Metal Substitution Modulates the Reactivity and Extends the Reaction Scope of Myoglobin Carbene Transfer Catalysts

Article abstract of DOI:10.1002/adsc.201700202

Engineered myoglobins have recently emerged as promising scaffolds for catalyzing carbene-mediated transformations. In this work, we investigated the effect of altering the metal center and first-sphere coordination residue on the carbene transfer reactivity of myoglobin. To this end, we first established an efficient protocol for the recombinant expression of myoglobin variants incorporating metalloporphyrins with non-native metals, including second- and third-row transition metals (ruthenium, rhodium, iridium). Characterization of the cofactor-substituted myoglobin variants across three different carbene transfer reactions (cyclopropanation, N–H insertion, S–H insertion) revealed a major influence of the nature of the metal center, its oxidation state and first-sphere coordination environment on the catalytic activity, stereoselectivity, and/or oxygen tolerance of these artificial metalloenzymes. In addition, myoglobin variants incorporating manganese- or cobalt-porphyrins were found capable of catalyzing an intermolecular carbene C–H insertion reaction involving phthalan and ethyl α-diazoacetate, a reaction not supported by iron-based myoglobins and previously accessed only using iridium-based (bio)catalysts. These studies demonstrate how modification of the metalloporphyrin cofactor environment provides a viable and promising strategy to enhance the catalytic properties and extend the reaction scope of myoglobin-based carbene transfer catalysts. (Figure presented.).

Full text of DOI:10.1002/adsc.201700202

Title: Metal Substitution Modulates the Reactivity and Extends the
Reaction Scope of Myoglobin Carbene Transfer Catalysts
Authors: Rudi Fasan, Gopeekrishnan Sreenilayam, Eric Moore, and
Viktoria Steck
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Adv. Synth. Catal. 10.1002/adsc.201700202
Link to VoR: http://dx.doi.org/10.1002/adsc.201700202

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Metal Substitution Modulates the Reactivity and
Extends the Reaction Scope of Myoglobin Carbene
Transfer Catalysts
a‡
a‡
a
a
Gopeekrishnan Sreenilayam , Eric J. Moore , Viktoria Steck , and Rudi Fasan *.
a
Department of Chemistry, University of Rochester, Rochester, New York 14627, USA
Email: rfasan@ur.rochester.edu; Ph: +1 (585) 273 3504; Fax: +1 (585) 276 0205
‡
These authors contributed equally to this work.
KEYWORDS: myoglobin • biocatalysis • carbene transfer • artificial metalloenzymes • protein
engineering • C‒H carbene insertion
1
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Abstract. Engineered myoglobins have recently emerged as promising scaffolds for catalyzing
carbene-mediated transformations. In this work, we investigated the effect of altering the metal
center and its first-sphere coordination environment on the carbene transfer reactivity of
myoglobin. To this end, we first established an efficient protocol for the recombinant expression
of myoglobin variants incorporating metalloporphyrins with non-native metals, including second-
and third-row transition metals (ruthenium, rhodium, iridium). Characterization of the cofactor-
substituted myoglobin variants across three different carbene transfer reactions (cyclopropanation,
N‒H insertion, S‒H insertion) revealed a major influence of the nature of metal center, its
oxidation state and first-sphere coordination environment on the catalytic activity,
stereoselectivity, and/or oxygen tolerance of these artificial metalloenzymes. In addition,
myoglobin variants incorporating manganese- or cobalt-porphyrins were found capable of
catalyzing an intermolecular carbene C‒H insertion reaction involving phthalan and ethyl -
diazoacetate, a reaction not supported by iron-based myoglobins and previously accessed only
using iridium-based (bio)catalysts. These studies demonstrate how modification of the
metalloporphyrin cofactor environment provides a viable and promising strategy to enhance the
catalytic properties and extend the reaction scope of myoglobin-based carbene transfer catalysts.
2
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Introduction
The transition metal-catalyzed insertion of carbenoid species into C=C, Y‒H (Y = N, S, O, Si),
and C‒H bonds constitutes a powerful approach to the construction of new carbon-carbon and
carbon-heteroatom bonds in organic chemistry.[1] While major efforts have been devoted to the
development of organometallic complexes for executing these transformations,[1] biocatalytic
platforms useful for promoting carbene transfer reactions have only recently begun to emerge.[2]
Arnold and co-workers have shown how cytochrome P450 enzymes can catalyze the
[
3]
cyclopropanation of styrene derivatives and carbene N‒H insertion with aniline derivatives. Our
group has demonstrated how engineered myoglobins can promote a wide range of carbene-
mediated transformations, including olefin cyclopropanation,[4] carbene N‒H and S‒H
insertions,[5] aldehyde olefination,[6] and [2,3]-sigmatropic rearrangement reactions.[7] More
recently, biocatalytic systems based on these[8] and other protein and DNA scaffolds have been
investigated in the context of carbene transfer reactions.[9]
Myoglobin (Mb) contains a heme cofactor (= iron-protoporphyrin IX), which is embedded
within the core of the protein via non-covalent interactions and coordination of the iron atom
through the side-chain imidazolyl group of a ‘proximal’ histidine residue (His93 in sperm whale
myoglobin).[10] During myoglobin-mediated carbene transfer catalysis, the α-diazoester carbene
donor reagents are predicted to react with the heme center to give rise to a reactive heme-bound
carbenoid intermediate.[4a] This species has electrophilic character[4a] and can engage a variety of
nucleophilic substrates in a reaction manifold that leads to the formation of cyclopropanes,[4a]
amines,[5a] thioethers,[5b, 7] and olefins.[6] Under this mechanistic scenario, we hypothesized that
the nature of the metal center as well as the presence/absence of first-sphere coordination via the
proximal residue could have a significant influence on the carbene transfer reactivity of these
3
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
biocatalysts. In this work, we have investigated this aspect through the preparation of metal- and
proximal ligand-substituted myoglobin variants and the characterization of their activity in a series
of model carbene insertion reactions under varying reaction conditions. To enable these studies,
we first implemented an efficient strategy for the recombinant production of myoglobin variants
incorporating first-, second- and third-row transition metals in bacterial cells. We demonstrate how
metal and proximal ligand substitution provides an effective means to modulate the catalytic
activity, selectivity, and oxygen tolerance of these myoglobin-based catalysts toward different
types of carbene-mediated transformations and extend the reaction scope of these carbene transfer
biocatalysts to include intermolecular carbene C‒H insertion reactivity.
Results and Discussion
Recombinant expression of cofactor-substituted myoglobins. Pioneering studies by
Yonetani, Hayashi and Watanabe have demonstrated the possibility to substitute the heme cofactor
in myoglobin with non-native metalloporphyrins and porphynoids.[11] These strategies have
involved extraction of the heme from the protein with organic solvents followed by dialysis and
refolding of apomyoglobin (= heme-free myoglobin) in the presence of the non-native cofactor.
More recently, Watanabe and coworkers reported the successful preparation of metallo-substituted
hemoproteins via reconstitution of apo-hemoproteins with metalloporphyrins directly in cell
lysate[12], whereas a related protocol was applied by Hartwig and coworkers to obtain various
metallo-substituted myoglobins, although only Ir-containing variants were isolated and
characterized.[8a] Despite this progress, in vitro reconstitution protocols tend to be laborious and
time-consuming, requiring extensive manipulation of the protein after expression and purification.
4
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
In addition, they are incompatible with the realization of biocatalytic transformations using whole-
cell systems.[4b]
To overcome these limitations, our group has investigated strategies to enable the
recombinant production of metal- and cofactor-substituted hemoprotein-based catalysts. To this
end, we previously developed and reported an efficient protocol for the recombinant production of
Mn- and Co-containing Mb, which relies on the use of E. coli cells expressing a heterologous outer
membrane heme transporter (ChuA)[13] and the chaperone complex GroEL/ES.[14] In this system,
ChuA mediates the cellular uptake of exogenous Mn- and Co-protoporphyrin IX (Mn(ppIX) and
Co(ppIX), respectively) supplemented to the growth medium. Cells are grown in iron-free minimal
medium to prevent the production of heme. Co-expression of the chaperone proteins GroEL/ES
was found to increase the expression levels of soluble cofactor-substituted Mb, possibly by
reducing aggregation of apomyoglobin prior to binding to the non-native cofactor.[14] While these
and other studies suggested a certain permissivity of the ChuA transporter toward the uptake of
heme analogs containing first-row transition metals,[9c, 14-15] its functionality in the context of
metalloporphyrin complexes containing second- and third-row transition metals remained to be
determined.
Given the promising activity of synthetic Ir(Me)-porphyrins as carbene transfer
catalysts,[16] we initially targeted the preparation of Mb incorporating Ir(Me)-protoporphyrin IX
(
Ir(Me)(ppIX), 3a). The latter was prepared in two steps and 22% overall yield via metalation of
protoporphyrin IX dimethyl ester 1a with [Ir(COD)Cl] , followed by methylation with MeI, and
2
ester hydrolysis (Scheme 1A, see Supplemental Information for further details). Upon
optimization of the expression conditions, Mb[Ir(Me)(ppIX)] was successfully produced and
5
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
-1
isolated from E. coli in good yields (3.5 mg L culture). The electronic absorption spectrum of
this protein shows a Soret band at 399 nm prior and after addition of dithionite, which differs
6
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Scheme 1. Synthetic routes for preparation of Ir(Me)-protoporphyrin IX and Ir(Me)-
mesoporphyrin IX (A) and Rh- and Ru(CO)-mesoporphyrins IX (B).
from the characteristic Soret signatures of ferric and ferrous Mb at 410 and 434 nm, respectively
(Figure 1A). Furthermore, no signal at 424 nm was observed upon addition of reductant and CO
to the protein, further confirming the absence of contaminant Mb. These results were achieved
upon introducing important modifications to our previously reported expression protocol,
including pre-incubation of the cells with the non-native cofactor prior to induction, culture
condensation, and optimization of cell induction conditions (see Experimental Section for further
details). The cell condensation strategy also reduces the amount of non-native cofactor required
during expression of the Mb variant.
Encouraged by these initial results, we tested the incorporation of Ir(Me)-mesoporphyrin
IX (Ir(Me)(mpIX), 3b), which is structurally analogous to Ir(Me)(ppIX) and synthetically more
accessible (48% yield from mesoporphyrin IX dimethyl ester (1b); Scheme 1A). Notably, the
Ir(Me)(mpIX) complex could be incorporated into Mb with comparable efficiency to the
-1
protoporphyrin IX-based counterpart (5 mg L culture). These results thus demonstrated the ability
of the ChuA transporter to recognize heme analogs with a modified metal center and porphyrin
ligand. Building upon these findings, Mb variants incorporating second-row transition metals in
the form of Ru(CO)(mpIX) and Rh(mpIX) (Scheme 1B) were also successfully expressed and
-1
isolated from E. coli (2-3 mg L culture). Based on the amount of apomyoglobin obtained under
same expression conditions in the absence of supplemented cofactor, the efficiency of in vivo
incorporation of the non-natural metalloporphyrin cofactor was estimated to range between 50%
7
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
and 90%. Overall, using the strategy outlined above, a panel of Mb catalysts based on the broadly
active Mb variant Mb(H64V,V68A)[4a] and containing different metal
Figure 1. Electronic absorption spectra for Mb(H64V,V68A) and Mb(H64V,V68A,H93F)
variants incorporating (A) Fe-protoporphyrin IX (heme), (B) Mn-protoporphyrin IX , (C) Co-
protoporphyrin IX, (D) Ru(CO)-mesoporphyrin IX, (E) Rh-mesoporphyrin IX, and (F) Ir(Me)-
mesoporphyrin IX, as the cofactor. Spectra were measured prior (solid line) and after (dashed line)
8
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
addition of sodium dithionite. The inset plots illustrate the Q band region of each spectrum. See
Table S1 for further details on the spectroscopic features of these proteins.
centers (Mn, Co, Ru, Rh, Ir) was obtained. As expected, each construct exhibits a distinctive
spectroscopic signature corresponding to the Soret band and Q bands (Figure 1 and Table S1).
The spectral properties of the recombinantly produced cofactor-substituted Mb variants were
identical to those of corresponding variants obtained via in vitro reconstitution with apomyoglobin
(see Experimental Section), thus further confirming the efficiency of cofactor incorporation
through the recombinant method. Altogether, these experiments demonstrated that the substrate
scope of ChuA extends beyond metalloporphyrins based on protoporphyrin IX and first-row
transition metals. Importantly, a convenient, time-effective, and technically straightforward
strategy was implemented to produce cofactor-substituted Mb catalysts.
Selection of Mb variants and carbene transfer reactions. We previously identified
Mb(H64V,V68A) as a highly active and stereoselective biocatalyst for olefin cyclopropanation.[4a]
This Mb variant also possesses high catalytic activity toward other carbene-mediated
[
5]
transformations such as N‒H and S‒H insertion. Based on these properties, Mb(H64V,V68A)
was selected as an optimal scaffold to examine the impact of metal substitution on Mb-mediated
carbene transfer reactivity and stereoselectivity. Accordingly, a first set of cofactor-substituted
Mb(H64V,V68A) variants containing a Mn, Co, Ru, Rh, and Ir metal center were prepared as
described earlier. A second set of Mb variants was then generated based on the triple mutant
Mb(H64V,V68A,H93F), in which the histidine residue involved in axial coordination of the heme
iron (His93) is substituted for another aromatic residue which lacks any side-chain Lewis basic
group for metal binding (Phe). These proximal ligand Mb variants were meant to provide insights
9
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
into the functional role of first-sphere metal coordination by the proximal histidine residue on the
carbene transfer activity of these metalloproteins. Perturbation of the cofactor environment as a
result of the H93F substitution is evidenced by the shift in the Soret bands in the corresponding
visible range spectra of these proteins, in particular for those containing first-row metals (Figure
1
A-C and Table S1).
The panel of ten different Mb catalysts (= 5 metal centers, with and without H93F
substitution) was initially characterized for activity in three model carbene transfer reactions,
namely cyclopropanation of styrene, N‒H insertion with aniline, and S‒H insertion with
thiophenol, in the presence of ethyl α-diazoacetate (EDA) as carbene donor. In each case, the
reactions were performed in the presence or absence of reductant (sodium dithionite) and under
anaerobic and aerobic conditions with the two-fold purpose of (a) evaluating the effect of these
parameters on catalyst performance, and (b) identifying Mb variants capable of operating under
aerobic conditions. Indeed, the application of iron-based myoglobins as carbene transfer catalysts
has so far required strictly anaerobic conditions, due to the inhibitory action of oxygen, i.e., Mb
native ligand, on this non-native reactivity. Oxygen tolerance is desirable as it enhances the
accessibility of these biocatalysts for synthetic applications.
Cyclopropanation activity. Table S2 reports the results for the cyclopropanation
reactions in the presence of the various Mb catalysts under the four different reaction conditions,
with the most relevant data being summarized in Table 1. These reactions were carried out under
catalyst-limited conditions to enable a direct comparison of the performance of each Mb variant
under these settings. Under ‘standard conditions’ (anaerobic + Na
2
S
2
O
4
or ‘-/red.’),
Mb(H64V,V68A) provides a 64% conversion of styrene (6) and EDA (7) to the trans-(1S,2S)
cyclopropane product (8) with excellent diastereo- and stereoselectivity (99.9% de and ee; Entry
1
0
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
1
, Table 1). Under the same conditions, low product yields (0-7%) were observed for the
Mb(H64V,V68A) and Mb(H64V,V68A,H93)-based variants containing a Co, Rh, or Ru center
Entries 5-7, Table 1). Compared the latter, Mb(H64V,V68A)[Mn(ppIX)] and
(
Mb(H64V,V68A)[Ir(Me)(mpIX)] exhibit noticeably higher cyclopropanation activity, as
indicated by the higher yields (29-36% vs. 0-7%) and catalytic turnovers (290-356 TON vs. 11-68
TON, Table S2) obtained with these Mb variants under anaerobic conditions (Entries 4 and 9,
Table 1). Interestingly, Mb(H64V,V68A,H93F)[Mn(ppIX)] showed negligible product
conversion (<2%) under identical conditions (Table S2), indicating that coordination of the
manganese center by the proximal histidine residue is critical for cyclopropanation activity. This
is contrast to the Fe-based variants, for which the H93F substitution has no effect on the catalytic
efficiency of the protein (Entry 3 vs. 1, Table 1). Even if undetectable spectroscopically (Figure
1
), the absence of contaminant Mb in the reactions with the cofactor-substituted variants was
further confirmed via control experiments in which these catalysts were incubated with sodium
dithionite and carbon monoxide prior to the reaction.
Other interesting observations derived from the analysis of the impact of the cofactor and
proximal ligand substitutions on the stereoselectivity of the biocatalyst. Indeed, the Mn-containing
Mb(H64V,V68A) variant, along with the less active Co- and Rh-containing counterparts, were
found to mirror the high stereoselectivity of the Fe-containing Mb(H64V,V68A) variant,
producing 8 with high diastereo- and enantioselectivity (96-99% de, 97-99% ee; Entries 4, 5, and
7
, Table 1). In contrast, the Mb variants incorporating the Ir(Me)- and Ru(CO)-mesoporphyrin IX
cofactors show significantly reduced diastereo- and enantioselectivity under identical reaction
conditions (42-56% de; 4-7% ee; Entries 6 and 8, Table 1). These values are similar, albeit not
identical, to those exhibited by the free cofactor (e.g., 57% de, 0% ee for Ir(Me)(mpIX)), raising
1
1
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
the question of whether these cofactors are released during the reaction. To address this point,
control experiments were conducted using 4-morpholinopyridine, which can bind to Ir(Me)(mpIX)
in free form but not when bound to myoglobin.[8a] In the presence of 50 mM 4-
morpholinopyridine[17], negligible product formation was observed in the presence of free
Ir(Me)(mpIX), whereas significant amounts of 8 accumulated in the reaction with
Mb(Mb(H64V,V68A)[Ir(Me)(mpIX)] (i.e., approx.. 2-fold lower yield and similar % de and ee
values compared to the reaction with no inhibitor). These results clearly indicate that the observed
carbene transfer activity arises from the metalloprotein and not from dissociated cofactor. Since
both Ir(Me)(ppIX) and Ru(CO)(ppIX) cofactor bears a metal-bound ligand, we ascribe the low
stereoselectivity exhibited by the Ir- and Ru-containing Mb variants compared to the Fe-, Mn, Co-
and Rh-based counterparts to a disruption of the chiral environment provided by the H64V/V68A
mutations due to the presence of the axial ligands. This outcome is perhaps not surprising
considering that these mutations were selected to favor trans-(1S,2S) selectivity in styrene
cyclopropanation with EDA by heme-containing Mb variants which naturally lack a metal-bound
ligand.[4a]
As shown by the data in Table 1 and Table S2, the stereochemical outcome of the
cyclopropanation reaction is also affected by the oxidation state of the metal (Entry 2 vs. 1, Table
1
) and nature of the proximal ligand (Entries 3 vs. 1, Table 1). For example, the His93Phe
substitution significantly affects the enantioselectivity of the Fe-, Mn-, and Co-based Mb variants
48-86% ee vs. 99% ee). This effect is less pronounced for the Ir- and Ru-based catalysts, likely
(
due to the lack of metal coordination via the proximal ligand (vide infra). Importantly, the results
above demonstrate that the stereoselectivity of these biocatalysts can be affected, and thus tuned,
by altering the first coordination sphere and oxidation state of the metal center.
1
2
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Finally, analysis of the aerobic reactions showed that both Mb(H64V,V68A) and
Mb(H64V,V68A)[Ir(Me)(mpIX)] can furnish the cyclopropanation product 8 in good yields (41-
4
3%) in the presence of oxygen (Entries 10 and 12, Table 1). In contrast, the cyclopropanation
activity of Mb(H64V,V68A)[Co(mpIX)] is drastically suppressed under aerobic conditions (<1%,
Table S2). Importantly, the Mb(H64V,V68A)-catalyzed transformation maintains the excellent
diastereo- and stereoselectivity observed under anaerobic conditions. Using Mb(H64V,V68A) at
0
.5 mol%, the cyclopropanation product 8 was obtained in 68% yield and >99% de and ee in an
open vessel reaction, thus demonstrating the viability of this biocatalytic transformation under
aerobic conditions.
Table 1. Catalytic activity and selectivity of cofactor-substituted Mb variants for cyclopropanation
of styrene with ethyl α-diazoacetate (EDA).a
Yb
Yieldc
% ded
% eed
Entry
Catalyst
Cofactor
1
2
3
4
5
6
7
8
9
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A,H93F)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A,H93F)
Mb(H64V,V68A)
Mb(H64V,V68A)
Fe(ppIX)
Fe(ppIX)
Fe(ppIX)
Mn(ppIX)
Co(ppIX)
-/red.
-/-
64%
14%
63%
29%
4%
>99
71
98
>99
99
42
96
56
75
98
>99
72
86
99
99
7
97
4
0
-/red.
-/red.
-/red.
-/red.
-/red.
-/red.
-/-
Ru(CO)(mpIX)
Rh(mpIX)
Ir(Me)(mpIX)
Ir(Me)(mpIX)
Fe(ppIX)
Fe(ppIX)
Ir(Me)(mpIX)
Fe(ppIX)
7%
3%
25%
36%
43%
15%
41%
68%
1
1
1
1
0
1
2
3
O
O
2
/red.
/red.
95
70
-1
2
94
59
>99
O
2
/-
O
2
/red.e
>99
1
3
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
a
Reaction conditions: 0.01 M styrene, 0.02 M EDA, 10 µM Mb catalyst, 50 mM phosphate
b
buffer (pH 7) containing 5% ethanol, room temperature. Variable parameter (Y): “-/red.” =
anaerobic, 0.01 M Na
2
S
2
O
4
; “-/-” = anaerobic, no reductant; “O
2
-red.” = aerobic, 0.01 M Na
2
S
2
O ;
4
c
“O
2
/-” = open vessel, no reductant. As determined by gas chromatography (GC) using calibration
d
e
curves with isolated 8. For trans-(1S,2S) stereoisomer as determined by chiral GC. Using 0.5
mol% catalyst.
N‒H Insertion activity. The carbene N‒H insertion reactivity of the Mb catalysts was
assessed using the conversion of aniline (9) to 10 in the presence of EDA as a model reaction
(Tables 2 and S3). Similarly to the cyclopropanation reactions, the highest yield in this reaction
were obtained with Mb(H64V,V68A) and its related variant bearing the H93F substitution,
Mb(H64V,V68A,H93F), under reducing and anaerobic conditions (47-56%, Entries 1 and 2,
Table 2). At the same time, the differential impact of metal substitution in the N-H insertion
reactions compared to the cyclopropanation became evident from analysis of these data. Indeed,
higher N-H insertion activity was observed with Mb catalysts containing first-row metals
following the order: Fe > Mn ≈ Co > Rh > Ir ≈ Ru, whereas a different order of reactivity
characterized the cyclopropanation reactions (Fe > Mn ≈ Ir > Ru > Rh ≈ Co). Other interesting
trends emerged from these experiments. For both types of reactions, the TON exhibited by the
first-row metal-containing Mb catalysts are generally higher (2- to 5-fold) in the presence of
reductant (Tables S2 and S3), indicating that higher catalytic efficiency is provided by lower
II
III
II
III
oxidation states of these metals (i.e., Fe (ppIX) > Fe (ppIX); Co (ppIX) > Co (ppIX),
II
III
Mn (ppIX) > Mn (ppIX)). An opposite trend is exhibited by the Mb catalysts incorporating
second- and third-row metals, in particular for the N‒H insertion reactions (Table S3).
Importantly, these experiments led to the successful identification of N‒H insertion
catalysts capable of operating under aerobic conditions. Specifically, both Mb(H64V,V68A,H93F)
and Mb(H64V,V68A,H93F)[Rh(mpIX)] were found to produce 10 in relatively good yields (34%
1
4
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
and 12%, respectively) in the presence of oxygen (Entries 10-11, Table 2). Since
Mb(H64V,V68A) shows only minimal product formation (3%) under identical conditions, it can
be derived that oxygen tolerance in the Fe-based catalyst is conferred by removal of axial
coordination on the heme iron by the proximal residue (Entry 9 vs 10, Table 2). In contrast, the
Rh-based catalyst shows no dependence on the nature of the proximal ligand (Table S3). Using
Mb(H64V,V68A,H93F) at 0.75 mol%, the N-H insertion product 10 was obtained in 78% yield in
an open flask reaction (Entry 12, Table 2).
Table 2. Catalytic activity of cofactor-substituted Mb variants for N‒H insertion reaction with
a
aniline and EDA. See Table S3 for additional data.
Yieldc
Entry
Catalyst
Cofactor
Y
TON
1
2
3
4
5
6
7
8
9
Mb(H64V,V68A)
Mb(H64V,V68A,H93F)
Mb(H64V,V68A)
Fe(ppIX)
Fe(ppIX)
Mn(ppIX)
Co(ppIX)
-/red.
-/red.
-/red.
-/red.
-/red.
-/red.
-/red.
-/-
56%
47%
14%
13%
5%
10%
6%
17%
3%
555
467
138
129
52
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Ru(CO)(mpIX)
Rh(mpIX)
Ir(Me)(mpIX)
Ir(Me)(mpIX)
Fe(ppIX)
97
65
168
33
342
114
100
O
O
O
2
2
2
/red.
/red.
/red.
1
1
1
0
1
2
Mb(H64V,V68A,H93F)
Mb(H64V,V68A)
Mb(H64V,V68A,H93F)
Fe(ppIX)
Rh(mpIX)
Fe(ppIX)
34%
12%
78%
O
/red.c
2
a
Reaction conditions: 0.01 M aniline, 0.01 M EDA, 10 µM Mb catalyst, 50 mM phosphate buffer
pH 8) containing 4% ethanol, room temperature. See legend of Table 1 for definition of variable
(
1
5
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
b
parameter Y.
As determined by GC using calibration curves with isolated 10.
c
Using 0.75 mol% catalyst.
S‒H insertion activity. The carbene S‒H insertion reactivity of the metalloproteins was
investigated using a model reaction involving thiophenol 11 and EDA to give 12 (Table 3). The
results from these experiments further evidenced the peculiar effect of metal and proximal ligand
substitution on the carbene transfer reactivity of the Mb catalysts (Tables 3 and S4). Under
anaerobic conditions, the order of reactivity derived from these data is: Fe > Ir ≈ Ru > Co ≈ Mn >
Rh. Thus, whereas the Ru-based Mb variants are poor catalysts for cyclopropanation and N-H
insertion reactions (<1-7% yield, corresponding to <10-70 TON), they represent effective catalysts
for carbene S‒H insertion reactions (70-79% yields, corresponding to 700-790 TON).
Furthermore, oxygen-tolerant S‒H insertion catalysts were identified. In this respect, the best
performing Mb variants were determined to be Mb(H64V,V68A,H93F)[Ir(Me)(mpIX)] (58%
yield; Entry 11, Table 3), followed by Mb(H64V,V68A)[Ru(CO)(mpIX)] (36% yield) and
Mb(H64V,V68A,H93F) (29% yield) (Entries7 and 3, respectively, Table 3). Further substitution
of the proximal ligand residue (H93A) in the Ir(Me)-containing variant (vide infra for further
details) resulted in a catalyst capable of providing quantitative conversion of thiophenol to the
desired S‒H insertion product 7 in an open vessel reaction (Entry 12, Table 3). To examine the
role of the protein matrix in influencing the reactivity of the metalloporphyrin cofactors, parallel
experiments were conducted using isolated Ir(Me)(mpIX) or Ru(CO)(mpIX) as the catalysts for
the S‒H insertion reaction under aerobic conditions. These experiments showed that the Ir- and
Ru-containing Mb variants exhibit a 3.5-fold and 6-fold higher S‒H insertion activity, respectively,
compared to corresponding cofactors when used in free form (Ir: >1,000 vs. 290 TON; Ru: 359
1
6
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
vs. 59 TON). These results demonstrate the beneficial effect of the protein scaffold toward
enhancing the S‒H transfer reactivity of the metalloporphyrins.
Table 3. Catalytic activity of cofactor-substituted Mb variants for S‒H insertion reaction with
a
thiophenol and EDA. See Table S4 for additional data.
Yieldc
Entry
Catalyst
Cofactor
Y
TON
1
2
3
4
5
6
7
8
9
Mb(H64V,V68A)
Mb(H64V,V68A,H93F)
Mb(H64V,V68A,H93F)
Mb(H64V,V68A)
Fe(ppIX)
Fe(ppIX)
Fe(ppIX)
Mn(ppIX)
Co(ppIX)
-/red.
-/red.
99%
89%
29%
13%
26%
79%
36%
9%
70%
82%
58%
>99%
985
885
293
133
260
795
359
95
703
819
579
>1000
O /red.
2
-/red.
-/red.
-/red.
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Ru(CO)(mpIX)
Ru(CO)(mpIX)
Rh(mpIX)
Ir(Me)(mpIX)
Ir(Me)(mpIX)
Ir(Me)(mpIX)
Ir(Me)(mpIX)
O /-
2
-/red.
-/red.
-/-
Mb(H64V,V68A)
1
1
1
0
1
2
Mb(H64V,V68A,H93F)
Mb(H64V,V68A,H93F)
Mb(H64V,V68A,H93A)
O
O
2
/-
/-
2
a
Reaction conditions: 0.01 M thiophenol, 0.02 M EDA, 10 µM Mb catalyst, 50 mM phosphate
buffer (pH 8) containing 5% ethanol, room temperature. See legend of Table 1 for definition of
b
variable parameter Y. As determined by GC using calibration curves with isolated 12.
1
7
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
C‒H insertion activity. Carbene insertion into C‒H bonds constitutes an attractive
strategy for C‒H functionalization[1c] and seminal work by Callot, Che, and Woo have shown how
porphyrin complexes of Rh and Ir can promote these reactions.[16b, 18] Inspired by these studies, we
hypothesized that metal substitution could provide a potential avenue to endow the Mb catalysts
with carbene C‒H insertion reactivity. Preliminary experiments indeed showed no detectable C‒
H insertion activity for the iron-based Mb variants Mb(H64V,V68A) and Mb(H64V,V68A,H93F),
as determined using a model reaction with phthalan (13) and EDA (Table 4). To our delight,
various cofactor-substituted Mb variants were found to produce the desired C‒H functionalized
product
14
(Table
4).
Among
them,
Mb(H64,V68A)[Co(ppIX)],
Mb(H64,V68A,H93F)[Mn(ppIX)], and Mb(H64,V68A)[Ir(Me)(mpIX)] emerged as the most
promising catalysts for this transformation, producing the C-H insertion product 9 in 20-28% yield
at a catalyst loading of 0.2 mol% (100-140 TON; Entries 7-9, Table 4). Moreover, whereas the
catalytic activity of the Ir-containing Mb variant decreases in the presence of oxygen (Entry 10 vs.
7
, Table 4), that of the Co- and Mn-based catalysts slightly increases under aerobic conditions
(Entries 8-9 vs. 2-3, Table 4). Parallel experiments using isolated metalloporphyrins showed that
Ir(Me)(mpIX) is a viable, albeit inferior catalyst for this reaction compared to
Mb(H64,V68A)[Ir(Me)(mpIX)] (95 vs. 140 TON). In contrast, both Mn(ppIX) and Co(ppIX)
show negligible C‒H insertion reactivity under the same reaction conditions (<5 TON).
Altogether, these results highlight the key role of the myoglobin scaffold in endowing the protein-
bound Co- and Mn-porphyrins with C‒H insertion reactivity.
1
8
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
While this work was in progress, Hartwig and coworkers have reported the directed
evolution of an Ir-containing P450 (CYP119) capable of converting 13 to 14 in the presence of
EDA (55% yield, 320 TON).[8b] Complementing and extending beyond these findings, the present
studies demonstrate the functionality of Ir-substituted Mb variants in promoting an intermolecular
carbene C‒H insertion reaction. More importantly, our results demonstrate for the first time that
artificial metalloproteins incorporating earth-abundant, first-row transition metals (i.e., Co and
Mn) can catalyze this challenging transformation. This is particularly noteworthy considering that,
as in the case of the aforementioned metalloporphyrins and CYP119(Ir) variants, catalytic systems
previously developed for promoting carbene C‒H insertions have largely relied on the use of
precious metals.[1c, 16b, 18-19] It is also worth noting that intermolecular C‒H insertions in the
presence of acceptor-only diazo compounds, such as EDA, have been notoriously difficult,
primarily due to the more facile carbene dimerization side reaction competing for the desired C‒
H insertion process.[19-20] As a result, these transformations have been most effectively carried out
in the presence of donor-acceptor diazo reagents, which are less prone to dimerization.[20a, 21] In
previous studies, we noted how carbene dimerization is largely disfavored in Mb-catalyzed
carbene transfer reactions with EDA,[4a, 5-6] possibly due to steric protection of the heme-bound
carbene intermediate by the protein active site. This feature is likely to provide a beneficial
advantage toward favoring the intermolecular carbene C‒H insertion reaction documented here.
Insights into the enantioselectivity of the Mb-catalyzed C‒H insertion reactions described in Table
4
could not be gained at this point due to the difficulty of resolving the enantiomers by chiral GC
or SFC. Future studies are thus warranted to further investigate this aspect as well as the scope of
these Mb-catalyzed C‒H insertion reactions.
1
9
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Table 4. Activity of cofactor-substituted Mb variants toward C‒H functionalization of phthalan in
a
the presence of EDA. See Table S5 for complete data.
Yieldc
Entry
Catalyst
Cofactor
Y
TON
1
2
3
4
5
6
7
8
9
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A,H93F)
Mb(H64V,V68A)
Mb(H64V,V68A)
Mb(H64V,V68A,H93A)
Fe(ppIX)
Mn(ppIX)
Co(ppIX)
-/red.
-/red.
-/-
-/red.
-/red.
-/red.
-/-
0%
0
54
60
23
59
80
140
101
108
98
11%
12%
5%
12%
16%
28%
20%
22%
20%
65%
Ru(CO)(mpIX)
Rh(mpIX)
Ir(Me)(mpIX)
Ir(Me)(mpIX)
Mn(ppIX)
Co(ppIX))
Ir(Me)(mpIX)
Ir(Me)(mpIX)
O
O
O
2
2
2
/-
/-
/-
1
1
0
1
c
O
2
/-
324
a
Reaction conditions: 0.01 M phthalan, 0.04 M EDA, 20 µM Mb catalyst, 50 mM phosphate
buffer (pH 6) containing 8% ethanol, room temperature. See legend of Table 1 for definition of
b
variable parameter Y. As determined by GC using calibration curves with isolated 14. Errors are
c
within 10%. With slow addition of EDA (10 equiv).
2
0
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Axial methyl ligand in Ir(Me)-myoglobins. The presence of a metal-bound methyl group
and carbonyl group in Ir(Me)(mpIX) and Ru(CO)(mpIX), respectively, implies that these axial
ligands could occupy either the proximal or distal side of the metalloporphyrin cofactor once the
latter is embedded into the protein. Assuming these ligands do not dissociate during catalysis, this
aspect has important implications for function since the catalyst configuration with the ligand-free
distal site is expected to be sterically more accessible for activating the diazo reagent and thus
promoting the carbene transfer reaction. To examine this point, the chemical environment of the
Ir-bound methyl group in the Mb catalysts was analyzed by NMR. For the free Ir(Me)(mpIX)
cofactor, a single singlet signal corresponding to the methyl ligand is visible at -7.78 ppm (Figure
2
A). In contrast, two signals are observed in the same region of the spectrum ( -8.05 and -8.17
ppm in ~ 1:1 ratio) for Mb(H64V,V68A)[Ir(Me)(mpIX)] (Figure 2B). These results are consistent
with the presence of two different chemical environments surrounding the methyl group, as given
by this group residing on either the ‘proximal’ or the ‘distal’ face of the protein-bound cofactor.
The same experiment was carried out with Mb(H64V,V68A,H93F)[Ir(Me)(mpIX)] and
Mb(H64V,V68A,H93A)[Ir(Me)(mpIX)], in which different steric demands are imposed at the
level of the proximal face of the porphyrin cofactor by virtue of the substitution of His93 with a
slightly larger (Phe) or smaller amino acid residue (Ala). Interestingly, the relative intensity of the
1
two signals in the relevant region of the H NMR spectra differed significantly for these Mb
variants, showing an approximately 20:80 ratio in the presence of the H93F mutation and a 65:35
ratio in the presence of the H93A substitution (Figure 2B). These results thus support a direct
effect of the proximal residue on the relative abundance of the two possible catalyst configurations
2
1
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
as relative to the orientation of the methyl axial ligand. Whereas a definite assignment cannot be
made on the basis of these data alone, steric arguments suggest that the predominant species in the
H93A variant could correspond to that in which the Ir-bound methyl group is located on the
proximal face of the protein-bound cofactor.
1
Figure 2. Partial H NMR spectra corresponding to (A) free Ir(Me)-mesoporphyrin IX, and (B)
the three Ir(Me)-containing variants. This region of the spectrum shows the signal(s)
corresponding to the Ir-bound methyl group.
2
2
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
To examine the impact of the H93A mutation on catalytic function,
Mb(H64V,V68A,H93A)-[Ir(Me)(mpIX)] was characterized for its activity toward the various
carbene transfer reactions discussed earlier. Interestingly, this Mb variant was found to exhibit
similar or slightly lower TON in the cyclopropanation, N‒H insertion, and C‒H insertion reactions
when compared to Mb(H64V,V68A)[Ir(Me)(mpIX)] or Mb(H64V,V68A,H93F)[Ir(Me)(mpIX)]
(Figure 3). The orders of reactivity emerging from these experiments are His > Phe ≈ Ala for
cyclopropanation, His > Phe > Ala for N‒H insertion, and His > Ala > Phe for C‒H insertion.
Departing from this trend are the S‒H insertion reactions, in which the Ala93 containing variant
consistently exhibits higher performance than the His93 and Phe93 counterparts under the different
reaction conditions (Ala > Phe >> His; Table S4). Furthermore, the C‒H insertion reaction
catalyzed by Mb(H64V,V68A,H93A)[Ir(Me)(mpIX)] could be applied to obtain 9 in 65% yield
(Entry 11, Table 4), suggesting a higher stability of this variant as compared to the related variants
containing Phe or His at the proximal site.
Altogether, these results reveal an interesting and complex role of the proximal ligand on
modulating the carbene transfer reactivity of these metalloproteins. Although the H93A mutations
is likely to favor a catalyst configuration with an open distal side, this translates in superior
catalytic performance for only one of the four different carbene transfer reactions investigated
here. Thus, it is possible that electronic and/or steric effects imposed by bulkier proximal residues
could provide an advantage in terms of reactivity for the sterically disfavored species bearing the
methyl ligand in the proximal side. While further studies are warranted to elucidate this aspect in
more detail, it is anticipated that these structure-activity insights will aid the future design and
2
3
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
development of hemoprotein-based artificial enzymes incorporating non-natural cofactors, and
especially those bearing metal-bound ligands.
Figure 3. Relative activity of H93 variants of Ir(Me)-containing Mb catalysts across different
carbene transfer reactions. For each reaction, activity is normalized to that of the most active
variant (cycloprop.: 356 TON; N-H insert.: 168 TON; S-H insert.: 1,000 TON; C-H insert.: 140
TON). Reactions conditions are as described in Tables 1-4 with Y = anaerobic, no reductant.
2
4
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Conclusion
In summary, this work demonstrates how modification of the cofactor environment
constitutes an effective strategy to modulate and expand the reactivity of myoglobin-based carbene
transfer catalysts. In particular, our results show how the catalytic activity, stereoselectivity, and
oxygen tolerance of this metalloprotein in various carbene transfer reactions can be tuned by
varying the nature of the metal, its oxidation state, and first-sphere coordination by the proximal
histidine residues. Of note, this approach proved viable to obtain Mb catalysts capable of
promoting cyclopropanation, S‒H and N‒H insertion reactions under aerobic conditions, which is
expected to enhance the accessibility and operational simplicity of these carbene transfer
biocatalysts for synthetic applications. Importantly, it also enabled expansion of the reaction scope
of engineered myoglobins to include intermolecular carbene C‒H insertion. In this regard, the
demonstrated reactivity of the Co- and Mn-based myoglobins toward carbene C‒H insertions is
particularly noteworthy, as previously reported catalytic systems useful for this type of carbene
transfer reactions have typically relied on the use of rare and expensive transition metals. Finally,
another relevant contribution of the present work is the implementation of a simple and technically
straightforward strategy for the recombinant production of hemoprotein-based catalysts bearing
non-native metalloporphyrin cofactors, also including second- and third-row transition metal
centers, in bacterial cells. These advances are expected to facilitate the application of these
artificial enzymes in whole cell systems and pave the way to the integration of carbene-mediated
transformations into biosynthetic pathways in living organisms. Work in these directions is
currently ongoing in our laboratory.
2
5
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
Experimental Section
Reagents. All the chemicals and reagents were purchased from commercial suppliers
(Sigma-Aldrich, Alfa Aesar) and used without any further purification, unless otherwise stated.
Mn- and Co-protoporphyrin IX, protoporphyrin IX dimethyl ester, and mesoporphyrin IX dimethyl
ester were purchased from Frontier Scientific. EDA was purchased from Sigma-Aldrich as 87%
solution in dichloromethane. All dry reactions were carried out under argon or nitrogen in oven-
1
dried glassware with magnetic stirring using standard gas-tight syringes, cannulae and septa. H
1
3
1
and C NMR spectra were measured on Bruker DPX-400 (operating at 400 MHz for H and 100
MHz for 13C) or Bruker DPX-500 (operating at 500 MHz for H and 125 MHz for 13C).
1
Tetramethylsilane (TMS) (0 ppm) or D
2
O (4.6 ppm) or d
6
-DMSO (2.50 ppm) served as the internal
1
standard for H NMR and CDCl
3
(77.0 ppm) or d
6
-DMSO (39.51 ppm) was used as the internal
1
3
standard for C NMR. Silica gel chromatography purifications were carried out using AMD Silica
Gel 60 230-400 mesh. Thin Layer Chromatography (TLC) and preparative TLC were carried out
using Merck Millipore TLC silica gel 60 F254 glass plates.
Analytical Methods. Gas chromatography (GC) analyses were carried out using a
Shimadzu GC-2010 gas chromatograph equipped with a FID detector and a Chiral Cyclosil-B
column (30 m x 0.25 mm x 0.25 μm film). Separation method for cyclopropanation reaction: 1 μL
injection, injector temp.: 200 ºC, detector temp: 300 ºC. Gradient: column temperature set at 120
o
ºC for 3 min, then to 150 ºC at 0.8 ºC/min, then to 245 C at 25 ºC/min. Total run time was 46.30
2
6
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
min. Separation method for N-H insertion reaction, S-H insertion reaction and C-H insertion
reaction: 1 μL injection, injector temp.: 200 ºC, detector temp: 300 ºC. Gradient: column
temperature set at 140 ºC for 3 min, then to 160 ºC at 1.8 ºC/min, then to 165 ºC at 1 ºC/min, then
o
to 245 C at 25 ºC/min. Total run time was 28.31 min.
Growth Media. Cell cultures were grown in enriched M9 medium which was prepared as
follows. For 1 L, 770 mL deionized H
glucose (20% v/v), 10 mL casamino acids (20% m/v), 1 mL MgSO
1 M). The M9 salts (5x) solution was prepared by dissolving 15 g Na
g NaH PO , 0.3 g KH PO , 1.5 g NaCl, 5 g NH Cl in 2 L deionized H
autoclaving. The casamino acids and MgSO solutions were autoclaved separately. The CaCl
2
O was added with 200 mL M9 salts (5x) solution, 20 mL
(2 M), and 100 L CaCl
HPO , 7.5 g K HPO , 0.3
O and sterilized by
and
4
2
(
2
4
2
4
2
4
2
4
4
2
4
2
glucose stock solutions were sterilized by filtration. Enriched M9 agar plates were prepared by
adding 17 g agar to 1 L of enriched M9 media containing all of the aforementioned components at
the specified concentrations with the exception of glucose and CaCl , which were added
2
immediately prior to plating. To media and plates, ampicillin was added to a final concentration
of 100 mg/L and chloramphenicol was added to a final concentration of 34 mg/L.
Expression of Cofactor-Substituted Myoglobin Variants. The Mb variants were
expressed from pET22-based vectors[14] containing the protein gene under a T7 promoter. The
recombinant protein contains a C-terminal poly-histidine tag. The heme transporter ChuA and
GroES/EL proteins were expressed from a pACYC-based plasmid (pGroES/EL-ChuA), in which
the ChuA gene is under the control of a T7 promoter and the GroES/EL genes are under the control
of an araBAD promoter.[14] E. coli C41(DE3) cells were co-transformed with the pGroES/EL-
2
7
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
ChuA plasmid and the pET22 vector for expression of the desired Mb variant and selected on M9
-1
-1
agar plates containing ampicillin (100 mg L ) and chloramphenicol (34 mg L ). Single colonies
of the co-transformed cells were used to inoculate 5 mL overnight cultures of enriched M9 media
(
37°C, 150 rpm, 12 hours). The overnight cultures were used to inoculate 1 L of enriched M9
medium. When OD600 reached 1.4, cell cultures were condensed by centrifugation (4,000 rpm,
°C, 20 min) followed by resuspension of the cell pellet in 200 mL of enriched M9 medium. A 6
4
mg sample of solid cofactor was dissolved in 50 µL of DMSO or less. The cofactor solution was
added directly to the condensed expression culture. In addition, arabinose was added to a final
-
1
concentration of 4 mM L from a 0.4 M stock solution. The expression culture was shaken at 37°C
-1
and 150 rpm for 30 minutes before induction with IPTG (final concentration 0.5 mM L ) from a
.5 M stock solution. The induced culture was shaken at 20°C and 150 rpm for additional 16 to 20
0
hours. The cells were pelleted by centrifugation (4,000 rpm, 4°C, 20 min) and then resuspended
in 20 mL of Ni-NTA Lysis Buffer (50 mM KPi, 250 mM NaCl, 10 mM histidine, pH = 8.0).
Resuspended cells were frozen and stored at -80°C until purification.
Protein Purification. Cell suspensions were thawed at room temperature, lysed by
sonication, and clarified by centrifugation (14,000 rpm, 20 min, 4°C). The clarified lysate was
transferred to a Ni-NTA column equilibrated with Ni-NTA Lysis Buffer. The resin was washed
with 50 mL of Ni-NTA Lysis Buffer and then 50 mL of Ni-NTA Wash Buffer (50 mM KPi, 250
mM NaCl, 20 mM Histidine, pH = 8.0). Proteins were eluted with Ni-NTA Elution Buffer (50 mM
KPi, 250 mM NaCl, 250 mM Histidine, pH = 7.0). After elution from the Ni-NTA column, the
protein was buffer exchanged against 50 mM KPi buffer (pH 7.0) using 10 kDa Centricon filters.
The concentration of the Mb variants was determined using the following extinction coefficients:
2
8
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201700202
−1
−1
−1
−1
ε
410 = 157 mM cm for Fe(ppIX)-containing variants; ε470 = 60 mM cm for Mn(ppIX)-
−1
−1
containing variants; ε424 = 152.5 mM cm
for Co(ppIX)-containing variants; ε398 =
−
1
−1
−1
−1
3
95 mM cm for Ru(CO)(mpIX)-containing variants; ε405 = 122 mM cm for Rh(mpIX)-
−
1
−1
containing variants; ε400 = 117.5 mM cm for Ir(Me)(mpIX)-containing variants.
Expression and Purification of Apomyoglobin. BL21(DE3) cells were transformed with
the pET22-based vector encoding for the appropriate Mb variant and transformed cells were
-1
selected on enriched M9 agar plates containing ampicillin (100 mg L ). Single colonies were used
-1
to inoculate 5 mL of enriched M9 media supplemented with ampicillin (100 mg L ), followed by
incubation at 37°C with shaking (180 rpm) for 10 to 15 hours. The overnight cultures were
transferred to 1 L enriched M9 medium containing ampicillin, followed by incubation at 37°C with
shaking (180 rpm). At an OD600 of 1.4, cells were induced with IPTG (final conc.: 0.5 mM) and
incubated at 20°C with shaking (180 rpm) for 20 to 24 hours. Cells were harvested by
centrifugation (4,000 rpm, 4°C, 20 minutes). Cell pellets were resuspended in 20 mL of Ni NTA
Lysis Buffer (50 mM KPi, 250 mM NaCl, 10 mM histidine, pH = 8.0) and stored at -80° C until
purification. Apomyoglobin was purified using the protocol described above. After elution from
the Ni-NTA column, the protein was buffer exchanged against 50 mM KPi buffer (pH 7.0) using
1
0 kDa Centricon filters. The apomyoglobin variants were stored at 4°C for no longer than 12
hours.
Chemical synthesis. Synthetic procedures and characterization data for the
metalloporphyrins Ir(CO)Cl protoporphyrin dimethyl ester, Ir(Me)-protoporphyrin IX, Ir(CO)Cl
mesoporphyrin dimethyl ester, Ir(Me) mesoporphyrin IX, Rh mesoporphyrin IX, Ru(CO)-
2
9
This article is protected by copyright. All rights reserved.