Exploring the Strength of the H-Bond in Synthetic Models for Heme Proteins: The Importance of the N?H Acidity of the Distal Base

Article abstract of DOI:10.1002/chem.201601505

The distal hydrogen bond (H-bond) in dioxygen-binding proteins is crucial for the discrimination of O₂with respect to CO or NO. We report the preparation and characterization of a series of Zn^IIporphyrins, with one of three meso-phenyl rings bearing both an alkyl-tethered proximal imidazole ligand and a heterocyclic distal H-bond donor connected by a rigid acetylene spacer. Previously, we had validated the corresponding Co^IIcomplexes as synthetic model systems for dioxygen-binding heme proteins and demonstrated the structural requirements for proper distal H-bonding to Co^II-bound dioxygen. Here, we systematically vary the H-bond donor ability of the distal heterocycles, as predicted based on pK_avalues. The H-bond in the dioxygen adducts of the Co^IIporphyrins was directly measured by Q-band Davies-ENDOR spectroscopy. It was shown that the strength of the hyperfine coupling between the dioxygen radical and the distal H-atom increases with enhanced acidity of the H-bond donor.

Full text of DOI:10.1002/chem.201601505

DOI: 10.1002/chem.201601505
Full Paper
&
Hydrogen Bonds
Exploring the Strength of the H-Bond in Synthetic Models for
Heme Proteins: The Importance of the NÀH Acidity of the
Distal Base
Mariza N. Alberti,[a] Yevhen Polyhach,[b] Manolis D. Tzirakis,[a] Laura Tçdtli,[b]
Gunnar Jeschke,*[b] and FranÅois Diederich*[a]
Dedicated to Professor James P. Collman
Abstract: The distal hydrogen bond (H-bond) in dioxygen-
binding proteins is crucial for the discrimination of O2 with
respect to CO or NO. We report the preparation and charac-
terization of a series of ZnII porphyrins, with one of three
meso-phenyl rings bearing both an alkyl-tethered proximal
imidazole ligand and a heterocyclic distal H-bond donor
connected by a rigid acetylene spacer. Previously, we had va-
lidated the corresponding CoII complexes as synthetic model
systems for dioxygen-binding heme proteins and demon-
strated the structural requirements for proper distal H-bond-
ing to CoII-bound dioxygen. Here, we systematically vary the
H-bond donor ability of the distal heterocycles, as predicted
based on pKa values. The H-bond in the dioxygen adducts of
the CoII porphyrins was directly measured by Q-band Davies-
ENDOR spectroscopy. It was shown that the strength of the
hyperfine coupling between the dioxygen radical and the
distal H-atom increases with enhanced acidity of the H-bond
donor.
been a matter of some controversy.[3] Pauling initially suggest-
ed that the distal histidine stabilizes bound dioxygen by a H-
bonding interaction.[4] A series of EPR,[5] resonance Raman,[6]
NMR,[7] neutron diffraction analysis,[8] X-ray,[6b,9] as well as rapid
mixing and laser photolysis[10] studies supported the presence
of such an interaction (“Pauling hypothesis”).
Introduction
Myoglobin (Mb) and hemoglobin (Hb) are heme proteins
whose physiological importance is principally related to their
ability to bind, transport, and store dioxygen. The binding
pocket of these proteins contains an iron protoporphyrin IX,
which is designated as heme (FeII) or hemin (FeIII).[1] The FeII
atom is coordinated to the heme through four pyrrole nitro-
gens and to the protein moiety through a proximal histidine.
On the other side of the heme, in the so-called distal pocket,
three strongly conserved residues are in close proximity: phe-
nylalanine, valine, and most importantly histidine. In this
pocket, the FeII atom can efficiently bind O2, CO, and NO,
which is of critical importance to physiological processes such
Previously, a suitable pulse EPR method was described to di-
rectly determine the nature, the orientation, as well as the
length of the H-bonding interaction between the distal side
residue and bound dioxygen in Co-Mb-O2 and in model com-
plex 1-Co-O2 (Figure 1).[11] As an extension of our studies in
this area, we also examined the geometrical prerequisites for
the H-bonding interaction in a series of synthetic models for
heme proteins.[12] These prerequisites were found to be narrow
and very specific. Among the studied complexes, 1-Co-O2 pos-
sessed ideal distal constitution to form this interaction and
hence is regarded as an excellent model for the binding
pocket of Mb and Hb. The aim of the present study is to shed
further light onto the strength of the vital H-bonding interac-
tion by systematically varying the NÀH acidity of the distal
base in the series 1–8-Co-O2 (Figure 1) and directly investigat-
ing the effect of these changes on the dioxygen-binding
pocket by NMR and Q-band Davies-ENDOR (Electron-Nuclear
DOuble Resonance) spectroscopies.
as respiration, neurotransmission, and vasodilation.[2]
A
common feature shared by all heme proteins is their enhanced
affinity for O2 over CO or NO. There is great interest in under-
standing the fundamental principles of how the heme distal
pocket favors O2 binding. In fact, the elucidation of the mecha-
nism responsible for this molecular recognition process has
[a] Dr. M. N. Alberti, Dr. M. D. Tzirakis, Prof. Dr. F. Diederich
Laboratorium fꢀr Organische Chemie, ETH Zurich
Vladimir-Prelog-Weg 3, 8093 Zurich (Switzerland)
E-mail: diederich@org.chem.ethz.ch
[b] Dr. Y. Polyhach, L. Tçdtli, Prof. Dr. G. Jeschke
Laboratory of Physical Chemistry, ETH Zurich
Vladimir-Prelog-Weg 2, 8093 Zurich (Switzerland)
E-mail: gunnar.jeschke@phys.chem.ethz.ch
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201601505.
Chem. Eur. J. 2016, 22, 1 – 10
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 1. Model compounds 1–9-Co-O2 for pulse EPR investigations. The values in parenthesis are the ACD-calculated pKa[13] values for the phenylacetylene
derivatives of the distal bases in 1–8-Co-O2.
ware[13] for the phenylacetylene derivatives of the distal bases,
in parenthesis. The last model complex, 9-Co-O2 (Figure 1), was
synthesized from the precursor porphyrin 9-Zn, a fully charac-
terized by-product of the Sonogashira cross-couplings under
the conditions assessed in this study (vide infra).
Results and Discussion
Porphyrin synthesis
For the purpose of the present study, we prepared a series of
seven novel model complexes 2–8-Co-O2 endowed with differ-
ent distal bases. Figure 1 depicts their structures along with
the estimated pKa values, calculated by the ACD/pKa soft-
The synthesis of porphyrins 1–9-Co is outlined in Scheme 1.
Porphyrin 10-Zn is the common precursor of these porphyrins
Scheme 1. Synthesis of model compounds 1–9-Co. DMF=dimethylformamide, FC=flash chromatography, GPC=gel permeation chromatography, TFA=tri-
fluoroacetic acid, THF=tetrahydrofuran.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Scheme 2. Synthesis of iodinated distal base precursors 12–16. TFA=trifluoroacetic acid, THF=tetrahydrofuran.
and its synthesis is described in the literature[11,12] and—upon
addition of new findings—in Section S2.1 of the Supporting In-
formation. Sonogashira cross-coupling of 10-Zn with com-
pounds 11–18 in the presence of [Pd(PPh3)4] as a catalyst af-
forded porphyrins 1–8-Zn in satisfactory yields (32–70%).[14] In-
terestingly, in all cases, in addition to the desired porphyrins
1–8-Zn, the unanticipated by-product 9-Zn was also formed
and fully characterized by 1D/2D NMR techniques as well as
mass spectrometry. Porphyrin 9-Zn is formed by phenyl-aryl
exchange during the cross-coupling reaction (aryl=distal base;
phenyl derives from [Pd(PPh3)4]);[15] a plausible mechanism[16]
for the formation of this porphyrin is provided in Section S2.2
of the Supporting Information. In the next step, the ZnII ion
was removed with trifluoroacetic acid (TFA) affording porphyr-
ins 1–9-2H in moderate to high yields (51–86%). Surprisingly,
these free-base porphyrins were stable at ambient conditions,
which allowed for their characterization by 1D/2D NMR tech-
niques as well as by mass spectrometry. Subsequently, CoII was
inserted into the free-base porphyrins 1–9-2H using CoCl2; this
reaction was conducted in a glovebox. In all cases, the CoII in-
sertion provided a mixture of both CoII and CoIII porphyrins, as
was evidenced by their UV/Vis spectra.[17] Thus, an additional
reduction with an aqueous solution of Na2S2O4 was performed
in the glovebox, affording the desired CoII-containing com-
plexes.
yield.[22,23] Finally, 5-iodo-2-bromo-1H-benzimidazole (16) was
synthesized by treatment of 5-iodo-1,3-dihydro-2H-benzimida-
zole-2-thione (21) with HBr/Br2 in MeOH.[24] Compound 21 was
synthesized by the thiocarbonylation of 4-iodobenzene-1,2-dia-
mine (20) using 1,1’-thiocarbonyldiimidazole.[25]
The synthesis of benzimidazole 17 proceeded in three steps
(Scheme 3). Iodination of 3-fluoro-2-nitroaniline (22)[26] with
iodine monochloride[18] gave a mixture of two regioisomers 23
and 24 in 19% and 63% yield, respectively. It is worth noting
that in the course of this reaction, a less polar compound was
also isolated (3–4%), the structure of which could not be de-
termined based on 1D/2D NMR techniques as well as HR-MS
spectrometry. Reduction of compound 24 with SnCl2·2H2O and
12 n HCl in MeOH[27] afforded 3-fluoro-4-iodobenzene-1,2-dia-
mine (25) in 88% yield. Due to the instability of diamine 25 at
258C, this compound was used directly in the subsequent re-
action without further purification. Accordingly, cyclization of
crude diamine 25 in the presence of formic acid[28] gave distal
base precursor 17 in 73% yield. The synthesis of precursor 18
The synthesis of the precursors to the distal bases, 12–16, is
presented in Scheme 2. Specifically, the preparation of 5-iodo-
2-methyl-1H-benzimidazole (13) was achieved in one step by
the Na2S2O4 reduction of 4-iodo-2-nitroaniline (19)[18] in the
presence of acetaldehyde.[19] Reduction of aniline 19,[18a,20] fol-
lowed by treatment with TFA in the presence of catalytic HCl,
gave 5-iodo-2-(trifluoromethyl)-1H-benzimidazole (14) in 62%
overall yield.[21] 5-Iodo-1,3-dihydro-2H-benzimidazol-2-one (12)
was obtained by treatment of 4-iodobenzene-1,2-diamine (20)
with 1,1’-carbonyldiimidazole.[22] Subsequently, distal base 12
was reacted with POCl3 and a catalytic amount of phenol to
afford 5-iodo-2-chloro-1H-benzimidazole (15) in 35% overall
Scheme 3. Synthesis of iodinated benzimidazoles 17 and 18.
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
was carried out in a similar manner (Scheme 3). Specifically, 6-
iodo-3-nitropyridin-2-amine (26)[29] was treated with SnCl2·2H2O
and 12 n HCl[18a,20] to afford 6-iodopyridine-2,3-diamine (27) in
45% yield. Subsequent cyclization of diamine 27 in the pres-
ence of formic acid[28] gave 18 in 50% yield.
the other hand, the aforementioned upfield-shifted signals are
1
not clearly visible in the H NMR spectra of 7-Zn and 8-Zn at
258C (Supporting Information, Figure S59 and S67). In the
1H NMR spectra of 7-Zn, a peak broadening throughout the
spectral range is observed. In addition, the comparison of the
1H NMR spectra of 7-Zn or 8-Zn at 408C and À408C (Support-
ing Information, Figure S65 or S72) showed new signals in the
range of d=1.8–2.6 and 4.5–6.0 ppm, which could be ex-
plained by the convergence of various conformers of 7-Zn or
8-Zn to more stable ones. In these two model complexes, it
seems reasonable to assume that there is a dynamic equilibri-
um between the ligation of hexyl-tethered imidazole and
alkyne-appended benzimidazole to the ZnII ion (with rates that
are comparable to the NMR timescale) and that the axial liga-
tion of proximal imidazole to the ZnII ion is favored at lower
temperatures.
Conformation of zinc complexes in CDCl3 solution at 258C
The conformations of the proximal and distal site residues of
the different models 1–9-Zn in solution were investigated by
NMR spectroscopy.
Regarding the proximal site of porphyrins, for model com-
1
plexes 1–6-Zn and 9-Zn the H NMR resonances of the hexyl-
tethered imidazole moieties are significantly upfield shifted,
appearing at d=1.97–2.09 (position 5), 2.22–2.59 (position 2),
and 4.91–5.30 ppm (position 4; Figure 2). This observation is
indicative of axial imidazole ligation to the ZnII ion;[30] upon li-
gation, the imidazole protons are exposed to the shielding
region of the strong aromatic ring current of the porphyrin. On
Regarding the preferred orientation and tautomeric state of
the distal bases, after vigorously shaking a solution of 1-Zn, 2-
Zn, and 4–6-Zn in CDCl3 with D2O for 10 s, 1H NMR spectra
Figure 2. Schematic view of the preferred conformations of model complexes 1–9-Zn in solution (CDCl3, 258C) as determined by 1H NMR spectroscopy. Indica-
tive proton chemical shifts are given in ppm (chemical shifts for complex 1-Zn were taken from Ref. [11]). Certain residues of the molecules were omitted for
clarity.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
showed the disappearance of NH signals (assignments of these
signals (ppm values) are shown in Figure 2). In all cases, these
NH signals are shifted upfield. The range of the upfield shift is
d=2.61–6.72 ppm, as compared to the NH resonances of the
corresponding iodinated bases (11, 12, and 14–16; Scheme 1).
This upfield shift indicates the proximity of the NH proton to
the porphyrin p-cloud. Thus, it can be concluded that the
distal base NH moieties in complexes 1-Zn, 2-Zn, and 4–6-Zn
are turned towards the porphyrin plane and, especially in com-
plexes 1-Zn and 4–6-Zn, only the tautomers with the NH
proton pointing towards the porphyrin plane are present in so-
lution. The preferred conformations of the aforementioned
complexes are shown in Figure 2. This structural assignment is
further supported by the resonances of the aromatic protons
in the distal bases of 2-Zn, and 4–6-Zn (Figure 2). Presumably,
the proposed conformations are stabilized by NÀH···p interac-
tions of the exchangeable distal protons with the porphyrin
chromophore.[31] Due to the broadness and weakness of benzi-
midazole peaks in complexes 3-Zn, 7-Zn and 8-Zn, direct as-
signment of these proton resonances was not possible. For
complexes 7-Zn and 8-Zn, the distal bases are probably rotat-
ed outwards, away from the porphyrin plane, because of elec-
trostatic repulsion between the porphyrin p-cloud and the F-
atom or the N4 lone pair of the substituted benzimidazoles.
H-Bond strength between the distal residue and bound
oxygen in 1–8-Co-O2 species by Q-band Davies-ENDOR spec-
troscopy
Prior to the Q-band Davies-ENDOR experiments, field-swept
ESE (electron spin echo)-detected spectra were taken for 1–8-
Co-O2 complexes. These ESE spectra were used to select the
observer positions for acquisition of the Davies-ENDOR spectra
and determine the g-tensors needed for the simulations of
Davies-ENDOR spectra. The ESE spectrum of 6-Co-O2 is shown
in Figure 3A, whereas the ESE spectra of 1–5-Co-O2 and 8-Co-
O2 complexes are included in the Supporting Information (Sec-
tion S5.2, Figures S175–S182 A, respectively); as mentioned
below, a stable, oxygenated 7-Co-O2 complex could not be ob-
tained.
Figure 3. Experimental EPR spectra for 6-Co-O2 (Q band, 10 K). A) Magnetic
field swept ESE spectrum (black) superimposed with the EasySpin simulation
(red). Black circles with numbers on the ESE spectrum denote observer posi-
tions on which Davies-ENDOR spectra were acquired. B) Experimental
Davies-ENDOR spectra (black) superimposed with the EasySpin simulation
(red). Arrows denote spectral regions were resonances due to the strongly
coupled proton occur.
Principal values of the electronic g-tensors of 1–6-Co-O2 and
8-Co-O2 complexes, determined by EasySpin simulations, are
presented in Table 1. The set of the Davies-ENDOR spectra for
6-Co-O2 acquired at many observer positions is shown in Fig-
ure 3B. Such sets of spectra for the remaining complexes are
shown in the Supporting Information (Section S5.2, Fig-
ures S175–S182). Note that these spectra are very rich in infor-
mation as they feature resonance bands stemming from many
coupled protons in 1–6-Co-O2 and 8-Co-O2. In the present
study, we focused only on the strongest coupling that corre-
sponds to the exchangeable proton of the distal base.[11,12]
Spectral features associated with that coupling are indicated
by arrows in Figure 3B. The set of Davies-ENDOR spectra ac-
quired at different observer positions contains information
about the strength of the hyperfine coupling as well as the rel-
ative orientation of the hyperfine coupling tensor and the elec-
tronic g-tensor, which can be extracted by means of rigorous
spectra simulations. A complete example of such analysis is
given for 1-Co-O2 (for which we observed the most prominent
resonance bands from the exchangeable proton among all the
samples) in the Supporting Information (Section S5.2, Fig-
ures S175–S176). With our new, higher quality experimental
data, we confirmed the conformation which has been previ-
ously found[11,12] for the distal base in 1-Co-O2: the Euler rota-
tion relating the g-tensor and the hyperfine coupling tensors
AH was found to be [08, 1058, 08] Æ108. The new data suggest
a purely dipolar nature of the hyperfine coupling between the
distal proton and the electron spin AH =[À6.5 À6.5 13.0] MHz
in contrast to the previous results in which a small isotropic
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
not contain information on relative orientation of the hyperfine
and the g-tensors, hence relative coupling strengths can be
readily compared.
Table 1. Principal values of g- and A- interaction tensors for 1–6-Co-O2
and 8-Co-O2 determined with EasySpin from the Q-band ESE and Davies-
ENDOR spectra. The distance between the distal proton and the electron
spin was computed using the point dipolar approximation.[34] The corre-
sponding data for 7-Co could not be produced because no stable 7-Co-
O2 complex could be obtained. The data are presented in the order of in-
creasing hyperfine coupling of the distal proton.
Regarding the 7-Co complex, no stable 7-Co-O2 species
could be obtained. Hence, neither ESE nor Davies-ENDOR ex-
periments could be performed in that case. With the 8-Co-O2
complex, only one Davies-ENDOR spectrum could be acquired
due to a very weak contribution from the strongly coupled
proton that required an extremely long acquisition time. This
prevented thorough characterization of the detected strong
coupling in 8-Co-O2 in terms of the strength and geometry
(see the Supporting Information; Figure S183 and the accom-
panying text).
Compound [gx gy gz]
[Ax Ay Az]
[MHz]
distal proton-e.
distance [ꢁ]
1-Co-O2
3-Co-O2
2-Co-O2
6-Co-O2
5-Co-O2
4-Co-O2
8-Co-O2
[2.0073 1.9920 2.0780] [À6.5 À6.5 13.0] 2.30
[2.0057 1.9920 2.0780] [À6.5 À6.5 13.0] 2.30
[2.0053 1.9925 2.0740] [À6.6 À6.6 13.2] 2.29
[2.0065 1.9910 2.0765] [À6.9 À6.9 13.8] 2.25
[2.0085 1.9920 2.0780] [À7.0 À7.0 14.0] 2.24
[2.0055 1.9910 2.0785] [À7.2 À7.2 14.4] 2.22
The present results show that the NÀH acidity of the distal
base has a decisive influence on the strength of the H-bond.
Initial theoretical estimates of NÀH acidity (more precisely—
ACD-calculated pKa[13] values) correlate well with the experi-
mentally determined H-bond strength. Both suggest existence
of two rather distinct groups of the compounds: one with
[2.0045 1.9900 2.0825]
–
–
contribution was reported.[11,12] Analysis of the experimental
Davies-ENDOR data revealed that the conformation of the
distal base in 1-Co-O2, which was found previously[11,12] and
confirmed here, was preserved in the 1–6-Co-O2 series (i.e. the
same Euler rotation applies throughout) while the strength of
the coupling varied. Principal values of the hyperfine coupling
tensor of the strongly coupled distal proton in the 1–6-Co-O2
series are summarized in Table 1. The data show that the
strength of the hyperfine coupling of the distal proton increas-
es as follows: 1-Co-O2 ꢀ 3-Co-O2 <2-Co-O2 <6-Co-O2 <5-Co-
O2 <4-Co-O2. Accordingly, the distance between the distal
proton and the electron spin, computed from the hyperfine
coupling tensor of the distal proton using the point dipolar ap-
proximation,[32] decreases along the series (Table 1). This trend
of the hyperfine coupling among the 1–6-Co-O2 complexes
was confirmed by adding up the Davies-ENDOR spectra ac-
quired at different observer positions and analyzing the result-
ing sum Davies-ENDOR spectra (Figure 4). Such sum spectra do
a
weaker coupling (1-Co-O2 ꢀ3-Co-O2 <2-Co-O2) and the
other one with a stronger coupling (6-Co-O2 <5-Co-O2 <4-Co-
O2). Even though the predictions do not perfectly agree with
experiment down to fine details (for instance, experimental
data suggest stronger coupling in 3-Co-O2 than in 2-Co-O2 as
well as the strongest coupling in 4-Co-O2), the inconsistencies
between predictions and experimental findings are minor.
Conclusion
A series of model complexes 1–8-Co-O2 for the dioxygen-bind-
ing site of Mb and Hb was synthesized in order to systemati-
cally study the strength of the H-bond between the distal resi-
due and bound dioxygen. The key step in their preparation
was the Sonogashira cross-coupling of porphyrin 10-Zn with
compounds 11–18 in the presence of [Pd(PPh3)4] as a catalyst.
During the course of this reaction, apart from the formation of
the desired model complex, porphyrin 9-Zn was formed by
transfer of a phenyl group from the PPh3 ligand of the palladi-
um catalyst to the metal and subsequent coupling to the acet-
ylene moiety.
Based on 1H NMR studies (CHCl3, 258C), the alkyl-tethered
imidazole (proximal residue) in complexes 1-Zn, 2-Zn, 4–6-Zn,
and 9-Zn is axially ligated to the ZnII ion and the distal residue
is preferably rotated towards the porphyrin plane. In contrast,
such conformational preferences could not be supported for
complexes 7-Zn and 8-Zn. It is likely that, for these two por-
phyrins, there is a dynamic equilibrium between ligation of the
proximal or the distal base to the central ZnII ion, with rates
that are comparable to the NMR timescale.
The hyperfine interaction between the dioxygen radical and
neighboring exchangeable H-nucleus in 1–6-Co-O2 was clearly
resolved by Q-band Davies-ENDOR spectroscopy. We have
shown that while the geometry of the hyperfine interaction
(accompanied by the conformation of the distal base in the
compounds) is preserved throughout the 1–6-Co-O2 series, the
strength of the interaction increases with increasing the NÀH
acidity of the distal base.
Figure 4. Summed Davies-ENDOR spectra for 1–6-Co-O2 species (Q band,
10 K). A) Complete spectra. B) Spectra are shown at increased scale. Vertical
dotted lines in (A) and (B) are added in order to facilitate visual comparison
of the spectra.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
ing spin system (high number of free parameters and complex line
broadening mechanism) and the high complexity of the experi-
mental spectra stemming from their rich content. The latter was
especially the case for the Davies-ENDOR spectra.
Experimental Section
General information
Porphyrins 1-Zn and 10-Zn,[11,12] benzimidazole 11,[18a,33] 4-iodo-2-
nitroaniline (19),[18] 3-fluoro-2-nitroaniline (22),[26] and 6-iodo-3-ni-
tropyridin-2-amine (26)[29] were prepared according to known liter-
ature procedures. The Supporting Information (SI) contains: syn-
thesis of porphyrin 10-Zn; proposed mechanism for the formation
of porphyrin 9-Zn; synthetic procedures and spectral data of zinc
porphyrins 2–9-Zn, free-base porphyrins 1–9–2H, cobalt porphyr-
ins 1–9-Co, iodinated distal base precursors 12–18, and precursor
compounds 19–27; copies of UV/Vis spectra, 1D and 2D NMR spec-
tra of all new compounds; X-band cw-EPR and Q-band Davies-
ENDOR spectra.
Preparation of 1–8-Co-O2 species
In order to prepare 1–8-Co-O2 complexes for the Davies-ENDOR
experiments, compounds 1–8-Co were dissolved in a CH2Cl2/tolu-
ene 4:1 mixture under oxygen-free atmosphere in a glovebox, and
were subsequently subjected to oxygenation. The process was
controlled by monitoring relative intensities of signals from the
oxygenated and the non-oxygenated species in the cw-EPR spectra
at X band. The signal from the non-oxygenated species (1–8-Co)
originates from a low-spin CoII species in axial local surrounding.[11]
After the oxygenation, the spin density is largely shifted on the di-
oxygen species of the oxygenated complex. The resulting EPR
signal is significantly narrower and exhibits different spectral fea-
tures. Importantly, it is located at higher magnetic fields (smaller g-
factor range) with only minor overlap with the signal from the
original non-oxygenated compound. Both types of signals were
characterized previously.[11,12] Our first attempts to perform oxygen-
ation by blowing a dioxygen stream from a balloon directly into
the 3 mm o.d. EPR tube as discussed previously[11] showed poor
control over the oxygenation process and led to oxidation beyond
the desired 1–8-Co-O2 species. In particular, a very strong drop of
intensity of the cw-EPR spectrum from the sample after the oxy-
genation was always observed. In order to overcome this problem,
milder oxygenation was performed with ambient air in smaller
1.6 mm o.d. EPR tubes with regular probing of the oxygenation
extent by cw-EPR (for more details see Section S5.1 in the Support-
ing Information). Upon exposure to dioxygen, most of the 1–8-Co
compounds were initially oxygenated to the respective 1–8-Co-O2
complexes (first phase) indicated by an increase in the signal inten-
sity stemming from the dioxygen species, and then they under-
went further transformations associated presumably with the for-
mation of other paramagnetic CoII complexes (second phase) ac-
companied by broadening and eventually deforming of the EPR
signal. In the present work we refrained from in-depth studying of
the oxygenation kinetics and focused on obtaining the primary
oxygenated 1–8-Co-O2 compounds in good quality.
EPR spectroscopy and spectral simulations
Continuous wave (cw) EPR experiments were performed at 120 K
at X-band frequencies with a commercial X/W-band Elexsys E680
spectrometer equipped with an EN4118X-MD4-W1 probehead
(both from Bruker). Spectra were acquired at 0.2012 mW incident
microwave power with 0.2 mT modulation amplitude; the micro-
wave resonator was critically coupled.
All pulse EPR experiments were performed at 10 K at Q band with
a commercial X/Q-band Elexsys E580 spectrometer equipped with
a EN 5107D2 Q-band resonator (both from Bruker). Electron spin
echo detected field swept spectra (ESE spectra) were taken with
the p/2 (40 ns) À p (80 ns) primary echo sequence. Q-band Davies-
ENDOR spectra were acquired using a conventional Davies-ENDOR
pulse sequence.[34] The length of the preparation (p) pulse was
80 ns; for the detection “p/2Àp” block 40 ns and 80 ns microwave
pulses were used. Such excitation enhances appearance of the
strongly coupled species in Davies-ENDOR spectra; it was found
optimal for detection of the exchangeable protons in the previous
studies.[11,12] To straighten the baseline, the stochastic acquisition
mode was chosen for recording the Davies-ENDOR spectra. For
every compound, Davies-ENDOR spectra were taken at many (13–
21) observer positions of the ESE spectrum to allow for an accurate
determination of the hyperfine interaction tensor of the strongly
coupled exchangeable proton of the distal base. For all EPR meas-
urements (both at X and Q bands) quartz sample tubes (Wilmad)
with 1.6 mm outer diameter were used.
Spectral simulations were performed with EasySpin.[35] The spin
system to simulate ESE spectra contained one electron (S=1/2)
and one 59Co nucleus (I=7/2). No further nuclei (e.g. protons) were
taken explicitly into account, as they contribute mainly to the over-
all linewidth due to their weak hyperfine couplings. An analogous
spin system was used previously.[11] The hyperfine coupling tensor
of 59Co was taken as ACo =[À54 À28 À28] MHz with the relative
orientation with respect to the g-tensor described by the Euler
angles (a, b, g)=(08, 708, 08). These parameters were determined
previously[11,12] and were left unchanged. For each compound, prin-
cipal values of the g-tensors and line broadening parameters (de-
fined as the HS-strain tensor) were varied in order to achieve the
best possible visual match with experimental spectra. The spin
system for simulating the Davies-ENDOR spectra contained one
electron (S=1/2) and one nucleus (I=1/2) representing a strongly
coupled proton. The hyperfine coupling tensor of the proton was
expressed as aiso + aT·[À1 À1 2] during the simulations. All experi-
mental spectra were background-corrected prior to simulations.
Best matching models were manually selected on the basis of the
visual analysis rather than automated fitting with an rmsd criterion.
We proceeded like this because of the complexity of the underly-
The results of the mild oxygenation procedure are summarized in
Figure S174 in the Supporting Information. For the 1–5-Co com-
pounds, the ratio between the signal intensities from the oxygen-
ated to the non-oxygenated species was gradually increasing upon
oxygenation (i.e. Figure S174 A) though a different oxygenating ki-
netics was observed in each case. It was generally difficult to reach
a full oxygenation.
Prolonged oxygenation led to oxidation beyond the 1–5-Co-O2
complex as indicated by a characteristic line broadening, Fig-
ure S174. Note that no signals from the strongly coupled protons
could later be detected by Q-band ENDOR from such secondary-
oxidized samples. Good 1-Co-O2, 2-Co-O2, 3-Co-O2, and 4–5-Co-O2
complexes were obtained after 3, <1, 15, and 10 min of oxygena-
tion, respectively. Unusually fast oxygenation kinetics was observed
for 6-Co; an undesired line broadening was seen already shortly
after the oxygenation begun (Figure S174B). Notably, a complete
oxygenated 6-Co-O2 sample was obtained by transferring the ma-
terial into the EPR tube in the air rather than in the glovebox. Fig-
ure S174 (part C) shows EPR spectra from the desired 1–6-Co-O2
complexes.
For compounds 7-Co and 8-Co (Figure 1), oxygenation/oxidation
kinetics was much faster than in the other cases. For example, for
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[9] a) S. E. V. Phillips, J. Mol. Biol. 1980, 142, 531–554; b) F. Yang, G. N. Phil-
lips, Jr., J. Mol. Biol. 1996, 256, 762–774; c) S.-Y. Park, T. Yokoyama, N.
Shibayama, Y. Shiro, J. R. H. Tame, J. Mol. Biol. 2006, 360, 690–701.
[10] I. Birukou, R. L. Schweers, J. S. Olson, J. Biol. Chem. 2010, 285, 8840–
8854.
[11] H. Dube, B. Kasumaj, C. Calle, M. Saito, G. Jeschke, F. Diederich, Angew.
Chem. Int. Ed. 2008, 47, 2600–2603; Angew. Chem. 2008, 120, 2638–
2642.
7-Co the secondary oxidation product was observed already in
samples that were not intentionally exposed to air or oxygen at
any time. For 8-Co, we could obtain the 8-Co-O2 complex only by
transferring the material into the EPR tube in the air, as was similar-
ly found in the case of 6-Co-O2. However, the relative content of
the species featuring the strongly coupled proton, as was later re-
vealed by Q-band ENDOR measurements, was extremely low. More
details on the oxygenation of 7-Co and 8-Co are provided in Sec-
tion S5.1 in the Supporting Information.
[12] H. Dube, B. Kasumaj, C. Calle, B. Felber, M. Saito, G. Jeschke, F. Dieder-
ich, Chem. Eur. J. 2009, 15, 125–135.
[13] ACD/Percepta, version 14.0.0, Advanced Chemistry Development, Inc.,
Toronto, ON, Canada, 2014, www.acdlabs.com (accessed April 1, 2015).
For predicting and tuning the pKa values of basic amine centers, see:
M. Morgenthaler, E. Schweizer, A. Hoffmann-Rçder, F. Benini, R. E.
Martin, G. Jaeschke, B. Wagner, H. Fischer, S. Bendels, D. Zimmerli, J.
Schneider, F. Diederich, M. Kansy, K. Mꢂller, ChemMedChem 2007, 2,
1100–1115.
[14] Due to the instability of 2–8-Zn, their purification proved to be very
challenging. In particular, these porphyrins were purified by FC in the
dark (SiO2; solvent, 1% v/v Et3N) and preparative GPC (Jaigel-2H and
NovoGROM 500 or Bio-Beads SX-1).
[15] For the formation of scrambled products under Sonogashira and Suzuki
cross-coupling conditions, see: a) R. W. Wagner, Y. Ciringh, C. Clausen,
J. S. Lindsey, Chem. Mater. 1999, 11, 2974–2983; b) M. Kozaki, S. Morita,
S. Suzuki, K. Okada, J. Org. Chem. 2012, 77, 9447–9457; c) M. N. Alberti,
S. Nowakowska, M. D. Tzirakis, J. Nowakowski, P. Fesser, W. B. Schweizer,
A. Shchyrba, C. Thilgen, T. A. Jung, F. Diederich, Eur. J. Org. Chem. 2014,
5705–5719.
Acknowledgements
This work was supported by the Swiss National Science Foun-
dation (SNF; grant number: 200021-140211). The research of
M.N.A. is implemented within the framework of the Action
“Supporting Postdoctoral Researchers” of the Operational Pro-
gram “Education and Lifelong Learning” (Action’s Beneficiary:
General Secretariat for Research and Technology), and is co-fi-
nanced by the European Social Fund (ESF) and the Greek
State. The authors are especially grateful to Prof. Dr. Willem H.
Koppenol and Dr. Reinhard Kissner for fruitful scientific discus-
sions and for generously providing laboratory facilities. Dr.
Henry Dube is gratefully acknowledged for valuable scientific
discussions and suggestions and Dr. Bruno Bernet for help
with the spectral interpretations. The authors also thank the
ETH NMR service for acquiring some of the 1D/2D NMR spec-
tra.
[16] F. E. Goodson, T. I. Wallow, B. M. Novak, J. Am. Chem. Soc. 1997, 119,
12441–12453.
[17] a) C.-z. Li, K. Nishiyama, I. Taniguchi, Electrochim. Acta 2000, 45, 2883–
2888; b) M. C. Marden, L. Kiger, C. Poyart, A. K. Rashid, J. Kister, F. Stetz-
kowski-Marden, G. Caron, M. Haque, L. Moens, FEBS Lett. 2000, 472,
221–224.
[18] a) J. G. Wilson, F. C. Hunt, Aust. J. Chem. 1983, 36, 2317–2325; b) F.
Maya, J. M. Tour, Tetrahedron 2004, 60, 81–92.
[19] For the synthesis of benzimidazoles from o-nitroanilines and aldehydes
in one step by a reductive cyclization, see: D. Yang, D. Fokas, J. Li, L. Yu,
C. M. Baldino, Synthesis 2005, 47–56.
Keywords: dioxygen binding · ENDOR spectroscopy · heme
proteins
·
hydrogen bond
·
model systems
·
NMR
spectroscopy · porphyrins
[20] M. Belema, A. C. Good, J. Goodrich, R. Kakarla, G. Li, O. D. Lopez, V. N.
Nguyen, J. Kapur, Y. Qiu, J. L. Romine, D. R. St. Laurent, M. Serrano-Wu,
L. B. Snyder, F. Yang (Bristol-Myers Squibb Company), PCT. Int. Appl. WO
2010/017401 A1, 2010.
[21] For a general method of the synthesis of 2-(trifluoromethyl)-1H-benzi-
midazoles, see: G. Navarrete-Vꢃzquez, M. M. Rojano-Vilchis, L. Yꢄpez-
Mulia, V. Melꢄndez, L. Gerena, A. Hernꢃndez-Campos, R. Castillo, F.
Hernꢃndez-Luis, Eur. J. Med. Chem. 2006, 41, 135–141.
ˇ
´
[1] a) J. Vojtechovsky, K. Chu, J. Berendzen, R. M. Sweet, I. Schlichting, Bio-
phys. J. 1999, 77, 2153–2174; b) E. L. Raven, A. G. Mauk, Adv. Inorg.
Chem. 2000, 51, 1–50; c) J. R. H. Tame, B. Vallone, Acta Crystallogr. Sect.
D 2000, 56, 805–811.
[2] a) M. D. Maines, Annu. Rev. Pharmacol. Toxicol. 1997, 37, 517–554; b) Y.
Zhao, P. E. Brandish, D. P. Ballou, M. A. Marletta, Proc. Natl. Acad. Sci. USA
1999, 96, 14753–14758; c) K. Shikama, Prog. Biophys. Mol. Biol. 2006,
91, 83–162.
[22] M. J. Wythes, M. J. Palmer, M. I. Kemp, M. C. Mackenny, R. J. Maguire,
J. F. Blake, Jr., PCT Int. Appl. WO 2000/05231A1, 2000.
[23] P. K. Dubey, A. Naidu, V. Anandam, G. Hemasunder, Indian J. Chem. Sect.
B 2005, 44, 1239–1242.
[24] M. Andrzejewska, M. A. Pagano, F. Meggio, A. M. Brunati, Z. Kazimierc-
zuk, Bioorg. Med. Chem. 2003, 11, 3997–4002.
[25] E. E. Simanek, A. Tsoi, C. C. C. Wang, G. M. Whitesides, M. T. McBride,
G. T. R. Palmore, Chem. Mater. 1997, 9, 1954–1961.
[26] C. J. Diaz, C. D. Haffner, J. D. Speake, C. Zhang, W. Y. Mills, P. K. Spearing,
D. J. Cowan, G. M. Green PCT Int. Appl. WO 2008/021851A2, 2008.
[27] a) D. A. Nugiel, K. Jacobs, L. Cornelius, C.-H. Chang, P. K. Jadhav, E. R.
Holler, R. M. Klabe, L. T. Bacheler, B. Cordova, S. Garber, C. Reid, K. A.
Logue, L. J. Gorey-Feret, G. N. Lam, S. Erickson-Viitanen, S. P. Seitz, J.
Med. Chem. 1997, 40, 1465–1474; b) M. von Wantoch Rekowski, A. Pyr-
iochou, N. Papapetropoulos, A. Stçßel, A. Papapetropoulos, A. Giannis,
Bioorg. Med. Chem. 2010, 18, 1288–1296.
[3] a) J. S. Olson, R. E. McKinnie, M. P. Mims, D. K. White, J. Am. Chem. Soc.
1983, 105, 1522–1527; b) B. A. Springer, S. G. Sligar, J. S. Olson, G. N.
Phillips, Jr., Chem. Rev. 1994, 94, 699–714; c) J. S. Olson, G. N. Phil-
lips, Jr., Biol. Inorg. Chem. 1997, 2, 544–552; d) T. G. Spiro, P. M. Kozlow-
ski, Acc. Chem. Res. 2001, 34, 137–144; e) G. S. Kachalova, A. N. Popov,
H. D. Bartunik, Science 1999, 284, 473–476; f) J. P. Collman, R. Boulatov,
C. J. Sunderland, L. Fu, Chem. Rev. 2004, 104, 561–588; g) J. S. Olson, A.
Ghosh, in The Smallest Biomolecules: Diatomics and their Interactions
with Heme Proteins (Ed.: A. Ghosh), Elsevier, London, 2008, pp. 3–17.
[4] L. Pauling, Nature 1964, 203, 182–183.
[5] a) T. Yonetani, H. Yamamoto, T. Iizuka, J. Biol. Chem. 1974, 249, 2168–
2174; b) M. Ikeda-Saito, T. Iizuka, H. Yamamoto, F. J. Kayne, T. Yonetani,
J. Biol. Chem. 1977, 252, 4882–4887; c) M. Ikeda-Saito, H. Hori, T. Inu-
bushi, T. Yonetani, J. Biol. Chem. 1981, 256, 10267–10271.
[6] a) T. Kitagawa, M. R. Ondrias, D. L. Rousseau, M. Ikeda-Saito, T. Yonetani,
Nature 1982, 298, 869–871; b) K. Nagai, B. Luisi, D. Shih, G. Miyazaki, K.
Imai, C. Poyart, A. De Young, L. Kwiatkowsky, R. W. Noble, S.-H. Lin, N.-T.
Yu, Nature 1987, 329, 858–860.
[28] K. F. Ansari, C. Lal, Eur. J. Med. Chem. 2009, 44, 4028–4033.
[29] B. C. Askew, M. J. Breslin, M. E. Duggan, J. H. Hutchinson, R. S. Meissner,
J. J. Perkins, T. G. Steele, M. A. Patane, PCT Int. Appl. WO 2001/53262A1,
2001.
[7] a) J. Mispelter, M. Momenteau, D. Lavalette, J.-M. Lhoste, J. Am. Chem.
Soc. 1983, 105, 5165–5166; b) J. A. Lukin, V. Simplaceanu, M. Zou, N. T.
Ho, C. Ho, Proc. Natl. Acad. Sci. USA 2000, 97, 10354–10358.
[8] S. E. V. Phillips, B. P. Schoenborn, Nature 1981, 292, 81–82.
[30] P. Weyermann, F. Diederich, Helv. Chim. Acta 2002, 85, 599–617.
[31] a) H. Adams, K. D. M. Harris, G. A. Hembury, C. A. Hunter, D. Livingstone,
J. F. McCabe, Chem. Commun. 1996, 2531–2532; b) C. Allott, H. Adams,
P. L. Bernad, Jr., C. A. Hunter, C. Rotger, J. A. Thomas, Chem. Commun.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1998, 2449–2450; c) E. A. Meyer, R. K. Castellano, F. Diederich, Angew.
Chem. Int. Ed. 2003, 42, 1210–1250; Angew. Chem. 2003, 115, 1244–
1287; d) L. M. Salonen, M. Ellermann, F. Diederich, Angew. Chem. Int. Ed.
2011, 50, 4808–4842; Angew. Chem. 2011, 123, 4908–4944.
[32] A. Schweiger, G. Jeschke, Principles of Pulse Electron Paramagnetic Reso-
nance, Oxford University Press, Oxford, 2001.
[34] a) C. Fan, P. E. Doan, C. E. Davoust, B. M. Hoffman, J. Magn. Reson. 1992,
98, 62–72; b) E. R. Davies, Phys. Lett. A 1974, 47, 1–2.
[35] S. Stoll, A. Schweiger, J. Magn. Reson. 2006, 178, 42–55.
Received: March 31, 2016
[33] Z. Kazimierczuk, L. Dudycz, R. Stolarski, D. Shugar, J. Carbohydr. Nucleo-
sides Nucleotides 1981, 8, 101–117.
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
The distal hydrogen bond in dioxygen-
binding proteins is crucial for the dis-
crimination of O2 with respect to CO or
NO. The H-bond donor ability of distal
heterocycles, as predicted based on pKa
values, has been varied systematically.
The H-bond in the dioxygen adducts of
the CoII porphyrins was directly mea-
sured by Q-band Davies-ENDOR spec-
troscopy, and it was shown that the
strength of the hyperfine coupling be-
tween the dioxygen radical and the
distal H-atom increases with enhanced
acidity of the H-bond donor.
&
Hydrogen Bonds
M. N. Alberti, Y. Polyhach, M. D. Tzirakis,
L. Tçdtli, G. Jeschke,* F. Diederich*
&& – &&
Exploring the Strength of the H-Bond
in Synthetic Models for Heme
Proteins: The Importance of the NÀH
Acidity of the Distal Base
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
10
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!