Hydrogen atom transfer reactions of iron-porphyrin-imidazole complexes as models for histidine-ligated heme reactivity

Article abstract of DOI:10.1021/ja711057t

Hydrogen atom transfer (HAT) reactions of the bis(histidine) cytochrome active site models (TPP)Fe^II(ImH)₂ (Fe^IIImH) and (TPP)Fe(Im)(ImH) (Fe^IIIIm) have been examined in acetonitrile solvent (TPP = tetraphenylporphyrin, ImH = 4-methylimidazole). The ascorbate derivative 5,6-isopropylidine ascorbate, hydroquinone, and the hydroxylamine TEMPOH all rapidly add H^? to Fe^IIIIm to give Fe^IIImH. Similarly, the phenoxyl radical 2,4,6-^tBu₃C₆H₂O^? and excess TEMPO^? each oxidize Fe^IIImH to give Fe^IIIIm. On the basis of redox potential, pKa, and equilibrium measurements, the N-H bond in Fe^IIImH was found to have a bond dissociation free energy (BDFE) of 70 ± 2 kcal mol^-1. A hydrogen atom transfer mechanism (concerted transfer of e- and H+) is indicated based on data for the ascorbate and TEMPO^? reactions. Copyright

Full text of DOI:10.1021/ja711057t

Published on Web 02/08/2008
Hydrogen Atom Transfer Reactions of Iron-Porphyrin-Imidazole Complexes
as Models for Histidine-Ligated Heme Reactivity
Jeffrey J. Warren and James M. Mayer*
UniVersity of Washington, Department of Chemistry, Box 351700, Seattle, Washington 98195
Received December 12, 2007; E-mail: mayer@chem.washington.edu
Scheme 1. Thermochemistry of the (TPP)Fe(ImH)2 System in
MeCN
Histidine-ligated hemes are key cofactors in a wide range of
biological redox processes. Many of these hemes are thought to
serve purely as electron-transfer cofactors; however their pH-
dependent redox potentials1 indicate that they can also react by
proton-coupled electron transfer (PCET2). Such PCET reactivity
has received relatively little attention. The clearest example is
perhaps the oxidation of ascorbate by cytochromes b561 by a
concerted proton-electron transfer (CPET) mechanism,3 as deter-
mined by Njus and co-workers using thermochemical consider-
ations.4 The heme b centers in the mitochondrial bc1 complex are
involved in the interconversion of quinones and hydroquinones,
which is inherently a PCET process. We report here that model
porphyrin-iron-bis(imidazole) complexes react readily by hydro-
gen atom transfer (HAT). HAT is, in our view,3,5 a type of CPET
reaction in which a proton and an electron are transferred in a single
kinetic step from one donor to one acceptor. These porphyrin-
imidazole complexes have been developed as structural, spectro-
scopic, and electron-transfer models for the growing family of
bis(histidine)-ligated heme proteins,6 which includes the cyt b561
and heme b centers mentioned above.7
Our model system uses iron complexes of tetraphenylporphyrin
(TPP) and 4-methylimidazole (ImH), following the studies of
Valentine et al.6 They showed that the use of 4-methylimidazole
prevents oligomerization of (TPP)Fe-imidazolate complexes. [(TPP)-
FeIII(ImH)2]PF6 (FeIIIImH, Scheme 1) was prepared by a modifica-
tion of their procedure.6 The new FeII derivative (TPP)FeII(ImH)2
(FeIIImH) was synthesized from (TPP)Fe8 and ImH, and was
characterized by 1H NMR and UV-vis spectroscopies and elemen-
tal analysis.9 The deprotonated FeIII and FeII complexes, (TPP)Fe(Im)-
(ImH)0/- (FeIIIIm6 and FeIIIm), were generated in situ from
FeIIIImH or FeIIImH with the base DBU (1,8-diazabicyclo(5.4.0)-
undec-7-ene). These complexes have not been isolated because of
weak binding of the second imidazole.6 Spectroscopic measurements
10.5 µM HAsc-), the reaction is 45% complete within the 2-4 ms
mixing time of our instrument. Global analysis of the available
spectral data using SpectFit13 gave good calculated spectra for
FeIIIIm and FeIIIIm and the rate constant k1 ) (3.5 ( 0.9) × 107
n
M-1 s-1.9 The deuterated Bu4N[DAsc] reacts more slowly under
similar conditions, in solutions containing 5 mM 4-methylimidazole-
ND. Global analysis gives k1D ) (7.5 ( 0.8) × 106 M-1 s-1 and
therefore kH/kD ) 4.6 ( 1.3.
in acetonitrile show that ImH binding to (TPP)FeIII(Im) has Keq
)
1300 ( 100 M-1, consistent with previous studies in toluene and
THF.6 All of the measurements below were performed in acetonitrile
containing 5 mM ImH, to ensure that only six-coordinate complexes
of both FeII and FeIII were present. Low-spin configurations are
indicated for FeIIIImH and FeIIIIm by the Evans method10 and for
n
The protonated derivative FeIIIImH is also reduced by Bu4N-
[HAsc] to FeIIImH, in a net electron transfer (ET) reaction.
Monitoring this reaction by stopped-flow UV-vis spectroscopy
shows that it is more than a factor of 10 slower than reaction 1
(Figure S3). The time course can be roughly fit to a second-order
rate law (with an apparent k of ∼2 × 106 M-1 s-1) but the kinetics
are clearly more complex, especially for the deuterated derivative
nBu4N[DAsc]. This may be due to rapid reactions of the ascorbyl
radical product, HAsc• or DAsc•, such as protonation of the HAsc-
reactant.14 DAsc- reduces FeIIIImH more slowly than HAsc- does
(Figure S4), but the kinetic complexity prevents obtaining a value
for kH/kD.
1
FeIIImH based upon its diamagnetic H NMR.9
The oxidation of ascorbate by cytochromes b561 has been modeled
by the reaction of FeIIIIm with the acetonitrile-soluble derivative
tetrabutylammonium 5,6-isopropylidene ascorbate (nBu4N[HAsc];
eq 1).11 This substrate should have similar reactivity to ascorbate
given the distance of the isopropylidene group from the enol
n
functionality.12 The reaction of Bu4N[HAsc] with a solution of
FeIIIIm results in UV-vis and 1H NMR spectral changes consistent
with quantitative conversion to FeIIImH (eq 1). This is a net
addition of H• to FeIIIIm. Stopped-flow kinetic measurements show
that the reaction proceeds very rapidly. Using second-order condi-
tions with the lowest practical concentrations (9.0 µM FeIIIIm +
Electron transfer from HAsc- to FeIIIImH cannot occur by an
inner-sphere (coordinated ascorbate) pathway because the exchange
9
2774
J. AM. CHEM. SOC. 2008, 130, 2774-2776
10.1021/ja711057t CCC: $40.75 © 2008 American Chemical Society

C O M M U N I C A T I O N S
1
of bound and free ImH in all these complexes is slow on the H
NMR time scale. Thus the reduction of FeIIIImH likely proceeds
via outer-sphere ET from HAsc-. Deprotonated FeIIIIm has a 0.36
V lower redox potential than FeIIIImH (see below), yet is reduced
substantially faster under the same conditions. This indicates that
the ascorbate reduction of FeIIIIm proceeds by a mechanism other
than outer-sphere ET, most likely by HAT. A HAT path is also
indicated by the primary H/D kinetic isotope effect of 4.6 ( 1.3.
FeIIIIm reacts rapidly with hydroquinone to give an equilibrium
mixture with FeIIImH and benzoquinone (eq 2). This equilibrium
can also be established from FeIIImH plus benzoquinone, as
observed in both directions by both UV-vis and 1H NMR
spectroscopies. Adding aliquots of benzoquinone to FeIIImH and
monitoring by optical spectroscopy yields K2 ) 180 ( 10 (∆G°2
) -3.1 ( 0.2 kcal mol-1). These reactions are potential models
for quinone/quinol interconversions by the b hemes in the bc1
complex.
Figure 1. Kinetic data for the reaction of FeIIIIm (15 µM) + 150 µM
TEMPOH in MeCN (eq 4): (a) visible spectra over 0.3 s; (b) plots of the
pseudo-first order kobs vs [TEMPOH/D].
FeIIImH is also rapidly and quantitatively oxidized by the stable
phenoxyl radical 2,4,6-tBu3C6H2O• to give FeIIIIm and the phenol
1
(eq 3; by UV-vis and H NMR spectroscopies).
of FeIIIIm with TEMPO-D proceeds more slowly, with k4D ) (2.2
( 0.1) × 104 M-1 s-1; the primary kinetic isotope effect (KIE) is
3.8 ( 0.4.
These reactions can be understood in terms of the thermochem-
istry of electron, proton, and hydrogen atom transfers in the iron
system.9 Cyclic voltammograms of FeIIIImH and FeIIImH show
a chemically reversible couple with EImH ) -0.585 ( 0.010 V
versus Cp2Fe0/+. Titrations of FeIIIImH and FeIIImH with Et3N
(pKa ) 18.5 in MeCN15) and DBU (pKa ) 24.315b), respectively,
were monitored by optical spectroscopy to give the pKa values
shown in Scheme 1. The FeII/IIIIm (EIm) redox potential is calculated
from these E° and pKa values, using the edges of Scheme 1 as a
closed thermochemical cycle [RT ln(KFeIII/KFeII) - F(EImH - EIm)
) 0]. Finally, eq 5 gives the bond dissociation free energy (BDFE)
for an N-H bond in FeIIImH as 70 ( 2 kcal mol-1 [CG(MeCN)
) 54.9 kcal mol-1 with E° versus Cp2Fe0/+].16 This value is also
independently determined from the equilibrium constants for
reactions 2 and 4 given previously. The BDFE of TEMPOH (66.5
To probe HAT processes in this system in more detail, the
reaction of FeIIIIm with the hydroxylamine TEMPOH in acetonitrile
has been examined. This reaction gives FeIIImH and the nitroxyl
radical TEMPO (eq 4), again by net H-atom (H+ + e-) transfer.
The equilibrium constant for reaction 4 was directly measured9 by
optical titration of FeIIImH with TEMPO, in MeCN at 298 K,
yielding K4 ) (4.5 ( 0.5) × 103 (∆G°4 ) -5.0 ( 0.2 kcal mol-1).
BDFE[X-H] )
23.06E° + 1.37pKa + CG (in kcal mol-1) (5)
( 1 kcal mol-1 17 and ∆G°4 ) -5.0 ( 0.2 kcal mol-1 give BDFE-
)
(FeIIImH) as 71.5 ( 1 kcal mol-1. Similarly, ∆G°2 ) -3.1 ( 0.2
kcal mol-1 and the average BDFE for the two hydroquinone O-H
Rate constants for reaction 4 have been determined in MeCN
using stopped-flow spectrophotometry (Figure 1).9 At 298 K, the
forward reaction has k4 ) (9.2 ( 0.7) × 104 M-1 s-1. k4 has been
measured with 1-16 equiv of TEMPOH, ranging from mixed
second-order approach to equilibrium conditions to pseudo-first-
order conditions. The reverse reaction between FeIIImH and
TEMPO has k-4 ) 20 ( 2 M-1 s-1 (measured using a large excess
of TEMPO). The ratio k4/k-4 ) (4.6 ( 0.5) × 103 is in excellent
agreement with the static equilibrium measurements above. Reaction
bonds (69 ( 2 kcal mol-1 18 gives a FeIIImH BDFE of 70.5 ( 2
)
kcal mol-1. These values are all in very good agreement.
The equilibrium constant for the TEMPOH reaction (eq 4) has
been measured from 276 to 331 K; van’t Hoff analysis gives ∆H°4
) -13.0 ( 1.0 kcal mol-1 and ∆S°4 ) -27 ( 3 cal K-1 mol-1
.
∆H°4 is the difference between the bond dissociation enthalpies
(BDEs) of FeIIImH and TEMPOH in MeCN; using BDE-
(TEMPO-H) ) 71.5 ( 1 kcal mol-1 19 the N-H BDE of FeIIImH
,
is 84.5 ( 2 kcal mol-1. The ground-state entropy change for reaction
9
J. AM. CHEM. SOC. VOL. 130, NO. 9, 2008 2775

C O M M U N I C A T I O N S
4 is quite substantial (T∆S°4 ) -8 kcal mol-1). We have previously
reported similarly large |∆S°| values for HAT reactions of non-
heme iron centers and found them to be an intrinsic property of
the FeIIIL/FeIILH redox couple.16c The nonzero entropies for these
reactions indicate that bond dissociation free energies need to be
used for HAT reactions, not the more commonly used BDEs.
The reaction of FeIIIIm and TEMPOH (eq 4) could in principle
occur by initial outer-sphere electron transfer followed by proton
transfer (ET/PT), initial PT followed by ET, or by concerted transfer
of the electron and proton (HAT/CPET).3,5 Using the thermochemi-
cal data in Scheme 1 and the properties of TEMPOH,20 the ∆G°ET
for initial electron transfer from TEMPOH to FeIIIIm to give FeIIIm
and TEMPOH•+ is +38 kcal mol-1. The barrier for initial ET must
2007, 46, 291-305. (c) Zu, Y.; Fee, J. A.; Hirst, J. J. Am. Chem. Soc.
2001, 123, 9906-9907. (d) Saraiva, L. M.; Fauque, G.; Besson, S.; Moura,
I. Eur. J. Biochem. 1994, 224, 1011-1017.
(2) (a) Huynh, M. H. V.; Meyer, T. J. Chem. ReV. 2007, 107, 5004-5064.
(b) Cukier, R. I.; Nocera, D. G. Ann. ReV. Phys. Chem. 1998, 49, 337-
369.
(3) Mayer, J. M. Ann. ReV. Phys. Chem. 2004, 55, 363-390.
(4) (a) Njus, D.; Jalukar, V.; Zu, J.; Kelley, P. M. Am. J. Clin. Nutr. 1991,
54, 1179S-1183S. (b) Njus, D.; Wigle, M.; Kelley, P. M.; Kipp, B. H.;
Schlegel, H. B. Biochemistry 2001, 40, 11905-11911.
(5) Reference 2a gives a different definition of HAT (p. 5024) that would
exclude the reactions described here, because the transferred H+ forms
an N-H σ bond while the e- formally adds to a different orbital, an iron
π-symmetry t2g-type orbital.
(6) Quinn, R.; Nappa, M.; Valentine, J. S. J. Am. Chem. Soc. 1982, 104,
2588-2595 and references therein.
(7) (a) Shikama, K.; Matsuoka, A. Crit. ReV. Biochem. Mol. 2004, 39, 217-
259. (b) Kundu, S.; Trent, J., III; Hargrove, M. S. Trends. Plant. Sci.
2003, 8, 387-393. (c) Su, D.; May, J. M.; Koury, M. J.; Asard, H. J.
Biol. Chem. 2006, 281, 39852-39839. (d) Schenkman, J. B.; Jansson, I.
Pharmacol. Therapeut. 2003, 97, 139-152.
(8) Collman, J. P.; Reed, C. A. J. Am. Chem. Soc. 1973, 95, 2048-2049.
(9) Full details are given in the Supporting Information.
be at least as large as this value: ∆Gq g ∆G°ET. Since this is
ET
much larger than the observed Eyring barrier, ∆Gq4 ) 10.3 ( 0.8
kcal mol-1, ET cannot be the pathway for reaction 4. Similarly,
initial PT to give TEMPO- and FeIIIImH has ∆Gq g ∆G°PT
)
PT
27 kcal mol-1, again much larger than the observed ∆Gq . Thus
4
(10) (a) Braun, S.; Kalinowski, H.-O.; Berger, S. 150 and More Basic NMR
Experiments Wiley-VCH: Weinheim, Germany, 1998. (b) Grant, D. H.
J. Chem. Ed. 1995, 72, 39-40.
neither stepwise path can be occurring. Initial HAT, where the
proton and electron are transferred in a single kinetic step, is much
more favorable (∆G°4 ) -5.0 ( 0.2 kcal mol-1) and is the only
one of these pathways that is thermodynamically viable. The
conclusion that reaction 4 proceeds via a HAT mechanism is
supported by the KIE of 3.8.
(11) From 5,6-isopropylidne ascorbic acid (Aldrich) + nBu4NOH.9
(12) For example: (a) Cabral, J.; Haake, P. J. Org. Chem. 1988, 53, 5742-
5750. (b) Zhang, L.; Lay, P. A. J. Am. Chem. Soc. 1996, 118, 12624-
12637.
(13) Binstead, R. A.; Zuberbu¨hler, A. D.; Jung, B. Specfit, version 3.0.38 (32-
bit Windows); Spectrum Software Associates: Chapel Hill, NC, 2006.
The ascorbate + FeIIIIm reaction (eq 1), as noted above, also
proceeds by an HAT mechanism. The BDFE for HAsc- in MeCN
has not been reported, but the aqueous BDFE for ascorbate is
calculated to be 74 ( 3 kcal mol-1 from eq 5, the aqueous
thermochemical data,21 and the aqueous CG of 57.5 ( 2 kcal
(14) The protonated radical HAsc• (pKa of HAsc• is -0.45 in water) is likely
to protonate HAsc- and disproportionate on the timescale of the reaction
of FeIIIImH + HAsc-. Ascorbic Acid: Chemistry, Metabolism, and Uses;
Seib, P. A., Tolbert, B. M., Eds.; Advances in Chemistry Series, 200;
American Chemical Society: Washington, D.C., 1982 (especially Bielski,
B. H. J., pp 81-100).
(15) (a) Isutzu, K. Acid-Base Dissociation Constants in Dipoloar Aprotic
SolVents; Blackwell Scientific: Oxford, U.K., 1990. (b) Kaljurand, I.; Ku¨tt,
A.; Soova¨li, L.; Rodima, T.; Ma¨emets, V.; Leito, I.; Koppel, I. A. J. Org.
Chem. 2005, 70, 1019-1028.
(16) (a) Bordwell, F. G.; Cheng, J.-P.; Harrelson, J. A. J. Am. Chem. Soc.
1988, 110, 1229-1231. (b) Tilset, M.; Parker, V. D. J. Am. Chem. Soc.
1989, 111, 6711-6717; 1990, 112, 2843. (c) Mader, E. A.; Davidson, E.
R.; Mayer, J. M. J. Am. Chem. Soc. 2007, 129, 5153-5166.
(17) (a) Semmelhack, M. F.; Chou, C. S.; Cortes, D. A. J. Am. Chem. Soc.
1983, 105, 4492-4494. (b) Mori, Y.; Sakaguchi, Y.; Hayashi, H. J. Phys.
Chem. A 2000, 104, 4896-4905. (c) Bordwell, F. G.; Liu, W.-Z. J. Am.
Chem. Soc. 1996, 118, 10819-10823. (d) Chantooni, M. K., Jr.; Kolthoff,
I. M. J. Phys. Chem. 1976, 80, 1306-1310.
(18) As described in the Supporting Information, the average hydroquinone
BDFE was calculated from the known gas-phase thermochemistry18a,b and
estimates of the free energies of solvation. (a) NIST Chemistry Webbook,
http://webbook.nist.gov/chemistry. (b) Sf°(g)(benzoquinone): Burcat, A.;
Ruscic, B. Third Millennial Ideal Gas and Condensed Phase Thermo-
chemical Database for Combustion with Updates from Active Thermo-
chemical Tables. ftp://ftp.technion.ac.il/pub/supported/aetdd/thermody-
namics (accessed 15 July 2007).
(19) (a) Mader, E. A. Ph.D. Thesis, University of Washington, Seattle, WA,
2007. (b) Mader, E. A.; Manner, V. W.; Wu, A.; Mayer, J. M., manuscript
in preparation.
(20) pKa ) 41 (converted from measurement in DMSO), E° ) 0.71 V.9
(21) Williams, N. H.; Yandell, J. K. Aust. J. Chem. 1982, 35, 1133-1144.
(22) (a) Bisby, R. H.; Parker, A. W. J. Am. Chem. Soc. 1995, 117, 5664-
5670. (b) Barclay, L. R. C.; Dakin, K. A.; Zahalka, H. A. Can. J. Chem.
1992, 70, 2148-2153.
mol-1 16
Our preliminary thermochemical data suggest that the
.
BDFE is lower in MeCN but not lower than the BDFE of
TEMPOH. In this light, the 3.5 × 107 M-1 s-1 rate constant for
reaction 1 is rapid, suggesting a small intrinsic barrier to HAT.
Ascorbate has been shown to be a competent H-atom donor in both
water and acetonitrile,22 and Njus et al. showed that cyt b561 reacts
with ascorbate by CPET.4 Njus did not suggest a deprotonated
histidinate ligand analogous to FeIIIIm but such ligands have been
implicated in proton-coupled ET reactions of cyt b56123 and Rieske
proteins,24 and discussed for other heme cofactors.25
In conclusion, the FeIIIIm and FeIIImH complexes that are
models for heme cofactors undergo facile reactions in acetonitrile
with an ascorbate derivative, hydroquinone and benzoquinone,
phenoxyl and nitroxyl radicals, and a hydroxylamine. These
reactions are potential models for biological reactions of histidine-
ligated hemes with oxyl radicals and with hydroxyl substrates. The
TEMPO•/TEMPOH and ascorbate reactions proceed by a hydrogen
atom transfer (HAT) pathway, a type of concerted proton-electron
transfer (CPET). On the basis of these results, HAT reactions should
be considered as part of the primary arsenal of reactivity of
histidine-ligated heme cofactors.
(23) Takigami, T.; Takeuchi, F.; Nakagama, M.; Hase, T.; Tsubaki, M.
Acknowledgment. We gratefully acknowledge support from
U.S. National Institutes of Health (GM50422) and the University
of Washington.
Biochemistry 2003, 42, 8110-8118.
(24) (a) Zu, Y.; Fee, J. A.; Hirst, J. J. Am. Chem. Soc. 2001, 123, 9906-9907.
(b) Hunsicker-Wang, L. M.; Heine, A.; Chen, Y.; Luna, E. P.; Todaro,
T.; Zhang, Y. M.; Williams, P. A.; McRee, D. E.; Hirst, J.; Stout, C. D.;
Fee, J. A. Biochemistry 2003, 42, 7303-7317.
(25) (a) Barker, P. D.; Butler, J. L; de Oliveira, P.; Hill, H. A. O.; Hunt, N. I.
Inorg. Chim. Acta. 1996, 252, 71-77. (b) Arnesano, F.; Banci, L.; Bertini,
I.; Ciofi-Baffoni, S.; Woodyear, T. L.; Johnson, C. M.; Barker, P. D.
Biochemistry 2000, 39, 1499-1514. (c) Marques, H. M.; Perry, C. B. J.
Inorg. Biochem. 1999, 75, 281-291. (d) Gattistuzzi, G.; Bellei, M.;
Borsari, M.; Di Rocco, G.; Ranieri, A.; Sola, M. J. Biol. Inorg. Chem.
2005, 10, 643-651.
Supporting Information Available: Experimental details for
synthetic, kinetic, and thermochemical studies. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Reid, L. S.; Lim, A. R.; Mauk, A. G. J. Am. Chem. Soc. 1986, 108,
8197-8201. (b) Reddi, A.; Reedy, C.; Mui, S.; Gibney, B. Biochemistry
JA711057T
9
2776 J. AM. CHEM. SOC. VOL. 130, NO. 9, 2008