Toward the rational design of superoxide dismutase mimics: Mechanistic studies for the elucidation of substituent effects on the catalytic activity of macrocyclic manganese(II) complexes

Article abstract of DOI:10.1021/ja964271e

Two new isomeric bis(trans-fused cyclohexano) substituted 1,4,7,10,13-pentaazacyclopentadecane ligands and their Mn(II) complexes, 3 and 4, have been synthesized, and their activity as superoxide dismutase (SOD) catalysts has been studied. Complex 3 is an

Full text of DOI:10.1021/ja964271e

6522
J. Am. Chem. Soc. 1997, 119, 6522-6528
Toward the Rational Design of Superoxide Dismutase Mimics:
Mechanistic Studies for the Elucidation of Substituent Effects
on the Catalytic Activity of Macrocyclic Manganese(II)
Complexes
Dennis P. Riley,* Patrick J. Lennon, William L. Neumann, and Randy H. Weiss
Contribution from the The Monsanto Company, 800 North Lindbergh BouleVard,
St. Louis, Missouri 63167
ReceiVed December 12, 1996X
Abstract: Two new isomeric bis(trans-fused cyclohexano) substituted 1,4,7,10,13-pentaazacyclopentadecane ligands
and their Mn(II) complexes, 3 and 4, have been synthesized, and their activity as superoxide dismutase (SOD) catalysts
has been studied. Complex 3 is an excellent SOD catalyst with a second-order rate constant at pH ) 7.4 of 1.2 ×
10+8 M-1 s-1. In contrast, the isomeric complex 4 has virtually no detectable catalytic SOD activity, implying the
need to understand the effect that the position, number, and stereochemistry of substituents exert on the catalytic
rate. The crystal structure of the complex 4 was determined and reveals that the Mn(II) ion is coordinated in a
pentagonal bipyramid array of the five nitrogens of the macrocyclic ligand and capped by two trans-chloro ligands.
Crystal data for MnC18H37Cl2N5 are as follows: triclinic at 20 °C, space group P1--Ci2 (no. 2); a ) 9.746(3) Å, b
) 12.631(6) Å, c ) 11.311(5) Å; R ) 73.14(4)°, â ) 76.39(3)°, γ ) 79.98(3)°, V ) 1287(1) Å3, and Z ) 2 (Fcalc
) 1.279 g/cm3; µa Mo KR ) 6.23 mm-1). Mechanistic studies with the complex 3 and the pentamethyl susbstituted
complex, 5, including D2O rate studies, are reported and are consistent with the existence of two pathways for the
rate-determining electron-transfer from Mn(II) to superoxide: (1) hydrogen atom transfer from a bound water on
Mn(II) to HO2• to yield a Mn(III) hydroxo intermediate and (2) the dissociative pathway in which superoxide anion
binds to a vacant coordination site on Mn(II) followed by protonation/oxidation to yield a Mn(III)hydroperoxo species.
Subsequent reduction of the intermediate Mn(III) with superoxide anion completes the catalytic cycle. Substituent
effects on the rates and relative contribution of the two pathways to the overall rate of SOD activity is ascribed to
the propensity of the ligand to fold around Mn(II) forming a pseudo-octahedral complex similar in geometry to the
oxidized Mn(III) complex. Folding of the pentaaza macrocyclic ligand is confirmed as a relevant structural motif
for this series of Mn(II) complexes by the X-ray structure determination of the bis(nitrate) derivative of 1, [Mn-
(C10H25N5)NO3]NO3, which reveals a six-coordinate structure with a folded conformation of the macrocyclic ligand.
Crystal data for [Mn(C10H25N5)NO3]NO3: orthorhombic at -100 °C, space group P212121; a ) 9.457(2) Å, b )
12.758(2) Å, c ) 13.834(2) Å, V ) 1669.1(5) Å3, and Z ) 4 (Fcalc ) 1.549 g/cm3).
Introduction
We have shown that the pentaaza macrocyclic Mn(II)
complex, 1 (Figure 1),1 and derivatives of this complex which
bear substituents on the carbon atoms of the macrocycle2 are
highly active and stable superoxide dismutase (SOD) mimics.
In numerous biological studies both in Vitro and in ViVo these
complexes display potent mechanism-based biological effects.3
Our goal has been to design low molecular weight, nonpeptidic
molecules possessing the function of human enzymes but
eliminating some of the negatives associated with the use of
peptide-based pharmaceuticals; e.g., cost, immunogenicity, tissue
penetration, oral bioavailability, etc. Successful implementation
of such metal-based drugs as human pharmaceutical agents
mandates that we construct molecules with high inherent
chemical stability while retaining high SOD activity. Recently,2
we reported that increasing the number of substituents on the
carbon atoms of the macrocycle invariably increases stability
(both kinetic and thermodynamic) of the Mn(II) complexes
derived from 1. Additionally, we observed that the number,
position, and stereochemistry of the substituents on the carbon
Figure 1. Mn(II) dichloro complexes of the carbon substituted pentaaza
macrocyclic ligands utilized in this study.
atoms of the macrocyclic ring exert a dramatic effect on the
catalytic SOD activity of the resultant complex.
In order to focus our synthetic efforts on the rational
construction of both highly stable (maximize substituents) and
highly SOD active Mn(II) complexes, we have studied the
superoxide dismutase catalytic reaction mechanism in detail.
This understanding is absolutely essential for the rational design
of these SOD catalysts as we have observed that the two effects
do not necessarily work in concert; i.e., simply increasing the
number of substituents does not guarantee that the resultant
Mn(II) complex will possess any SOD activity. Herein, we
X Abstract published in AdVance ACS Abstracts, June 15, 1997.
(1) Riley, D. P.; Weiss, R. H. J. Am. Chem. Soc., 1994, 116, 387.
(2) Riley, D. P.; Henke, S. L.; Lennon, P. J.; Weiss, R. H.; Neumann,
W. L., Rivers Jr, W. J.; Aston, K. W.; Sample, K. R.; Rahman, H.; Ling,
C,-S.; Shieh, J.-J.; Busch, D. H.; Szulbinski, W. Inorg. Chem. 1996, 35,
5213.
S0002-7863(96)04271-0 CCC: $14.00 © 1997 American Chemical Society

Superoxide Dismutase Mimics
J. Am. Chem. Soc., Vol. 119, No. 28, 1997 6523
N-Tosylglycyl-(1R,2R)-diaminocyclohexane. (1R,2R)-Diaminocy-
clohexane (10.0 g, 87.57 mmol) was dissolved in dry DMF (150 mL)
under argon and cooled to -10 °C. Separately, N-tosylglycine (10.04
g, 43.62 mmol), 1-hydroxybenzotriazole (6.75 g, 44.08 mmol), and
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (8.45 g,
44.05 mmol) were dissolved in dry DMF (150 mL) and cooled to -10
°C under argon. The latter solution was added to the diaminocyclo-
hexane solution at -10 °C via cannula. After 2 h at this temperature,
water (8 mL) was added and the reaction was allowed to warm to 0 °C
over 1 h and then to room temperature over the next half hour. The
solvent was removed on the rotary evaporator under reduced pressure.
The residue was heated to 40-42 °C with water (150 mL) added in
small portions with stirring. After 25 min this solution was filtered.
The white precipitate was largely the bis adduct (5.55 g). Exactly 68
mL of the filtrate was worked up by repeated extraction with
dichloromethane (9 × 50 mL). The combined organic phase was dried
(sodium sulfate) and filtered, and the solvent was removed. The
resulting white solid which contained some residual DMF was
redissolved in dichloromethane (30 mL) and added dropwise to a stirred
solution of 9:1 ether:hexane (250-300 mL) giving an immediate
precipitate which was stirred overnight and then filtered. This procedure
was repeated, stirring for 3 h instead of overnight. After drying the
white product on the vacuum line, 2.36 g, 7.25 mmol were obtained,
equivalent to a 36.7% yield for the entire reaction. 1H NMR (CDCl3,
400 MHz) δ 1.10-1.34 (m, 4H), 1.70 (d, J ) 9.7 Hz, 2H), 1.81-1.97
(2 m, 2H), 2.41 (s, 3H), 2.51 (td, J ) 10.2, 3.8 Hz, 1H), 3.53 (m +
ABq, J ) 16.9 Hz, δν ) 51.6 Hz, 3H), 3.69 (br s, 3H), 6.84 (d, J )
9.1 Hz, 1H), 7.30 (d, J ) 8.3 Hz, 2H), 7.73 (d, J ) 8.3 Hz, 2H); 13C
NMR (CDCl3, 100 MHz) δ 21.48, 24.87, 24.97, 32.08, 35.16, 46.09,
54.85, 55.78, 127.15, 129.85, 136.02, 143.84, 168.69; MS (GT HCl):
m/z 326 (100) [M + H]+.
(2R,3R,8R,9R)-Dicyclohexano-13-p-toluenesulfonyl-1,4,7,10,13-
pentaazacyclopentadecan-5,11,15-trione. N-p-Toluenesulfonylglycyl-
(1R,2R)-diaminocyclohexane (1.11 g, 3.42 mmol) and N,N′-bis-
(chloroacetyl)-(1R,2R)-diaminocyclohexane (0.913 g, 3.42 mmol) were
combined in a 1-L flask and dry N,N-dimethylacetamide (650 mL) was
added. The flask was inerted. After 10 min, the sodium hydride (0.19
g, 95%) was added directly to the homogeneous mixture. The reaction
flask was placed in a 70 °C oil bath. After the internal temperature
reached 45-50 °C, gas evolution became constant. The oil bath
temperature was stabilized at about 65 °C with some excursions from
about 60 to 75 °C. Overnight, the reaction mixture became homoge-
neous. After heating for 17 h the reaction flask was removed from the
bath and allowed to cool. The solvent was removed under reduced
pressure, and the yellowish oil was placed on the vacuum line. The
residue was treated with dichloromethane (300 mL) and washed with
water (40 mL) and twice with saturated sodium chloride (40 mL each).
After combining, the aqueous layers were backwashed with dichlo-
romethane (100 mL). The combined organic layers were dried over
sodium sulfate, filtered, and stripped down to a viscous yellow oil which
was placed on the vacuum line, 2.14 g. This residue was chromato-
graphed using 0.5% NH4OH/9% CH3OH/90.5% CH2Cl2. On TLC on
silica using the same system, Rf ) 0.25. Fractions containing the correct
spot were combined and evaporated down to a slightly off white solid,
0.89 g, 1.71 mmol, 50.1% yield. 1H NMR (CDCl3, 300 MHz) δ 0.92-
2.1 (several m, 16H), 2.27 (m, 1H), 2.41 (s, 3H), 2.57 (m, 1H), 3.10
(ABq, J ) 16 Hz, δν ) 34.2 Hz, 2H), 3.39 (m, 1H), 3.58 (m, 3H),
3.83 (m, 1H), 4.08 (d, J ) 17.6 Hz, 1H), 4.39 (d, J ) 17.4 Hz, 1H),
7.30 (m, 3H), 7.44 (d, J ) 5.9 Hz, 1H), 7.76 (d, J ) 7.8 Hz, 2H), 8.05
(d, J ) 8.4 Hz, 1H); 13C NMR (CDCl3, 100 MHz) δ 21.39, 24.20,
24.69, 24.87 (double intensity), 31.49, 31.54, 31.58, 32.43, 47.01, 52.19,
52.25, 52.49, 52.97, 55.63, 58.36, 127.65, 129.67, 135.28, 143.97,
167.52, 170.04, 172.84; MS (FAB, NBA-LiCl matrix): m/z (rel
intensity) 526 (100) [M + Li]+, 370 (29) [M + Li - Ts]+. HRMS
(NBA-LiCl) exact mass calcd for [C25H37N5O5S], 520.2594, found
520.2659.
report our efforts to understand the details of the electron-transfer
process and describe two new complexes incorporating two
trans-fused cyclohexano backbones on the macrocyclic ring
ligands4 sa structural motif which enhances both the stability
and activity; e.g., complex 2, which possesses one trans-
cyclohexano substituent is more than twice as active as 1 as a
SOD catalyst and is both kinetically (2-fold enhancement) and
thermodynamically (log K ) 11.6 vs 10.8 for 1) more stable
than the parent unsubstituted complex, 1.
We also report here the X-ray crystal structure of two new
complexes. One of the new complexes is the bis(trans)-
cyclohexano derivatives, 4, which crystallized as the seven-
coordinate dichloro derivative. The second complex is the
bis(nitrato) derivative of complex 1, which was crystallized as
the six-coordinate nitrato nitrate complex. The relevance of
this new six-coordinate nitrato derivative to these studies will
be discussed.
Experimental Section
Syntheses of Ligands: N,N′-Bis(chloroacetyl)-1R,2R-diaminocy-
clohexane. 1R,2R-(-)-Diaminocyclohexane (6.98 g, 61.11 mmol) was
dissolved in 75 mL of alcohol free CHCl3 in a four-neck, 2000 mL
round bottom flask along with 37 mL of H2O under argon. Two
Normag dropping funnels were connected to the reaction flask and
charged separately with chloroacetyl chloride (15 mL, 188.3 mmol) in
alcohol free CHCl3 (88 mL) and K2CO3 (24.1 g, 174.4 mmol) in 918
mL of H2O. An internal thermometer was inserted into the reaction
flask. After cooling the two phase mixture in the reaction flask to 0
°C in an ice bath, the additions from the dropping funnels were started
in such a way as to keep the proportion of each solution added
approximately equal over a 1 h 20 min period. During the addition,
an ice salt bath was used to moderate the temperature, keeping it
between 3 and -3 °C. A shell of ice formed on the inside of the
reaction flask which did not seem to impede the stirring. The reaction
flask was removed from the ice bath at the end of the addition and
was stirred for 2 h 20 min. The lower chloroform layer appeared to
have a considerable quantity of a light solid in it at ice bath temperature,
but it dissolved as the reaction warmed. The reaction mixture was
placed in a separatory funnel, some additional chloroform added, and
the layers were separated. The aqueous layer was extracted with another
portion of CHCl3, and the combined chloroform layers were washed
with water, then saturated NaCl, dried (Na2SO4), and stripped down to
a brownish white solid. This solid was stirred overnight with about
450 mL of ether and then filtered, much of the color staying in the
ether, giving a beige solid, 13.68 g, 51.60 mmol, 84.4% yield. 1H NMR
(CDCl3, 400 MHz) δ 1.34 (m, 4H), 1.80 (m, 2H), 2.08 (m, 2H), 3.74
(m, 2H), 3.99 (ABq, J ) 15.1 Hz, δν ) 8.2 Hz, 4H), 6.78 (br s, 2H);
13C NMR (CDCl3, 100 MHz) δ 24.59, 32.07, 42.45, 53.94, 166.65;
MS (FAB, NBA-LiCl matrix): m/z (rel intensity) 273 (100) [M + Li]+,
275 (71) [M + Li]+.
(3) a) Weiss, R. H.; Flickinger, A. G.; Rivers, W. J.; Hardy, M. M.;
Aston, K. W.; Ryan, U. S.; Riley, D. P. J. Biol. Chem. 1993, 268, 23049-
23054. (b) Weiss, R. H.; Riley, D. P.; Rivers, W. J.; Aston, K. W.;
Flickinger, A. G.; Hardy, M. M.; Ryan, U. S. Manganese-Based Superoxide
Dismutase Mimics: Design, Discovery and Pharmacologic Efficacies In
The Oxygen Paradox; Davies, K. J. A., Ursini, F., Eds.; CLEUP University
Press: Padova, Italy, 1995; pp 641-651. (c) Weiss, R. H.; Fretland, D. J.;
Baron, D. A.; Ryan, U. S.; Riley, D. P. J. Biol. Chem. 1996, 271, 26149.
(d) Kasten, T. P.; Settle, S. L.; Misko, T. P.; Riley, D. P.; Weiss, R. H.;
Currie, M. G.; Nickols, G. A. Proc. Soc. Exp. Biol. Med. 1995, 208, 170-
177. (e) Hardy, M. M.; Flickinger, A. G.; Riley, D. P.; Weiss, R. H.; Ryan,
U. S. J. Biol. Chem. 1994, 269, 18535-18540. (f) Kilgore, K. S.; Friedrichs,
G. S.; Johnson, C. R.; Schasteen, C. S.; Weiss; R. H.; Riley, D. P.; Ryan,
U. S.; Lucchesi, B. R. J. Mol. Cell. Cardiol. 1994, 26, 995-1006. (g) Black,
S. C.; Schasteen, C. S.; Weiss, R. H.; Riley, D. P.; Driscoll, E. M.; Lucchesi,
B. R. J. Pharmacol. Exp. Therapeut. 1994, 270, 1208-1215. (h) Venturini,
C. M.; Sawyer, W. B.; Smith, M. E.; Palomo, M. A.; McMahon, E. G.;
Weiss R. H.; Riley, D. P.; Schasteen, C. S. In The Biology of Nitric Oxide;
Moncada, S., Feelisch, M., Busse, R., Higgs, E. A., Eds.; Portland Press:
London, 1994; Vol. 3, pp 65-9.
(2R,3R,8R,9R)-Dicyclohexano-1,4,7,10,13-pentaazacyclopentade-
cane. (2R,3R,8R,9R)-Dicyclohexano-13-p-toluenesulfonyl-1,4,7,10,
13-pentaazacyclopentadecan-5,11,15-trione (4.072 g, 7.84 mmol) was
placed in a 1-L flask under an argon atmosphere, and dry 1,2-
dimethoxyethane (dme, 220 mL) was added. The powder fused and
did not appreciably dissolve. It was partially broken up with a spatula
and stirred in a cold water bath, while lithium aluminum hydride (0.5
(4) (a) Lennon, P. J.; Rahman, H.; Aston, K. W.; Henke, S. L.; Riley,
D. P. Tetrahedron Lett. 1994, 35, 853. (b) Neumann, W. L.; Franklin, G.
N.; Sample, K. R.; Riley, D. P.; Rath, N. Tetrahedron Lett. 1997, 38, 3143.

6524 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Riley et al.
M in dme, 140 mL, 70 mmol) was added in portions over a 10 min
period. Initially, the solution became cloudy, and undissolved chunks
of compound were present. After about 70 mL had been added, the
solution was fairly homogeneous, with only a few undissolved pieces
remaining, which appeared to dissolve with gas evolution. Heating
was started after a few minutes, and the solution rapidly became
heterogeneous and yellow. The reaction mixture was refluxed over-
night. Reflux was ended after 16.5 h. The reaction mixture was cooled
in a cold water bath and then in a -18 °C bath. Water (2.2 mL) was
added cautiously in small quantities over a 5-10 min period, followed
more rapidly by 15% NaOH (2.2 mL) and then by water (6.6 mL).
Stirring was continued for 2 h in the ice bath. Tetrahydrofuran (THF,
210 mL) was added, and stirring was continued for about an hour. The
thick white suspension was allowed to settle and was filtered with a
filter transfer device (#1 Whatman paper) under argon. The filtrate
was stripped. The white residue was stirred with THF (150 mL) and
filtered into the stripped first filtrate. The solvent was removed under
reduced pressure, and the residue was placed on the vacuum line. The
resulting yellow-white solid was extracted with hot dry hexane (initially
70 mL, 65 °C; then an additional 15 mL) and filtered through a filter
transfer device (#50 Whatman paper) under argon, and the solvent was
removed under reduced pressure. This crude product, weight about
1.5 g, was dissolved in hot (>70 °C) dry acetonitrile (about 60 mL),
filtered (filter transfer device, #50 Whatman paper), concentrated by
more than half, reheated to dissolve all of the white solid, and then
allowed to cool slowly to room temperature all under argon. White
crystals were obtained, 0.923 g, 2.85 mmol, 36.4% yield. 1H NMR
(C6D6, 300 MHz) δ 0.75-1.21 (several m, 8H), 1.23-2.19 (several
m, 17H), 2.36-2.61 (several H, 6H), 2.61-2.73 (m, 2H), 2.74-2.85
(m, 2H), 2.90 (d, J ) 7.5 Hz, 2H); 13C NMR (C6D6, 75 MHz) δ 25.48,
25.56, 32.41, 32.48, 46.50, 47.82, 49.56, 61.86, 62.88. Anal. Calcd
for C18H37N5: C, 66.83; H, 11.54; N, 21.65. Found: C, 66.66; H,
11.46; N, 21.78.
became constant. About 1 h later, the internal temperature was 64 °C.
After stirring overnight, the reaction mixture became homogeneous.
Heating was stopped after 20 h and allowed to cool. Some white solid
product was present, which was filtered off. After drying, this portion
weighed 1.34 g. The filtrate was concentrated under reduced pressure
until a yellow solid was obtained. This solid was treated with
dichloromethane (600 mL) and washed with water (100 mL, twice)
and then with saturated aqueous sodium chloride (twice, 100 mL). The
combined aqueous layers were backwashed with dichloromethane (150
mL). The combined organic layers were dried (sodium sulfate), filtered,
and stripped down to a pale yellow solid. This solid was stirred in
methanol at 50-55 °C for 1.5 h and then allowed to reach room
temperature overnight. After filtration 1.0 g of a white product was
recovered by filtration, and a subsequent 0.071 g was recovered from
the filtrate, giving a total of 2.41 g (44.4% yield), mp 305-306 °C. 1H
NMR (CDCl3 + CD3OD (four drops), 400 MHz) δ 1.0-1.5 (several
m, 7H), 1.67-2.07 (several m, 9H), 2.81 (td, J ) 10.2, 3.2 Hz, 1H),
2.91-3.06 (d and m, J ) 17.2 Hz, 2H), 3.4-3.74 (several m, 5H),
3.92 (d, J ) 17.7 Hz, 1H), 4.05 (d, J ) 17.0 Hz), 7.35 (d, J ) 8.1 Hz,
2H), 7.75 (d, J ) 8.3 Hz, 2H); 13C NMR (CDCl3 + CD3OD (four
drops), 100 MHz) δ 21.56, 24.59, 24.93, 25.01, 25.19, 30.35, 31.31,
32.21, 33.21, 51.49, 52.81, 53.32, 53.99, 54.49, 57.95, 58.51, 127.69,
129.97, 134.76, 144.49, 168.54, 169.35, 173.01; HRMS (FAB+, GT-
HCl) exact mass calcd for [C25H37N5O5S + H+], 520.2594; found
520.2605.
(2S,3S,8R,9R-Dicyclohexano-1,4,7,10,13-pentaazacyclopentade-
cane. The (2S,3S,8R,9R)-dicyclohexano-13-p-toluenesulfonyl-1,4,7,
10,13-pentaazacyclopentadecan-5,11,15-trione (2.176 g, 4.187 mmol)
was suspended in dry 1,2-dimethoxyethane (dme, 175 mL) in a 1000
mL round bottom flask and inerted with argon. While this solution
was stirred in an ice water bath, lithium aluminum hydride in dme
(0.5 M, 86 mL, 43 mmol) was rapidly added with a syringe over a
2-3 min period. There was considerable gas evolutio,n and the solution
became heterogeneous. After stirring for about 15 min, the ice water
bath was removed and replaced with a heating mantle. Over the next
15 min the reaction mixture was heated to reflux becoming bright
(opaque) yellow. After refluxing for several hours the yellow color
faded somewhat. After refluxing for 24 h, the faded yellow mixture
was allowed to cool. It was placed in an ice bath, and water (1.4 mL)
was cautiously added dropwise over about 2 min with considerable
gas evolution. Aqueous 15% NaOH (1.4 mL) was added, followed
by more water (4.0 mL). This thick mixture was stirred for about 30
min, and then tetrahydrofuran (100 mL, passed through a column of
basic alumina) was added. The mixture was stirred for about 2 h and
then filtered using a filter transfer device under argon. The filtrate
was removed under reduced pressure and placed on the vacuum line,
giving 1.13 g of a white solid. This was dissolved in hot dry hexanes
(100 mL) and filtered using a filter transfer device. The small amount
of solid left behind was washed with 40 mL more hexanes. The hexane
was removed under reduced pressure, leaving a cream colored solid
which was recrystallized from hot acetonitrile under argon, 0.622 g,
1.92 mmol, 46% yield. 1H NMR (CDCl3, 400 MHz) δ 0.96 (m, 4H),
1.22 (m, 4H), 1.63-2.18 (several m, 17H), 2.47-2.69 (several m, 6H),
2.75 (m, 2H), 2.87 (m, 2H), 2.95 (m, 2H); 13C NMR (CDCl3, 100 MHz)
δ 25.13, 25.20, 31.99, 32.15, 45.78, 45.94, 48.66, 61.17, 61.65; HRMS
(FAB+, glycerol; TFA; PEG-300) exact mass calcd for [C18H37N5 +
H+], 324.3127; found, 324.3142.
N,N ′-Bis(chloroacetyl)-(1S,2S)-diaminocyclohexane. (1S,2S)-Di-
aminocyclohexane (7.0 g, 61.13 mmol) was dissolved in 75 mL of
alcohol free CHCl3 in a four-neck, 2000 mL round bottom flask along
with 38 mL of H2O under argon. Two Normag dropping funnels were
connected to the reaction flask and charged separately with chloroacetyl
chloride (15 mL, 188.2 mmol) in alcohol free CHCl3 (88 mL) and K2-
CO3 (24.2 g, 174.4 mmol) in H2O (918 mL). An internal thermometer
was inserted into the reaction flask. After cooling the two phase mixture
in the reaction flask to 0 °C in an ice bath, the additions from the
dropping funnels were started in such a way as to keep the proportion
of each solution added approximately equal over a 1 h 30 min period.
During the addition, an ice salt bath was used to moderate the
temperature, keeping it between 3 and 0 °C. A shell of ice formed on
the inside of the reaction flask which did not seem to impede the stirring.
The reaction flask was removed from the ice bath at the end of the
addition and was stirred for 2 h 15 min. The lower chloroform layer
appeared to have a considerable quantity of a light solid in it at ice
bath temperature, but it dissolved as the reaction warmed. The reaction
mixture was placed in a separatory funnel, some additional chloroform
was added, and the layers were separated. The aqueous layer was
extracted with another portion of CHCl3 (150 mL), and the combined
chloroform layers were washed with water (100 mL), then saturated
aqueous NaCl (150 mL), dried (Na2SO4), and stripped down to an off-
white solid. This solid was stirred for 2 days with about 500 mL of
ether and then filtered, much of the color staying in the ether, giving
a white solid, 14.14 g, 53.33 mmol, 87.0% yield, mp, 230-231 °C. 1H
NMR (CDCl3, 400 MHz) δ 1.34 (m, 4H), 1.80 (m, 2H), 2.08 (m, 2H),
3.75 (m, 2H), 3.99 (ABq, J ) 15.0 Hz, δν ) 8.1 Hz, 4H), 6.78 (br s,
2H); 13C NMR (CDCl3, 100 MHz) δ 24.53, 31.99, 42.41, 53.85, 166.66;
HRMS (FAB+, GT-HCl) exact mass calcd for [C10H16Cl2N2O2 + H+],
267.0667; found 267.0652.
Syntheses of Complexes: Manganese(II) Dichloride (2R,3R,8R,9R-
Dicyclohexano-1,4,7,10,13-pentaazacyclopentadecane (3). A metha-
nol solution (35 mL) of the free base (2R,3R,8R,9R)-dicyclohexano-
1,4,7,10,13-pentaazacyclopentadecane, (5.80 g, 17.93 mmol) was added
to a hot methanolic solution (100 mL) containing 2.25 g (17.9 mmol)
of anhydrous manganese(II) chloride. The solution was refluxed under
a dry nitrogen atmosphere for 1 h and then stirred for an additional 12
h at room temperature. After cooling, the solution was taken to dryness,
and the solid mass was dissolved in 50 mL of hot (∼60 °C) THF. The
solution was filtered through Celite and concentrated in Vacuo to
dryness. Ethyl ether was added to the solid mass which was then
vigorously stirred. White crystals were collected by filtration, washed
with ether, and dried to give 7.30 g (91% yield) of 3: [R]20D ) -96.97°
(c ) 0.02, methanol); MS (FAB, NBA matrix) m/z (rel intensity): 448.1
(M+, 2), 413.2 [(M - Cl, 100)+. Anal. Calcd for C18H37Cl2-
(2S,3S,8R,9R)-Dicyclohexano-13-p-toluenesulfonyl-1,4,7,10,13-
pentaazacyclopentadecan-5,11,15-trione. N-p-Toluenesulfonylglycyl-
(1R,2R)-diaminocyclohexane (3.4 g, 10.44 mmol) and the N,N′-
bis(chloroacetyl)-(1S,2S)-diaminocyclohexane (2.79 g, 10.44 mmol)
were combined in a 3-L flask, and dry N,N′-dimethylacetamide (1.95
L) was added. After inerting the flask, sodium hydride (95%, 0.559
g, 22.1 mmol) and an additional portion of N,N′-dimethylacetamide
(100 mL) were added. The reaction flask was placed in an oil bath at
71 °C. After the internal temperature reached 60 °C, gas evolution

Superoxide Dismutase Mimics
J. Am. Chem. Soc., Vol. 119, No. 28, 1997 6525
MnN5: C, 48.11; H, 8.30; N, 15.59; Cl, 15.78. Found: C, 48.23; H,
8.32; N, 15.47; Cl, 15.73.
Manganese(II) Dichloride (2R,3R,8S,9S-Dicyclohexano-1,4,7,10,
13-pentaazacyclopentadecane (4). To a hot methanol solution
containing 186 mg (1.474 mmol) of anhydrous manganese(II) chloride
was added the free base (477 mg, 1.474 mmol) under a dry N2
atmosphere. The reaction was carried out identically to that described
for complex 3. After drying in Vacuo a yield of 0.56 g (84%) was
obtained; MS (FAB, NBA matrix) m/z (rel intensity): 448.1 (M+, 1),
413.2 [(M - Cl, 100)+. Anal. Calcd for C18H37Cl2MnN5: C, 48.11;
H, 8.30; N, 15.59; Cl, 15.78. Found: C, 48.34; H, 8.17; N, 15.65; Cl,
15.70.
Manganese(II) 1,4,7,10,13-Pentaazacyclopentadecane Bis(nitrate)
(6). To 50 mL of warm methanol solution containing 0.605 g (0.233
mmol) of Mn(NO3)2‚xH2O (21% Mn) was added 0.50 g of the ligand
1,4,7,10,13-pentaazacyclopentadecane. The solution was refluxed for
1 h and reduced in volume to ∼10 mL. At this point diethyl ether
was added to the cloud point (∼20 mL). The solution was allowed to
sit undisturbed for 2 days whereupon very large clear white crystals
had formed. The solid was collected by filtration, washed with diethyl
ether, and dried overnight in vacuo. The yield of the complex 6 was
0.80 g (73% of theory). MS (FAB, NBA matrix) m/z (rel intensity):
332 [(M(L)(NO3), 100)+. Anal. Calcd for C10H25MnN7O6: C, 30.46;
H, 6.39; N, 24.87. Found: C, 30.70; H, 6.33; N, 24.66.
Figure 2. ORTEP drawing for manganese(II) dichloride (2R,3R,8S,9S)-
dicyclohexano-1,4,7,10,13-pentaazacyclopentadecane (4) showing the
labeling scheme and the 50% probability ellipsoids for non-hydrogen
atoms.
squares techniques with hydrogen atoms positionally refined with
common isotropic U.
Results and Discussion
Previously we reported that complex 2, containing a single
trans-fused cyclohexano ring, is both a more active SOD catalyst
than 1, and it possesses an enhanced chemical stability toward
dissociation of the Mn(II) ion as compared to 1. The point of
this prior work was to identify those substituents and substituent
patterns which would allow us to maximize the stability of the
corresponding Mn(II) complex and at the same time retain the
catalytic activity. From this exercise the trans-cyclohexano
moiety emerged as the best substituent for accomplishing this
dual function. In this paper we describe our efforts to probe
the additivity of the beneficial effects arising from the trans-
fused cyclohexano moiety by synthesizing ligands containing
multiple trans-fused cyclohexano substituents. We report here
two new isomeric bis(trans-fused cyclohexano) ligands
(2R,3R,8R,9R and 2R,3R,8S,9S stereochemistries) and their
Mn(II) complexes, 3 (2R,3R,8R,9R) and 4 (2R,3R,8S,9S). The
new complexes, 3 and 4, were synthesized by reacting the
ligands with anhydrous MnCl2 in MeOH under dry N2.
Additionally, the new bis(nitrate) complex 6, derived from the
ligand of complex 1, was synthesized by reacting the ligand
with anhydrous Mn(NO3)2 in dry methanol under N2. The molar
conductances of each of the white complexes, 3, 4, and 6, were
measured in methanol, and all were consistent with a 1:1
electrolyte type implying either a six-coordinate [Mn(L)X]+ or
solvated seven-coordinate structure, [Mn(L)(MeOH)X]+ in
solution. Magnetic susceptibilities measured by the Evan’s
method in methanol solution were consistent with these
complexes existing as high-spin d5 complexes.7 The cyclic
voltammetry of each of these new complexes, 3 and 4, exhibits
a well-defined reversible one-electron oxidation at E1/2 at 0.74
V vs SHE in anhydrous methanol containing 0.25 M n-Bu4-
NBF4.
Physical Methods. Stability constants for the two new dicyclo-
hexano) complexes, 3 and 4, were obtained by utilizing standard
potentiometric titration methods as first described by Martell et al.,5
and further described in detail in our previous studies.2 Kinetic
stabilities were determined by the Cu(II) ion competition method
described in detail in our previous studies.2 The method utilized to
determine the superoxide dismutase catalytic activity of the Mn(II)
complexes reported here is based on the stopped flow kinetic assay
described previously.1,2,3a,6 The work described here utilized the same
methodology, and the kcat for each complex was assessed over the pH
range of 7.4-8.3.
X-ray Crystal Structures. Single crystals of 4‚C2H5OH were
collected on a computer-controlled Nicolet autodiffractometer using
Θ-2Θ scans and nickel-filtered CuKR radiation with a total of 4050
independent absorption-corrected reflections having 2Θ (CuKR) < 120°
(the equivalent of 0.65 limiting CuKR spheres). The structure was
solved using “Direct Methods” techniques with the Siemens SHELXTL-
PC software package as modified at Crystalytics Company. The
resulting structural parameters were refined to convergence {R1
(unweighted, based on F) ) 0.058 for 1611 independent absorption-
corrected reflections having 2 Θ (CuKR) < 120° and I > 3s(I)} using
counter-weighted full-matrix least-squares techniques and a structural
model which incorporated anisotropic thermal parameters for all but
three (O1s′, C1s′, and C2s′) non-hydrogen atoms and isotropic thermal
parameters for these three non-hydrogens and all included hydrogen
atoms. The hydrogen atoms included in the structural model were fixed
at idealized sp3-hybridized positions with C-H and N-H bond lengths
of 0.96 and 0.90 Å, respectively. The ethanol solvent molecule of
crystallization is disordered in the lattice with (at least) two alternate
positions for each of its nonhydrogen atoms. The oxygen atoms are
distinguished from the carbons based on the favorable hydrogen-
bonding interactions in the lattice; the following factors were employed
in all structure factor calculations: O1s, 0.80; O1s′, 0.20; C1s, 0.65; C1s′,
0.35; C2s, 0.60; C2s′, 0.40. Since none of these solvent non-hydrogen
atoms could be satisfactorily refined as independent atoms, they were
constrained to have common C-O and C-C bond lengths which
refined to final values of 1.336 and 1.439 Å, respectively. The
intramolecular C‚‚‚O separations between terminal non-hydrogen
solvent molecule atoms were also constrained to be 1.633 times the
value of the common C-O bond length.
X-ray Diffraction and Crystal Structures. Complex 4 was
shown by single-crystal X-ray diffraction to have a seven-
coordinate structure (Figure 2) analogous to other Mn(II)
complexes of this class.1,2,8 In Tables 1 and 2 are reported some
pertinent bond length and bond angle data, respectively.
Additional data are reported in the Supporting Information. The
structure is fairly typical for this class of pentaaza macrocyclic
complexes of Mn(II). The ClMnCl bond angle is 177.2°, and
Single crystals of 6 were collected on a Siemens R3m/V diffracto-
meter at -100 °C using 2Θ-Θ scans and MoKR radiation with a total
of 2464 independent semiempirical absorption corrected reflections
collected with a 2Θ range from 3.5 to 55.0 °C. The structure was
solved using “Heavy Atom Methods” with the Siemens SHELXTL
PLUS (VMS) software package. The resulting structural parameters
were refined to convergence {R1(unweighted) ) 2.09% for 2192
independent absorption-corrected reflections} using full matrix least-
(5) Martell, A. E.; Motekaitis, R. J. Determination and Use of Stability
Constants, 2nd ed.; VCH Publishers, Inc.: New York, 1992; p 29.
(6) Riley, D. P.; Rivers, W. J.; Weiss, R. H. Anal. Biochem. 1991, 196,
344.
(7) Evans, D. F. J. Chem. Soc. 1959, 2003.
(8) Neumann, W. L.; Franklin, G. W.; Sample, K. R.; Aston, K. W.;
Weiss, R. H.; Riley, D. P. Tetrahron Lett. 1997, 38, 779.

6526 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Table 1. Crystal Data for Complex 4
Riley et al.
specimen
colorless, rectangular paralellopiped,
0.15 × 0.22 × 0.85 mm
crystal system
space group
formula
a, Å
b, Å
c, Å
triclinic, at 20 °C
P1-Ci2 (no. 2)
C18H37Cl2MnN5‚C2H5OH
9.746(3)
12.631(6)
11.311(5)
R, deg
73.14(4)
â, deg
76.39(3)
γ, deg
79.98(3)
V, Å
1287(1)
F
calc, g/cm3
1.279
2
Z
fw
λ, Å
diffractometer
495.43
1.54184
Four-Circle Nicolet (Siemens)
Autodiffractometer
radiation (Å)
nickel-filtered CuKR
data collection
Θ-2Θ
4050
1611
6.23
reflectionstot
obsd data, I > 3σ(I)
abs. coeff, mm-1
goodness of fit
Figure 3. ORTEP drawing for manganese(II) (1,4,7,10,13-pentaaza-
cyclopentadecane Bis(nitrate) (6) showing the labeling scheme and the
50% probability ellipsoids for the non-hydrogen atoms.
1.59
R) ∑||F0| - |Fc||/∑|F0|
0.058
0.069
Rw) {∑w(|F0| - |Fc|)2/
2
∑w|F0| }1/2
Table 4. Selected Bond Angles (deg) for complexes 4 and 6
complex 4
complex 6
Table 2. Crystal Data for Complex 6
type
angle
type
angle
specimen
colorless prism,
Cl1MnCl2
Cl1MnN1
Cl1MnN2
Cl1MnN3
Cl1MnN4
Cl1MnN5
Cl2MnN1
Cl2MnN1
Cl2MnN3
Cl2MnN4
Cl2MnN5
177.2(1)
85.4(2)
96.4(2)
85.2(2)
93.5(2)
97.4(2)
93.9(2)
80.9(2)
93.9(2)
88.7(2)
84.9(2)
O1MnN1
O1MnN2
O1MnN3
O1MnN4
O1MnN5
N1MnN2
N2MnN3
N3MnN4
N4MnN5
N5MnN1
128.6(1)
93.3(1)
78.0(1)
132.0(1)
92.5(1)
76.5(1)
77.7(1)
76.1(1)
76.7(1)
76.9(1)
0.22 × 0.42 × 0.64 mm
orthorhombic, at 20 °C
P212121
crystal system
space group
formula
a, Å
.
C10H25MnN6O3 NO3
9.457(2)
12.758(2)
13.834(2)
1669.1(5)
1.549
b, Å
c, Å
V, Å3
F
calc, g/cm3
Z
4
fw
λ, Å
389.3
0.710 73 Å
Siemens R3m/V
Θ-2Θ
2652
octahedral coordination geometry with the N4 occupying a site
which approximates a trans position in an octahedral geometry.
The resultant N4MnO bond angle is 132.0°. This structure is
seemingly intermediate between the planar macrocyclic ligand
conformation and a folded array with the idealized N4MnO angle
of 180°. Thus, while the structure is distorted from true
octahedral, it nevertheless shows that the ligand has adopted a
folded geometry in which one of the nitrogens is considerably
displaced from the plane of the Mn(II) ion and the remaining
four nitrogens. The Mn-N bond lengths are somewhat shorter
than those observed in both complex 1 and 4; e.g., the average
Mn-N bond length in 6 (2.28 Å) is less than that observed
with either 4 (2.33 Å) or 1 (2.36 Å).
Stability Studies. The stabilities of the two new bis(trans-
fused) cyclohexano complexes, 3 and 4, were assessed for both
their thermodynamic and kinetic stabilities via the potentiometric
titration method and the Cu(II) ion competition method,
respectively (see Experimental Section). The log K value for
3 was slightly greater than that measured for 4: 13.6 versus
13.5 and the corresponding kinetic stability of 3 was also slightly
greater than that observed with 4. The second-order rate
constant, kdiss, at 21 °C for the proton-assisted dissociation of
each complex (first-order in [complex] and first-order in [H+]),
was measured to be 28 M-1 s-1 for 3 and 31.5 M-1 s-1 for 4.
If one considers the kinetic stability at a given pH, then the
first-order rate of loss of ligand from the metal is calculated by
multipying kdiss by the [H+]; e.g., for 3 at pH ) 7 the rate is
2.8 × 10-7 s-1 which gives a half-life to equilibrium at this pH
of about 690 h.9 For comparison the kinetic stability enhance-
diffractometer
data collection
reflectionstot
obsd data, I > 3σ(I)
abs. coeff
2192
semiempirical
1.20
0.0283
0.0381
goodness of fit
R ) ∑||F0| - |Fc||/∑|F0|
Rw ) {∑w(|F0| - |Fc|)2/∑w|F0| }1/2
2
Table 3. Selected Bond Lengths (Å) for Complexes 4 and 6
complex 4 complex 6
type
length
type
length
Mn-Cl1
Mn-Cl2
Mn-N1
Mn-N2
Mn-N3
Mn-N4
Mn-N5
2.560(4)
2.631(4)
2.366(7)
2.308(9)
2.313(7)
2.319(8)
2.338(8)
Mn-O1
Mn-N1
Mn-N2
Mn-N3
Mn-N4
Mn-N5
2.222(2)
2.277(2)
2.281(3)
2.273(2)
2.308(2)
2.275(3)
this very nearly linear angle is similar to that observed in the
structure for 1 indicating that the steric bulk of the trans-fused
cyclohexano substituents is minimally sensed by the axial
ligands. In contrast the ClMnCl angle in the 2,2,3,3-tetramethyl
complex2 is 167° indicating that the axial methyl groups do
indeed exert an axial steric effect. The five nitrogen atoms and
the Mn(II) ion are also very nearly coplanar. Additionally, the
Mn-N and Mn-Cl bond lengths are similar to those values
observed in 1 and other complexes in this class.
The crystal structure of complex 6 is illustrated in Figure 3
and is in marked contrast to other structures in this series.
Replacing the chloride counterions with nitrate has made it
possible to crystallize a six-coordinate complex with the
unsubstituted pentaaza ligand of structure 1. The resultant
complex 6 has a Mn(II) ion in a six-coordinate distorted
(9) The kinetics of dissociation of the ligand at any fixed pH is first-
order in complex; thus, at a given pH the second-order rate constant can be
converted to a half-life value by the following equation: t1/2 ) 0.694/(kdiss
× [H+]).

Superoxide Dismutase Mimics
J. Am. Chem. Soc., Vol. 119, No. 28, 1997 6527
ment for 3 over complex 1 is approximately ∼1125-fold. From
the potentiometric titration data, the complex 3 is completely
intact down to a pH as low as 6.1; thus, at pH ) 7.4 there is no
free Mn(II) in the stoichiometric presence of this ligand and
hence this complex is quite stable under normal physiological
conditions. It is also important to compare the log K for 3 to
the parent unsubstituted complex 1 and the mono trans-fused
cyclohexano complex 2. Clearly, the two cyclohexano rings
are acting synergistically to enhance the stability of the complex,
since 3 is nearly 600 times more stable than 1 and nearly 60
times more stable than 2.
Kinetic Studies. Each complex was evaluated as an SOD
catalyst over a range of pH’s.6 While complex 6 was identical
to that of the dichloro analogue 1, the kcat observed for the two
bis (trans-fused) cyclohexano derivatives, 3 and 4, provided us
with some very intriguing results; namely, 3 is over three times
some important insights. First, the rate of the pH independent
path is consistent with an SN1 (bond-breaking) process in which
creation of the vacant axial coordination site on Mn(II) is rate-
limiting, followed by rapid binding of superoxide anion to
Mn(II) (eq 2) and then rapid protonation/oxidation of the
resultant Mn(II)-superoxo complex yielding the Mn(III)hydro-
peroxo complex. This pH independent pathway exhibits a
modest deuterium isotope effect for both 3 and 5, suggesting
that creation of the vacant coordination site by loss of
coordinated water is involved in the rate-determining step. The
sign and magnitude are not readily rationalized, as solvent
exchange rates on metal ions can be both faster and slower in
D2O than in H2O,12 but that there is an effect indicates the
involvement of aquo ligand exchange in the rate-determining
step. The rate of loss of water from the inner-coordination
sphere of aquo Mn(II) has been measured to be 0.8 × 107 s-1 13
,
faster than 1 at pH ) 7.4: kcat ) 1.21 × 10+8 M-1 s-1
,
and such a value similar to that observed here for k2 provides
additional support for this mechanistic interpretation. Finally,
a weakly binding counterion such as chloride should act as a
competitive inhibitor and hence decrease the rate in a linear
fashion with added chloride. When the kcat for complex 5 was
measured at pH ) 7.8 with [Cl-]added over the range from 0-250
µM, the kcat decreased linearly from 3.82 × 107 to 2.61 × 107
M-1 s-1 consistent with competitive binding with superoxide
to Mn(II) in this pH independent pathway.
functioning at a rate equivalent to native mitochondrial Mn-
SOD enzyme,6 while complex 4 exhibits no detectable SOD
activity. This very striking contrast in the catalytic rate of
superoxide dismutation for these two steroisomeric complexes
dramatically demonstrates the need to understand how substit-
uents affect the rate.
For complex 3 the catalytic rate of dismutation of superoxide,
kcat, is first-order in [O2•-] and [3]added. The [H+] dependence
for SOD catalysis exhibits a linear dependence on the [H+] from
pH ) 7.4-8.6 (Supporting Information).1 A [H+] independent
pathway is observed (kind ) 3.16 × 107 s-1, k2 ) 1.58 × 107
s-1) such that the observed rate law is shown in eq 1
A second conclusion is that the near diffusion-controlled rate
of the pH dependent electron-transfer process requires that the
barrier to electron transfer to HO2• be minimal.14 This implies
that no charge separation develops in the transition state and
that the reorganizational energy be minimal;15 i.e., the Mn(III)
and Mn(II) complexes are similar in structure. This means that
the precursor Mn(II) complex must possess a six-coordinate
pseudo-octahedral geometry resembling the six-coordinate
octahedral geometry of Mn(III), regardless of the oxidation
pathway involved.
V ) -d[O2•-]/dt ) [Mn]1tot[O2•-]1{kH+[H+]1 + k } (1)
ind
+
where kH ) 2k1/Ka, kind ) 2k2 and Ka is the acidity constant
for superoxide (Ka ) 2.04 × 10-5).10
At 21 °C, kH was found from the slope of the kcat vs [H+]
+
plot to be 2.24 × 1015 M-1 s-1, which affords a calculated value
of k1 ) 2.25 × 1010 M-1 s-1. These results indicate that 3
operates via a mechanism similar to the other complexes in this
series with oxidation of MnII(L) being rate-limiting in the
catalytic process (eqs 2-4),1 although 3 does so with faster
rates: e.g., for 1: k2 ) 0.91 × 10-7 s-1 and k1 ) 0.29 × 1010
M-1 s-1, for 2: k2 ) 1.44 × 107 s-1 and k1 ) 0.78 × 1010 M-1
The results observed for the pH dependence on the rate in
D2O are consistent with these conclusions. The rate drop
observed for 3 in D2O is similar to that observed for the FeII/
FeIII self-exchange in water in which an isotope effect of 2 is
observed for the process involving a H-atom transfer from a
bound water on Fe(II).16-18
While the small magnitude of the
s-1 2
.
effect does not provide an overwhelmingly compelling case for
H-atom transfer,19 it is important to note that this apparent
kinetic isotope effect in our system has an equilibrium isotope
effect component. At any pH weak acids generally are
MnII(L) + HO2• 9k 8 MnIII(L) + HOO-
(2)
1
dissociated to a greater extent in light water than in D2O (e.g.,
MnII(L) + O2•- 9k 8 MnIII(L)(OO2-) (inner-sphere) (3)
•
K
HOAc/KDOAc ∼ 3.3),20 thus, the actual [DO2 ] will be greater
2
•
than [HO2 ] by a factor of approximately 3. Factoring in this
MnIII(L)(OO2-) + H+ 8 MnIII(L)(O2H)
(4)
•
•
increased concentration of DO2 relative to HO2 means that
fast
(10) Sawyer, D. T.; Valentine, J. S. Acc. Chem. Res. 1981, 14, 393.
(11) Covington, A. K.; Paabo, M.; Robinson, R. A.; Bates, R. G. Anal.
Chem. 1968, 40(3), 501.
(12) (a) Reidler, J.; Silber, H. B. J. Phys. Chem. 1973, 77, 1275. (b)
Sutin, N.; Rowley, J. K.; Dodson, J. Phys. Chem. 1961, 65, 1248. (c) Bunn,
D.; Dainton, F. S.; Duckworth, S. Trans. Faraday Soc. 1961, 57, 1131.
(13) Eigen, M. Pure Appl. Chem. 1963, 6, 105.
(14) Marcus, R. A. Annu. ReV. Phys. Chem. 1964, 15, 155.
(15) a) Macartney, D. H.; Thompson, D. W. Inorg. Chem. 1989, 28,
2199. (b) Endicott, J. F.; Kumar, K.; Ramasami, T.; Rotszinger, F. P. In
Progress in Inorganic Chemistry; Lippard, S. J., Ed., Wiley: New York,
1993; Vol. 30, p 141. (c) Creaser, I. I.; Harrowfield, J. M.; Herlt, J.;
Sargeson, A. M.; Springborg, J.; Geue, R. J.; Snow, M. R. J. Am. Chem.
Soc. 1977, 99, 3181.
(16) Hudis, J.; Dodson, R. W. J. Am. Chem. Soc. 1956, 78, 911.
(17) Friedman, H.; Newton, M. J. Electroanal. Chem. 1986, 204, 21.
(18) Reynolds, W. L.; Lumry, R. W. J. Chem. Phys. 1955, 23, 2460.
(19) Kresge, A. J. Sixth Steenbock Symposium on Isotope Effects on
Enzyme Catalyzed Reactions; Cleland, W. W., O’Leary, M. H., Northrop,
D. B., Eds.; University Park Press: Baltimore, 1977; pp 37-63. (b)
Westheimer, F. H. Chem. ReV. 1961, 61, 265.
To aid design efforts, we set out to gain added insight into
the details of the rate-limiting electron-transfer steps by studying
•-
the kinetics of the dismutation of O2 in D2O with 3 and the
2S,5R,8S,11R,14S-pentamethyl complex, 5,2 which exhibits no
[H+] dependence in its SOD activity (k2 ) 1.91 × 107 s-1).
•-
The catalytic rate for O2 dismutation with both 3 and 5 in
D2O over the pH range of 7.4-8.611 exhibits similar kinetic
behavior to that oberved in H2O; namely, 5 exhibits a pH
independent rate which was decreased ∼45%: k2 ) 1.01 ×
107 s-1. In contrast, 3 also displays a rate change in D2O, with
k1 decreased by ∼50%, but k2 increased by ∼10%: k1 ) 1.07
× 1010 M-1 s-1 and k2 ) 1.83 × 107 s-1
.
These results, combined with the observations that 4 is not a
catalyst, that 5 has no pH dependent component to its kcat, and
that the proton-dependent electron-transfer rate constants for
1, 2, and 3 are approaching diffusion controlled limits, lead to

6528 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Riley et al.
Scheme 1
the actual kinetic isotope effect is ∼ 6sa value more in keeping
for H-atom transfer.19
may be the reason why complex 5, for example, shows no rate
for this pH dependent pathway. It is intriguing to speculate
that the steric constraints imposed on this complex by the
presence of a methyl substituent symmetrically positioned on
one carbon of each chelate ring may inhibit folding and thereby
block this pathway for electron-transfer. This implies that the
two pathways may operate by different folding motifs. We are
continuing to explore this substituent effect via a combination
of molecular mechanics calculations and synthesis with the goal
of rational design of highly substituted chemically stable
synthetic enzymes.
Folding of the macrocyclic ring so that one of its secondary
amines occupies an axial site generating a pseudo-octahedral
complex, [MnII(L)X]n+, is a structural motif which accom-
modates this mechanistic picture. It is for this reason that the
crystal structure of complex 6 is significant. This six-coordinate
complex exhibiting the partially folded ring motif confirms that
stable pseudo-octahedral geometries are to be expected within
this class of complex and can conceivably play a major role in
the redox activity of the Mn center. Thus, if the macrocyclic
[15]aneN5 ligand possesses C-substituents which, due to in-
tramolecular steric repulsions and angle strains, could force the
ligand to adopt a folded pseudo-octahedral geometry about the
spherically symmetrical Mn(II) ion, then the Mn(II) complex
would be poised to undergo facile electron-transfer. In the
scheme the two pathways for the rate-determining oxidation of
Mn(II) to Mn(III) are depicted. The important features are that
the outer-sphere path proceeds through an intermediate pseudo-
octahedral [MnII(L)(H2O)]2+ complex with a folded macrocyclic
ligand, while the inner-sphere substitution pathway would
Acknowledgment. The authors would like to thank Mr. Karl
Aston, Mr. Hayat Rahman, Mr. Willie Rivers, and Mr. Kirby
Sample for their outstanding technical assistance.
Note Added in Proof: Recent studies provide additional
support for a proton-coupled electron transfer (as discussed
herein) as an important pathway of electron transfer in biological
systems such as photosystem II,21 as this path does not require
charge separation to build up in the transition state. Addition-
ally, a recent report indicates that the thermodynamics of an
•
involve a superoxo [MnII(L)(O2 )]+ six-coordinate intermediate
•
H-atom transfer from a Mn-bound water to HO2 will be
complex, containing the folded macrocyclic ligand, which would
be protonated to generate the [MnIII(L)(HO2)]2+ pseudo-
octahedral complex.
Thus, by affecting the degree of folding of the macrocyclic
ligand, the position, number, and stereochemistry of the sub-
stituents could reasonably be expected to exert a major effect
on the rate of electron-transfer via the pH dependent H-atom
transfer pathway. This need to fold the ligand into a conforma-
tion which will stabilize a pseudo-octahedral geometry on Mn(II)
favorable; i.e., the bond dissociation energy (BDE) of an O-H
bond of a water bound to a Mn(III)(Salen) derivative is 86 kcal/
mol.22 Thus, H-atom transfer to HO2• is energetically allowed
as the BDE of the O-H bond of hydrogen peroxide is reported
to be 87.2 kcal/mol.23
Supporting Information Available: Figure showing the
measured kcat as a function of pH for the complex 3 and
crystallographic details for complexes 4 and 6 (including atomic
coordinates, anisotrophic thermal parameters, bond lengths, and
bond angles for non-hydrogen atoms) (15 pages). See any
current masthead page for ordering and Internet access
instructions.
(20) Swain, C. G.; Bader, R. F. W.; Thorton, E. R. Tetrahedron 1960,
10, 200.
(21) Haganson, C. W.; Lydakis-Simantris, N.; Tang, X.; Tommos, C.;
Warncke, K.; Babcock, G. T.; Diner, B. A.; McCracken, J.; Styring, S.
Photosynth. Res. 1995, 46, 177.
(22) Caudle, M. T.; Pecoraro, V. L. J. Am. Chem. Soc. 1997, 119, 3415.
(23) Howard, C. J. J. Am. Chem. Soc. 1980, 102, 6937.
JA964271E