Stability of a ferryl-peptide conjugate is controlled by a remote substituent

Article abstract of DOI:10.1021/ic9013653

The formation of a synthetic ferryl-peptide conjugate and mechanistic studies that elucidate its mode of decomposition are presented. A ferryl species is generated from a ligand-dipeptide conjugate 4. The ferryl species [Fe ^IV(4)(0)]²⁺

Full text of DOI:10.1021/ic9013653

8078 Inorg. Chem. 2009, 48, 8078–8080
DOI: 10.1021/ic9013653
Stability of a Ferryl-Peptide Conjugate Is Controlled by a Remote Substituent
Nitinkumar D. Jabre, Lew Hryhorczuk, and Jeremy J. Kodanko*
Department of Chemistry and Central Instrumentation Facility, Wayne State University, 5101 Cass Avenue,
Detroit, Michigan 48202
Received July 13, 2009
The formation of a synthetic ferryl-peptide conjugate and me-
chanistic studies that elucidate its mode of decomposition are
presented. A ferryl species is generated from a ligand-dipeptide
conjugate 4. The ferryl species [FeIV(4)(O)]2þ, noted as com-
pound 5, was characterized by UV-vis spectroscopy and by high-
resolution electrospray mass spectrometry. The ferryl-peptide
conjugate 5 is stable for over 1 h at room temperature. Ester
derivatives of 5 decay at different rates, consistent with the remote
ester group controlling the stability of the ferryl. The kinetic isotope
effect value (4.5) and F = -1.3 observed with ester derivatives
suggest that the mechanism for decomposition of 5 follows a
hydrogen-atom-transfer pathway. The formation and decay of 5
was fit to a two-step process, with the decay being unimolecular
with respect to the ferryl 5.
artifical oxygenases that generate ferryl species. In this
Communication, we report the interesting observation that
the stability of a ferryl-peptide conjugate can be controlled
by a remote functional group that is 11 atoms removed from
the iron center. In addition, reported mechanistic studies
support an intramolecular decay pathway for the ferryl-
peptide conjugate.
In order to synthesize the ferryl-peptide conjugate, syn-
thetic methodology was developed for attaching pentaden-
tate ligands like N4Py to an alanine side chain. Toward this
goal, the unnatural amino acid 2, hydroxymethylpyridylala-
nine (HPA), was synthesized in racemic form in two steps
starting from bromide 1 using phase-transfer alkylation as
the key step (Scheme 1).9 After acetylation, the unnatural
amino acid was incorporated into dipeptide 3 upon coupling
with H-Gly-OBn. Attaching the remainder of the metal-
binding group was accomplished by transforming the sily-
loxy group of 3 into a chloride, followed by displacement
with the secondary amine NH(CH2Py)(CHPy2) (where Py=
2-pyridyl). Installation of the metal-binding unit in the last
step was ideal because it maximized the divergence of the
synthesis. Furthermore, this approach holds the advantage
that the amino acid 2 can be placed at varied positions along a
peptide chain.
To form the ferrous complex of the peptide, a solution of 4 in
H2O/CH3CN (1:1) was treated with 1 equiv of FeII(ClO4)2 to
generate the species in situ. Spectral data for [FeII(4)-
(CH3CN)]2þ are in good agreement with those for the parent
ligand N4Py in the same solvent mixture (Table S1 in the
Supporting Information, SI). The complex [FeII(4)(CH3-
CN)]2þ was characterized by UV-vis (ε380=4200 M-1 cm-1
and ε455=3600 M-1 cm-1) and NMR spectroscopy, as well as
by electrospray mass spectrometry (ESMS; Figures S1 and
S2 in the SI). All resonances in the NMR spectrum of
[FeII(4)(CH3CN)]2þ lie between 10 and 0 ppm, consistent with
the ferrous complex being in the low-spin state.
High-valent iron(IV) oxo species, also known as ferryls,
are found in enzymes that carry out oxidation reactions in
nature.1,2 Ligands that mimic the active sites of these enzymes
can be used to generate biomimetic ferryl species3,4 that
oxidize a variety of organic molecules in an intermolecular
fashion.5,6 In contrast, reactions of ferryls tethered to organic
substrates have received little attention.7,8 In fact, it is not
currently understood which functional groups are compati-
ble with a ferryl in a complex molecule. This is unfortunate
because these data are relevant to the understanding of which
functionalities would be tolerated in synthetic catalysts or
*To whom correspondence should be addressed. E-mail: jkodanko@
chem.wayne.edu.
(1) Krebs, C.; Fujimori, D. G.; Walsh, C. T.; Bollinger, J. M. Jr. Acc.
Chem. Res. 2007, 40, 484–492.
(2) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L. Jr. Chem. Rev. 2004,
104, 939–986.
(3) Que, L. Acc. Chem. Res. 2007, 40, 493–500.
(4) Nam, W. Acc. Chem. Res. 2007, 40, 522–531.
(5) Kaizer, J.; Klinker, E. J.; Oh, N. Y.; Rohde, J.-U.; Song, W. J.;
::
The ferrous complex [FeII(4)(CH3CN)]2þ was used to
generate a ferryl with impressive stability, considering the
weak C-H bonds present in the peptide placed in close proxi-
mity to the iron center. Treatment of [FeII(4)(CH3CN)]2þ
with KHSO5 at 25 °C generated a pale-green species, noted
as compound 5, with a maximum absorption wavelength at
Stubna, A.; Kim, J.; Munck, E.; Nam, W.; Que, L. Jr. J. Am. Chem. Soc.
2004, 126, 472–473.
(6) van den Berg, T. A.; de Boer, J. W.; Browne, W. R.; Roelfes, G.;
Feringa, B. L. Chem. Commun. 2004, 2550–2551.
(7) Jensen, M. P.; Lange, S. J.; Mehn, M. P.; Que, E. L.; Que, L. Jr. J. Am.
Chem. Soc. 2003, 125, 2113–2128.
(8) For characterization of an FeIII(OOH) species derived from N4Py
bound to a peptide, see: (a) Choma, C. T.; Schudde, E. P.; Kellogg, R. M.;
Robillard, G. T.; Feringa, B. L. J. Chem. Soc., Perkin Trans. 1 1998, 769–774.
(b) van den Heuvel, M.; van den Berg, T. A.; Kellogg, R. M.; Choma, C. T.;
Feringa, B. L. J. Org. Chem. 2004, 69, 250–262.
(9) The enantioselective synthesis of 2 will be reported elsewhere.
r
pubs.acs.org/IC
Published on Web 07/31/2009
2009 American Chemical Society

Communication
Inorganic Chemistry, Vol. 48, No. 17, 2009 8079
Scheme 1. Synthesis of Ligand-Peptide Conjugate 4a
Table 1. Observed and Relative Rate Constants for the Decomposition of
Ferryl-Peptide Conjugates Generated from Various Ester Derivatives
entry
R
rate (s-1 a
)
krel
1.0
0.086
0.22
2.2
1
2
3
4
5
6
-Bn (4)
-Me (6)
7.3(6) ꢀ 10-4
6.3(3) ꢀ 10-5
1.6(2) ꢀ 10-4
1.6(3) ꢀ 10-3
1.2(2) ꢀ 10-3
3.6(2) ꢀ 10-4
a (a) Benzyl 2-(diphenylmethyleneamino)ethanoate, CH2Cl2/toluene
(3:7), TBAH, 50% NaOH, 4 h, 0 °C, 88%; (b) Pd/C, MeOH, rt, 6 h,
77%;(c) Ac2O, Et3N, CH2Cl2, rt, 4 h, quantitative; (d) H-Gly-OBn, EDAC,
HOBt, Et3N, CH2Cl2, rt, 18 h, 72%; (e) TBAF, AcOH, THF, rt, 1.5 h, 76%;
(f) LiCl, iPr2EtN, TsCl, CH3CN, 0 °C to rt, 6 h, 70%; (g) NH(CH2Py)-
(CHPy2), iPr2EtN, NaI (catalyst), CH3CN, 55 °C, 24 h, 90%.
-CD2C6D5 (7)
-CH2C6H4-p-OMe (8)
-CH2C6H4-p-Me (9)
-CH2C6H4-p-Cl (10)
1.6
0.49
a Rates were determined from the slope of ln(Abs) vs time plots. The
number reported is the average of at least three runs; the error as the
standard deviation is given in parentheses.
maximizes after approximately 10 min and then decays via a
first-order process, as evidenced by the linearity of the
ln(Abs) vs time plot (Figure 1b). The induction period
(∼10 min) indicates that formation of 5 occurs along with
decay at early time points. Therefore, 5 is never fully
generated. Extrapolation of the linear portion of the plot in
Figure 1b to t = 0 allows the calculation of an extinction
coefficient for 5 (ε680=200 M-1 cm-1), which is consistent
with data for [FeIV(O)(N4Py)]2þ under the same conditions
(Table S1 in the SI). The high-resolution ESMS spectrum for
5 has a dominant molecular ion at m/z 357.6093, along with a
suitable isotopic pattern, which matches the formula
[FeIV(O)(4)]2þ (5; Figure S2 in the SI). This molecular ion
appears and then disappears on the same time scale as the
absorption at 680 nm, consistent with its assignment as the
ferryl species.
Although an isosbestic point at 550 nm was observed, the
FeII species (λmax at 450 nm) was not fully regenerated after
the decay process, indicating that partial destruction of the
metal-binding unit occurred. MS and NMR analysis of the
decomposition products revealed that a complex mixture was
formed. Several products were identified, including dipyridyl
ketone (25%), from oxidative decomposition of the N4Py
metal-binding unit,12 the free acid derived from 4 (32%) and
benzaldehyde (8%), consistent with oxidation of the ester
substituent (Scheme S1 in the SI). The reactivity of 5
encouraged us to investigate the mechanism of decomposi-
tion further in order to pinpoint the group responsible for
quenching the ferryl species on the peptide.
Derivatives of the ferryl-peptide conjugate 5 were synthe-
sized in order to determine if a functional group(s) contained
in 5 controlled the rate of decomposition (Table 1). The ferryl
generated from methyl ester derivative 6 was found to be 12
times more stable than 5, clearly implicating the role of
the benzyl moiety in controlling the decomposition. To test
this hypothesis further, the benzyl group was replaced by
benzyl-d7 (7). The species [FeIV(O)(7)]2þ decomposed much
slower than 5, giving a kinetic isotope effect (KIE) of 4.5,
Figure 1. (a) Decomposition of the ferryl-peptide conjugate
[FeIV(O)(4)]2þ (5) monitored by UV-vis spectroscopy. The decrease in
λ
max = 680 nm and increase in λmax = 450 nm corresponds to the
decomposition of FeIV and regeneration of FeII, respectively. (b) Plot of
ln(Abs) at 680 nm vs time (red) and linear fit (blue) for species 5 ([Fe] =
1.67 mM).
λ
max=680 nm(Figure 1a), identical with that obtainedfor the
species [FeIV(O)(N4Py)]2þ 10,11
.
This absorption at 680 nm
(10) Sastri, C. V.; Seo, M. S.; Park, M. J.; Kim, K. M.; Nam, W. Chem.
Commun. 2005, 1405–1407.
(11) Klinker, E. J.; Kaizer, J.; Brennessel, W. W.; Woodrum, N. L.;
Cramer, C. J.; Que, L. Jr. Angew. Chem., Int. Ed. 2005, 44, 3690–3694.
(12) Abouelatta, A. I.; Campanali, A. A.; Ekkati, A. R.; Shamoun, M.;
Kalapugama, S.; Kodanko, J. J. Inorg. Chem. 2009, ASAP.

8080 Inorganic Chemistry, Vol. 48, No. 17, 2009
Jabre et al.
Figure 3. Proposedmechanism for the generation anddecomposition of
the ferryl-peptide conjugate 5.
Figure 2. Hammett plot derived from decomposition rates of benzyl
ester derivatives 4 and 8-10 (F = -1.3; r = 0.99).
was considered a unimolecular reaction (Figure 3). Accord-
ingly, second- and first-order rate constants of 5.2(4) M-1 s-1
and 5.9(1) ꢀ 10-4 s-1 were determined for k1 and k2,
respectively. The first-order rate constant k2 determined by
this method was within reasonable error of the rate constant
measured from the slope of the ln(Abs) vs time plot shown in
Figure 1b (Table 1, entry 1), which confirms that the slope
can be used to estimate k2. The proposal that the decay of 5 is
an intramolecular process is further supported by the fact
that the addition of 1 equiv of benzyl acetate or benzyl
alcohol to 5 has no effect on its rate of decomposition
(Table S2 in the SI). However, rates do change slowly
between 5 and 10 equiv of external substrate added, consis-
tent with a unimolecular reaction becoming competitive with
a bimolecular pathway only at higher concentrations of
external substrate.
In conclusion, ferryls can be conjugated to peptides, and
their stability can be controlled by changing the nature of a
remote functional group. An intramolecular pathway is
proposed for the decay of a ferryl-peptide conjugate. Studies
are now underway in our laboratory to understand and
define the functional group compatibility of ferryl groups
in larger peptides, in terms of both their identity and their
spatial relationship to the ferryl center. Furthermore, the
methodology disclosed herein is expected to be useful for the
synthesis of peptides or proteins that contain these metal-
binding units.
strikingly similar to the KIE of 4.4 observed for oxidative
dealkylation of ethyl esters by the heme protein cytochrome
P-450.13 Three additional para-X-substituted benzyl ester
derivatives (8-10) were tested, and their rates of decomposi-
tion were used to construct a Hammett plot (Figure 2). A
strong linear correlation (r=0.99) between σp and log krel was
obtained, giving a F value of -1.3. The small F value suggests
abstraction of a benzylic hydrogen atom by the ferryl group
of 5 rather than reaction of the ferryl with the aromatic ring,
where a more negative F value would be expected.14 The
F value is similar to that observed in the oxidation of
para-substituted toluene derivatives by cytochrome P-450
(F = -1.6) and signifies that there is a substantial positive
charge built up on the benzylic center during homolytic
cleavage.15 Regarding the nature of the bond cleavage event,
the magnitude of the F value is in good agreement with a
hydrogen-atom-transfer (HAT) model rather than an elec-
tron-transfer, proton-transfer reaction, where a larger sub-
stituent effect (more negative F value) would be expected.
However, the observed KIE was smaller than what is
typically expected for HAT reactions with similar ferryls.5
The first-order decay of the ferryl 5 at later time points is
consistent with the decomposition reaction being unimole-
cular and not bimolecular with respect to 5. In order to probe
this process further, data were collected over a variable
concentration range ([Fe]=0.8-1.7 mM),16 and these data
were evaluated against several theoretical models using the
program DynaFit (see the SI for further details).17 The data
fit best to a two-step process, where formation of the ferryl
was treated as a bimolecular reaction and decay of the ferryl
Acknowledgment. We thank Alyssa Yousif for the
preparation of synthetic intermediates, Dr. Domenico
Gatti for assistance with the DynaFit software, and
Wayne State University for its generous support of this
research.
(13) Guengerich, F. P. J. Biol. Chem. 1987, 262, 8459–8462.
(14) de Visser, S. P.; Oh, K.; Han, A.-R.; Nam, W. Inorg. Chem. 2007, 46,
4632–4641.
(15) Blake, R. C. II; Coon, M. J. J. Biol. Chem. 1981, 256, 12127–12133.
(16) Data were not collected at lower concentrations because of the low
extinction coefficient of 5.
Supporting Information Available: Experimental procedures
for the preparation of 1-10, including characterization data and
kinetic fits. This material is available free of charge via the
Internet at http://pubs.acs.org.
(17) Kuzmic, P. Anal. Biochem. 1996, 237, 260–273.