A Dissociative Mechanism for Reactions of Nitric Oxide with Water Soluble Iron(III) Porphyrins

Full text of DOI:10.1021/ja972448e

J. Am. Chem. Soc. 1997, 119, 12663-12664
12663
parameters ∆Hq, ∆Sq, and ∆Vq determined from temperature
(T) and hydrostatic pressure (P) effects on reactions with the
A Dissociative Mechanism for Reactions of Nitric
Oxide with Water Soluble Iron(III) Porphyrins
3-
water soluble complexes FeIII(TPPS)(H2O)2 (I) and FeIII-
n-
Leroy E. Laverman,† Mikio Hoshino,‡ and Peter C. Ford*,†
(TMPS)(H2O)2 (II) (TPPS ) tetra(4-sulfonatophenyl)por-
phine; TMPS ) tetra(sulfonatomesityl)porphine).12 These point
to a ligand dissociative pathway as dominating the “on”
mechanism and offer insight into the nature of such pathways
in bioregulatory signaling.
Department of Chemistry, UniVersity of California,
Santa Barbara, California 93106, and Institute of Physical
and Chemical Research, Wako, Saitama 351-01, Japan
Laser flash photolysis13 of aqueous solutions of I or II
under defined NO pressures gave transient spectra consistent
with the spectral differences between FeIII(Por)(H2O)(NO)n- and
ReceiVed July 21, 1997
Nitric oxide has significant roles in mammalian biology as
an intercellular signaling agent and in cytotoxic immune
response.1 The principal targets for NO under bioregulatory
conditions are metal centers, primarily iron.2 Although the
ferroheme enzyme soluble guanylyl cyclase (sGC) is the best
characterized example,3 various reports point to roles in inhibi-
tion of metalloenzymes such as cytochrome oxidase,4 nitrile
hydrase5 and catalase6 and in vasodilator properties of a salivary
ferriheme protein of blood-sucking insects.7 Concentrations
generated for bioregulation are low ([NO] < 1 µM reported in
endothelium cells for blood pressure control8 ); thus, reactions
with targets such as sGC must have very high rate constants
(kon) to compete effectively with other physical and chemical
processes leading to NO depletion.
Rates of NO reactions with various metal targets can be
determined by laser flash techniques,9,10 where NO is photo-
labilized from a M-NO precursor, and subsequent relaxation
to equilibrium is followed spectroscopically. Indeed, reactions
of NO with heme centers were thus investigated before the
bioregulatory functions were postulated.10 Despite this activity,
the mechanisms by which NO reacts with heme iron, i.e., the
“on” reaction for eq 1, (Por ) a porphyrin moiety), have not
n-
FeIII(Por)(H2O)2 (e.g., Figure 1). The transients decayed
exponentially (kobsd) to regenerate the equilibrium mixture of
solvated and nitrosyl complexes (see Figure 1 inset). No
permanent photoproducts were observed; thus the decay rep-
resents the relaxation of the FeIII(Por)/NO system according to
eq 1 and kobsd ) kon[NO] + koff. Accordingly, a plot of kobsd
versus [NO] was linear with a slope (kon) of 3.0 × 106 M-1 s-1
and a nonzero intercept (koff) of 7.3 × 102 s-1 for Por ) TMPS
at 25 °C giving a kon/koff ratio (4.1 × 103 M-1) within
experimental uncertainty of the equilibrium constant (3.93 ×
103 M-1) determined spectroscopically.
Temperature and hydrostatic pressure effects were evaluated
by determining kobsd at several [NO] and extracting the kon and
koff values at individual T or P over the respective ranges of
25-45 °C and 0.1-250 MPa. Eyring plots for kon and koff gave
sizable activation enthalpies ∆Hq and, more dramatically, Very
positiVe activation entropies ∆Sq for both the “on” and “off”
reactions for each complex. Similarly, plots of ln(kon) and ln-
‡
(koff) vs P were found to be linear,13 and the calculated ∆Von
‡
and ∆Voff values are substantially positive (Table 1).
The large and positive ∆Sq and, more emphatically, ∆Vq
values for kon and koff represent signatures for a substitution
mechanism dominated by ligand dissociation,14 i.e., eqs 2 and
3:
kon
Fe(Por) + NO y z Fe(Por)(NO)
(1)
koff
FeIII(Por)(H2O)2 yk1z FeIII(Por)(H2O) + H2O
(2)
(3)
been systematically probed. While it is often assumed that
reaction of a ligand with heme iron requires an open coordina-
tion site, the quantitative basis for such assumption, especially
for reaction with NO, is more legendary than factual. Further-
more, the possibility remains that NO, a stable free radical, may
react by pathway(s) different than other Lewis bases.11 To
address these mechanistic questions, we report the activation
k-1
FeIII(Por)(H2O) + NO yk2z FeIII(Por)(H2O)NO
k-2
Consistent with this mechanism is the report by Hunt et al.15
n-
that H2O exchange between solvent and FeIII(TPPS)(H2O)2
† University of California.
occurs at a first-order rate (kex ) 1.4 × 107 s-1 in 25 °C water)
far exceeding the kobsd values determined here for any [NO]. If
the steady-state approximation were taken with regard to
intermediate FeIII(Por)(H2O), the kobsd for the exponential
relaxation of the nonequilibrium mixture generated by the flash
photolysis experiment would be
‡ Institute of Physical and Chemical Research.
(1) (a) Moncada, S.; Palmer, R. M. J.; Higgs, E. A. Pharmacol. ReV.
1991, 43, 109-142. (b) Feldman, P. L.; Griffith, O. W.; Stuehr, D. J. Chem.
Eng. News 1993, 71, 10, 26-38.
(2) (a) Radi, R. Chem. Res. Toxicol. 1996, 9, 828-835. (b) Traylor, T.
G.; Sharma, V. S. Biochemistry 1992, 31, 2847-2849.
(3) Kim, S.; Deinum, G.; Gardner, M. T.; Marletta, M. M.; Babcock, G.
T. J. Am. Chem. Soc. 1996, 118, 8769-8770 and references therein.
(4) Cleeter, M. W. J.; Cooper, J. M.; Darley-Usmar, V. M.; Moncada,
S.; Scapira, A. H. V. FEBS Lett. 1994, 345, 50-54.
k1k2[NO] + k-1k-2[H2O]
(5) Noguchi, T.; Honda, J.; Nagamune, T.; Sasabe, H.; Inoue, Y.; Endo,
I. FEBS Lett. 1995, 358, 9-12.
kobsd
)
(4)
k-1[H2O] + k2[NO]
(6) Brown, G. C. Eur. J. Biochem. 1995, 232, 188-191.
(7) Ribiero, J. M. C.; Hazzard, J. M. H.; Nussenzveig, R. H.; Champagne,
D. E.; Walker, F. A. Science 1993, 260, 539-541.
Under the experimental conditions, one may conclude that
k-1[H2O] . k2[NO] since both steps involve nearly diffusion-
(8) Malinski, T.; Czuchajowski, C. In Methods in Nitric Oxide Research;
Feelish, M., Stamler, J. S., Eds.; J. Wiley and Sons: Chichester, England,
1996; Chapter 22 and references therein.
(12) (a) Barkey, M. H.; Takeuchi, K. J.; Meyer, T. J. J. Am. Chem. Soc.
1986, 108, 5876-5885. (b) Cheng, S.; Chen, Y.; Su, Y. O. J. Chin. Chem.
Soc. 1991, 38, 15-22.
(9) Hoshino, M.; Ozawa, K.; Seki, H.; Ford, P. C. J. Am. Chem. Soc.
1993, 115, 9568-9575
(10) (a) Tamura, M.; Kobayashi, K.; Hayashi, K. FEBS Lett. 1978, 88,
124-126. (b) Rose, E. J.; Hoffman, B.; J. Am. Chem. Soc. 1983, 105, 2866-
2873. (c) Cornelius, P. A.; Hochstrasser, R. M.; Steele, A. W. J. Mol. Biol.
1983, 163, 119-128. (d) Jongeward, K. A; Magde, D.; Taube, D. J.;
Marsters, J. C.; Traylor, T. G.; Sharma, V. S. J. Am. Chem. Soc. 1988,
110, 380-387.
(13) (a) Flash photolysis was performed in the pump-probe mode as
described previously.13b,c The activation volume was calculated from ∆Vq
i
) -RT(d ln(ki)/dP)T, where ki is the rate constant at a particular P.13c (b)
Crane, D. R.; Ford, P. C. J. Am. Chem. Soc. 1991, 113, 8510-8516. (c)
Traylor, T. G.; Luo, J.; Simon; J. A.; Ford, P. C. J. Am. Chem. Soc. 1992,
114, 4340-4345.
(11) (a) For example, reaction of NO with Ru(NH3)63+ to give Ru(NH3)5-
(14) Ducommun, Y.; Merbach, A. E. In Inorganic High Pressure
Chemistry; van Eldik, R., Ed.; Elsevier: Amsterdam, 1986; Chapter 2, pp
69-114.
(NO)3+ occurs via an associative mechanism at a second-order rate (kon
)
0.19 M-1 s-1) exceeding the lability of the coordinated ammines. (b) Armor,
J. N.; Scheidegger, H. A.; Taube, H. J. Am. Chem. Soc. 1968, 90, 5928-
5929.
(15) Ostrich, I. J.; Gordon, L.; Dodgen, H. W.; Hunt, J. P. Inorg. Chem.
1980, 19, 619.
S0002-7863(97)02448-7 CCC: $14.00 © 1997 American Chemical Society

12664 J. Am. Chem. Soc., Vol. 119, No. 51, 1997
Communications to the Editor
parameters in this case are largely defined by a dissociative
mechanism, the limiting example being eq 2. The key point is
n-
that, for the “on” reaction with the FeIII(Por)(H2O)2 com-
plexes, NO apparently displays reaction kinetics properties quite
similar to those for H2O owing to the dominance of ligand
dissociation in determining the behaVior of kon.
The principle of microscopic reversibility argues that the koff
pathway proceeds via the same intermediates. However,
activation parameters similar to those of kon are not required,
since the energetically dominant k-2 step involves NO dissocia-
tion. Formation of the nitrosyl complex is accompanied by
charge transfer to give a species often represented as FeII(NO+);
furthermore, the iron is no longer high-spin FeIII but (formally)
low-spin FeII. Thus, the reaction coordinate of NO dissociation
must incorporate the entropic and volume differences ac-
companying the solvation changes as charge redistributes and
Figure 1. Transient difference spectrum 50 µs after 355 nm flash
photolysis (15 mJ/pulse) of FeIII(TMPS)(H2O)(NO) in pH 6.0 aqueous
solution (25 °C). Inset: decay of transient bleaching at 426 nm ([NO]
) 1.7 × 10-3 M).
the spin crossover. This may explain why ∆Vq is more
off
positive than ∆Vq in both cases, but regardless of such
on
speculation, these activation parameters remain consistent with
the limiting mechanism described by eqs 2 and 3.
Table 1. Activation Parameters for Reaction of NO with
What do these results signify with regard to the reactivity of
NO with other heme centers? Hoshino et al.9 have summarized
the “on” rates for several FeII and FeIII heme proteins and noted
a range of more than 8 orders of magnitude in kon values. These
rates are in part a function of protein structure. For example,
“on” rates are very slow with both FeII and FeIII cytochrome c,
where the protein limits the access of NO to the metal site, even
though the equilibrium constants for complex formation are
large. When such access is available, facile reaction requires
either a very labile coordination site, such as the high-spin FeIII
heme centers I and II and catalase (kon ) 3.0 × 107 M-1 s-1),9
or, better, a vacant coordination site such as high-spin FeII heme
centers FeII(TPPS)(H2O)4- (1.8 × 109),9 myoglobin (1.7 ×
107),16 and, presumably, soluble guanylyl cyclase.3,17 This
suggests that the free radical nature of NO, which clearly has
utmost importance in determining the stability and chemical
properties of biologically relevant metal nitrosyl complexes, may
have but minor influence on the reaction dynamics to form such
complexes. Since the odd electron resides in a NO π* orbital,
its involvement with the metal center is unlikely to be significant
except at short distances where coordination is largely ac-
complished. Thus, in terms of its “on” reactions with FeII and
FeIII hemes, NO acts as a normal two-electron donor ligand in
the initial stages of its interaction with the metal center. We
are currently extending these activation parameter measurements
to NO reactions with relevant heme proteins.
FeIII(TMPS) and FeIII(TPPS) (eq 1)
FeIII(TMPS)(NO)
FeIII(TPPS)(NO)
kon
koff
kon
koff
∆Hq (kJ mol-1
∆Sq (J mol-1 K-1
∆Vq (cm3 mol-1
)
62 ( 2
83 ( 3
70 ( 3
78 ( 4
)
+86 ( 2 +89 ( 3 +100 ( 4
+67 ( 3
)
+13 ( 1 +18 ( 4
+8.3 ( 1.5 +17.9 ( 1.4
limited trapping of an unsaturated metal center and [H2O] .
[NO]. Accordingly, kon ) k1k2/k-1[H2O] and koff ) k-2. In
this context, the apparent activation parameters for kon would
be sums of terms, e.g.,
∆Sqon ) ∆Sq1 + ∆Sq2 - ∆Sq
-1
∆Vqon ) ∆Vq1 + ∆Vq2 - ∆Vq
(5)
-1
Since the k2 and the k-1 steps represent similar (very fast)
reactions of the unsaturated intermediate FeIII(Por)(H2O) with
an incoming ligand (NO and H2O, respectively), the differences
in their activation parameters (e.g., ∆Sq2 - ∆Sq-1 and ∆Vq
-
2
∆Vq-1) should be small. In such a case the principal contributor
to ∆Sq would be ∆Sq , the activation entropy for the H2O
on
1
dissociative step. The k1 step should thus display a positive
q
∆H1 reflecting the energy necessary to break the FeIII-OH2
bond, a large, positive ∆S1q owing to formation of two species
Acknowledgment. This work was supported by the National
Science Foundation (CHE 9400919). Jon Bridgewater and Brian Lee
helped with initial flash photolysis experiments.
from one without a significant change in solvation and a
q
substantially positive ∆V1 for the same reason. These condi-
tions are met for the kon activation parameters for both I and II
(Table 1). Furthermore, the values of ∆Hqex (57 kJ mol-1) and
∆Sqex (+84 J K-1 mol-1) for the H2O exchange15 on I are very
similar to the respective kon activation parameters. Thus, the
factors that determine H2O exchange kinetics for the di-aquo
species apparently dominate the kinetics of the NO kon step for
the same species. We therefore conclude that the reaction
Supporting Information Available: Absorption spectra and kobs
plots (3 pages). See any current masthead page for ordering and Internet
access instructions.
JA972448E
(16) Moore, E. G.; Gibson, Q. H. J. Biol. Chem. 1976, 251, 2788-2794.
(17) Burstyn, J. N.; Yu, A. E.; Dierks, E. A.; Hawkins, B. K.; Dawson,
J. H. Biochemistry 1995, 34, 5896-5903.