was purified by column chromatography (silica gel, 5%
Hexa-tert-butyl 3,3′-[Pyridine-2,6-diylbis(methylene)]bis(oxy)-
bis(1,5,9-triazacyclododecane-1,5,9-tricarboxylate) (18). 2,6-Bis-
(tosyloxymethyl)pyridine24 (0.72 mmol, 0.32 g) was used. Compound
1
18 was obtained in 76% yield (0.59 g). H NMR (CDCl3, 50 °C): δ
7.56 (t, J ) 7.7 Hz, 1H), 7.26 (d, J ) 7.7 Hz, 2H), 4.60 (s, 4H), 3.84
(quint., J ) 5.8 Hz, 2H), 3.54 (dd, J ) 6.2, 14.9 Hz, 4H), 3.35 (m,
12H), 1.94 (m, 4H), 1.70 (m, 4H), 1.36 (s, 36H), 1.35 (18H). 13C NMR
(CDCl3, 50 °C): δ 157.76, 156.10, 155.84, 137.08, 119.84, 79.72, 79.45,
76.43, 72.59, 49.30, 46.96, 44.76, 29.39, 28.34, 28.31. HR-MS for
(C55H95N7O14 + H)+: requires 1078.7015, found 1078.6977.
1
MeOH in DCM). Yield: 7.7 g (83.2%). H NMR (CDCl3, 50 °C): δ
4.04 (quint., J ) 5.8 Hz, 1H), 3.44 (m, 12H), 1.85 (quint., J ) 6.7 Hz,
4H), 1.50 (s, 18H), 1.48 (s, 9H). 13C NMR (CDCl3, 50 °C): δ 155.77,
80.65, 79.55, 73.55, 53.56, 47.72, 43.43, 28.44, 28.36. HR-MS for
(C24H45N3O7 + H)+: requires 488.3336, found 488.3328.
Nona-tert-butyl 3,3′,3′′-[Benzene-1,3,5-triyltris(methylene)]tris-
(oxy)tris(1,5,9-triazacyclododecane-1,5,9-tricarboxylate) (19). 1,3,5-
Tris(bromomethyl)benzene25 (0.48 mmol, 0.17 g) was used. Compound
1
19 was obtained in 80% yield (0.56 g). H NMR (CDCl3, 50 °C): δ
7.07 (s, 3H), 4.49 (s, 6H), 3.77 (quint., J ) 5.9 Hz, 3H), 3.49 (dd,
J ) 6.1, 14.9 Hz, 6H), 3.32 (m, 18H), 3.07 (m, 12H), 1.92 (m, 6H),
1.68 (s, 6H), 1.35 (s, 54H), 1.34 (s, 27H). 13C NMR (CDCl3, 50 °C):
δ 156.05, 155.80, 138.78, 125.69, 79.61, 79.36, 75.79, 71.71, 49.62,
46.85, 44.82, 29.36, 28.34, 28.32. HR-MS for (MH - C5H8O2)+,
C81H141N9O21: requires 1476.9796, found 1476.9840.
General Procedure for Preparation of Ligands 1-6. Compounds
14-19 were dissolved in MeOH (30 mL). Concentrated aqueous HCl
(0.8 mL) was added. The mixture was incubated at 40-45 °C for 2 h.
Upon the solution cooling to 0 °C, the deprotected product precipitated
as a hydrochloride. The hydrochlorides were converted to free bases
by passing their aqueous solutions through a strong anion-exchange
resin (Dowex 1X2, OH- form).
General Procedure for Preparation of the Boc-Protected Ligands
(14-19). R,R′-Bis(halomethyl)arene (0.72 mmol, to obtain 1-4),
pyridine-2,6-diyl ditosylate (0.72 mmol, to obtain 5), or 1,3,5-tris-
(bromomethyl)benzene (0.48 mmol, to obtain 6) was dissolved in dry
DMF (20 mL), and tri-tert-butyl 1,5,9-triazacyclododecan-3-ol-1,5,9-
tricarboxylate (13) (1.44 mmol, 0.7 g) was added. Dry sodium hydride
(2.1 mmol, 50 mg) was added, and the mixture was stirred for 1 h at
room temperature. The unreacted sodium hydride was destroyed with
MeOH (2 mL), and the volatiles were removed under reduced pressure.
The residue was dissolved in water (40 mL) and extracted with DCM
(5 × 20 mL). The combined extracts were dried with Na2SO4 and
evaporated to dryness under reduced pressure. The product was purified
by column chromatography (silica gel, 25-40% ethyl acetate in DCM).
Hexa-tert-butyl 3,3′-[1,2-Phenylenebis(methylene)bis(oxy)]bis-
(1,5,9-triazacyclododecane-1,5,9-tricarboxylate) (14). 1,2-Bis(bro-
momethyl)benzene (0.72 mmol, 0.19 g) was used as a starting material.
1,2-Bis[(1,5,9-triazacyclododecan-3-yloxy)methyl]benzene Hexahy-
1
drochloride (1). Yield: 94%. H NMR (D2O): δ 7.42 (m, 2H), 7.34
(m, 2H), 4.73 (s, 4H), 4.04 (m, 2H), 3.23 (m, 24H), 2.03 (m, 8H). 13
C
1
NMR (D2O): δ 135.13, 129.38, 128.90, 70.26, 68.80, 47.06, 44.05,
41.63, 19.97. HR-MS for (C26H48N6O2 + H)+: requires 477.3917, found
477.3892.
Compound 14 was obtained in 76% yield (0.59 g). H NMR (CDCl3,
50 °C): δ 7.37 (m, 2H), 7.24 (m, 2H), 4.67 (s, 4H), 3.84 (quint., J )
4.9 Hz, 2H), 3.59 (dd, J ) 6.1, 14.8 Hz, 4H), 3.42 (m, 12H), 3.11 (m,
8H), 2.01 (m, 4H), 1.74 (m, 4H), 1.45 (s, 54H). 13C NMR (CDCl3, 50
°C): δ 156.22, 155.95, 136.42, 128.64, 127.59, 79.79, 79.54, 75.33,
69.13, 49.53, 47.02, 44.79, 29.43, 28.43, 28.40. HR-MS for (MH -
C5H8O2)+, C56H96N6O14: requires 977.6538, found 977.6503.
Hexa-tert-butyl 3,3′-[1,3-Phenylenebis(methylene)bis(oxy)]bis-
(1,5,9-triazacyclododecane-1,5,9-tricarboxylate) (15). 1,3-Bis(bro-
momethyl)benzene (0.72 mmol, 0.19 g) was used. Compound 15 was
obtained in 83% yield (0.64 g). 1H NMR (CDCl3, 50 °C): δ 7.24 (m,
1,3-Bis[(1,5,9-triazacyclododecan-3-yloxy)methyl]benzene Hexahy-
drochloride (2). Yield: 92%. 1H NMR (D2O, 50 °C): δ 7.68 (m, 4H),
4.93 (s, 4H), 4.31 (m, 2H), 3.68 (dd, J ) 5.4, 13.8 Hz, 4H), 3.51 (m,
16H), 3.35 (m, 4H), 2.33 (m, 8H). 13C NMR (D2O, 50 °C): δ 138.05,
129.77, 128.69, 128.45, 71.75, 71.37, 48.94, 45.41, 43.50, 21.52. HR-
MS for (C26H48N6O2 + H)+: requires 477.3917, found 477.3897.
1,4-Bis[(1,5,9-triazacyclododecan-3-yloxy)methyl]benzene Hexahy-
drochloride (3). Yield: 96%. 1H NMR (D2O, 50 °C): δ 7.70 (s, 4H),
9
J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006 10727

A R T I C L E S
Wang and Lo¨nnberg
4.93 (s, 4H), 4.25 (m, 2H), 3.50 (m, 24H), 2.32 (m, 8H). 13C NMR
(D2O, 50 °C): δ 137.75, 128.98, 71.65, 71.45, 49.73, 45.73, 44.14,
22.09. HR-MS for (C26H48N6O2 + H)+: requires 477.3917, found
477.3899.
Reactions were carried out in sealed tubes immersed in a water bath,
the temperature of which was adjusted to 90.0 ( 0.1 °C. The aliquots
withdrawn at appropriate intervals during 3 half-lives of the disap-
pearance of the starting material were immediately cooled in an ice-
water bath and stored in a freezer before analysis. The composition of
the aliquots was analyzed by RP-HPLC on a Hypersil ODS column
C18 (250 × 4.6 mm, particle size 5 µm) or a Waters Atlantis column
dC18 (250 × 4.6 mm, particle size 5 µm). The samples were eluted
with a linear gradient from 0.1 mol L-1 KH2PO4/KOH buffer (pH 6.0,
EDTA 2.0 mmol L-1) to the same buffer containing 8% MeCN. The
detection wavelength was 260 nm, and the bandwidth was 4 nm
throughout the work. The initial concentration of the starting material
was 50 µmol L-1. In other words, the cleaving agent was present in
10-fold excess (or more). The pseudo-first-order rate constants of the
disappearance of the starting material were calculated by means of the
integrated first-order rate law.
UV Absorption Titration. Two milliliters of a 50 µmol L-1 solution
of either ApA, ApU, or UpU (HEPES buffer, 0.1 mol L-1, I ) 0.1
mol L-1, pH 7.5, 7.0, or 6.5) was placed in a cell having an optical
path of 1 cm. The temperature of the cell housing block was adjusted
to 25.0 ( 0.1 °C. A much more concentrated solution of 3 (7.19 mmol
L-1) and Zn2+ (14.4 mmol L-1) was added portionwise, and the UV
spectrum was measured immediately after each addition.
4,4′-Bis[(1,5,9-triazacyclododecan-3-yloxy)methyl]biphenyl Hexahy-
drochloride (4). Yield: 89%. 1H NMR (D2O, 50 °C): δ 7.97 (d, J )
8.1 Hz, 4H), δ 7.78 (d, J ) 8.1 Hz, 4H), 4.98 (s, 4H), 4.30 (m, 2H),
3.51 (m, 24H), 2.33 (m, 8H). 13C NMR (D2O, 50 °C): δ 140.60, 137.25,
129.43, 127.71, 71.81, 71.51, 50.05, 45.95, 44.43, 22.34. HR-MS for
(C32H52N6O2 + H)+: requires 553.4230, found 553.4218.
2,6-Bis[(1,5,9-triazacyclododecan-3-yloxy)methyl]pyridine Hep-
tahydrochloride (5). Yield: 91%. 1H NMR (D2O): δ 8.47 (t, J ) 8.0
Hz, 1H), 7.95 (d, J ) 8.0 Hz, 2H), 5.04 (s, 4H), 4.10 (m, 2H), 3.55
(dd, J ) 4.6, 13.3 Hz, 4H), 3.25 (m, 12H), 2.97 (m, 4H), 2.05 (m,
8H). 13C NMR (D2O): δ 151.99, 147.07, 124.63, 72.66, 66.60, 48.90,
44.62, 43.08, 21.35. HR-MS for (C25H47N7O2 + H)+: requires
478.3869, found 478.3848.
1,3,5-Tris[(1,5,9-triazacyclododecan-3-yloxy)methyl]benzene Non-
1
ahydrochloride (6). Yield: 93%. H NMR (D2O, 50 °C): δ 8.11 (s,
3H), 4.97 (s, 6H), 4.31 (m, 3H), 3.71 (dd, J ) 5.2, 13.7 Hz, 6H), 3.54
(m, 24H), 3.36 (m, 6H), 2.36 (m, 12H). 13C NMR (D2O, 50 °C): δ
138.36, 128.16, 71.27, 71.25, 48.77, 45.12, 43.25, 21.33. HR-MS for
(C36H69N9O3 + H)+: requires 676.5602, found 676.5587.
Kinetics Experiments. The pH of the reaction solutions was adjusted
with a HEPES buffer (0.1 mol L-1) to a desired value at 25 °C (I )
0.1 mol L-1 with NaNO3). Sufficient amounts of the ligand and
Zn(NO3)2 were added to give the final concentration of 0.5 mmol L-1
for the complex (in some cases 1.5 or 2.0 mmol L-1). In other words,
the concentration of the ligand (1-6) was 0.5 mM, and the concentra-
tion of Zn2+ was 1.0 mmol L-1 in solutions of 1-5 and 1.5 mmol L-1
in the solution of 6. The solutions were allowed to stand for a few
hours to be sure that the equilibrium had settled. The pH was then
checked and, when needed, adjusted back to the original value with
sodium hydroxide. No precipitation of zinc hydroxide was observed,
even at pH < 7, indicating that the binding of Zn2+ to the ligand was
virtually quantitative. The pH values at 90 °C were obtained by
extrapolation based on the known temperature dependence of the pKa
value of HEPES.26
1H NMR Spectroscopy Experiments. 3′,5′-ApA or 3′,5′-UpU was
dissolved in D2O to give a concentration of 4.0 mmol L-1, and the
1
pD was adjusted to 7.1 with NaOD in D2O. The H NMR spectrum
was recorded at 600 MHz. Ligand 3 and Zn2+ in D2O were added to
the solution in a concentration ratio of 1:2. The pH was adjusted to
1
pD 7.1, and the H NMR spectrum was again measured. The addi-
tions and adjustment of pH were repeated until 1 equiv of 3 had been
added.
A similar technique was used to study the ability of ligands 1-6 to
bind Zn2+. The measurements were carried out at 35 °C and pD 7.1
and with ligand 3 additionally at 90 °C. The concentration of 3 was 2
mmol L-1, and altogether 4 equiv of Zn2+ compared to 3 were added.
JA058806S
9
10728 J. AM. CHEM. SOC. VOL. 128, NO. 33, 2006