A Facially Coordinating Tris-Benzimidazole Ligand for Nonheme Iron Enzyme Models

Article abstract of DOI:10.1002/ejic.202000984

Herein, we report a new tripodal tris-benzimidazole ligand (Tbim) that structurally mimics the 3-His coordination environment of certain nonheme mononuclear iron oxygenases. The coordination chemistry of Tbim was explored with iron(II) revealing a diverse set of coordination modes. The aerobic oxidation of biomimetic model substrate diethyl-2-phenylmalonate was studied using the Tbim?Fe and Fe(OTf)₂.

Full text of DOI:10.1002/ejic.202000984

Communications
doi.org/10.1002/ejic.202000984
1
2
3
4
5
6
7
8
9
A Facially Coordinating Tris-Benzimidazole Ligand for
Nonheme Iron Enzyme Models
Parami S. Gunasekera,[a] Preshit C. Abhyankar,[a] Samantha N. MacMillan,[b] and
David C. Lacy*[a]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Herein, we report a new tripodal tris-benzimidazole ligand
(Tbim) that structurally mimics the 3-His coordination environ-
ment of certain nonheme mononuclear iron oxygenases. The
coordination chemistry of Tbim was explored with iron(II)
revealing a diverse set of coordination modes. The aerobic
oxidation of biomimetic model substrate diethyl-2-phenylmalo-
nate was studied using the TbimÀ Fe and Fe(OTf)2.
Figure 1. Natural 2H1 C enzyme binding site (left) compared to tridentate
The mononuclear iron dependent nonheme subclass commonly
ligands with representative “mono-dentate” ligand donor groups and their
respective [NH]+ pKa in water (see Figure S19 and Table S1).
have an iron center coordinated within a two-histidine one-
carboxylate (2-His-1-C) facial triad binding pocket.[1,2] Additional
binding modes have also been discovered, and these include
the facial three-histidine (3-His), see-saw three-histidine one-
carboxylate, and see-saw four histidine binding modes.[3]
Modelling of the facial 3-His coordination has been accom-
plished through various nitrogen donor ligands such as 1,4,7-
triazacyclononane (tacn),[4] tris(2-pyridyl)methane,[5] trispyrazolyl
variants namely trispyrazolylborates,[6] trispyrazolylmethanes[7]
and trisimidazolylphosphines (TIP).[8] Some of these facially
coordinating ligands have been used to prepare Fe-based O2
derived oxidants, such as superoxo and oxo species[9,10,11,12,13,14]
and in catalysis.[15,16]
Despite the successes of these ligands in modeling 3-His
coordination,[17,18] we noted that most ligands contain donor
groups that are not represented in nature (Figure 1). For
instance, histidine donors are aromatic nitrogen groups with
sp2 hybridized N-atoms. In contrast, tacn has unconjugated sp3
hybridization. Accordingly, we noted that facially coordinating
ligands with imidazole and benzimidazole N donors have been
used to mimic facial 2-His-1-C and 3-His coordination modes
with iron.[19,20] Gebbink and coworkers reported a 2-His-1-C
facial triad using imidazole and benzimidazole ligands (2bim1C)
with iron.[19] In another report Fielder and coworkers use TIP,
which contains imidazole nitrogen donors, to model the 3-His
facial triad in an iron complex.[21] It is also noteworthy that, in
addition to the electronic similarity (i.e. sp2 hybridized and
aromatized), benzimidazole has excellent pKa similarities to
histidine. For example, the pKa of the nitrogen donor of
histidine is 6.0 in water,[22] which is close to the pKa of 6.6 for
1,2-dimethylbenzimidazole that is the donor moiety in Geb-
bink’s 2bim1C ligand.[23,24] This is also in contrast to a pKa of 7.69
for 4-methylimidazole.[25] While differences may be superficial,
ligands can impart a certain set of appropriate thermodynamic
properties required for dioxygen reactivity.[26] Herein, we report
the synthesis of a novel 3-His model with a tris-benzimidazole
ligand,
2,2’-(2-(1-ethylbenzimidazol-2yl)ethane-1,1-diyl)bis(1-
methylbenzimidazole) (Tbim), its coordination with iron, and a
brief foray into catalysis.
Synthesis of the new ligand Tbim used a strategy inspired
by the one Gebbink used to prepare 2bim1C (Scheme 1).[20,27,28]
With ligand in hand, we explored the coordination chemistry of
Tbim using a variety of FeX2 salts (X=OAc, Cl, OTf) (Scheme 2
[a] P. S. Gunasekera, P. C. Abhyankar, Prof. Dr. D. C. Lacy
Department of Chemistry, University at Buffalo, State University of New
York,
Buffalo, New York 14260, United States
E-mail: dclacy@buffalo.edu
[b] Dr. S. N. MacMillan
Department of Chemistry and Chemical Biology, Cornell University,
Ithaca, New York 14853, United States
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/ejic.202000984
Scheme 1. Ligand synthesis.
Eur. J. Inorg. Chem. 2021, 654–657
654
© 2020 Wiley-VCH GmbH

Communications
doi.org/10.1002/ejic.202000984
1
2
3
4
5
6
7
8
9
Scheme 2. Synthesis of 1 and 2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
and Scheme 3). Treatment of Tbim with Fe(OAc)2 or Fe(Cl)2 in
an acetonitrile solvent mixture afforded new complexes from
which crystals suitable for diffraction revealed the mono-ligated
complexes [Fe{Tbim}(X)2] (1, X=OAc; 2, X=Cl) (Figure 2). The
complexes show paramagnetically shifted 1H NMR signals in the
range À 20 ppm to 90 ppm; Evans method was conducted and
is consistent with an S=2 ground state for 1 and 2 (μeff =5.31
and 5.79). Both acetates in 1 are k-2 and Tbim is bound through
the two benzimidazole arms that form six-membered chelate
rings; the third benzimidazole arm, if bound to the metal, would
give a seven-membered ring. Salt metatheses with NaBPh4 on 1
or 2 were performed in MeOH in an attempt to remove a single
acetato or chlorido ligand and coordinate the third benzimida-
zole arm to iron. However, the reaction produced a yellow
precipitate from which we obtained colorless crystals of the
formulation [Fe{Tbim}2][BPh4]2 ([3][BPh4]2), which is a bis-ligated
metal complex salt whose connectivity was confirmed through
XRD (Figure S16).
A reaction of Tbim with Fe(OTf)2·2MeCN in acetonitrile also
produced the bis-ligated metal complex [3][OTf]2. However, if a
different workup procedure was used for the same in situ
prepared 1:1 ligand:metal mixture, a different product was
obtained. Namely, if the acetonitrile reaction mixture was
removed in vacuo to near dryness and the resulting residue
dissolved in dichloromethane the mono-ligated metal complex
[Fe{Tbim}(MeCN)2(OTf)][OTf] ([4][OTf]) was obtained in moder-
ate yields. The presence of acetonitrile ligands is confirmed
from ATR-FTIR spectroscopy (νCN =2279 and 2285 cmÀ 1) and X-
ray crystallography (Figure 2). Owing to a Schlenk-like equili-
brium between the two complexes and free Fe(OTf)2, [4][OTf]
was inseparable from [3][OTf]2 under the conditions studied
here.
Figure 2. Molecular structure of (top to bottom) 1, 2 and [4]+ with ellipsoids
shown at 50% probability; H-atoms, counterions, and solvent molecules are
not shown. Color scheme: orange=Fe; blue=N; red=O; yellow/green=F;
yellow=S; green=Cl; grey=C.
A 1:1 reaction mixture of Tbim with Fe(OTf)2·2MeCN in
1
acetonitrile followed by H NMR and 19F {1H} NMR indicated the
presence of both the mono-ligated and the bis-ligated com-
plexes in solution (Figure S15). The 19F {1H} NMR spectrum of a
2:1 ligand:metal ratio contains a sharp peak at À 80 ppm
consistent with an unbound triflate ion for [3][OTF]2.[20] In
contrast, when the ratio is less than 2:1, a broad signal is
apparent at À 73 ppm indicative of an equilibrium between
bound and unbound triflate ions in solution implicating the
presence of mono-ligated species [4]+ (Figure S12). This is
consistent with a Schlenk-like equilibrium between [3]2+ and
[4]+ at room temperature in MeCN; using density functional
theory (DFT), the calculated equilibrium lies toward the bis-
1
ligated complex with a free energy of À 3.2 kcal/mol (see SI). H
NMR spectroscopy was used to construct a kind of “Job plot” to
determine the optimal ratio of Fe(OTf)2:Tbim to prepare in situ
[4]+ for the catalysis studies later (Figure S17), the optimal ratio
Scheme 3. Synthesis of [3]2+ and [4]+.
Eur. J. Inorg. Chem. 2021, 654–657
www.eurjic.org
655
© 2020 Wiley-VCH GmbH

Communications
doi.org/10.1002/ejic.202000984
to achieve the highest concentration of [4]+ in the 10 mM
regime was found to be 3:2.
1
2
In nature, Fe(II) centers coordinated by the three-histidine
facial triad nominally occupy the face of a pseudo octahedron
similar to the coordination observed in [4]+ (Figure 3). However,
there are some notable differences between [4]+ and enzymatic
coordination. The average benzimidazole FeÀ N distance of [4]+
is 2.15 Å (average distance for all FeÀ N/O bonds in [4]+ is
2.16 Å), whereas the average Fe–N/O distance in resting state
mammalian cysteine dioxygenase (CDO) determined through K-
edge EXAFS is 2.04 Å.[33] Additionally, the protein structures’
3
4
5
6
7
8
9
Scheme 4. Catalytic 1,3-Diester Oxidation Studies.
Table 1. Results from catalytic aerobic oxidation of lithium diethyl 2-
phenylmalonate (Li[Phmal]).[a]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
cat[b]
H-a [%]
b [%]
c [%]
°
°
NÀ MÀ N angles are about 98 (average angle 95.7 for 2atf,
[31]
°
°
Fe/Tbim (3:2); “[4]+”
0
0
0
22
17
16
0
41
51
41
17
4
100.6 for 2b5 h, 94.5 for 4fbf, and 102.4 for 3bal), whereas
[4]+ has an average N–Fe–N angle of 89.9 for the benzimida-
Fe/Tbim(1:4); “[3]2+
Fe
”
°
zole nitrogen atoms. Also, it is to be noted that, unlike the
binding mode in Tbim and other synthetic ligands (e.g., Tp),
the protein active site imidazoles twist into a paddle wheel
conformation.
Fe+Ph2NH
no iron or ligand
61
77
0
[a] Conditions: Substrate added dropwise, 5 mol% catalyst, dry O2, 1 h;
data reported average of two runs, see SI for full data. [b] Fe=Fe
(OTf)2·2MeCN; Fe/L represents that complex was prepared in situ.
Following from the structural comparison of [4]+ with 3-His
coordination at enzyme active sites, we next tested [4]+ in
biomimetic oxidation reactions and chose the substrate lithium
diethyl 2-phenylmalonate (Li[Phmal], or Li[a]), a model substrate
often used in biomimicry of the 3-His enzyme diketone
dioxygenase (Dke1).[15,34] The use of Li[a] in Dke1 model studies
is common because the natural substrate acac is difficult to
used, H-a was obtained as the major final product when no
catalyst was present. For runs that contained a catalyst a
mixture of both b and c were obtained (Table 1); if water was
not rigorously excluded, c was obtained as the major product.[37]
When the iron triflate control oxidation contained the oxygen
radical scavenger diphenylamine,[38] H-a was obtained as the
major product accompanied by a small amount of c indicating
that radical oxygen species are responsible for the oxidation.
However, KO2 (either with or without O2) did not oxidize Li[a]
under the parent conditions we used (Scheme 4).
In conclusion, we synthesized the biomimetic ligand Tbim
and prepared coordination complexes with iron that structurally
mimics the 3-His active site in nonheme iron enzymes, such as
Dke1. We also demonstrated catalytic oxidation chemistry using
the substrate Li[a], but the simple salt Fe(OTf)2 had comparable
performance and so the role of ligand was not inferred.
Therefore, despite the common use of Li[a], it is not advisable
for biomimetic studies where it could give a “false positive” of
ligand-induced biomimicry. Novel ligand platforms are still
required to achieve the selectivity and rates achieved in
enzymes. In particular, designing systems that do not form bis-
ligated complexes and can oxidize difficult substrates like acac
are required and are ongoing.
oxidize; even Dke1 has a sluggish rate for acac oxidation (kcat
=
6.5 sÀ 1).[35] The expected product distribution for aerobic
oxidation of diethyl 2-phenylmalonate (H-a) has been studied
using a system with O2 and electrochemically generated
superoxide.[36] These products are ethylbenzoylformate (b) and
HOPhmal (c) and their relative distribution depends on the
concentration of substrate owing to competing reaction paths
from a common alkylperoxo intermediate.
Catalytic oxidation studies were performed by treating a
solution of Li[a] and 5 mol% catalyst with bubbling O2 for
1 hour (Scheme 4). After an aqueous work-up (see SI) the
products were analyzed by GCMS and compared against
authentic samples (Figures S20–S24). Under the conditions we
Deposition Numbers 1990453 (for 1), 1990454 (for 2), and
1990455 (for [4]+) contain the supplementary crystallographic
data for this paper. These data are provided free of charge by
the joint Cambridge Crystallographic Data Centre and Fachin-
formationszentrum Karlsruhe Access Structures service
www.ccdc.cam.ac.uk/structures.
Figure 3. Simplified primary coordination spheres of three-histidine iron
enzymes and [4]+. Clockwise from top left: PDB 2atf (Ni-bound CDO),[29] PDB
3bal (Zn-bound Dke1),[30] 4fag (Fe-bound gentisate dioxygenase),[31] PDB
2b5h (Fe-bound CDO).[32] Color scheme: grey=C; blue=N; red=O;
orange=Fe; green=Ni; blue-grey=Zn.
Eur. J. Inorg. Chem. 2021, 654–657
www.eurjic.org
656
© 2020 Wiley-VCH GmbH

Communications
doi.org/10.1002/ejic.202000984
[14] M. N. Blakely, M. A. Dedushko, P. C. Y. Poon, G. Villar-Acevedo, J. A.
Acknowledgements
1
2
3
4
5
6
7
8
9
Kovacs, J. Am. Chem. Soc. 2019, 141, 1867–1870.
[15] I. Siewert, C. A. Limberg, Angew. Chem. Int. Ed., 2008, 47, 7953–7956.
[16] S. Debobrata, T. K. Paine, Chem. Sci. 2016, 7, 5322–5331.
[17] S. Sahu, D. P. Goldberg, J. Am. Chem. Soc. 2016, 138, 11410–11428.
[18] L. Que, W. B. Tolman, Nature 2008, 455, 333–340.
[19] P. C. A. Bruijnincx, M. Lutz, A. L. Spek, W. R. Hagen, B. M. Weckhuysen, G.
Van Koten, R. J. M. K. Gebbink, J. Am. Chem. Soc. 2007, 129, 2275–2286.
[20] P. C. A. Bruijnincx, M. Lutz, A. L. Spek, E. E. Van Faassen, B. M. Weckhuy-
sen, G. Van Koten, R. J. M. Klein Gebbink, Eur. J. Inorg. Chem. 2005, 4,
779–787.
Financial support was provided by the University at Buffalo (UB).
This work was completed using the resources of the Chemistry
Instrument Center (CIC) and Center for Computational Research
(CCR) at UB. The FTICR Mass spectrometer used in this study was
purchased with funding from NIH S10 RR029517.
[21] H. Park, J. S. Baus, S. V. Lindeman, A. T. Fiedler, Inorg. Chem. 2011, 50,
11978–11989.
Conflict of Interest
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
[22] A. Moser, K. Range, D. M. York, J. Phys. Chem. B. 2010, 114, 13911–
13921.
[23] The pKa of 1,2-dimethylbenzimidazole was measured as part of this
work (see SI).
The authors declare no conflict of interest.
[24] C. H. R. Martínez, C. Dardonville, ACS Med. Chem. Lett. 2013, 4, 142–145.
[25] B. Lenarcik, P. J. Ojczenasz, Heterocyclic Chem. 2002, 39, 287.
[26] D. C. Lacy, Inorg. Chem. Front. 2019, 6, 2396–2403.
[27] S. Elgafi, L. D. Field, B. A. Messerle, P. Turner, T. W. J. Hambley, Organo-
met. Chem. 1999, 588, 69–77.
[28] I. I. Sahay, P. S. Ghalsasi, Synth. Commun. 2017, 47, 825–834.
[29] J. G. McCoy, L. J. Bailey, E. Bitto, C. A. Bingman, D. J. Aceti, B. G. Fox,
G. N. Phillips, Proc. Natl. Acad. Sci. USA 2006, 103, 3084–3089.
[30] G. R. Stranzl, U. G. Wagner, G. Straganz, W. Steiner, C. Kratky, Protein
Data Bank in Europe, https://www.ebi.ac.uk/pdbe/entry/pdb/3bal, (ac-
cessed June 2020), unpublished work.
[31] M. Ferraroni, L. Steimer, I. Matera, S. Bürger, A. Scozzafava, A. Stolz, F.
Briganti, J. Struct. Biol. 2012, 180, 563–571.
[32] C. R. Simmons, Q. Huang, Q Hao, T. P. Begley, P. A. Karplus, M. H.
Stipanuk, J. Biol. Chem. 2006, 281, 18723–18733.
[33] S. C. Chai, J. R. Bruyere, M. J. Maroney, J. Biol. Chem. 2006, 281, 15774–
15779.
[34] a) C. J. Allpress, K. Grubel, E. Szajna-Fuller, A. M. Arif, L. M. Berreau, J. Am.
Chem. Soc. 2013, 135, 659–668; b) H. Park, M. M. Bittner, J. S. Baus, S. V.
Lindeman, A. T. Fiedler, Inorg. Chem. 2013, 135, 659–668; c) R. Ramasu-
bramanian, K. Anandababu, M. Kumar, R. Mayilmurugan, Dalton Tran.
2018, 47, 4049–4053.
[35] G. D. Straganz, B. Nidetzky, J. Am. Chem. Soc. 2005, 127, 12306–12314.
[36] P. M. Allen, U. Hess, C. S. Foote, M. M. Baizer, Synthetic Communications
1982, 12, 123–129.
Keywords: Nonheme iron
Oxygenase · Enzyme model · Ligand design
·
Coordination chemistry
·
[1] P. C. A. Bruijnincx, G. Van Koten, R. J. M. K. Gebbink, Chem. Soc. Rev.
2008, 37, 2716–2744.
[2] K. D. Koehntop, J. P. Emerson, L. Que, J. Biol. Inorg. Chem. 2005, 10, 87–
93.
[3] D. Buongiorno, G. D. Straganz, Coord. Chem. Rev. 2013, 257, 541–563.
[4] D. W. Blakesley, S. C. Payne, K. S. Hagen, Inorg. Chem. 2000, 39, 1979–
1989.
[5] K. Anandababu, R. Ramasubramanian, H. Wadepohl, P. Comba, N. J.
Britto, M. Jaccob, R. Mayilmurugan, Chem. Eur. J. 2019, 25, 9540–9547.
[6] M. Sallmann, C. Limberg, Acc. Chem. Res. 2015, 48, 2734–2743.
[7] D. L. Reger, C. A. Little, A. L. Rheingold, R. Sommer, G. J. Long, Inorganica
Chim. Acta, 2001, 316, 65–70.
[8] C. G. J. Tazelaar, J. C. Slootweg, K. Lammertsma, Coord. Chem. Rev. 2018,
356, 115–126.
[9] F. Oddon, Y. Chiba, J. Nakazawa, T. Ohta, T. Ogura, S. Hikichi, Angew.
Chem., Int. Ed., 2015, 54, 7336–7339.
[10] A. A. Fischer, S. V. Lindeman, A. T. Fiedler, Chem. Commun. 2018, 54,
11344.
[11] J. B. Gordon, A. C. Vilbert, I. M. DiMucci, S. N. MacMillan, K. M. Lancaster,
P. Moëmme-Loccoz, D. P. Goldberg, J. Am. Chem. Soc. 2019, 141, 17533–
17547.
[12] D. Kass, T. Corona, K. Warm, B. Braun-cula, U. Kuhlmann, E. Bill, S. Mebs,
M. Swart, H. Dau, M. Haumann, P. Hildebrandt, K. Ray, J. Am. Chem. Soc.
2020, 142, 5924–5928.
[13] C. W. Chiang, S. T. Kleepspies, H. D. Stout, K. K. Meier, P. Y. Li, E. L.
Bominaar, L. Que, E. Münk, W. Z. Lee, J. Am. Chem. Soc. 2014, 136,
10846.
[37] See SI for drying procedure.
[38] P. Comba, Y. M. Lee, W. Nam, A. Waleska, Chem. Commun. 2014, 50, 412.
Manuscript received: October 23, 2020
Revised manuscript received: November 30, 2020
Accepted manuscript online: December 11, 2020
Eur. J. Inorg. Chem. 2021, 654–657
www.eurjic.org
657
© 2020 Wiley-VCH GmbH