The Thiolate Trans Effect in Heme {FeNO}6 Complexes and Beyond: Insight into the Nature of the Push Effect

Article

pubs.acs.org/IC

Cite This: Inorg. Chem. XXXX, XXX, XXX −XXX

The Thiolate Trans Eﬀect in Heme {FeNO}⁶Complexes and Beyond:

Insight into the Nature of the Push Eﬀect

Andrew P. Hunt and Nicolai Lehnert*

Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109, United States

S

* Supporting Information

ABSTRACT: Cyt P450 nitric oxide (NO) reductase (P450nor) is an

important enzyme in fungal denitriﬁcation, responsible for the large-scale

production of the greenhouse gas N₂O. In the ﬁrst step of catalysis, the

ferric heme-thiolate active site of P450nor binds NO to produce a ferric

heme-nitrosyl or {FeNO}⁶intermediate (in the Enemark−Feltham

notation). In this paper, we present the low-temperature preparation of

six new heme-thiolate {FeNO}⁶model complexes, [Fe(TPP)(SPh*)-

(NO)], using a unique series of electron-poor thiophenolates (SPh*⁻),

and their detailed spectroscopic characterization. Our data show

experimentally, for the ﬁrst time, that a direct correlation exists between

the thiolate donor strength and the Fe−NO and N−O bond strengths,

evident from the corresponding stretching frequencies. This is due to a σ-trans eﬀect of the thiolate ligand, which manifests itself

in the population of an Fe−N−O σ-antibonding (σ*) orbital. Via control of the thiolate donor strength (using hydrogen

bonds), nature is therefore able to exactly control the degree of activation of the FeNO unit in P450nor. Vice versa, NO can be

used as a sensitive probe to quantify the donor strength of a thiolate ligand in a model system or protein, by simply measuring

the Fe−NO and N−O frequencies of the ferric NO adduct and then projecting those data onto the correlation plot established

here. Finally, we are able to show that the σ-trans eﬀect of the thiolate is the electronic origin of the “push” eﬀect, which is

proposed to mediate O−O bond cleavage and Compound I formation in Cyt P450 monooxygenase catalysis.

∼75% of the atmospheric N₂O now being linked to

agriculture.⁶On the basis of these circumstances, there is

currently a great interest in determining the sources of N₂O

emissions from soils and in elucidating the mechanisms of the

bacterial and fungal enzymes that catalyze its production. In

this regard, P450nor is the key fungal enzyme responsible for

N₂O generation, and it is currently the subject of intense

scientiﬁc investigations.⁷⁻¹⁰

Analogous to other members of the cytochrome P450 (Cyt

P450) enzyme family, the central features of the P450nor

active site are the presence of an iron porphyrin, heme b

center, with a proximally bound, deprotonated cysteinate

(thiolate) ligand (Figure 1).^5,11,12The presence of the thiolate

ligand is central to catalysis for Cyt P450s and presents a

hallmark of their active sites. Additionally, these enzymes

contain conserved active site residues around the proximal

cysteinate ligand which form hydrogen bonds (H-bonds) with

the sulfur atom (see Figure 1), making up what is referred to as

the Cys pocket. These H-bonds are proposed to play

important roles in ﬁne tuning the reactivity of the heme Fe

for the desired catalytic reaction¹³and for stabilizing the

thiolate ligand to prevent undesired side reactions, for example

protonation or reaction with O₂or NO, which would lead to

enzyme deactivation.¹⁴Two particular roles for the thiolate

INTRODUCTION

■

Nitric oxide (NO) is well recognized as an important

intercellular signaling molecule that is involved in blood

pressure control and nerve signal transduction in humans and

other mammals at nanomolar concentrations.¹At increased

micromolar concentrations, however, NO is toxic to most

organisms, resulting in nitrosative stress, which can cause

detrimental health conditions such as nerve damage, organ

degradation, and cartilage disintegration, to name a few.¹NO

is also a key component of septic shock.²Owing to the

important functions and the potential threats NO poses, the

presence of diﬀerent enzymes that mediate its production,

regulation, and detoxiﬁcation is crucial in biological systems.³

One such enzyme, cytochrome P450 nitric oxide reductase

(P450nor), is utilized by soil-dwelling fungi during denitriﬁ-

cation, which is an anaerobic form of respiration where nitrate

−

(NO₃) and nitrite (NO₂) are stepwise reduced to nitrous

oxide (N₂O).⁴In these organisms, P450nor catalyzes the

reduction of NO to N₂O in order to prevent the accumulation

of this toxic metabolite,⁵according to the following overall

equation: 2 NO + 2e⁻+ 2 H⁺→ N₂O + H₂O. The product of

fungal denitriﬁcation, N₂O, is a potent greenhouse gas, which

in turn contributes to the eﬀects of global warming, with a 100

year warming potential approximately 300 times higher than

that of CO₂.⁶N₂O is also an ozone-depleting agent. As farmers

in the developed world tend to overfertilize their agricultural

soils, denitriﬁcation and N₂O production are stimulated, with

Received: January 10, 2019

A

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

having an Fe^II−NO⁺electronic structure.²¹⁻²³This electronic

description is reﬂected by the reactivity of the P450nor

{FeNO}⁶complex, where the formally NO⁺ligand is

electrophilic, which explains its ability to abstract a hydride

from NAD(P)H to ultimately generate Intermediate I.

One important question in this regard is how the thiolate

ligand aﬀects the geometric and electronic structure of the

{FeNO}⁶unit and how this relates to catalysis. Whereas

imidazole-coordinated {FeNO}⁶complexes in model systems

and proteins (especially globins have been investigated) show

linear Fe−N−O units and N−O and Fe−NO stretching

frequencies in the 1900−1920 and ∼590 cm⁻¹ranges,²⁴

respectively, interesting deviations from these properties have

been observed in the presence of axial thiolate coordination,

but the electronic−structural reasons for these deviations are

currently unknown. For example, the crystal structure of the

P450nor {FeNO}⁶complex displays a bent Fe−N−O unit

with an angle of 161°,⁵and the N−O and Fe−NO stretching

modes are observed at 1851 and 530 cm⁻¹, respectively, for

this enzyme.²⁵The {FeNO}⁶complexes of other heme-thiolate

enzymes have also been successfully generated, including Cyt

P450cam, chloroperoxidase (CPO), and nitric oxide synthase

(NOS), which show overall similar properties.²⁶⁻²⁹Computa-

tional work has further indicated that there is a direct

correlation between the thiolate donor strength and the

strengths of the Fe−NO and N−O bonds (evident from the

corresponding stretching frequencies).²¹However, other than

computational results, no experimental data are available that

directly probe the relation between the thiolate donor strength

and the properties of the Fe−N−O unit in thiolate-

coordinated {FeNO}⁶complexes. In this regard, making

comparisons between diﬀerent proteins is problematic, since

the active sites of diﬀerent proteins also diﬀer greatly in (a) the

exact environment of the cysteinate axial ligand and the heme,

(b) heme distortions, and (c) steric restrictions in the active

site, which makes it diﬃcult to derive exact trends from these

data. This is a scientiﬁc question that is ideally addressed with

model complexes, but past studies have shown that thiolate-

coordinated {FeNO}⁶complexes are extremely unstable, and

in fact, only two corresponding model complexes have been

reported, with limited characterization available.³⁰⁻³²The data

presented in this study ﬁll this void and provide, for the ﬁrst

time, experimental data for a series of model complexes,

[Fe(TPP)(SR)(NO)] (TPP = tetraphenylporphyrin dianion),

where the exact same porphyrin ligand is used and only the

donor strength of the axial thiolate ligand (SR⁻) is varied. This

is accomplished by using a series of thiophenolates (SPh*⁻)

that contain diﬀerent electron-withdrawing substituents (see

Scheme 1). The corresponding {FeNO}⁶complexes are again

very unstable but, as we demonstrate here, can be prepared at

low temperature (−80 °C) and then studied using IR and

resonance Raman spectroscopy. In this way, we are able to

establish the direct correlation between the thiolate donor

strength and the Fe−NO and N−O bond strengths, which

implies that the thiolate ligand imparts a σ-trans eﬀect

(interaction) on the Fe−N−O unit. This eﬀect is then further

analyzed using DFT calculations. On the basis of the ν(Fe−

NO)/ν(N−O) correlation that we establish here, it is now

possible to quantify the donor strength of a thiolate (in model

systems and new Cyt P450 enzymes) by simply binding NO to

the ferric form of the complex and then determining the Fe−

NO and N−O stretching frequencies. Finally, we investigated

whether a similar thiolate σ-trans eﬀect is at work in the ferric

Figure 1. Crystal structure of the Cyt P450nor ferric heme-nitrosyl

complex (PDB code: 1CL6).²The strongest hydrogen bond to the

proximally bound cysteinate (thiolate) ligand from the amide

backbone is shown as a yellow dashed line. The other two H-bonds

present in the P450nor Cys pocket have been omitted for clarity.

ligand in the Cyt P450 catalytic cycle have been identiﬁed: (a)

the so-called “push eﬀect”, where it has been proposed that the

electron donation from the thiolate accelerates O−O bond

cleavage in the hydroperoxo intermediate,¹⁵and (b) the

increase in the basicity of Compound II,¹⁶which assists in the

H atom abstraction step central to the reaction of many Cyt

P450s.

In the mechanism of P450nor, the ferric heme-thiolate

resting state binds 1 equivalent (equiv) of NO to form a ferric

heme-nitrosyl complex,¹⁷which constitutes the ﬁrst key

intermediate of the reaction. This species subsequently reacts

with NAD(P)H directly (by abstracting a hydride) to form a

ferrous HNO complex.¹⁸Either this species or the

corresponding, protonated species that contains an axial [Fe-

NHOH]³⁺unit constitutes the key “Intermediate I”, which is

then reactive toward the second molecule of NO and induces

N−N bond formation and N₂O generation. DFT calculations

predict that the initially formed HNO ligand is quite basic, due

to the presence of the strongly donating, proximal cysteinate

ligand and therefore becomes protonated,^19,7which would

identify the (formally) hydroxylamide (HNOH⁻) adduct as

Intermediate I. However, the true nature and electronic

structure of Intermediate I is currently unknown, due to

limited characterization of this short-lived species in the

enzyme.

To elucidate the mechanism and gain a better understanding

of how the P450nor active site is tailored for NO reduction, we

use synthetic model complexes to determine the geometric and

electronic properties associated with the reactive species

involved in P450nor catalysis. Speciﬁcally, in this work we

model the ﬁrst step in the P450nor mechanism in which NO

binds to the ferric heme resting state, generating a thiolate-

coordinated ferric heme-nitrosyl or {FeNO}⁶complex in the

Enemark−Feltham notation²⁰(Figure 1), where the super-

script “6” represents the sum of the electrons in the Fe(d) and

NO(π*) orbitals. Since NO is a “noninnocent” (or redox-

active) ligand, this notation is useful for keeping track of the

oxidation state of the Fe−N−O unit, without having to be

speciﬁc about the exact electron distribution in a given

complex. For example, while heme {FeNO}⁶complexes are

referred to as ferric heme-nitrosyls, spectroscopic studies have

concluded that the Fe−N−O unit is actually best described as

B

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

University). For all simulations, a D value of 7 cm⁻¹was used, as the

D values of other high-spin ferric heme-thiolate complexes have been

reported to vary roughly between 5 and 10 cm⁻¹.^34,35

Scheme 1. Series of Thiophenolate Ligands (SPh*⁻) Used

in This Work to Prepare Ferric Heme-Thiolate Complexes,

in Order of Decreasing Donor Strength (from Left to

Resonance Raman Spectroscopy. Resonance Raman (rRaman)

measurements were performed using the 458 nm excitation line from

a mixed gas Kr−Ar Coherent Innova 70C Spectrum laser or the 457

nm excitation line from a Cobolt 08-DPL diode pumped laser, on 1/1

CH₂Cl₂/toluene frozen glass samples in liquid N₂. For measurements

with the 458 nm laser, the scattered light was collected at a 135°

backscattering geometry by a 55 mm FD f/1.2 mounted Canon lens

and focused onto the 100 μm entrance slit of a spectrograph using a

Newport Corporation 50 mm, f/4 achromatic lens. Rayleigh

scattering was rejected by use of a Semrock RazorEdge long-pass

edge ﬁlter. The Raman scattered light was dispersed in an f/4.6, 0.32

m imaging spectrograph (Princeton Instruments, IsoPlane SCT 320)

equipped with an 1800 gr/mm holographic grating and imaged onto a

Peltier-cooled CCD detector (Princeton Instruments, Pixis 100B).

For data collection with the 457 nm laser, the scattered light was

focused onto an Acton two-stage TriVista 555 monochromator and

detected by a liquid-N₂-cooled Princeton Instruments Spec-10:400B/

LN CCD camera. Samples were prepared at a concentration of 2 mM,

and the laser power was adjusted to 10 mW or lower for data

collection of the heme-thiolate {FeNO}⁶samples to prevent NO

photolysis. Single-pixel spikes due to cosmic rays recorded by the

detector were removed manually from the spectra after data

collection. All spectra were calibrated to the toluene peaks at 522

and 623 cm⁻¹(see Figure S8) as an internal standard.

a

Right)

a

The order in thiolate donor strength to the heme indicated here is

largely based on experimental thiol/thiolate pK_avalues (SPh⁻, 6.62;⁶⁶

SPh-pNO₂, 4.72; SPhF₄, 2.75 ) and the results of the DFT

66

67

−

calculations shown in Table 2.

heme-hydroperoxo intermediate in Cyt P450s that would

provide an orbital-based explanation for the “push” eﬀect of

the thiolate (which assists in O−O bond cleavage), as

proposed by Dawson.¹⁵These results are summarized in this

paper.

EXPERIMENTAL SECTION

■

General Methods. The preparation, handling, and reaction of all

O₂- and H₂O-sensitive materials was carried out under inert

conditions (N₂or Ar gas) using standard Schlenk techniques or in

an N₂-atmosphere MBraun glovebox equipped with a circulating

puriﬁer (O₂, H₂O <0.1 ppm). All reagents were purchased from

commercial sources (Fisher Scientiﬁc, Sigma-Aldrich, Acros Organics,

Alfa Aesar, Cambridge Isotope Laboratories Inc.) and were used as

received, unless noted below. Nitric oxide (Cryogenic Gases Inc.,

99.5%) was puriﬁed by passage through an Ascarite II column

(NaOH on silica) followed by a cold trap at −80 °C to remove

higher-order nitrogen oxide impurities. ¹⁵N¹⁸O-labeled nitric oxide

(Sigma-Aldrich) was used without further puriﬁcation. All solvents

used here were distilled from CaH₂under N₂, degassed via ﬁve

freeze−pump−thaw cycles, and stored over appropriately sized (3 or

4 Å) activated molecular sieves in a glovebox until used. All low-

temperature experiments and sample preparations were carried out in

a glovebox cold well, externally cooled by the use of dry ice/acetone,

or liquid N₂as needed to reach the desired temperature.

[Fe(TPP)(Cl)] was synthesized as previously reported.³³

Syntheses. Synthesis of [Fe(TPP)(OCH₃)]. This complex has been

previously reported and was prepared using a modiﬁed procedure

developed here which diﬀers from that reported by Sato and co-

workers.^36,37In the glovebox, 1.05 g (1.49 mmol) of [Fe(TPP)(Cl)]

and 323 mg (5.98 mmol) of NaOCH₃were dissolved in

approximately 125 mL of CH₂Cl₂and 12 mL of anhydrous methanol.

The solution was vigorously stirred for 1 h, at which time the solution

changed color from dark brown to dark green. The reaction mixture

was then tested by UV−vis, which indicated complete conversion of

the chloro complex to [Fe(TPP)(OCH₃)]. All solvent was then

removed under reduced pressure with gentle heating on a Schlenk

line, resulting in a purple solid. Once it was returned to the glovebox,

the solid was again dissolved in CH₂Cl₂and ﬁltered through a frit by

vacuum ﬁltration to remove insoluble salts. All solvent was then

removed, resulting in a dark purple solid. The solid was then

suspended in anhydrous methanol and collected on a frit by vacuum

ﬁltration, washed with additional, fresh methanol to ensure removal of

any excess NaOCH₃followed by washing with hexanes, and ﬁnally

dried under vacuum. Yield: 988 mg (1.41 mmol, 94% yield). UV−vis

Physical Measurements. UV−Vis Spectroscopy. All UV−vis

spectra were recorded using an Analytik Jena Specord S600

spectrometer. In situ dip probe experiments were carried out with

the same spectrometer connected through a ﬁber optics cable to a

Hellma Analytics Excalibur Standard All-Quartz Immersion Probe

with a 10 mm path length (Model 661-202-10-S-46). The

temperature of the solution for low-temperature UV−vis dip probe

experiments was directly monitored by use of an Omega HH91

Digital Thermometer equipped with a PFA Coated Thermocouple

Probe.

IR Spectroscopy. All IR spectra were taken on a Bruker Alpha-E

FTIR spectrometer. Solution samples were measured in a thin-layer

solution cell equipped with CaF₂windows.

NMR Spectroscopy. Proton and ﬂuorine NMR spectra were

recorded on a Varian MR 400 MHz instrument or a Varian NMRS

500 or 700 MHz spectrometer at room temperature (typically 20−22

°C). All spectra were referenced to internal solvent peaks (e.g.,

benzene-d₆at 7.16 ppm).

EPR Spectroscopy. Electron paramagnetic resonance (EPR)

spectra were recorded on a Bruker X-band EMX spectrometer

equipped with Oxford Instruments liquid-nitrogen and liquid-helium

cryostats. EPR spectra were typically obtained on frozen-solution

samples (∼1−2 mM) using 20 mW microwave power and 100 kHz

ﬁeld modulation with the amplitude set to 1 G. EPR spectra were ﬁt

to determine the g values and degree of rhombicity (E/D) using the

program SpinCount (by Prof. M. P. Hendrich, Carnegie Mellon

1

(CH₂Cl₂): 416, 578, 630 nm. H NMR (C₆D₆): 80.78 (br, β-pyrrole

H), 10.59 (s, m-Ph), 9.82 (s, o-Ph), 6.66 (br, p-Ph). Anal. Calcd for

C₄₅H₃₁FeN₄O: C, 77.26; H, 4.47; N, 8.01. Found: C, 77.13; H, 4.31;

N, 7.91; Cl, 0.0.

General Procedure for the Synthesis of [Fe(TPP)(SPh*)]

Complexes. In the glovebox, [Fe(TPP)(OCH₃)] (typically 50−100

mg) was mixed with 1.05 equiv of the desired thiol in dry CH₂Cl₂,

resulting in an acid/base ligand exchange between the basic

methoxide ligand and the more acidic thiol, generating the desired

ferric heme-thiolate complex and methanol as the only byproduct.

Upon addition of the thiol, the dark green solution of [Fe(TPP)-

(OCH₃)] changed color to dark brown-red. The reaction was

monitored by UV−vis spectroscopy to determine when the reaction

had gone to completion, which was extremely quickly (the color

change was visually complete within the mixing time; less than 1 min

by UV−vis spectroscopy). Once the reaction was completed, all

solvent was removed via reduced pressure on a Schlenk line. The

product was then collected on a frit inside the glovebox via vacuum

ﬁltration by suspending the resulting solid in a mixture of hexanes and

methanol (3/1). The collected solid was then washed with additional

methanol, followed by hexanes, and dried under reduced pressure.

The product was tested by UV−vis and NMR spectroscopy to ensure

purity. If necessary, the product was further puriﬁed by recrystalliza-

tion by dissolving it in a minimal amount of CH₂Cl₂, layering the

C

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

solution with hexanes, and placing it in a freezer overnight. Of the

ferric heme-thiolate complexes prepared and studied here, [Fe(TPP)-

(SPhF₄)],³⁸[Fe(TPP)(SPh-pNO₂)],^39,40and [Fe(TPP)(SPh)]⁴¹⁻⁴³

have been previously reported (see references).

characterization, the {FeNO}⁶complexes were prepared in CH₂Cl₂

and then transferred into a solution IR cell using a stainless steel pipet,

both of which had been cooled in the cold well along with the sample

solution. IR spectra were then continually recorded until all signals of

the {FeNO}⁶complex had diminished.

For rRaman characterization, all samples were prepared in a 1/1

mixture of CH₂Cl₂and toluene at −80 to −90 °C, transferred to

precooled EPR tubes using a precooled stainless steel pipet, capped,

and immediately frozen in liquid N₂upon removal from the glovebox

(using a precooled copper block to keep the samples cold in the

process).

Density Functional Theory Calculations. All density functional

theory (DFT) calculations were performed with the Linux version of

the program Gaussian 09.⁴⁶All geometry optimizations and frequency

calculations were performed at the BP86/TZVP level, using Becke’s

1988 exchange functional⁴⁷and the gradient corrections of Perdew,

along with his 1981 local correlation functional P86 (BP86).⁴⁸TZVP

is a triple-ζ polarized basis set by Ahlrich and co-workers.⁴⁹Molecular

orbitals were obtained from subsequent single-point calculations using

the program ORCA at the same level of theory (BP86/TZVP).⁵⁰The

heme-thiolate enzyme mimics with two, one, and no hydrogen bonds

were generated by starting with the crystallographic coordinates of the

heme-thiolate enzyme nNOS (PDB code: 2G6K)⁵¹and keeping only

the coordinates of the Fe−NO protoporphyrin IX complex, Cys415,

Trp409, and Gly417. The Cys415 and Trp409 residues were

truncated to their side chains, an ethanethiolate and a 3-methylindole,

respectively. Gly417 was converted to an N-ethylacetamide to better

mimic the relevant form in the Cyt P450 active site: i.e., as a backbone

amide group. The coordinates of the terminal carbon atoms of the

ethanethiolate, the N-ethylacetamide, and the protoporphyrin IX

propionic acid groups and of the methyl group of 3-methylindole were

frozen to conserve their native positions. All other atoms were left

unconstrained, and the geometry of the structure was fully optimized.

The 3-methylindole and the N-ethylacetamide were subsequently

removed to generate models with no and one hydrogen bond. The

{FeNO}⁶complexes of these models were then fully optimized

preceding frequency calculations. All of these calculations were again

performed at the BP86/TZVP level.

Synthesis of [Fe(TPP)(SPhF₄CF₃)]. UV−vis: (toluene) 384, 413,

516, 578, 607, 675, 716 nm; (CH₂Cl₂) 380, 413, 516, 581, 608, 673,

1

717 nm. H NMR (C₆D₆): 71.8 (br, β-pyrrole H), 13.2 and 11.9 (d,

meta-Ph TPP), 9.0 and 5.8 (br, ortho-Ph TPP), 6.4 (s, para-Ph TPP)

ppm. ¹⁹F NMR (C₆D₆): 123.9 (br, ortho-SPhF₄CF₃), 86.2 (s, para-

SPhF₄CF₃), −199.1 (s, meta-SPhF₄CF₃). Anal. Calcd for

C₅₁H₂₈F₇FeN₄S: C, 66.75; H, 3.08; N, 6.11; S, 3.49. Found: C,

66.56; H, 2.96; N, 6.10, S, 3.48. EPR (1/1 CH₂Cl₂/toluene): g_x=

1.945, g_y= 1.96, g_z= 1.975, E/D = 0.03. Yield: 55.0 mg, 82% (starting

[Fe(TPP)(OCH₃)] mass 51.0 mg).

Synthesis of [Fe(TPP)(SPh(NO₂)₂-pCH₃)]. The thiol used for the

preparation of this complex, 2,6-dinitro-p-thiocresol (HSPh(NO₂)₂-

pCH₃), was synthesized as previously reported.⁴⁴UV−vis (toluene):

1

384, 414, 517, 583, 611, 670, 712 nm. H NMR (C₆D₆): 112.35 (s,

para-CH₃-SPh), 72.2 (br, β-pyrrole H), 68.4 (s, meta-SPh), 12.8 and

11.6 (s, meta-Ph TPP), 9.4 and 5.8 (br, ortho-Ph TPP), 6.4 (s, para-

Ph TPP) ppm. EPR (1/1 CH₂Cl₂/toluene): g_x= 1.977, g_y= 1.97, g_z=

1.98, E/D = 0.031. Yield: 54.5 mg, 92% (starting [Fe(TPP)(OCH₃)]

mass 46.9 mg).

Synthesis of [Fe(TPP)(SPhF₄)]. UV−vis (toluene): 412, 513, 575,

605, 671, 715 nm. ¹H NMR (C₆D₆): 72.4 (br, β-pyrrole H), 12.7 and

11.8 (s, meta-Ph TPP), 8.8 and 5.9 (br, ortho-Ph TPP), 6.4 (para-Ph,

TPP), −93.2 (s, para-SPhF₄) ppm. ¹⁹F NMR (C₆D₆): 140.1 (br,

ortho-SPhF₄), −197.8 (s, meta-SPhF₄) ppm. EPR (1/1 CH₂Cl₂/

toluene): g_x= 1.971, g_y= 1.971, g_z= 1.97, E/D = 0.038. Yield: 57.4

mg, 84% (starting [Fe(TPP)(OCH₃)] mass 56.4 mg).

Synthesis of [Fe(TPP)(SPh-3,5-CF₃)]. UV−vis (toluene): 408, 513,

575, 668, 710 nm. ¹H NMR (C₆D₆): 72.16 (br, β-pyrrole H), 11.9 (s,

meta-Ph TPP), 8.8 (br, ortho-Ph TPP), 6.8 (para-Ph TPP), −90.4 (s,

ortho-SPh-3,5-CF₃), −92.8 (s, para-SPh-3,5-CF₃). ¹⁹F NMR (C₆D₆):

−87.8 (s) ppm. Anal. Calcd for C₅₂H₃₁F₆FeN₄S: C, 68.35; H, 3.42; N,

6.13. Found: C, 68.33; H, 3.57; N, 6.24. EPR (1/1 CH₂Cl₂/toluene):

g_x= 1.98, g_y= 1.98, g_z= 1.97, E/D = 0.053. Yield: 26.6 mg, 30%

(starting [Fe(TPP)(OCH₃)] mass 67.7 mg). Note: the complex was

observed to be soluble in hexanes to a much higher degree in

comparison to any of the other [Fe(TPP)(SPh*)] complexes

prepared here. This resulted in a lower yield of the complex due to

a greater loss of compound upon recrystallization from CH₂Cl₂/

hexanes.

For the calculations on the [Fe(P)(SPh)(OOH₂)] complex, the

O−O bond was frozen at the optimized distance for the hydroperoxo

complex [Fe(P)(SPh)(OOH)]⁻(1.4664 Å), as optimization of the

protonated structure leads to spontaneous O−O bond cleavage,

release of H₂O, and generation of a Compound I type species (an

[Fe^IVO](Por^•+)(SPh)] complex). This is in agreement with the

mechanism of Cyt P450 monoxygenases. The O−H distances were

also ﬁxed (at 0.9857 Å) to prevent proton transfer to the pyrrole

nitrogens of the heme. All other parameters where left unconstrained

and were allowed to fully optimize.

Synthesis of [Fe(TPP)(SPh-pNO₂)]. UV−vis (toluene): 407, 514,

1

575, 670, 710 nm. H NMR (C₆D₆): 66.9 (br, β-pyrrole H), 60.2 (s,

meta-SPh-pNO₂), 12.0 (meta-Ph TPP), 8.7 (ortho-Ph TPP), 7.0

(para-Ph TPP), −92.5 (s, ortho-SPh-pNO₂). Anal. Calcd for

C₅₀H₃₂FeN₅O₂S: C, 72.99; H, 3.92; N, 8.51. Found: C, 72.42; H,

4.32; N, 8.22. EPR (1/1 CH₂Cl₂/toluene): g_x= 2.1, g_y= 2.15, g_z= 2.2,

E/D = 0.035. Yield: 56.8 mg, 78% (starting [Fe(TPP)(OCH₃)] mass

67.4 mg).

Synthesis of [Fe(TPP)(SPh-pCF₃)]. UV−vis (toluene): 406, 510,

570, 602, 670, 710 nm. ¹H NMR (C₆D₆): 62.0 (br, β-pyrrole H), 60.4

(s, meta-SPh-pCF₃), 11.9 (s, meta-Ph TPP), 9.8 (br, ortho-Ph TPP)

7.6 (s, para-Ph TPP), −94.0(s, ortho-SPh-pCF₃) ppm. ¹⁹F NMR

(C₆D₆): 94.8 (s) ppm. EPR (1/1 CH₂Cl₂/toluene): g_x= 1.976, g_y=

1.976, g_z= 1.98, E/D = 0. Yield: 51.9 mg, 64% (starting

[Fe(TPP)(OCH₃)] mass 67.2 mg).

RESULTS AND ANALYSIS

■

Hydrogen Bonding versus Electron-Withdrawing

Substituents: Finding the Right Thiolate Ligands to

Model P450nor. Previous DFT calculations performed by

our group predict that the donor strength of the thiolate ligand,

as modulated by its hydrogen-bonding network (in Cyt P450

enzymes), is directly reﬂected in the properties of the Fe−N−

O unit in {FeNO}⁶complexes.²¹In particular, a direct

correlation between the thiolate donor strength and the Fe−

NO and N−O bond strengths, reﬂected by the corresponding

ν(N−O) and ν(Fe−NO) stretching frequencies, is predicted,

but an experimental validation of this claim is missing so far.

To better understand Nature’s use of hydrogen bonds to

control the thiolate donor strength in heme enzymes and to

gauge the appropriate type of synthetic thiolate ligand to use

that models the {FeNO}⁶complex of P450nor, we ﬁrst

performed DFT calculations. In particular, since our goal is to

use thiophenolate derivatives with electron-withdrawing

Synthesis of [Fe(TPP)(SPh)]. This complex was prepared as

previously described.⁴⁵UV−vis (toluene): 408, 514, 567, 613, 700

1

nm. H NMR (C₆D₆): 81.4 (br, β-pyrrole H), 63.0 (s, meta-SPh),

11.9 (s, meta-Ph TPP), 8.9 (br, ortho-Ph TPP) 6.6 (s, para-Ph TPP),

−94.9 (s, para-SPh) ppm. Yield: 49.0 mg, 56%.

Preparation of the Heme-Thiolate {FeNO}⁶Complexes. The

temperature-sensitive heme-thiolate {FeNO}⁶samples were prepared

for characterization by UV−vis, IR, and rRaman spectroscopy as

follows. Approximately 1 equiv of NO(g) was syringed into the

headspace of a septum-sealed vial containing a −80 to −90 °C

solution of the desired ﬁve-coordinate heme-thiolate precursor, mixed,

and allowed to react at this temperature for at least 10 min. For IR

D

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 2. Structures of the DFT enzyme models 0H-B, 1H-B, and 2H-B (from left to right). All three structures were fully optimized using BP86/

TZVP. Additional details about these models are provided in the Experimental Section. All non-polar hydrogen atoms have been omitted for clarity.

Table 1. Comparison of Key Parameters of the 0H-B, 1H-B, and 2H-B DFT Models (BP86/TZVP) to Those of the Cyt

P450nor {FeNO}⁶Complex from Crystallographic and Spectroscopic Data

source

complex

ν(NO) (cm⁻¹

)

Fe−N−O angle (deg)

Fe−N(O) (Å)

N−O (Å)

Fe−S (Å)

DFT

enzyme

DFT

0H-B {FeNO}⁶

1826

1837

1856

1851

1855

166

167

169

161

169

1.687

1.680

1.672

1.63

1.163

1.161

1.159

1.16

2.315

2.328

2.365

2.3

1H-B {FeNO}⁶

2H-B {FeNO}⁶

P450nor {FeNO}⁶

[Fe(TPP)(SPhF₄CF₃)(NO)]

1.666

1.159

2.374

groups to model the axial thiolate ligand of P450nor, it is a

priori not clear what substitution pattern would result in the

best agreement with experiment.

stronger Fe−S bond induces weaker Fe−NO and N−O bonds.

This trend is evident from Table 1 and has also previously

been predicted by DFT for synthetic model systems.²¹The

Fe−N−O angle is additionally predicted to become more

linear with weaker thiolate donors, going from 166° to 169° for

the 0H-B and 2H-B models. A comparison of the calculated

ν(N−O) values for these models to the experimentally

determined N−O stretching frequency for the {FeNO}⁶

adduct of P450nor shows excellent agreement for the 2H-B

model, where the calculated ν(N−O) value of 1856 cm⁻¹is

close to the experimental value of 1851 cm⁻¹for P450nor. In

summary, the DFT calculations predict that a weaker thiolate

donor actually better represents the P450nor {FeNO}⁶

complex.

In order to accomplish this goal, we used DFT calculations

(with BP86/TZVP) on a set of simple Cyt P450 active site

models with a diﬀerent number of H-bonds, as shown in

Figure 2, and then optimized the structures of the

corresponding {FeNO}⁶adducts and calculated their vibra-

tional frequencies. The simplest model (0H-B; Figure 2, left)

just contains the protoporphyin IX and ethanethiolate as

proximal ligand, but no hydrogen bond donors, and provides a

baseline with a very strong thiolate donor. In the next step, we

added one hydrogen bond, originating from an amide group,

similar to those that are provided by the peptide chain in the

Cys pocket (1H-B; Figure 2, middle). Note that backbone

amides are the most common hydrogen-bond donors to the

cysteinate ligand in Cyt P450s and other heme-thiolate

enzymes. The ﬁnal model contains the amide hydrogen bond

(in the same relative position), plus an additional hydrogen

bond from a tryptophan side chain (2H-B; Figure 2, right),

since this amino acid has been reported to be a very strong

hydrogen bond donor (inspired by nitric oxide synthase, where

this combination of an amide and Trp hydrogen bond to the

axial cysteinate ligand is observed).²⁸The most important

results from these calculations are summarized in Table 1.

By analysis of the DFT results of the heme-thiolate

{FeNO}⁶models with zero, one, and two H-bonds, clear

trends can be derived from the calculations. On comparison of

the Fe−S bond lengths, it can be seen that 0H-B has the

shortest predicted bond length of 2.315 Å, i.e. the strongest

Fe−S bond, and that the Fe−S bond decreases in strength with

the addition of hydrogen bonds, resulting in Fe−S distances of

2.328 Å for 1H-B and 2.365 Å for 2H-B. This increase in the

Fe−S bond distance along this series has signiﬁcant eﬀects on

the calculated Fe−NO and N−O bond distances, which are

both predicted to decrease as the number of hydrogen bonds

increases. Accordingly, the ν(N−O) and ν(Fe−NO) frequen-

cies increase along the series. The DFT calculations therefore

predict a direct correlation of the Fe−NO and N−O bond

strengths as a function of the thiolate donor strength, where a

In terms of generating {FeNO}⁶model complexes with

electron-poor thiophenolate ligands, we then used DFT

calculations to predict which thiolate ligand would lead to an

{FeNO}⁶complex with predicted properties that most closely

match 2H-B. Our calculations show excellent agreement

between 2H-B and the complex with SR⁻= SPhF₄CF₃(see

−

Scheme 1), as shown in Table 1, which made this complex the

ﬁrst target for our experimental work. The optimized structure

of this complex is shown in Figure 3. The calculated properties

Figure 3. Image of the fully optimized DFT structure of [Fe(TPP)-

(SPhF₄CF₃)(NO)] (BP86/TZVP). Hydrogen atoms of the porphyrin

have been omitted for clarity.

E

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Table 2. Comparison of Key DFT-Calculated Properties (BP86/TZVP) of the Ferric Heme-Nitrosyl Complexes with the

a

Thiophenolate Ligands (SPh*⁻) Shown in Scheme 1

thiolate

ν(NO) (cm⁻¹

)

ν(Fe−NO) (cm⁻¹

)

Fe−N−O angle (deg)

Fe−N(O) (Å)

N−O (Å)

Fe−S (Å)

SPh⁻

SPh-pCF₃

SPh-pNO₂

SPh-3,5-CF₃

SPhF₄

1825

1841

1848

1850

1860

1863

581

590

589

590

592

598

600

163

166

167

168

170

173

1.687

1.680

1.676(6)

1.676(8)

1.675

1.163

1.161

2.342

2.346

2.353

2.350

2.360

2.370

2.386

−

1.159(7)

1.159(6)

1.159(7)

1.158(3)

1.158(5)

−

SPhF₄CF₃

SPh(NO₂)₂-pCH₃

1.670

1.665

−

a

The porphine dianion (P²⁻) was used for all calculations shown here.

for this complex closely match a number of the key parameters

of the 2H-B model, including the Fe−N−O angle (169°), the

N−O bond distance (1.159 Å), and the very similar calculated

ν(N−O) (1856 vs 1855 cm⁻¹). The advantage of using

thiophenolates as the basis to develop a series of {FeNO}⁶

complexes with thiolates of diﬀerent donor strength is the fact

that many thiophenols with a broad spectrum of electron-

donating and -withdrawing substituents are commercially

available. It is surprising though that the very “electron poor”

−

thiophenolate SPhF₄CF₃is predicted to give a {FeNO}⁶

complex that closely matches the electronic properties of the

NO adduct of ferric P450nor. Table 2 gives the DFT-

calculated geometric and vibrational properties for the

complete series of thiolate-coordinated model complexes

investigated here, using the thiophenolate ligands shown in

Scheme 1.

Having these predictions from DFT calculations in hand, we

then set oﬀ to test our hypothesis by ﬁrst synthesizing the ﬁve-

coordinate heme-thiolate complex [Fe(TPP)(SPh₄CF₃)] and

then reacting this compound with NO(g) at −80 °C to form

the corresponding {FeNO}⁶complex.

Figure 4. UV−vis spectra of [Fe(TPP)(OCH₃)] (black) and

[Fe(TPP)(SPhF₄CF₃)] (red), in CH₂Cl₂at room temperature.

observed by UV−vis spectroscopy with λ_maxat 413 and 384

nm. In the Q-band region, a unique pattern with bands at 515,

581, 608, 675, and 716 nm is observed, which is characteristic

for high-spin ferric [Fe(TPP)(SR)] type complexes. Any

excess thiol left after the reaction using this method can then

be easily removed by evaporation of the solvent under reduced

pressure, followed by collecting solid heme-thiolate complex

on a frit by vacuum ﬁltration and washing the solid with

methanol. After drying of the obtained complex [Fe(TPP)-

(SPhF₄CF₃)] under vacuum to remove residual solvent, the

Synthesis of Five-Coordinate Ferric Heme-Thiolate

Model Complexes. To prepare ﬁve-coordinate ferric heme-

thiolate complexes of the type [Fe(TPP)(SPh*)] (TPP²⁻

=

tetraphenylporphyrin dianion; SPh*⁻= thiophenolate deriva-

tive) directly from the free thiols in high purity, we have

further developed a synthetic method that starts from the ferric

methoxide complex, where the methoxide ligand acts as an

internal, strong base. For example, the synthesis of [Fe(TPP)-

(SPh₄CF₃)] was accomplished by reacting [Fe(TPP)(OCH₃)]

with a slight excess of 4-triﬂuoromethyl-2,3,5,6-tetraﬂuorothio-

phenol (HSPhF₄CF₃), in which an acid−base ligand exchange

occurs that generates the desired ﬁve-coordinate heme-thiolate

complex and methanol as the products, according to the

equation:

1

complex was then further characterized by H and ¹⁹F NMR

spectroscopy to ensure purity (see Figures S1 and S2).

1

Characteristic paramagnetic shifts in the H NMR spectra of

[Fe(TPP)(SPhF₄CF₃)] are observed, which demonstrate that

this is a high-spin (S = 5/2) ferric complex. In particular, the β-

pyrrole protons of the heme are shifted to 71.8 ppm. The ¹⁹F

NMR spectra exhibit three paramagnetically shifted signals for

the three unique ﬂuorine environments of the bound

[Fe(TPP)(OCH₃)] + HSPhF CF

4

3

−

SPhF₄CF₃ligand (Figure S2). All other ﬁve-coordinate ferric

→ [Fe(TPP)(SPh₄CF )] + CH₃OH

3

heme-thiolate complexes were prepared in an analogous way.

By comparison, the paramagnetic signals not belonging to the

high-spin ferric [Fe(TPP)]⁺unit can be identiﬁed, and in this

This type of acid/base thiol/methoxide ligand exchange

reaction has previously been used by Tonzetich and co-

workers for the synthesis of pure ferric heme-silanethiolate

complexes⁵²and for the preparation of [Ru(TTP)(NO)(SR)]

by reaction of [Ru(TTP)(NO)(OMe)] with the correspond-

ing thiol.⁵³This reaction is extremely fast, and a characteristic

color change from dark green to dark red-brown is observed

immediately after the thiol is added to a solution of the ferric

TPP methoxide complex. This is further documented in Figure

4: here, the methoxide complex shows the Soret band at 416

nm, and after addition of HSPhF₄CF₃, a split Soret band is

1

way, the paramagnetically shifted H and ¹⁹F signals of the

bound thiophenolate ligands can be assigned in a systematic

1

way (see Figure 5). Assignments for all H chemical shifts are

summarized in Table 3.

Additionally, all of the precursor complexes were charac-

terized by EPR spectroscopy. The EPR spectrum of [Fe-

(TPP)(SPhF₄CF₃)] (Figure S3) displays a rhombic S = 5/2

signal with eﬀective g values of 6.6, 5.1, and 1.95, again

conﬁrming that the complex is a ﬁve-coordinate high-spin

F

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 5. (A) Overlay of ¹H NMR spectra of selected high-spin ferric [Fe(TPP)(SPh*)] complexes, indicating the assignments of the ortho (o-),

meta (m-), and para (p-) phenyl protons of the bound thiophenolate ligands. (B) Overlay of the ¹⁹F NMR spectra of [Fe(TPP)(SPhF₄CF₃)],

[Fe(TPP)(SPhF₄)], and [Fe(TPP)(SPh-pCF₃)], showing the signals of the bound thiophenolate ligands.

1

Table 3. Assignments of the H NMR Chemical Shifts of the Seven Ferric Heme-Thiolate Complexes Shown in Scheme 1,

Recorded in C₆D₆at Room Temperature

complex

β-pyrrole H

m-Ph, TPP

o-Ph, TPP

p-Ph, TPP

m-Ph, SPh

68.4

o-Ph, SPh

p-Ph, SPh

[Fe(TPP)(SPhF₄CF₃)]

[Fe(TPP)(SPh(NO₂)₂-pCH₃)]

[Fe(TPP)(SPhF₄)]

[Fe(TPP)(SPh-3,5-CF₃₎]

[Fe(TPP)(SPh-pNO₂)]

[Fe(TPP)(SPh-pCF₃)]

[Fe(TPP)(SPh)]

71.8

72.2

72.4

72.2

66.9

81.3

81.4

13.2, 11.9

12.8, 11.6

12.7, 11.8

11.9

12.0

11.9

9.0, 5.8

9.4, 5.8

8.8, 5.9

8.8

8.7

9.8

6.4

6.8

7.0

7.6

6.6

−93.2

−92.8

−90.4

−92.5

−94.0

N/D

60.2

60.4

63.0

11.9

8.9

−94.9

Scheme 2. Potential Reaction Pathways of Ferric Heme-Thiolate Model Complexes with NO, Which Ultimately Generate

Ferrous Heme-Nitrosyl ({FeNO}⁷) Decomposition Products

heme-thiolate complex. A simulation of the EPR data yields an

E/D value of 0.03, as shown in Figure S3.

was prepared by an alternative method as described

previously.⁴⁵

As mentioned above, all other [Fe(TPP)(SPh*)] complexes

with the thiolate ligands shown in Scheme 1 were synthesized

using the same method. Figures S10 −S30 in the Supporting

Information show the basic spectroscopic characterization of

all of these complexes, using UV−vis absorption, NMR, EPR,

and vibrational spectroscopy. It is noted that this thiol/

methoxide ligand exchange method works best when more

acidic thiols are used, such as those used here. On the other

hand, the reaction of [Fe(TPP)(OCH₃)] with the more basic

thiol, p-thiocresol (HSPh-pCH₃), leads to complete reduction

to the ferrous [Fe(TPP)] complex, as is evident from UV−vis

spectroscopy (data not shown). Reaction of [Fe(TPP)-

(OCH₃)] with HSPh leads to formation of a mixture of

[Fe(TPP)(SPh)] and [Fe(TPP)], as previously observed by

Tonzetich and co-workers.⁵²For this reason, [Fe(TPP)(SPh)]

Reaction of Ferric Heme-Thiolate Model Complexes

with Nitric Oxide. Whereas {FeNO}⁶complexes with

thiolate coordination are relatively stable in protein active

site environments, the generation of the corresponding model

complexes has remained highly elusive. One challenge in

generating ferric heme-nitrosyl complexes in general is their

decomposition into the corresponding ﬁve-coordinate, ferrous

heme-nitrosyl complexes [Fe(Porph)(NO)] (Porph²⁻

=

general porphyrin dianion), or {FeNO}⁷species in the

Enemark-Feltham notation, in the presence of excess NO

and various nucleophiles, including alcohols, water, amines,

and thiols. This is due to a process known as “reductive

nitrosylation”.⁵⁴⁻⁵⁶Simple [Fe(Porph)(SR)(NO)] model

complexes suﬀer from additional instability in comparison to

{FeNO}⁶complexes coordinated by N-donor ligands, due to

G

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

the instability of the thiolate ligand in the presence of NO.

Here, NO can directly nitrosylate the coordinated thiolate

ligand, leading to decomposition. Finally, homolytic cleavage

of the Fe−S bond after NO coordination is also possible,

depending on the nature of the thiolate.⁴⁴Ultimately, both of

these latter pathways again lead to generation of the

corresponding ferrous [Fe(Porph)(NO)] complex. These

potential decomposition pathways are summarized in Scheme

2.

the disappearance of the Soret band at 413 nm of the precursor

and the appearance of a new, intense Soret feature at 440 nm

and two new Q bands at 553 and 591 nm. Additionally,

isosbestic points are observed at ∼345, 425, 465, 537, and 645

nm, indicating that a simple one-step reaction between NO

and [Fe(TPP)(SPhF₄CF₃)] takes place. This new species is

stable at −80 °C, and no signs of decomposition into the

{FeNO}⁷complex are observed for ∼2 h when the solution is

kept at this temperature (we did not monitor the solution past

2 h, and so it is not clear at what point decomposition would

occur). Upon removal of the solution from the cold well and

warming it to room temperature, signals for the {FeNO}⁶

complex start to slightly decrease when the solution temper-

ature reaches approximately −60 °C. Further warming then

leads to signiﬁcant decomposition to the ﬁve-coordinate

{FeNO}⁷complex [Fe(TPP)(NO)] (Figure S4). By −10 °C

the solution has almost fully converted to the {FeNO}⁷

complex.

In summary, reaction of [Fe(TPP)(SPhF₄CF₃)] with

NO(g) at −80 °C results in a new stable species, until the

solution is warmed, suggesting that this new species is the

corresponding ferric heme-nitrosyl complex [Fe(TPP)-

(SPhF₄CF₃)(NO)]. In support of this idea, the UV−vis

spectrum observed for this complex matches quite well with

those reported for heme-thiolate {FeNO}⁶adducts in enzymes.

For example, the ferric-nitrosyl adduct of nNOS is reported to

have a Soret band at 440 nm and two Q bands at 549 and 580

nm, with the same intensity ordering as observed in our new

species (where these features are located at 553 and 591

nm).⁶⁰Next, to conﬁrm the exact nature of the reaction

product between [Fe(TPP)(SPhF₄CF₃)] and NO at −80 °C,

we performed solution IR and resonance Raman (rRaman)

spectroscopy to measure the ν(N−O) and ν(Fe−NO)

vibrations, respectively.

Solution IR characterization of the new species observed in

the UV−vis experiments was conducted by addition of ∼1

equiv of NO(g) to a −80 to −90 °C solution of

[Fe(TPP)(SPhF₄CF₃)] in CH₂Cl₂, stirring of the solution

for 10 min, and then quickly transferring the reaction mixture

to a precooled solution IR cell, followed by immediate

collection of the IR spectra as the solution warmed up. Upon

doing so, a large signal at 1867 cm⁻¹was observed in the ﬁrst

scan (Figure 7). Additional spectra of the solution were

continually recorded as the solution gradually warmed up

inside the spectrometer, until the signal at 1867 cm⁻¹had

completely disappeared. As this signal decreased in intensity, a

new signal was observed at 1675 cm⁻¹that proportionally

increased in intensity. Repeating this experiment with

isotopically labeled ¹⁵N¹⁸O(g) shifted the observed band at

1867 to 1792 cm⁻¹. Similar to the experiment with natural

abundance isotopes (n.a.i.) NO, the signal at 1792 cm⁻¹

decreased in intensity as the solution warmed up, with the

appearance of a new band at 1607 cm⁻¹(Figure 8). The

isotope sensitivity of the band at 1867 cm⁻¹conﬁrms the

assignment of this feature to the N−O stretch of the new six-

coordinate {FeNO}⁶heme-thiolate complex [Fe(TPP)-

(SPhF₄CF₃)(NO)], observed in the UV−vis experiments at

−80 °C. The appearance of the 1675 cm⁻¹feature (1607 cm⁻¹

with ¹⁵N¹⁸O) further conﬁrms that [Fe(TPP)(NO)] is formed

as the decomposition product of [Fe(TPP)(SPhF₄CF₃)(NO)],

as the reaction mixture warms up. An overlay of the ﬁrst scan

of the IR experiment with n.a.i. NO and ¹⁵N¹⁸O is shown in

Figure S5.

Correspondingly, whereas many (>8) {FeNO}⁶complexes

with bound axial imidazole or other neutral N-heterocyclic

ligands have been generated and structurally character-

ized,^{24,36,57−59}there is only one reported crystal structure of

a thiolate-coordinated {FeNO}⁶complex, [Fe(OEP)(SR-

−

H₂)(NO)] (SR-H₂= S-2,6-(CF₃CONH)₂C₆H₃; OEP²⁻

=

octaethylporphyrin dianion), available to this date.³⁰Here, the

thiolate ligand utilizes two strong amide hydrogen bonds for

stabilization, and this compound is in fact able to reversibly

bind NO at low temperatures.³¹Still, the crystal structure of

this complex could only be obtained by diﬀusion of NO gas

into the crystal lattice of the ﬁve-coordinate heme-thiolate

precursor, again emphasizing the stability challenges for

generating [Fe(Porph)(SR)(NO)] model complexes.

As indicated above, previous work in our group has found

that reaction of ferric Fe(TPP) complexes with bound

thiophenolates, such as [Fe(TPP)(SPh)], with NO at room

temperature or at −40 °C results in direct conversion to the

ferrous heme-nitrosyl complex [Fe(TPP)(NO)].⁴⁴Starting

from these results, we speculated that application of electron-

poor thiophenolates in combination with improved exper-

imental conditions (especially conducting experiments at −80

°C) might allow us to ﬁnally observe and characterize the

corresponding {FeNO}⁶complexes. Since the DFT results

(see above) indicated that [Fe(TPP)(SPhF₄CF₃)(NO)] is a

good electronic−structural model for the {FeNO}⁶adduct of

P450nor, we ﬁrst investigated the reaction of the correspond-

ing ferric precursor with NO. Figure 6 shows the spectral

changes observed by UV−vis spectroscopy when [Fe(TPP)-

(SPhF₄CF₃)] is reacted with NO gas at −80 °C, monitored in

situ using a Hellma immersion probe. Addition of NO causes

Figure 6. UV−vis spectra monitoring the reaction between

[Fe(TPP)(SPhF₄CF₃)] (black) and NO, upon addition of NO(g)

to a solution of the complex at −80 °C in toluene. Insert: expanded

view of the Q-band region between 450 and 800 nm. Arrows indicate

the intensity changes of the absorption features over time.

H

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 7. Solution IR spectra of [Fe(TPP)(SPhF₄CF₃)(NO)]

recorded in CH₂Cl₂. Scan 1 (black) is the ﬁrst IR spectrum recorded

upon injection of the solution of [Fe(TPP)(SPhF₄CF₃)(NO)] into

the solution IR cell at −80 °C. Additional scans are shown in gray as

the solution warmed up to room temperature. By scan 5 (green), the

solution has been fully converted to the ﬁve-coordinate ferrous heme-

nitrosyl complex [Fe(TPP)(NO)].

Figure 9. Overlay of the low-energy region of the resonance Raman

spectra (458 nm excitation) of [Fe(TPP)(SPhF₄CF₃)(NO)] and

[Fe(TPP)(SPhF₄CF₃)(¹⁵N¹⁸O)], each one prepared in a 1/1

CH₂Cl₂/toluene mixture. Data were collected at 77 K. The solvent

bands of toluene are marked with S (see Figure S8 for the full

spectrum of the solvent mixture).

of a heme-thiolate {FeNO}⁶complex with TPP²⁻as the

porphyrin coligand to date. To further vary the electronic

nature of the thiolate ligand in heme-thiolate {FeNO}⁶model

complexes, we then applied the same method used to generate

[Fe(TPP)(SPFh₄CF₃)(NO)] to the other ﬁve-coordinate

ferric heme-thiolate precursors with the thiolate ligands

shown in Scheme 1. First, the additional thiolate complexes

were synthesized by reaction of the corresponding thiols

2,3,5,6-tetraﬂuorobenzenethiol (HSPhF₄), 4-(triﬂuoromethyl)-

thiophenol (HSPh-pCF₃), 4-nitrothiophenol (HSPh-pNO₂),

2,6-dinitrothiocresol (HSPh(NO₂)₂-pCH₃), and 3,5-bis-

(triﬂuoromethyl)thiophenol (HSPh-3,5-CF₃) with [Fe(TPP)-

(OCH₃)] as described above. After successful synthesis and

characterization (Figures S10 −S28), the ﬁve-coordinate heme-

thiolate complexes were then individually reacted with NO(g)

at −80 °C while the reaction was monitored with in situ UV−

vis spectroscopy. For [Fe(TPP)(SPh(NO₂)₂-pCH₃)] and

[Fe(TPP)(SPhF₄)], stable new species were observed, similar

to the results for [Fe(TPP)(SPhF₄CF₃)], with intense Soret

bands close to 440 nm and two new Q-bands in the 550 and

590 nm region (Figures S31 and S34). Warming the solutions

to room temperature resulted in decomposition of these new

species, generating [Fe(TPP)(NO)]. The {FeNO}⁶adducts

were then further characterized by IR and rRaman spectros-

copy to determine their N−O and Fe−NO stretching

frequencies. These were observed at 1863 and 547 cm⁻¹for

[Fe(TPP)(SPh(NO₂)₂-pCH₃)(NO)] and 1860 and 538 cm⁻¹

for [Fe(TPP)(SPhF₄)(NO)] (Figures S32 and S33 and

Figures S35 and S36, respectively).

Figure 8. Solution IR spectra of [Fe(TPP)(SPhF₄CF₃)(¹⁵N¹⁸O)]

recorded in CH₂Cl₂. Scan 1 (black) is the ﬁrst IR spectrum recorded

upon injection of the solution of [Fe(TPP)(SPhF₄CF₃)(¹⁵N¹⁸O)]

into the solution IR cell at −80 °C. Additional scans are shown in gray

as the solution warmed up to room temperature. By scan 13 (purple),

the solution has been fully converted to the ﬁve-coordinate ferrous

heme-nitrosyl complex [Fe(TPP)(¹⁵N¹⁸O)].

Next, [Fe(TPP)(SPhF₄CF₃)(NO)] was characterized by

rRaman spectroscopy at 77 K to determine the Fe−NO stretch

of this complex, using 458 nm laser excitation. This laser

wavelength was chosen due to its proximity in energy to the

Soret band of the complex at 440 nm. An overlay of the spectra

in the low-energy region obtained for [Fe(TPP)(SPhF₄CF₃)-

(NO)] and the corresponding ¹⁵N¹⁸O complex is shown in

Figure 9. Here, an isotope-sensitive band is observed at 543

cm⁻¹that shifts to 530 cm⁻¹in the ¹⁵N¹⁸O complex (diﬀerence

13 cm⁻¹), allowing us to directly assign this feature to the Fe−

NO stretch of [Fe(TPP)(SPhF₄CF₃)(NO)]. The complete

rRaman spectra of the n.a.i. and isotopically labeled NO

complexes, of the 1/1 CH₂Cl₂/toluene solvent mixture, and of

the ﬁve-coordinate ferric heme-thiolate precursor are provided

in Figures S7 −S9 in the Supporting Information.

Encouraged by these positive results with the very electron

−

poor thiolate ligands SPFh₄CF₃, SPhF₄, and SPh(NO₂)₂-

pCH₃, we then tested whether the {FeNO}⁶adduct could

−

also be obtained for the complex [Fe(TPP)(SPh-pCF₃)],

which contains a distinctively more electron rich (and hence

more donating) thiolate. Upon reaction of [Fe(TPP)(SPh-

pCF₃)] with low equivalents of NO(g) at −80 °C, a new

species is observed to form with UV−vis spectral features

similar to those of the other three {FeNO}⁶complexes, with a

new Soret band at 440 nm and two new prominent Q-bands at

556 and 597 nm (Figure 10A). After addition of a total of 1

equiv of NO(g), this new species, [Fe(TPP)(SPh-pCF₃)-

Synthesis and Testing of Additional Ferric Heme-

Thiolate Complexes. The successful generation of [Fe-

(TPP)(SPFh₄CF₃)(NO)] represents the ﬁrst characterization

I

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 10. UV−vis spectra monitoring the reaction between [Fe(TPP)(SPh-pCF₃)] (black) and NO(g) at −80 °C in toluene: (A) initial spectral

changes observed for the reaction with ∼1 equiv of NO(g) added to the solution; (B) changes observed upon addition of another ∼0.25 equiv of

NO.

(NO)], is observed to form in approximately 40−50% yield

(by comparison to the intensity of the Soret band of

[Fe(TPP)(SPhF₄CF₃)(NO)]; see Figure 6). After no further

changes were observed in the spectra, an additional ∼0.25

equiv of NO(g) was added to the solution, resulting in a brief,

initial increase in the Soret band intensity at 440 nm, followed

by a subsequent intensity drop and formation of a new band at

∼409 nm, representing the generation of [Fe(TPP)(NO)]

(Figure 10B). On the basis of these results, [Fe(TPP)(SPh-

pCF₃)] represents the borderline in terms of thiolate donor

strength that still allows for the formation of the {FeNO}⁶

adduct at −80 °C. This incomplete conversion to [Fe(TPP)-

(SPh-pCF₃)(NO)] in the UV−vis experiments can be

rationalized to occur via a direct attack of NO at the bound

thiolate ligand in some of the [Fe(TPP)(SPh-pCF₃)]

precursor, generating the nitrosothiol and [Fe(TPP)], which

subsequently binds another NO molecule. Hence, 2 equiv of

NO are consumed in this process without generating any

[Fe(TPP)(SPh-pCF₃)(NO)], lowering the amount of NO gas

in solution. Upon the addition of more NO, the remaining

[Fe(TPP)(SPh-pCF₃)] can be transformed into the corre-

sponding {FeNO}⁶complex. Once all of the precursor has

been consumed, excess NO in solution might then be able to

attack the thiolate ligand in the product complex, [Fe(TPP)-

(SPh-pCF₃)(NO)] (top right in Scheme 2). At this point,

more [Fe(TPP)(NO)] is generated and becomes visible in the

UV−vis spectra. Regardless of this, since a signiﬁcant amount

of [Fe(TPP)(SPh-pCF₃)(NO)] is still formed at −80 °C

(especially with substoichiometric amounts of NO), we were

able to characterize this compound by solution IR and rRaman

spectroscopy, showing its N−O and Fe−NO stretches at 1847

and 532 cm⁻¹, respectively (Figures S37 −S39). On the basis of

these experimental results, it can be concluded that [Fe(TPP)-

(SPh-pCF₃)(NO)] actually shows the best agreement with the

ν(N−O) and ν(Fe−NO) vibrational energies of the P450nor

{FeNO}⁶adduct (1851 and 530 cm⁻¹), and hence, this

complex is a very good electronic−structural model for the NO

adduct of P450nor. Note that this is at odds with the DFT

results (see above), which predict that [Fe(TPP)(SPhF₄CF₃)-

(NO)] should be a better electronic model for the enzyme’s

ferric NO complex. On the basis of the experimental results,

[Fe(TPP)(SPhF₄CF₃)(NO)] is actually a better electronic−

structural model for the NO adducts of the heme-thiolate

enzymes CPO and NOS (see Table 4).

Table 4. Comparison of ν(N−O) and ν(Fe−NO) Stretching

Frequencies of Synthetic and Enzymatic Heme-Thiolate

a

{FeNO}⁶Complexes

{FeNO}⁶complex

ν(N−O) ν(Fe−NO)

ref

[Fe(TPP)(SPhF₄CF₃)(NO)]

[Fe(TPP)(SPh(NO₂)₂-pCH₃)(NO)]

[Fe(TPP)(SPhF₄)(NO)]

[Fe(TPP)(SPh-3,5-CF₃)(NO)]

[Fe(TPP)(SPh-pNO₂)(NO)]

[Fe(TPP)(SPh-pCF₃)(NO)]

iNOSoxy (+H4B)

iNOSoxy

CPO

P450nor (proto)

P450nor (meso)

1867

1863

1860

1855

1852

1847

1872

1870

1868

1853

1852

1851

1806

543

547

538

537

535

532

541

537

538

530

532

533

530

528

TW

29, 68

27

69

25

P450nor (deutero)

P450nor (WT)

Cyt P450cam

25, 26

a

TW = this work.

Finally, the complexes [Fe(TPP)(SPh-pNO₂)(NO)] and

[Fe(TPP)(SPh-3,5-CF₃)(NO)] with two additional thiolates

of intermittent donor strength were also prepared and

spectroscopically characterized (Figures S40 −S45). In the

case of the p-nitro-substituted thiophenolate ligand, the

reaction with NO would again not go to completion, but

enough of the NO adduct formed for full vibrational

characterization.

We had previously attempted to generate an {FeNO}⁶

complex with [Fe(TPP)(SPh)] at −40 °C, but conversion to

[Fe(TPP)(NO)] was solely observed under those condi-

tions.⁴⁴We therefore retested the reaction of [Fe(TPP)(SPh)]

with NO(g) at −80 °C. As shown in Figure S46, upon addition

of ∼1 equiv of NO a small amount of a new species is observed

J

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 11. (A) Correlation plot comparing the experimentally determined ν(Fe−NO) and ν(N−O) stretching frequencies from rRaman and IR

spectroscopy, respectively. The red line is the best linear ﬁt of the data of the {FeNO}⁶complexes with the four ﬂuoro-containing thiolates and

−

SPh-pNO₂. The blue line is the linear ﬁt for all six complexes (including the SPh(NO₂)₂-pCH₃⁻compound). (B) The same plot, but now shown

as a relative change of the Fe−NO and N−O stretching frequencies, setting the vibrational frequencies of [Fe(TPP)(SPh-pCF₃)(NO)] to zero.

under these conditions with a Soret band at 439 nm and Q-

bands at 558 and 598 nm, indicating formation of a small

amount of the {FeNO}⁶adduct. However, these signals

quickly decrease in intensity, and conversion of the spectrum

to that of the {FeNO}⁷complex [Fe(TPP)(NO)] is observed,

even without a subsequent NO addition. Importantly, these

results indicate that with more electron-rich thiolates such as

SPh⁻, either a direct attack of NO at the thiolate ligand to form

PhS-NO and the ferrous porphyrin (which subsequently binds

NO and forms [Fe(TPP)(NO)]) is preferred over the reaction

of NO with the open coordination site of the ferric heme, or

homolytic cleavage of the Fe−SPh bond occurs upon

generation of the {FeNO}⁶complex. This attack of NO at

the Fe−S bond in heme-thiolate model systems has previously

been reported,⁶¹and S-nitrosylation has also been observed

crystallographically in the heme-thiolate type nitrophorins.⁶²

With respect to Scheme 2, this indicates that the electron

richness of the thiolate is one factor that determines whether

the reaction proceeds to the left or right of the scheme and that

[Fe(TPP)(SPh-pCF₃)] and [Fe(TPP)(SPh-pNO₂)] are in fact

right on the borderline, as pointed out above. To explore

whether temperature could be a factor in improving the

formation of [Fe(TPP)(SPh)(NO)] or the stability of

[Fe(TPP)(SPh-pCF₃)(NO)], we then repeated the reactions

of [Fe(TPP)(SPh)] and [Fe(TPP)(SPh-pCF₃)] with NO(g)

at −125 °C in 2-methyltetrahydrofuran (2-MeTHF). How-

ever, these experiments resulted in direct conversion of

[Fe(TPP)(SPh)] to the corresponding {FeNO}⁷complex

[Fe(TPP)(NO)] (data not shown). In the case of [Fe(TPP)-

(SPh-pCF₃)], even less of the {FeNO}⁶complex accumulated

before decomposition to [Fe(TPP)(NO)] occurred (Figure

S47). On the basis of the UV−vis spectra, we speculate that in

2-MeTHF the corresponding six-coordinate complexes [Fe-

(TPP)(SPh*)(2-MeTHF)] form, which actually inhibits NO

coordination to the Fe center and instead promotes attack at

the thiolate. Hence, the decrease in temperature likely oﬀers no

additional beneﬁt toward the successful generation of [Fe-

(TPP)(SPh*)(NO)] type complexes in 2-MeTHF.

demonstrate experimentally that the Fe−NO and N−O bond

strengths (and correspondingly, the stretching frequencies) are

indeed directly correlated to the thiolate donor strength. For

example, as the thiolate donicity decreases (going from left to

right in Figure 11A), both the N−O and Fe−NO stretching

frequencies continually increase. Correspondingly, the complex

with the strongest thiolate donor, [Fe(TPP)(SPh-pCF₃)-

(NO)], has the lowest observed ν(N−O) and ν(Fe−NO)

stretching frequencies in the series. Figure 11A also shows two

diﬀerent linear ﬁts of the data. When only the ﬂuorine-

−

substituted thiolates and SPh-pNO₂are considered, a perfect

correlation of the vibrational data is observed, as shown by the

red line in Figure 11A. When the dinitro-substituted

compound is included in the ﬁt, the blue correlation line in

Figure 11A is obtained. The deviation of the vibrational

properties of the dinitro-thiophenolate complex from the

correlation line is likely caused by eﬀects other than those

purely electronic in nature. We speculate that the bulky o-

dinitro substituents also impart steric eﬀects on the resulting

NO complex, causing the observed deviations. These diﬀer-

ences aside, the correlation plot provides experimental evidence,

for the first time, that the thiolate ligand is in fact directly

responsible for modulating the N−O and Fe−NO bond strengths

in heme-thiolate {FeNO}⁶complexes.

This is a very important ﬁnding, as the direct correlation

between the Fe−NO and N−O bond strengths now proven

experimentally provides direct evidence for the electronic

mechanism by which the thiolate donor is able to exert this

eﬀect on the Fe−N−O unit. This ﬁnding is in contrast to the

inverse correlation observed for ferrous heme-carbonyl

complexes.⁶³In this case, variation of the trans ligand to CO

changes the Fe−CO π back-bond. A weakening of the π back-

bond, for example, would decrease the Fe−CO stretch

(because of the weaker metal−CO bond) and, at the same

time, increase the C−O stretch (due to the decrease in the

occupation of the CO(π*) orbitals). As established by Spiro,

such an inverse correlation is therefore a hallmark of a variation

in metal−ligand π back-bonding.⁶³Previous work has shown

that heme {FeNO}⁶complexes have Fe^II−NO⁺type electronic

structures, irrespective of the trans ligand to NO, and that in

this case the Fe^II−NO⁺interaction is dominated by π back-

bonding (as in the case of the isoelectronic Fe^II−CO

complexes). It is therefore very surprising that the variation

Correlation between the ν(N−O) and ν(Fe−NO)

Stretching Frequencies. Figure 11A compares the values

of the ν(N−O) and ν(Fe−NO) stretching frequencies (Table

4) for our series of thiolate-coordinated {FeNO}⁶complexes

visually, in the form of a correlation plot. These results

K

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

of the axial thiolate ligand in our {FeNO}⁶series does NOT

lead to the established inverse correlation of the Fe−NO and

N−O vibrational frequencies, implying that the variation in the

thiolate ligand does not signiﬁcantly alter the Fe^II−NO⁺π

back-bond. Instead, the direct correlation observed here implies

that there is a change in the Fe−NO σ-bond and, hence, that the

axial thiolate imparts a σ-trans effect on the Fe−N−O unit. In

order to further characterize the nature of this σ interaction, we

performed DFT calculations, as described in the next section.

Interestingly, the obtained slope of 1.20 ( 0.31 is the

standard error) from the linear ﬁt in Figure 11A suggests that

the thiolate’s donor strength has close to a proportional eﬀect

in decreasing both the Fe−NO and N−O stretching

frequencies (on an absolute wavenumber scale), with a slightly

greater eﬀect on the N−O stretching frequency. However,

comparison of the relative percent change of the Fe−NO and

N−O stretching frequencies of the six complexes, as shown in

Figure 11B, paints a diﬀerent picture. From this comparison,

the total variation observed for the Fe−NO stretching

frequency (∼3%) is seen to be about 3 times larger than the

total percentage change observed for the N−O stretching

frequency (∼1%). This suggests that the thiolate’s donor

strength actually has a greater eﬀect on modulating the Fe−

NO bond in heme-thiolate {FeNO}⁶complexes in comparison

to the N−O bond, which further supports the involvement of a

thiolate σ-trans eﬀect on the Fe−N−O unit to explain the

experimental trend.

indicates that the σ-trans eﬀect of the anionic ligand, here

chloride, is not responsible for the bending of the Fe−N−O

unit as originally proposed.^21,64However, as previously

proposed, our new results show again the mixing of an Fe−

N−O antibonding (σ*) orbital into the Cl(p_z) donor orbital,

2

mediated by the empty d_zorbital of the iron center. The

resulting, occupied molecular orbital (MO), labeled Cl(p_z)

2

_d_z, is shown in Figure 12 (bottom left). This orbital is

strongly Fe−NO and weakly N−O antibonding, which

explains the experimental observation that the Fe−NO bond

is more strongly aﬀected by changes in the donor strength of

the axial thiolate ligand in comparison to the N−O bond (see

Figure 13). We further investigated whether the changes in the

Computational Insight into the Nature of the

Thiolate Trans Eﬀect. With the new experimental trend in

hand, we then reinvestigated the Fe−N−O bonding in

{FeNO}⁶complexes using DFT calculations. Here, we invoked

(a) the [Fe(TPP)(Cl)(NO)] complex, which has slightly

weaker Fe−NO and N−O bonds in comparison to imidazole-

bound complexes with no trans eﬀect,^21,24,64,65(b) our

thiophenolate complexes, and (c) a hypothetical sulﬁde

complex, [Fe(P)(S)(NO)]⁻. For all of these calculations, the

porphine dianion (P²⁻) was used as the porphyrin ligand.

The optimized structure of the chloro complex [Fe(P)(Cl)-

(NO)] is shown in Figure 12 (top left). Importantly, this

compound shows a linear Fe−N−O unit (in agreement with

the crystal structure of [Fe(TPP)(Cl)(NO)]),²⁴which

Figure 13. Correlation plot comparing the experimentally determined

ν(Fe−NO) and ν(N−O) stretching frequencies of our {FeNO}⁶

model complexes and the NO adducts of ferric heme-thiolate

enzymes. Note the data point for [Fe(TPP)(Cl)(NO)] added at the

top right (ν(Fe−NO) 563 cm⁻¹and ν(N−O) 1880 cm⁻¹).²⁴A slope

of 0.99( 0.14) represents the best linear ﬁt of the data for the six

synthetic heme-thiolate and the TPP chloro {FeNO}⁶complexes. The

−

blue box represents a data point estimate for the SR-H₂(=S-2,6-

(CF₃CONH)₂C₆H₃, structure shown on the bottom right) {FeNO}⁶

complex, on the basis of the correlation line, giving an estimated

ν(N−O) ∼1870 cm⁻¹( 5 cm⁻¹).

Fe−NO and N−O bond strengths in [Fe(P)(Cl)(NO)] do

indeed correlate with the amount of Fe−N−O (σ*) character

2

that is mixed into the Cl(p_z)_d_zMO. For this purpose, we

ﬁxed the Fe−Cl distance in [Fe(P)(Cl)(NO)] at diﬀerent

values larger and smaller than the optimized Fe−Cl bond

length and then reoptimized the structure. As is evident from

Table S1, a decrease in the Fe−Cl distance causes a weakening

of both the Fe−NO and N−O bonds (and a lowering of the

corresponding stretching frequencies), and, at the same time,

an increase in the admixture of Fe−N−O (σ*) character into

2

Cl(p_z)_d_zis observed. The opposite trend is observed when

the Fe−Cl distance is increased from the optimized value.

These results therefore conﬁrm the previous observation that

the σ-trans eﬀect of the axial anionic ligand is indeed mediated

by back-bonding into the Fe−N−O antibonding σ* orbital,

which in turn is proportional to the donor strength of the

anionic ligand bound trans to NO.

Since in all of the structures obtained for the [Fe(P)(Cl)-

(NO)] system the Fe−N−O unit remains linear, this poses the

question then of what causes the bending of the Fe−N−O unit

in the thiolate coordinated complexes. In order to determine

whether this requires an anisotropic π donation from the axial

ligand, we then investigated the hypothetical, sulﬁde-bound

Figure 12. (top) DFT optimized (BP86/TZVP) structures of

[Fe(P)(Cl)(NO)] (left) and [Fe(P)(S)(NO)]⁻(right). (bottom)

MO contour plots that show the admixture of the Fe−N−O σ-

2

antibonding (σ*) orbital into the occupied (bonding) Cl(p_z)_d_z

2

(left) and S(p_z)_d_z(right) MOs, respectively.

L

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

complex [Fe(P)(S)(NO)]⁻. Surprisingly, although sulﬁde is an

isotropic π donor like chloride, the Fe−N−O unit is bent in

the optimized structure of this complex, as shown in Figure 12

(top right). The shape of the Fe−N−O antibonding (σ*)

orbital is slightly diﬀerent in this case, as shown in Figure 12

(bottom right) due to the Fe−N−O bending, but the

fundamental nature of this orbital being strongly Fe−NO

and weakly N−O antibonding remains. On the basis of these

results, we conclude that the bending of the Fe−N−O unit is

likely caused by strong π donation from the axial anionic

ligand. Due to the fact that iron is in the ferrous low-spin state

in heme {FeNO}⁶complexes (due to their Fe^II−NO⁺type

electronic structure), this causes strong Coulomb repulsion

between the occupied ligand π and Fe(d_π) orbitals, which is

Despite this electronic structure, however, the observed direct

correlation of the Fe−NO and N−O bond strengths rules out

this possibility and demonstrates that the thiolate ligand does

NOT modulate the Fe−NO π back-bond (since this would

lead to an inverse correlation of the Fe−NO and N−O

vibrational frequencies and, hence, bond strengths). Our

results therefore directly support previous computational

results which show that the strong σ donicity of the thiolate

(and other axial anionic ligands bound trans to NO) leads to

the occupation of an Fe−N−O σ-antibonding (σ*) orbital and

that the degree to which this orbital becomes occupied (via

2

mixing into the thiolate(p_z)_d_zbonding MO) directly

correlates with the donor strength of the trans-thiolate ligand.

In other words, the thiolate imparts a σ-trans eﬀect (or

interaction, since it is a thermodynamic eﬀect) on the bound

nitrosyl ligand and, in this way, is able to control the degree of

activation of the FeNO unit for P450nor catalysis.

2

decreased by mixing of one of the d_πorbitals with d_z, which

eﬀectively leads to a rotation of the d_πorbital and a bending of

the Fe−N−O unit in the corresponding plane. On the other

hand, for an anisotropic π-donor such as a thiolate, one would

then predict that the Fe−N−O unit should bend more or less

in the plane of the thiolate π-donor lone pair.

In summary, our results show that the σ-trans eﬀect and the

bending of the Fe−N−O unit in thiolate-coordinated heme

{FeNO}⁶complexes have two diﬀerent origins that can be

traced back to two diﬀerent orbital interactions. With respect

to the bonding properties of the Fe−N−O unit, the trans eﬀect

mediated by the σ* orbital of the Fe−N−O unit is the more

important interaction for the electronic structure of the heme

{FeNO}⁶complexes.

Figure 13 shows additional protein data superimposed on

our ν(N−O)/ν(Fe−NO) correlation plot, including the

P450nor, CPO, and iNOS {FeNO}⁶complexes. As is evident

from this ﬁgure, the vibrational properties observed for our

model complexes are reasonably well matched with the protein

data. The small deviations observed in the plot are likely due to

variations in the positioning of hydrogen bonds to the axial

cysteinate, diﬀerences in heme conformation, electrostatic and

steric eﬀects, etc. between the diﬀerent proteins, whereas, on

the other hand, our data are all derived from electron-poor

thiophenolates with the same heme coligand, in the same

relative electrostatic environment. Importantly, the exper-

imental data show that it is actually the [Fe(TPP)(SPh-

pCF₃)(NO)] complex with the most electron rich thiolate in

our series that exhibits the best agreement with the vibrational

properties of the {FeNO}⁶complex of P450nor and, hence,

constitutes an excellent electronic model for the enzyme. In

this regard, it should also be noted that the data point for the

ferric NO adduct of P450nor is located very close to the

correlation line. Thus, the prediction from DFT in this regard

is incorrect. Notably, it has recently been highlighted that CPO

shows much stronger hydrogen bonds to its axial cysteinate

ligand in comparison to typical Cyt P450 enzymes and, thus,

has a weaker thiolate donor in comparison to Cyt P450s.¹³The

same is true for NOS enzymes, which have a highly conserved

tryptophan residue that makes a strong H-bond to the

proximal cysteinate.²⁸Thus, the fact that the CPO and NOS

{FeNO}⁶complexes have higher N−O and Fe−NO stretching

frequencies in comparison to those of P450nor is in full

agreement with the results obtained for our {FeNO}⁶

complexes. In conclusion, the characterization of the model

complexes presented in this work shows clearly that it is in fact

the thiolate donor strength that is a major factor in dictating

the N−O and Fe−NO bond strengths in heme-thiolate

{FeNO}⁶complexes and, hence, provides a means of how

nature can control the electronic structure and the degree of

activation of the Fe−N−O unit in these types of species.

Furthermore, the fact that we can model the electronic eﬀects

of the hydrogen bonds to the Cys ligand in heme-thiolate

proteins with electron-poor thiolates further conﬁrms that the

main electronic eﬀect of the hydrogen bonds in the proteins is

indeed a reduction of the donor strength of the axial thiolate

(Cys) ligand.

DISCUSSION

■

In this work, we accomplished the synthesis and character-

ization of the ﬁrst series of heme-thiolate {FeNO}⁶complexes

where the donor strength of the thiolate is systematically

varied. This was achieved using [Fe(TPP)(SPh*)] type

complexes with a series of thiophenolate ligands (SPh*⁻)

that carry diﬀerent electron-withdrawing substituents, accord-

ing to Scheme 1. The ﬁve-coordinate high-spin ferric precursor

complexes were ﬁrst characterized using diﬀerent spectroscopic

methods and then reacted with NO gas at −80 °C. Here, the

application of such low temperatures is key in order to obtain

the corresponding NO adducts of type [Fe(TPP)(SPh*)-

(NO)], which decompose to the ﬁve-coordinate ferrous

nitrosyl complex [Fe(TPP)(NO)] upon warming above −60

°C. Still, our studies also show that when more highly donating

thiophenolates are used, such as the base compound SPh⁻,

then the thiolate either becomes more reactive toward NO

than the iron center (forming a nitrosothiol), or the NO

adduct decomposes immediately via homolytic Fe−S bond

cleavage (see Scheme 2). In any case, a stable {FeNO}⁶

complex cannot be formed, even at −80 °C. Nevertheless,

using six ﬂuoro- and nitro-substituted thiophenolates, we were

able to generate the ﬁrst series of thiolate-coordinated heme

{FeNO}⁶complexes and determine their Fe−NO and N−O

stretching frequencies. Our data demonstrate experimentally,

for the ﬁrst time, that the thiolate donor strength directly

controls the strength of the Fe−NO and N−O bonds in the

{FeNO}⁶complexes. Here, as shown in Figure 13, a direct

correlation of the Fe−NO and N−O bond strengths is

observed as a function of the thiolate donor strength. Since the

heme {FeNO}⁶complexes are best described as Fe^II−NO⁺

systems, where the Fe−NO interaction is dominated by π

back-bonding, one would a priori expect that the thiolate

ligand modulates the strength of the Fe−NO π back-bond.

Vice versa, this also means that NO can serve as a sensitive

probe to measure the thiolate trans eﬀect and, hence, thiolate

donor strength in a model complex or a new Cyt P450 enzyme,

M

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

which is not possible in any other direct way. This could be

accomplished by simply reacting the ferric form of the enzyme

or model complex with NO (at low temperature, in the case of

an inorganic complex) and then measuring the Fe−NO and

N−O stretching frequencies as demonstrated here. A

comparison of these data with the correlation plot in Figure

13 then provides a direct measure for the donor strength of the

thiolate in the new system. This approach can be directly

demonstrated for the only structurally characterized thiolate-

coordinated {FeNO}⁶complex, [Fe(OEP)(SR-H₂)(NO)]

Figure 14. (left) DFT-optimized structure of [Fe(P)(SPh)(OOH₂)].

(right) MO contour plot that shows the admixture of the Fe−O−O σ

antibonding (σ*) orbital into the (occupied) Fe−S σ bonding MO

(see the Experimental Section for additional details).

(SR-H₂= S-2,6-(CF₃CONH)₂C₆H₃; OEP²⁻= octaethylpor-

−

phyrin dianion), where SR-H₂⁻is a thiophenolate ligand that is

stabilized by two intramolecular amide hydrogen bonds.³⁰In

the initial communication, the N−O stretch of this complex

was reported to be about 1850 cm⁻¹, which would match very

well with that reported for the {FeNO}⁶complex of Cyt

P450nor, making it an excellent model for the enzyme.³⁰

However, a closer inspection of the IR data provided in the

paper shows that no well-deﬁned signal was obtained for the

N−O stretch in the IR spectrum of this complex (due to its

instability).³⁰Hence, the N−O stretch of this complex is not

really known. On the other hand, the Fe−NO stretch for this

complex has been accurately determined to be 549 cm⁻¹by

resonance Raman spectroscopy.³¹As indicated in Figure 13,

applying an Fe−NO stretch of about 550 cm⁻¹to the

correlation line predicts the N−O stretch of this complex to

actually occur around 1870 cm⁻¹. This result indicates that the

two hydrogen bonds of SR-H₂⁻are too strong, greatly reducing

the thiolate donor strength of the SR-H₂⁻ligand outside of the

biologically relevant range of Cyt P450 enzymes. In other

words, [Fe(OEP)(SR-H₂)(NO)] is not a great electronic

model for Cyt P450s. On the plus side, the two hydrogen

is shown in Figure 14 (left). Excitingly, in this case, we indeed

observe the admixture of an Fe−O−O σ* orbital into the σ-

bonding MO between the S(p_z) and Fe(d) orbitals, S(p_z)_d,

as shown in Figure 14 (right).

Hence, the σ-trans eﬀect described here and quantiﬁed using

NO as a probe provides an electronic−structural explanation

for Dawson’s push eﬀect and, therefore, provides further

validation that this is a real electronic eﬀect. Importantly, this

also means that Nature can control the magnitude of the push

eﬀect in an ideal way via adjustment of the hydrogen-bonding

network to the proximal cysteinate ligand of Cyt P450s. This

explains why typical Cyt P450 monooxygenases have overall

weaker hydrogen bonds to the thiolate or, in other words,

stronger cysteinate donors in comparison to P450nor. This

manifests itself in lower N−O stretching frequencies of 1800−

1820 cm⁻¹for Cyt P450 monoxygenases in comparison to

ν(N−O) of 1851 cm⁻¹in P450nor. In this indirect way, the

vibrational data of the {FeNO}⁶complexes, which measure the

donor strength of the thiolate, also serve as a sensitive probe

for the magnitude of the “push” eﬀect in Cyt P450

monooxygenase enzymes.

−

bonds present in the SR-H₂ligand greatly contribute to the

stability of the corresponding heme-thiolate complexes, making

this an attractive model system on the basis of stability.

In summary, our work provides fundamental insight into the

nature of the thiolate σ-trans eﬀect, which we propose is of

general importance in Cyt P450 chemistry. By adjusting the

strength of the hydrogen bonds, Nature is able to accurately

ﬁne tune the donor strength of the thiolate, which directly

aﬀects the electronic properties of the FeNO unit in the

corresponding {FeNO}⁶complexes. In our work, we show that

the same tuning of the Fe−NO and N−O bonds can be

accomplished using electron-poor thiophenolate donors. On

the other hand, the bending of the Fe−N−O unit is caused by

the strong π donation of the thiolate; thus, this eﬀect has a

diﬀerent orbital origin. Since the donor strength of the thiolate

aﬀects the FeNO unit, the question arises whether the σ-trans

eﬀect of the thiolate could be the underlying orbital interaction

that is responsible for the so-called “push eﬀect”, ﬁrst proposed

by Dawson.¹⁵The idea is that the strong donicity of the

proximal cysteinate ligand would assist in cleaving the O−O

bond in the ferric-hydroperoxo intermediate (called “Com-

pound 0”) of Cyt P450 catalysis, in comparison to the

analogous species in globins and other histidine-coordinated

proteins. We used DFT calculations to obtain insight into this

issue. First, we optimized the structure of the Cyt P450 model

[Fe(P)(SPh)(OOH)]⁻, but in this case no population of an

Fe−O−O σ antibonding (σ*) orbital is observed. However, it

is also known that Compound 0 is further protonated in order

to induce O−O bond cleavage. We therefore constructed a

model of Compound 0 that is protonated at the distal O atom,

[Fe(P)(SPh)(OOH₂)]. The optimized structure of this model

ASSOCIATED CONTENT

* Supporting Information

The Supporting Information is available free of charge on the

ACS Publications website at DOI: 10.1021/acs.inorg-

chem.9b00091.

■

S

Additional spectroscopic characterization of the ferric

thiolate precursor complexes, NO reactivity data for all

complexes, EPR data and ﬁts, and coordinates of DFT-

optimized structures (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail for N.L.: lehnertn@umich.edu.

ORCID

Andrew P. Hunt: 0000-0003-0565-9050

Nicolai Lehnert: 0000-0002-5221-5498

Notes

■

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

This work was supported by a grant from the National Science

Foundation (CHE-1464696 to N.L.). We thank Dr. Hannah

Shafaat and Dr. Anastasia Manesis from The Ohio State

University for granting us access to their resonance Raman

spectrometer and for their assistance in the data collection.

N

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

(20) Enemark, J. H.; Feltham, R. D. Principles of structure, bonding,

and reactivity for metal nitrosyl complexes. Coord. Chem. Rev. 1974,

13, 339−406.

REFERENCES

■

(1) Bredt, D. S.; Snyder, S. H. Nitric Oxide: A Physiologic

Messenger Molecule. Annu. Rev. Biochem. 1994, 63, 175−195.

(2) Vincent, J.-L.; Zhang, H.; Szabo, C.; Preiser, J.-C. Effects of

Nitric Oxide in Septic Shock. Am. J. Respir. Crit. Care Med. 2000, 161,

1781−1785.

(3) Lehnert, N.; Berto, T. C.; Galinato, M. G. I.; Goodrich, L. E. In

Handbook of Porphyrin Science; World Scientiﬁc: 2010; pp 1−247.

(4) Daiber, A.; Shoun, H.; Ullrich, V. Nitric oxide reductase

(P450nor) from Fusarium oxysporum. J. Inorg. Biochem. 2005, 99,

185−193.

(5) Shimizu, H.; Obayashi, E.; Gomi, Y.; Arakawa, H.; Park, S. Y.;

Nakamura, H.; Adachi, S.-i.; Shoun, H.; Shiro, Y. Proton delivery in

NO reduction by fungal nitric-oxide reductase: Cryogenic crystallog-

raphy, spectroscopy, and kinetics of ferric-NO complexes of wild-type

and mutant enzymes. J. Biol. Chem. 2000, 275, 4816−4826.

(6) US EPA Overview of Greenhouse Gases [Online Early Access];

published online 2016; https://www.epa.gov/ghgemissions/overview-

greenhouse-gases.

(21) Paulat, F.; Lehnert, N. Electronic Structure of Ferric Heme

Nitrosyl Complexes with Thiolate Coordination. Inorg. Chem. 2007,

46, 1547−1549.

(22) Praneeth, V. K. K.; Paulat, F.; Berto, T. C.; George, S. D.;

̈

Nather, C.; Sulok, C. D.; Lehnert, N. Electronic Structure of Six-

Coordinate Iron(III)-Porphyrin NO Adducts: The Elusive Iron(III)-

NO(radical) State and Its Influence on the Properties of These

Complexes. J. Am. Chem. Soc. 2008, 130, 15288−15303.

(23) Soldatova, A. V.; Ibrahim, M.; Olson, J. S.; Czernuszewicz, R.

S.; Spiro, T. G. New Light on NO Bonding in Fe(III) Heme Proteins

from Resonance Raman Spectroscopy and DFT Modeling. J. Am.

Chem. Soc. 2010, 132, 4614−4625.

(24) McQuarters, A. B.; Kampf, J. W.; Alp, E. E.; Hu, M.; Zhao, J.;

Lehnert, N. Ferric Heme-Nitrosyl Complexes: Kinetically Robust or

Unstable Intermediates? Inorg. Chem. 2017, 56, 10513−10528.

(25) Obayashi, E.; Tsukamoto, K.; Adachi, S. i.; Takahashi, S.;

Nomura, M.; Iizuka, T.; Shoun, H.; Shiro, Y. Unique Binding of Nitric

Oxide to Ferric Nitric Oxide Reductase from Fusarium oxysporum

Elucidated with Infrared, Resonance Raman, and X-ray Absorption

Spectroscopies. J. Am. Chem. Soc. 1997, 119, 7807−7816.

(26) Hu, S.; Kincaid, J. R. Resonance Raman characterization of

nitric oxide adducts of cytochrome P450cam: the effect of substrate

structure on the iron-ligand vibrations. J. Am. Chem. Soc. 1991, 113,

2843−2850.

(27) Hu, S.; Kincaid, J. R. Heme active-site structural character-

ization of chloroperoxidase by resonance Raman spectroscopy. J. Biol.

Chem. 1993, 268, 6189−6193.

(28) Couture, M.; Adak, S.; Stuehr, D. J.; Rousseau, D. L. Regulation

of the Properties of the Heme-NO Complexes in Nitric-oxide

Synthase by Hydrogen Bonding to the Proximal Cysteine. J. Biol.

Chem. 2001, 276, 38280−38288.

(29) Li, D.; Stuehr, D. J.; Yeh, S. R.; Rousseau, D. L. Heme

Distortion Modulated by Ligand-Protein Interactions in Inducible

Nitric-oxide Synthase. J. Biol. Chem. 2004, 279, 26489−26499.

(30) Xu, N.; Powell, D. R.; Cheng, L.; Richter-Addo, G. B. The first

structurally characterized nitrosyl heme thiolate model complex.

Chem. Commun. 2006, 2030−2032.

(31) Goodrich, L. E.; Paulat, F.; Praneeth, V. K. K.; Lehnert, N.

Electronic Structure of Heme-Nitrosyls and Its Significance for Nitric

Oxide Reactivity, Sensing, Transport, and Toxicity in Biological

Systems. Inorg. Chem. 2010, 49, 6293−6316.

(32) Suzuki, N.; Higuchi, T.; Urano, Y.; Kikuchi, K.; Uchida, T.;

Mukai, M.; Kitagawa, T.; Nagano, T. First Synthetic NO-Heme-

Thiolate Complex Relevant to Nitric Oxide Synthase and

Cytochrome P450nor. J. Am. Chem. Soc. 2000, 122, 12059−12060.

(33) Adler, A. D.; Longo, F. R.; Kampas, F.; Kim, J. On the

preparation of metalloporphyrins. J. Inorg. Nucl. Chem. 1970, 32,

2443−2445.

(34) Tsai, R.; Yu, C. A.; Gunsalus, I. C.; Peisach, J.; Blumberg, W.;

Orme-Johnson, W. H.; Beinert, H. Spin-State Changes in Cytochrome

P-450cam on Binding of Specific Substrates. Proc. Natl. Acad. Sci. U. S.

A. 1970, 66, 1157−1163.

(7) Riplinger, C.; Bill, E.; Daiber, A.; Ullrich, V.; Shoun, H.; Neese,

F. New Insights into the Nature of Observable Reaction Intermediates

in Cytochrome P450 NO Reductase by Using a Combination of

Spectroscopy and Quantum Mechanics/Molecular Mechanics Calcu-

lations. Chem. - Eur. J. 2014, 20, 1602−1614.

(8) Higgins, S. A.; Welsh, A.; Orellana, L. H.; Konstantinidis, K. T.;

Chee-Sanford, J. C.; Sanford, R. A.; Schadt, C. W.; Loffler, F. E.

̈

Detection and Diversity of Fungal Nitric Oxide Reductase Genes

(p450nor) in Agricultural Soils. Appl. Environ. Microbiol. 2016, 82,

2919−2928.

(9) Novinscak, A.; Goyer, C.; Zebarth, B. J.; Burton, D. L.;

Chantigny, M. H.; Filion, M. Novel P450nor Gene Detection Assay

Used To Characterize the Prevalence and Diversity of Soil Fungal

Denitrifiers. Appl. Environ. Microbiol. 2016, 82, 4560−4569.

(10) Chen, H.; Shi, W. Opening up the N₂O-producing fungal

community in an agricultural soil with a cytochrome p450nor-based

primer tool. Applied Soil Ecology 2017, 119, 392−395.

(11) McQuarters, A. B.; Wolf, M. W.; Hunt, A. P.; Lehnert, N.

1958−2014: After 56 Years of Research, Cytochrome P450 Reactivity

Is Finally Explained. Angew. Chem., Int. Ed. 2014, 53, 4750−4752.

(12) Denisov, I. G.; Makris, T. M.; Sligar, S. G.; Schlichting, I.

Structure and Chemistry of Cytochrome P450. Chem. Rev. 2005, 105,

2253−2278.

(13) Krest, C. M.; Silakov, A.; Rittle, J.; Yosca, T. H.; Onderko, E. L.;

Calixto, J. C.; Green, M. T. Significantly shorter Fe−S bond in

cytochrome P450-I is consistent with greater reactivity relative to

chloroperoxidase. Nat. Chem. 2015, 7, 696.

(14) Galinato, M. G. I.; Spolitak, T.; Ballou, D. P.; Lehnert, N.

Elucidating the Role of the Proximal Cysteine Hydrogen-Bonding

Network in Ferric Cytochrome P450cam and Corresponding Mutants

Using Magnetic Circular Dichroism Spectroscopy. Biochemistry 2011,

50, 1053−1069.

(15) Sono, M.; Roach, M. P.; Coulter, E. D.; Dawson, J. H. Heme-

Containing Oxygenases. Chem. Rev. 1996, 96, 2841−2888.

(16) Yosca, T. H.; Rittle, J.; Krest, C. M.; Onderko, E. L.; Silakov, A.;

Calixto, J. C.; Behan, R. K.; Green, M. T. Iron(IV)hydroxide pKa and

the Role of Thiolate Ligation in C−H Bond Activation by

Cytochrome P450. Science 2013, 342, 825−829.

(17) Shiro, Y.; Fujii, M.; Iizuka, T.; Adachi, S. i.; Tsukamoto, K.;

Nakahara, K.; Shoun, H. Spectroscopic and Kinetic Studies on

Reaction of Cytochrome P450nor with Nitric Oxide: Implication for

its Nitric Oxide Reduction Mechanism. J. Biol. Chem. 1995, 270,

1617−1623.

(18) Oshima, R.; Fushinobu, S.; Su, F.; Zhang, L.; Takaya, N.;

Shoun, H. Structural Evidence for Direct Hydride Transfer from

NADH to Cytochrome P450nor. J. Mol. Biol. 2004, 342, 207−217.

(19) Lehnert, N.; Praneeth, V. K. K.; Paulat, F. Electronic structure

of iron(II)−porphyrin nitroxyl complexes: Molecular mechanism of

fungal nitric oxide reductase (P450nor). J. Comput. Chem. 2006, 27,

1338−1351.

(35) Tang, S. C.; Koch, S.; Papaefthymiou, G. C.; Foner, S.; Frankel,

R. B.; Ibers, J. A.; Holm, R. H. Axial ligation modes in iron(III)

porphyrins. Models for the oxidized reaction states of cytochrome P-

450 enzymes and the molecular structure of iron(III) protoporphyrin

IX dimethyl ester p-nitrobenzenethiolate. J. Am. Chem. Soc. 1976, 98,

2414−2434.

(36) Li, J.; Peng, Q.; Oliver, A. G.; Alp, E. E.; Hu, M. Y.; Zhao, J.;

Sage, J. T.; Scheidt, W. R. Comprehensive Fe−Ligand Vibration

Identification in {FeNO}6 Hemes. J. Am. Chem. Soc. 2014, 136,

18100−18110.

(37) Kobayashi, H.; Higuchi, T.; Kaizu, Y.; Osada, H.; Aoki, M.

Electronic Spectra of Tetraphenylporphinatoiron(III) Methoxide.

Bull. Chem. Soc. Jpn. 1975, 48, 3137−3141.

O

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

(38) Miller, K. M.; Strouse, C. E. Two-dimensional short-range

order in the crystal structure of (tetraphenylporphinato)iron(III)

2,3,5,6-tetrafluorobenzenethiolate. Analysis of the x-ray diffuse

scattering and a simulation of crystal growth. Inorg. Chem. 1984, 23,

2395−2400.

(56) Roncaroli, F.; Videla, M.; Slep, L. D.; Olabe, J. A. New features

in the redox coordination chemistry of metal nitrosyls {M−NO+; M−

NO; M−NO-(HNO)}. Coord. Chem. Rev. 2007, 251, 1903−1930.

(57) Ellison, M. K.; Scheidt, W. R. Synthesis, Molecular Structures,

and Properties of Six-Coordinate [Fe(OEP)(L)(NO)]⁺Derivatives:

Elusive Nitrosyl Ferric Porphyrins. J. Am. Chem. Soc. 1999, 121,

5210−5219.

(58) Abucayon, E. G.; Khade, R. L.; Powell, D. R.; Zhang, Y.;

Richter-Addo, G. B. Hydride Attack on a Coordinated Ferric Nitrosyl:

Experimental and DFT Evidence for the Formation of a Heme

Model−HNO Derivative. J. Am. Chem. Soc. 2016, 138, 104−107.

(59) Abucayon, E. G.; Khade, R. L.; Powell, D. R.; Shaw, M. J.;

Zhang, Y.; Richter-Addo, G. B. Over or under: hydride attack at the

metal versus the coordinated nitrosyl ligand in ferric nitrosyl

porphyrins. Dalton Transactions 2016, 45, 18259−18266.

(60) Wang, J.; Rousseau, D. L.; Abu-Soud, H. M.; Stuehr, D. J.

Heme coordination of NO in NO synthase. Proc. Natl. Acad. Sci. U. S.

A. 1994, 91, 10512−10516.

(61) Franke, A.; Stochel, G.; Suzuki, N.; Higuchi, T.; Okuzono, K.;

van Eldik, R. Mechanistic Studies on the Binding of Nitric Oxide to a

Synthetic Heme-Thiolate Complex Relevant to Cytochrome P450. J.

Am. Chem. Soc. 2005, 127, 5360−5375.

(62) Weichsel, A.; Maes, E. M.; Andersen, J. F.; Valenzuela, J. G.;

Shokhireva, T. K.; Walker, F. A.; Montfort, W. R. Heme-assisted S-

nitrosation of a proximal thiolate in a nitric oxide transport protein.

Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 594−599.

(39) Oshio, H.; Ama, T.; Watanabe, T.; Nakamoto, K. Iron·sulfur

vibrations in model compounds of cytochromes P-450 and c. Inorg.

Chim. Acta 1985, 96, 61−66.

(40) Arasasingham, R. D.; Balch, A. L.; Cornman, C. R.; De Ropp, J.

S.; Eguchi, K.; La Mar, G. N. Proton NMR studies of iron(III)

porphyrins with axial phenoxide or thiophenoxide ligands. Inorg.

Chem. 1990, 29, 1847−1850.

(41) Collman, J. P.; Sorrell, T. N.; Hoffman, B. M. Models for

cytochrome P-450. J. Am. Chem. Soc. 1975, 97, 913−914.

(42) Collman, J. P.; Sorrell, T. N.; Hodgson, K. O.; Kulshrestha, A.

K.; Strouse, C. E. Molecular dynamics in the solid state. A dynamic

model of the low-spin iron(III) to high-spin iron(III) transformation

in P450 enzymes. J. Am. Chem. Soc. 1977, 99, 5180−5181.

(43) Byrn, M. P.; Strouse, C. E. Porphyrin sponges. Inversion

disorder and inversion twinning in lattice clathrates based on five-

coordinate metallotetraarylporphyrin complexes. J. Am. Chem. Soc.

1991, 113, 2501−2508.

(44) Goodrich, L. E. Model Complexes of Cytochrome P450 Nitric

Oxide Reductase. Ph.D. Dissertation, University of Michigan, 2012.

(45) Das, P. K.; Samanta, S.; McQuarters, A. B.; Lehnert, N.; Dey, A.

Valence tautomerism in synthetic models of cytochrome P450. Proc.

Natl. Acad. Sci. U. S. A. 2016, 113, 6611−6616.

(63) Spiro, T. G.; Wasbotten, I. H. CO as a vibrational probe of

heme protein active sites. J. Inorg. Biochem. 2005, 99, 34−44.

(64) Xu, N.; Goodrich, L. E.; Lehnert, N.; Powell, D. R.; Richter-

Addo, G. B. Preparation of the Elusive [(por)Fe(NO)(O-ligand)]

Complex by Diffusion of Nitric Oxide into a Crystal of the Precursor.

Angew. Chem., Int. Ed. 2013, 52, 3896−3900.

(46) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;

Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,

B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.

P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;

Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.;

Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.;

Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.;

Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi,

R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar,

S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox,

J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.;

Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;

Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.;

Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A.

D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J.

Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2009.

(65) Xu, C.; Spiro, T. G. Ambidentate H-bonding by heme-bound

NO: structural and spectral effects of − O versus − N H-bonding.

JBIC, J. Biol. Inorg. Chem. 2008, 13, 613−621.

́

(66) Perez-Gonzalez, A.; Castaneda-Arriaga, R.; Verastegui, B.;

̃

́

Carreon-Gonzalez, M.; Alvarez-Idaboy, J. R.; Galano, A. Estimation of

empirically fitted parameters for calculating pKa values of thiols in a

fast and reliable way. Theor. Chem. Acc. 2018, 137, 5.

(67) Hupe, D. J.; Pohl, E. R. A Demonstration of the Reactivity

Selectivity Principle for the ThiolDisulfide Interchange Reaction.

Isr. J. Chem. 1985, 26, 395−399.

(68) Horn, M.; Nienhaus, K.; Nienhaus, G. Fourier transform

infrared spectroscopy study of ligand photodissociation and migration

in inducible nitric oxide synthase. F1000Research 2014, 3, 290.

(69) Singh, U. P.; Obayashi, E.; Takahashi, S.; Iizuka, T.; Shoun, H.;

Shiro, Y. The effects of heme modification on reactivity, ligand

binding properties and iron-coordination structures of cytochrome

P450nor. Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 1998,

1384, 103−111.

(47) Becke, A. D. Density-functional exchange-energy approxima-

tion with correct asymptotic behavior. Phys. Rev. A: At., Mol., Opt.

Phys. 1988, 38, 3098−3100.

(48) Perdew, J. P. Density-functional approximation for the

correlation energy of the inhomogeneous electron gas. Phys. Rev. B:

Condens. Matter Mater. Phys. 1986, 33, 8822−8824.

̈

(49) Schafer, A.; Huber, C.; Ahlrichs, R. Fully optimized contracted

Gaussian basis sets of triple zeta valence quality for atoms Li to Kr. J.

Chem. Phys. 1994, 100, 5829−5835.

(50) Neese, F. The ORCA program system. Wiley Interdisciplinary

Reviews: Computational Molecular Science 2012, 2, 73−78.

(51) Li, H.; Igarashi, J.; Jamal, J.; Yang, W.; Poulos, T. L. Structural

studies of constitutive nitric oxide synthases with diatomic ligands

bound. JBIC, J. Biol. Inorg. Chem. 2006, 11, 753−768.

(52) Meininger, D. J.; Caranto, J. D.; Arman, H. D.; Tonzetich, Z. J.

Studies of Iron(III) Porphyrinates Containing Silanethiolate Ligands.

Inorg. Chem. 2013, 52, 12468−12476.

(53) Bohle, D. S.; Goodson, P. A.; Smith, B. D. Synthesis, structure

and ligand exchange reactions of Ru(TTP)(NO)(OMe). Polyhedron

1996, 15, 3147−3150.

(54) Addison, A. W.; Stephanos, J. J. Nitrosyliron(III) hemoglobin:

autoreduction and spectroscopy. Biochemistry 1986, 25, 4104−4113.

(55) Hoshino, M.; Maeda, M.; Konishi, R.; Seki, H.; Ford, P. C.

Studies on the Reaction Mechanism for Reductive Nitrosylation of

Ferrihemoproteins in Buffer Solutions. J. Am. Chem. Soc. 1996, 118,

5702−5707.

P

DOI: 10.1021/acs.inorgchem.9b00091

Inorg. Chem. XXXX, XXX, XXX−XXX

Article Doi

The Thiolate Trans Effect in Heme {FeNO}6 Complexes and Beyond: Insight into the Nature of the Push Effect

DOI: 10.1021/acs.inorgchem.9b00091

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.inorgchem.9b00091

Full text of DOI:10.1021/acs.inorgchem.9b00091

Products guided by the article

R&D Labs maybe for 126460-04-6

Relevant to this article

Hot Product