Synthesis of Cytochrome c Oxidase Models Bearing a Tyr244 Mimic

Article abstract of DOI:10.1021/jo0499625

A close structural analogue of the metal-free cytochrome c oxidase active site has been synthesized. This model has a proximal imidazole tail and three distal imidazole pickets attached to a porphyrin. One distal imidazole is cross-linked to a phenol, mim

Full text of DOI:10.1021/jo0499625

Two regioisomers are presented: N-methylimidazole
rings may lie either on cis meso-positions 10 and 15 which
makes the porphyrin asymmetric (1-cis) or on trans
meso-positions 10 and 20 which makes the porphyrin
symmetric (1-tr a n s). The free base porphyrins 1-cis and
-tr a n s share many structural features with our previous
reported models, such as R3â-atropisomerism, two distal
N-methyl-substituted imidazoles, and a proximal CF3Ph-
imidazole tail (1F ) or a nonsubstituted H-tail (1H).4a-c
Herein, we describe the synthesis of 1F -cis, the closest
structural analogue of the metal-free cytochrome c oxi-
dase active site. Its key feature is a phenol-containing
imidazole 2 in the distal area.
The synthesis of 2 starts from the key intermediate 4
which we described previously.3d A new procedure using
the H2WO4/H2O2 couple5 was adopted to desulfurize 4.
This led to 5 with an improved yield and a simple workup
compared with the earlier HNO3/NaNO2 method.3 The
desired phenoxy-substituted imidazole 2 was obtained
after acid hydrolysis of 5 and subsequent formation of
the acid chloride 6. The synthesis of 2 requires nine steps
and gives an 8% overall yield starting from 2-aminophe-
nol (Scheme 1).
Although the synthetic route to 1 appears simple, the
first challenge in its preparation relies on developing
methods to introduce two different imidazoles onto the
R3 face of the porphyrin 3 in good yield. Since the three
distal amines in 3 are not identical, we envisioned that
the introduction of one or two imidazoles would lead to
two regioisomers. The second challenge is separation of
the very polar porphyrin regioisomers.
We developed two synthetic routes to 1 starting from
either of two porphyrin synthons R3AâF3T (3F )4a-c or
R3AâT (3H):4a,b (1) introduction of the phenol substituted
imidazole 2 first and subsequently the two simple imi-
dazoles 7, the “[1 + 2] approach” (route A, Scheme 2);
(2) introduction of the two simple imidazoles 7 first and
subsequently the phenol-substituted imidazole 2, the
“[2 + 1] approach” (route B, Scheme 2).
Syn th esis of Cytoch r om e c Oxid a se Mod els
Bea r in g a Tyr 244 Mim ic
J ames P. Collman,* Richard A. Decre´au, and Chi Zhang
Department of Chemistry, Stanford University,
Stanford, California 94305-5080
jpc@stanford.edu
Received J anuary 5, 2004
Abstr a ct: A close structural analogue of the metal-free
cytochrome c oxidase active site has been synthesized. This
model has a proximal imidazole tail and three distal imi-
dazole pickets attached to a porphyrin. One distal imidazole
is cross-linked to a phenol, mimicking Tyr244. The strategy
behind the successful synthesis of this regioisomerically pure
model involved discovering the best sequence to introduce
the phenol-substituted imidazole and employing a fluori-
nated substituent.
Cytochrome c oxidase is the terminal enzyme in the
respiratory chains of mitochondria and aerobic bacteria.1a-e
One of the three His coordinating CuB in cytochrome c
oxidase (CcO) is cross-linked to a phenol residue from
tyrosine. This phenol that our previous CcO models
lacked is thought to play a key role in the 4H+, 4e-
reduction of O2 to H2O and has been the subject of model
studies.2a-j,3a-c Recently, this Tyr244-His240 moiety has
been studied in some non-heme models,2a-j where other
organic components were missing (heme, proximal imi-
dazole). To obtain a closer analogue of the enzyme active
site, we have designed 1, which has one of the distal
imidazole rings N-substituted with a phenol (Figure 1).
(1) (a) Ferguson-Miller, S.; Babcock, G. T. Chem. Rev. 1996, 96, 2889.
(b) Iwata, S.; Ostermeier, C.; Ludwig, B.; Michel, H. S. Nature 1995,
376, 660. (c) Tsukihara, T.; Aoyama, H.; Yamashita, E.; Tomizaki, T.;
Yamaguchi, H.; ShinzawaItoh, K.; Nakashima, R.; Yaono, R.; Yoshika-
wa, S. Science 1996, 272, 1136. (d) Tsukihara, T.; Aoyama, H.;
Yamashita, E.; Tomizaki, T.; Yamagushi, H.; Shinzawaitoh, K.; Na-
kashima, R.; Yaono, R.; Yoshikawa, S. Science, 1995, 269, 1069. (e)
Yoshikawa, S.; Shinzawa-Itoh, K.; Nakashima, R.; Yaono, R.; Ya-
mashita, E.; Inoue, N.; Yao, M.; Fei, M. J .; Libeu, C. P.; Mizushima,
T.; Yamagushi, H.; Tomizaki, T.; Tsukihara, T. Science, 1998, 280,
1723.
(2) (a) Gennis, R. B. Biophys. Biochim. Acta 1998, 1365, 241. (b)
MacMillan, F.; Kannt, A.; Behr, J .; Prisner, T.; Michel, H. Biochemistry
1999, 38, 9179. (c) Proshlyakov, D. A.; Pressler, M. A.; Babcock, G. T.
Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 8020. (d) Proshlyakov, D. A.;
Pressler, M. A.; DeMaso, C.; Leykam, J . F.; DeWitt, D. L.; Babcock, G.
T. Science 2000, 290, 1588. (e) Sucheta, A.; Szundi, I.; Einarsdottir,
O. Biochemistry 1998, 37, 17905. (f) Collman, J . P.; Zhong, M.;
Constanzo, S.; Zhang, C. J . Org. Chem. 2001, 66, 8252. (g) Collman,
J . P.; Wang, Z.; Zhong, M.; Zeng, L. J . Chem. Soc., Perkin Trans. 1
2000, 8, 1217. (h) Aki, M.; Ogura, T.; Naruta, Y.; Le, TH.; Sato, T.;
Kitagawa, T. J . Phys. Chem. A 2002, 106, 3436. (i) Cappuccio, J . A.;
Ayala, I.; Elliott, G. I.; Szundi, I.; Lewis, J .; Konopelski, J . P.; Barry,
B. A.; Einarsdottir, O. J . Am. Chem. Soc. 2002, 124, 1750. (j) Kamaraj,
K.; Kim, E.; Galliker, B.; Zakharov, L. N.; Rheingold, A. L.; Zuberbu-
hler, A.; Karlin, K. D. J . Am. Chem. Soc. 2003, 125, 6028.
These two strategies have advantages and disadvan-
tages regarding (a) yields of introducing the first imida-
zole leading to 8m on o ([1 + 2] approach) or 9bis ([2 + 1]
approach), (b) byproduct formation and recycling, (c)
yields of introducing the second imidazole, (d) chromato-
graphic separation of polar porphyrin intermediates from
the polar porphyrin starting materials, and (e) chromato-
graphic separation of regioisomers of polar porphyrins.
All of these issues are compared for each approach in
order to determine which offers the best compromise to
these problems and gives an optimum pathway to obtain
1.
(4) (a) Collman, J . P.; Broring, M.; Fu, L.; Rapta, M.; Schwenniger,
R.; Straumanis, A. J . Org. Chem. 1998, 63, 8082. (b) Collman, J . P.;
Broring, M.; Fu, L.; Rapta, M.; Schwenniger, R. J . Org. Chem. 1998,
63, 8084. (c) Collman, J . P.; Sunderland, C.; Boulatov, R. Inorg. Chem.
2002, 41, 2282.
(5) Chang, J . H.; Lee, K. W.; Nam, D. H.; Kim, W. S.; Shin, H. Org.
Process Res. Dev. 2002, 6, 674.
(3) (a) Boulatov, R.; Collman, J . P.; Shiryaeve, I. M.; Sunderland,
C. J . J . Am. Chem. Soc. 2002, 124, 11923. (b) Collman, J . P.; Boulatov,
R. Angew. Chem., Int. Ed. 2002, 41, 3487. (c) Collman, J . P.; Boulatov,
R.; Sunderland, C. J . In The Porphyrin Handbook; Academic Press:
Boston, 2003; Vol. 11, Chapter 63, pp 1-49. (d). Collman, J . P.;
Fudickar, W. Inorg. Chem. 2003, 42, 3384.
10.1021/jo0499625 CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/09/2004
3546
J . Org. Chem. 2004, 69, 3546-3549

F IGURE 1. (A) Chemical structure of the target models. (B) Heme a3/CuB of bovine cytochrome c oxidase.1e The C atoms are
gray, the N atoms are blue, the O and Fe atoms are red, and Cu is pink.
SCHEME 1 a
a
Reagents and conditions: (a) H2WO4, H2O2, Et2NC6H6, rt, 5 h, 95%; (b) HCl, 100 °C, 92%; (c) (COCl)2, CH3CN, cat. DMF, 4 h, rt,
quantitative.
In the first step of the [1 + 2] approach (route A), the
target intermediate is the monoimidazole porphyrin 8m on o
which is obtained from reaction of N-methoxyphenyl-
containing imidazole 2 with 3 (Scheme 2). A survey of
the 2/3 ratio vis-a`-vis dilution conditions showed that a
large excess of picket 2 leads to high conversion; however,
the byproducts 8bis and 8tr is were obtained in 17% and
14% yield, respectively, while the yield of the target 8m on o
is only 21%. Since 8bis-tr is are not useful for the prepara-
tion of 1, fewer equivalents of 2 were added. This led to
a low conversion (20%) but to a more satisfactory ratio
mono/bis/tris (4:1:1). However, this approach is unsatis-
factory because purification is difficult: the Rf values of
8m on o and the recovered starting material 3 are so close,
0.55 and 0.60, respectively, that it makes the isolation
of 8m on o difficult. This became even more troublesome
when fewer equivalents of 2 were used.
In the first step of the [2 + 1] approach (route B), the
target intermediate is the bis imidazole porphyrin species
9bis obtained from reaction between 7 and 3 (Scheme 2).
The intermediate 9bis is obtained together with the mono-
and tris-condensation products 9m on o and 9tr is. In contrast
to the [1 + 2] approach where two useless byproducts
8bis and 8tr is are formed, the useful intermediate 9m on o can
8tr is). Moreover, the [2 + 1] approach is more satisfactory
in terms of isolating the target porphyrin from the
statistical mixture, compared with the [1 + 2] approach
because the Rf values of bis(N-methylimidazole)porphy-
rins 9bis are very different from the starting material 3
in contrast to what we found with monomethoxyphenyl-
containing-imidazole porphyrins 8m on o
.
Another criteria for comparing the [1 + 2] and [2 + 1]
approaches is the separation of regioisomeric intermedi-
ates. Regardless of the type of imidazole introduced,
separation of regioisomers is more facile with monoimi-
dazole porphyrins than with bis-imidazole porphyrins.
Therefore, the [1 + 2] approach is superior to the [2 + 1]
approach for cleanly separating the regioisomers.6 Also,
we found that the nature of the tail appears to influence
the ease of regioisomer separation. For example, regioi-
somer separation was found to be more straightforward
with the CF3Ph-tailed (F) than with H-tailed porphyrins
(H), especially with the bis-imidazole porphyrin species
9bis. Finally, we found that the ratio of cis/trans com-
pounds obtained was about 2:1 for the mono- as well as
the bis-porphyrins. It should also be noted that in
contrast to the -tr a n s regioisomers, the -cis regioisomers
of 8m on o or 9bis are chiral, but we made no attempt to
resolve the enantiomers.
be recycled and reacted again with 7 to give 9bis
.
Therefore, in an effort to conserve the porphyrin synthon
3, the first step in the [2 + 1] approach appears to be
more practical than the one in the [1 + 2] because only
a single side-product porphyrin (9tr is) is formed whereas
two are generated from the [1 + 2] approach (8bis and
The second step in either the [1 + 2] or the [2 + 1]
approach starts with a pure regioisomer of mono- or bis-
imidazole porphyrin and introduces the second, different
imidazole. It is worth mentioning that when a mixture
of regioisomers 8m on o(cis + tr a n s) or 9bis (cis + tr a n s)
J . Org. Chem, Vol. 69, No. 10, 2004 3547

SCHEME 2. Dir ect In tr od u ction of Tw o Differ en t Typ es of Im id a zoles on to th e r3 F a ce of th e r3AâTa il
P or p h yr in 3: [1 + 2] a n d [2 + 1] Ap p r oa ch esa
a
Reagents and conditions: (a) 2, CH3CN, rt, 30 min, 21%; (b) 7, CH3CN, Et2NC6H5, rt, 30 min, 25%; (c) (1) HCl/Et2O, CH3CN-THF,
rt, 10 min, (2) 7, CH3CN, Et2NC6H5, rt 30 min, 60%; (d) 1) HCl/Et2O, CH3CN-THF , rt, 10 min, (2) 2, CH3CN, Et2NC6H5, rt 30 min,
35-40%; (e) (1) BBr3, CH3Cl2, -78 °C, 30 min, (2) 0 °C, 2 h, (3) CH3OH, rt 30 min, 80-85%.
was used for this second condensation reaction, regioi-
somer separation of the tris-imidazole porphyrin 10 was
tedious. As we had shown that regioisomer separation
is more difficult with bis-imidazole porphyrins than with
mono-imidazole porphyrins, it was not surprising to find
that regioisomer separation of the more polar tris-imid-
azole species 10 was even more difficult than with 9bis
species, even when an F-tail was present. Because of this
problem, we always performed the regioisomer separation
of each intermediate 8m on o or 9bis prior any condensation
reaction leading to porphyrin 10. We also discovered that
the second condensation reaction could not be performed
using the same experimental conditions as those de-
scribed for the first step; otherwise, low yields of target
molecule 10 are obtained together with a large amount
of unidentified fractions. Optimum conditions required a
homogeneous solution of the porphyrins, obtained by us-
ing cosolvents such as dichloromethane or tetrahydrofu-
ran. Also the porphyrin mixture must be acidified with
3 to 4 equiv of HCl before adding the acyl chlorides in
order to reprotect the nucleophilic imidazole nitrogen in
intermediates 8m on o or 9bis which were no longer proto-
nated after workup (NaHCO3) and chromatography
(NH3-saturated eluent). As was pointed out earlier,4b the
formation of unidentified fractions may be explained by
the fact that unprotected imidazoles quaternize by reac-
tion with acyl chloride to produce oligomeric acyl imida-
zolium species.4b During the condensation, some mini-
mum equivalent of base was added allowing the desired
transformation to occur. After taking these precautions,
the yields of 10 from 8m on o ([1 + 2] approach) or 9bis ([2
+ 1] approach) rose to 70% and 35-40%, respectively.
Without preliminary acid treatment of porphyrins the
yields dropped to 18% and 9%, respectively. In fact, better
yields in the second condensation reaction were always
obtained from the [1 + 2] approach compared with from
the [2 + 1] approach. This confirms our assumption that
the more nonprotected imidazoles that are present in the
starting material (up to three in 9bis in the [2 + 1]
approach, up to two in 8m on o in the [1 + 2] approach) the
lower are the yields in the second condensation reaction.
Alternatively, when the condensation reaction in acetic
acid/sodium acetate system was run following our earlier
procedures, only 30% yield was obtained.4b
A comparison between [1 + 2] and [2 + 1] approaches
is relevant for the first and most difficult step of the
synthetic scheme. With porphyrins bearing F-tail, al-
(6) For all compounds presented here, successful separation of
regioisomeric porphyrins depends on using thin silica layer (500 µm),
saturating the eluent with ammonia and low loading of porphyrin.
These criteria became more crucial for the separation of regioisomeric
intermediates of 9F extremely dependent on ammonia (with H-tail
bis
porphyrins, the separation of regioisomers of 9H is not possible).
bis
(7) (a) Differential protection of the amines using our previously
reported protective groups was attempted. The purification of mono-
and bis-trifluoroacetamido or mono- and bis-trityl porphyrins is even
more troublesome than the purification of mono- and bis-imidazole
porphyrins; for that reason, these protective groups were not used here.
(b) Given the difficulties to introduce only one or two imidazoles, we
tried the selective deprotection of one phenol among three available
in 8tr is. Eventually, some target mono-phenolic compound was obtained,
but the yields were low and hardly reproducible, and mostly the tris-
deprotected species was obtained.
3548 J . Org. Chem., Vol. 69, No. 10, 2004

though the separation of regioisomeric intermediates is
less straightforward with [2 + 1] (9bis) rather than with
[1 + 2] (8m on o), the [2 + 1] approach was preferred
because it was easier to control the introduction of the
first kind of picket, the possibility to recycle one byprod-
uct and the efficient separation of target from the
statistical mixture.7a However, with porphyrins bearing
the H-tail, the separation of regioisomeric intermediates
in the [2 + 1] approach was not possible; therefore, only
the [1 + 2] approach was considered.
phyrins (cis) changes noticeably with respect to the sym-
metric porphyrins (tr a n s). Because cis-porphyrins pre-
sented here are chiral, the methylene protons in the prox-
imal tail are diastereotopic as shown by their 1H NMR
signals. But after introduction of the second type of imid-
azole, bringing more symmetry to the system, the benzyl
CH2 signal of cis-porphyrins again appears as a singlet.
1F -cis is to be used for mechanistic and biomimetic
catalytic studies of reduction of O2 to H2O, after insertion
of Fe and Cu. Larger amounts of 1F -cis will need to be
prepared in the future for which we intend to bring
improvements to these methods.
The final step of our synthetic scheme is deprotection
of the phenol in 10F -cis. This was achieved using 18-20
equiv of BBr3 at low temperature (-78 to 0 °C) and gave
1F -cis in 85% yield. When more equivalents of BBr3 are
used or when the reaction is run at room temperature
for a longer time, several decomposition products are
observed together with a lower yield of 1. Except for this
point, this last step in the synthetic scheme is less
difficult and chromatographic separation is reasonable.7b
However, the target 1F -cis appears to be more air- and
light-sensitive than the other porphyrin intermediates
described above. This free base porphyrin is a good model
for the natural CcO active site because all of the key
groups are present: porphyrin, proximal base, three
distal imidazoles including one imidazole linked to a
phenol-protected residue.
Ack n ow led gm en t. This material is based upon
work supported by the National Institutes of Health
under Grant No. 017880. R.A.D. thanks the French
Foreign Ministry for a Lavoisier fellowship. We also
thank the Mass Spectrometry Facility, University of
California, San Francisco, and the Stanford University
Mass Spectrometry.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and characterization data for new compounds 1-2,
6, and 8-10. This material is available free of charge via the
Internet at http://pubs.acs.org.
1
Finally, it is interesting to note that the H NMR spec-
trum of the chiral mono- and bis-imidazole-bearing por-
J O0499625
J . Org. Chem, Vol. 69, No. 10, 2004 3549