Characterization of covalent Ene adduct intermediates in "hydride equivalent" transfers in a dihydropyridine model for NADH reduction reactions

Article abstract of DOI:10.1016/j.bioorg.2011.10.002

A study of the reactions of an NADH model, 1,4-di(trimethylsilyl)-1,4- dihydropyridine, 7, with a series of α,β-unsaturated cyano and carbonyl compounds has produced the first direct evidence for an obligatory covalent adduct between a dihydropyridine and substrate in a reduction reaction. The reactions were monitored by NMR spectroscopy. In all reactions studied, the covalent adduct was the first new species detected and its decomposition to form products could be observed. Concentrations of adducts were sufficiently high at steady-state that their structures could be determined directly from NMR spectra of the reaction mixtures; adduct structures are those expected from an Ene reaction between 7 and the substrate. This first reaction step results in transfer of the C₄ hydrogen nucleus of 7 to the substrate and formation of a covalent bond between C₂ of the dihydropyridine ring and the substrate α-atom. Discovery of these Ene-adduct intermediates completes the spectrum of mechanisms observed in NADH model reactions to span those with free radical intermediates, no detectable intermediates and now covalent intermediates. The geometry of the transition state for formation of the Ene adduct is compared with those of theoretical transition state models and crystal structures of enzyme-substrate/inhibitor complexes to suggest a relative orientation for the dihydropyridine ring and the substrate in an initial cyclic transition state that is flexible enough to accommodate all observed mechanistic outcomes.

Full text of DOI:10.1016/j.bioorg.2011.10.002

Bioorganic Chemistry 40 (2012) 57–66
Contents lists available at SciVerse ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
Characterization of covalent Ene adduct intermediates in ‘‘hydride
equivalent’’ transfers in a dihydropyridine model for NADH reduction reactions
1
⇑
R. Daniel Libby , Ryan A. Mehl
Moravian College, Chemistry Department, Bethlehem, PA 18018, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 August 2011
Available online 20 October 2011
A study of the reactions of an NADH model, 1,4-di(trimethylsilyl)-1,4-dihydropyridine, 7, with a series of
a,b-unsaturated cyano and carbonyl compounds has produced the first direct evidence for an obligatory
covalent adduct between a dihydropyridine and substrate in a reduction reaction. The reactions were
monitored by NMR spectroscopy. In all reactions studied, the covalent adduct was the first new species
detected and its decomposition to form products could be observed. Concentrations of adducts were suf-
ficiently high at steady-state that their structures could be determined directly from NMR spectra of the
reaction mixtures; adduct structures are those expected from an Ene reaction between 7 and the sub-
strate. This first reaction step results in transfer of the C4 hydrogen nucleus of 7 to the substrate and for-
Keywords:
NADH model
Dihydropyridine
Hydride transfer
Ene reaction
Retro-Ene reaction
Ene-adduct
mation of a covalent bond between C2 of the dihydropyridine ring and the substrate a-atom. Discovery of
these Ene-adduct intermediates completes the spectrum of mechanisms observed in NADH model reac-
tions to span those with free radical intermediates, no detectable intermediates and now covalent inter-
mediates. The geometry of the transition state for formation of the Ene adduct is compared with those of
theoretical transition state models and crystal structures of enzyme–substrate/inhibitor complexes to
suggest a relative orientation for the dihydropyridine ring and the substrate in an initial cyclic transition
state that is flexible enough to accommodate all observed mechanistic outcomes.
Mechanism
Transition state structure
Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction
reaction. Thus, the currently prevailing enzymatic mechanistic
model is a one-step hydride transfer process, often referred to as
Since Westheimer et al. [1] demonstrated that there is a direct
hydrogen transfer from ethanol to NAD+ in the alcohol dehydroge-
nase catalyzed reaction, much work has been focused on the chem-
ical mechanism of this transfer [2]. Both radical and ionic
mechanisms have been proposed but to date no direct conclusive
evidence for any covalent intermediate species has been presented
for any enzymatic or model nicotinamide hydride transfer
a ‘‘hydride equivalent transfer’’, which is generally written as in
Eq. (1) with a transition state (TS) that has C4 of the nicotinamide
ring (1), the H nucleus being transferred, and the carbon and oxy-
gen atoms of the substrate (2) collinear.
In 1971 Hamilton proposed an ionic mechanism in which the
hydrogen transfer from substrate to NAD+ is accomplished through
an electrocyclic reaction [3]. Hamilton’s mechanism (Eq. (2))
requires an intermediate, 5 or 6, with a covalent bond between
the oxygen atom of the substrate (2) and either C2 or C6 of NAD+ (1).
The first step of Hamilton’s mechanism is a nucleophilic addi-
tion of 2 to the iminium function of the NAD+ (1). The second step
⇑
Corresponding author. Fax: +1 610 625 7918.
E-mail address: rdlibby@chem.moravian.edu (R.D. Libby).
Present address: Department of Biochemistry & Biophysics, 2011 Agricultural Life
1
Sciences Building, Oregon State University, Corvallis, OR 97331-7305, United States.
0045-2068/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.bioorg.2011.10.002

58
R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
is a cyclic intramolecular hydrogen transfer known as a Retro-Ene
reaction, a well-documented pathway for decomposition of allyl
ethers and thioethers [4].
When the reaction of Eq. (2) is viewed in the reverse direction
(Eq. (3)), the first step is seen to be an electrocyclic Ene reaction,
a type of reaction first characterized by Alder et al. in 1943 [5].
We will refer to the mechanism proposed by Hamilton as the
Ene mechanism for hydride transfer in nicotinamide reactions.
Ene reactions involve transition states with aromatic character,
completing the spectrum of mechanisms observed in NADH model
reactions from those with single electron transfers, to apparently
concerted hydride transfers (no detectable intermediates) [2] and
now to Ene reactions (covalent intermediate Ene-adducts charac-
terized). The Ene mechanism also suggests a relative orientation
for the dihydropyridine ring and the substrate in an initial cyclic
transition state that is flexible enough to accommodate all of the
observed mechanisms.
which contributes to lowering their energies (Eq. (4)) [4].
2. Experimental
Because of a lack of evidence for covalent intermediates in
NADH reactions, the Ene mechanism for hydride transfer has not
received much consideration in the literature. We now report the
direct NMR observation of covalent intermediates in the reduction
2.1. General
Acrylonitrile (2-propenitrile, 8a) (Sigma–Aldrich) was distilled
while protected from moisture with CaCl2 drying tubes and was
stored under argon until used, tetrahydrofuran (Sigma–Aldrich)
was purified by drying with LiAlH4 and distillation under a positive
argon pressure with protection by CaCl2 drying tubes [6], pyridine
(Sigma–Aldrich) was distilled and stored under argon until used,
of a series of
a,b-unsaturated cyano and carbonyl compounds
(8a–e) by 1,4-di(trimethylsilyl)-1,4-dihydropyridine, 7 (Fig. 1).
Although our model is quite different from NADH, our observations
support the fundamental possibility of involvement of the Ene
mechanism for hydride transfer in dihydropyridine reactions
Fig. 1. 1,4-ditrimethylsilyl-1,4-dihydropyridine, 7 and
a,b-unsaturated cyano and carbonyl electron accepting reagents: acrylonitrile, 8a; methacrylonitrile, 8b; (E,Z)
crotonitrile, 8c; methyl vinyl ketone, 8d; methyl acrylate, 8e.

R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
59
98% chlorotrimethylsilane (Sigma–Aldrich) was distilled and
stored under argon until used. 1,4-Di(trimethysilyl)-1,4-dihydro-
pyridine (7) was synthesized according to the method of Sulzbach
and characterized by 1H NMR [7]. (5.66 d doublet J = 7.4 Hz, 4.25 d
doublet of doublets J = 4.3 and 7.6 Hz, 2.11 d triplet J = 4.3 Hz, 0.09
d singlet, 0.00 d singlet) The air sensitive product was purified by
vacuum distillation and stored under argon. It had minor impuri-
ties of 4-trimethylsilylpryridine, 12 (1H NMR: 7.40 d doublet of
doublets, 8.45 d doublet of doublets, 0.24 d singlet) and a hydrocar-
bon possibly from pump oil. Neither impurity caused any problem
with the reactions or analyses. Other reagents were from Sigma–
Aldrich and used as supplied.
addition of 8a showed a kmax at 265 nm. Acrylonitrile showed no
significant absorption in the region 190–440 nm. Over time the
265 nm absorption band decreased and a band at higher wave-
length appeared. After 24 h of reaction when there was no further
change in the absorption spectrum, it exhibited a kmax at 326 nm.
2.8. Reactions of methyl vinyl ketone(3-buten-2-one) (8d) with 7 in
chloroform-d
The reaction mixture consisted of 200
l
L (0.6
lmol) of 7, 60 lL
(0.7 mol) of methyl vinyl ketone and 500
l
lL chloroform-d. The
mixture was maintained at room temperature for 198 h and mon-
itored periodically by 1H NMR.
2.2. Reactions of 1,4-di(trimethysilyl)-1,4-dihydropyridine (7) with
a,b-unsaturated compounds
2.9. Reactions of methyl vinyl ketone(3-buten-2-one) (8d) with 7 in
acetone-d6
All reactions were carried out in 5 mm NMR tubes under an ar-
gon atmosphere. NMR spectra were obtained at 200 MHz in a Var-
ian XL200 Spectrometer. Ultraviolet–Visible absorption spectra
were obtained in a Perkin–Elmer Lambda 6 spectrometer using
1 cm quartz cells.
The reaction mixture, consisting of 200
l
L (0.6
lmol) of 7, 50 lL
(0.6 mol) of methyl vinyl ketone and 500
l
l
L acetone-d6, was
maintained at room temperature for 139 h and monitored period-
ically by 1H NMR.
2.3. Reactions of acrylonitrile (propenenitrile) (8a) with 7 in
chloroform-d
2.10. Reactions of methyl vinyl ketone(3-buten-2-one) (8d) with 7 in
acetone-d6/H2O
The reaction mixture consisted of 200
lL (0.6 lmol) of 7, 60 lL
The reaction mixture, consisting of 200
(0.6 mol) of methyl vinyl ketone, 500
H2O (0.7
l
L (0.6
l
mol) of 7, 50
l
l
L
L
(0.9 mol) of acrylonitrile and 500 L chloroform-d. The reaction
l
l
l
lL acetone-d6 and 12
mol), was maintained at room temperature for 144 h
mixture was maintained at room temperature and analyzed by
l
1H NMR periodically for 71 h.
and monitored periodically by 1H NMR.
2.4. Reactions of acrylonitrile (propenenitrile) (8a) with 7 in acetone-
d6/H2O
2.11. Reactions of methyl acrylate(methyl propenoate) (8e) with 7 in
acetone-d6/H2O
The reaction mixture consisted of 200
l
L (0.6
l
mol) of 7, 40
lL
The reaction mixture, consisting of 200
l
L (0.6
L acetone-d6 and 22
mol), was maintained at room temperature for 10 min and
l
mol) of 7, 54
lL
(0.6
(0.7
l
l
mol) of acrylonitrile, 500
l
L acetone-d6 and 12 l
L H2O
(0.6
(1.2
l
l
mol) of methyl acrylate, 500
l
lL H2O
mol). The reaction mixture was maintained at 50 °C for
210 h and monitored periodically by 1HNMR.
then heated at 50 °C for 196 h. During the course of the reaction
the mixture was monitored periodically by 1H NMR.
2.5. Reactions of methacrylonitrile (2-methyl-propenenitrile) (8b)
with 7 in chloroform-d
3. Results
The reaction mixture, consisting of 200
lL (0.6 lmol) of 7, 60 lL
All of the reactions of 7 with acceptors substrates, 8a–8e
(Fig. 1), were monitored by 1HNMR allowing direct observation
of intermediates as they formed in the reaction mixture. For each
reaction the initial 1HNMR spectrum for the reaction mixture con-
sisted of a combination of the spectra of 7 and 8 and a small
amount of irremovable impurities, 4-trimethylsilylpyridine, 12,
and an inert hydrocarbon. All reaction NMR spectra not included
as stacked time point sets and all NMR spectral characteristics of
reactants, products and reaction intermediates are provided in
Figs. A1–A6 and Tables A1–A7.
(0.9 mol) of methacrylonitrile 500 L chloroform-d, was main-
l
l
tained at room temperature for 124 h. The mixture was then
heated at 48 °C for an additional 94 h. During the course of the
reaction the mixture was monitored periodically by 1H NMR.
2.6. Reactions of (E,Z) crotononitrile ((E,Z) 2-butenenitrile) (8c) with 7
in chloroform-d
The reaction mixture, consisting of 200
lL (0.6 lmol) of 7, 60 lL
(0.9 mol) of (E,Z) crotononitrile 500 L chloroform-d, was main-
l
l
tained at room temperature for 120 h. The mixture was then
heated at 48 °C for an additional 70 h; total reaction time was
190 h. During the course of the reactions the mixture was moni-
tored periodically by 1HNMR.
3.1. Reactions of a,b unsaturated cyano substrates
3.1.1. In chloroform-d solution
Reaction of 7 with acrylonitrile, 8a, in chloroform-d yields dia-
stereomeric Ene-adducts 9a(1&2) (Fig. 2). As the reaction proceeds
the signals from 7 to 8a decrease, new series of signals for
9a(1&2) continually increase and no other new signals could be ob-
served. Either the formation of 9a(1&2) from 7 to 8a is a concerted
process or any intermediates formed produced concentrations
too low to be detectable by NMR.
The new signals consist of two sets that differ in intensity be-
cause 9a consists of two diastereomeric pairs (9a1 and 9a2) due
to creation of two adjacent stereogenic centers in its formation
2.7. UV analysis of 9a
Compound 9a was produced by reaction of acrylonitrile (8a)
with 7 in THF. The reaction mixture consisted of 5.0 mL (15
of 7, 5.0 mL (83 mol) of acrylonitrile and was magnetically stirred
at room temperature under an argon atmosphere. The reaction
progress was monitored periodically by removal of 2.0 L aliquots
lmol)
l
l
of reaction mixture and diluting them into 3 mL of THF. The spectra
were scanned from 190 to 440 nm. The initial spectrum of 7 before

60
R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
Fig. 2. Reaction of 1,4-di(trimethylsilyl)-1,4-dihydropryridine, 7, with acrylonitrile, 8a, in chloroform-d. Spectra at different time points: starting materials 7 and 8a and
irremovable impurities 12 and (I) from 7 are labeled in the first time point. Products 9a1 and 9a2 are labeled at times when they were first characterizable. Chemical shifts: 7:
H(a&e) 5.66d d, H(b&d) 4.25d dd, Hc 2.11d t; 8a: H2 5.63 d dd, H3cis 6.04d dd, H3trans 6.19d dd, I: 1.24d s; 9a1: Ha 3.81d dd, Hb 5.56d d, Hd 5.37d d, He 6.09d d, H2 2.78d dq, H3
1.12d d; 9a2: Ha 3.86d dd, Hb ꢀ5.5d d, Hd 5.29d d, He 6.20d d, H2 2.90d dq, H3 1.14d d; 12: H(a&e) 8.45d dd, H(b&d) 7.40d dd. (See Tables A1–A3 for coupling constants.)
(Fig. 2, positions a and 2). Product 9a1 was first barely detected
after 30 min of reaction and was characterizable after 4 h. The less
intense set of peaks associated with 9a2 was detectable after 4 h
and was characterizable after 17 h. The 3.86 d and 2.90 d signals
of 9a2 partially overlap with the 9a1 signals at 3.81 d and 2.78 d,
and the 9a2 signal in the 5.5 d region is likely obscured by the
5.56 d signal of 9a1. Similar newly formed chiral centers were also
detected in reaction products from the reactions of 7 with 8b and
8c (Figs. A1 and A2 and Table A3; see 9b and 9c).
The 9a diastereomers were formed in a ratio of approximately
9a1:9a2 = 5:1 based on the relative integrations of the 1.14 and
1.12 d signals. After 17 h, the spectrum of 7 was no longer detect-
able and there were no further changes in the reaction spectrum
for the additional 54 h that the reaction was monitored. Through-
out the course of the reaction, the 1HNMR signals from 12 did not
change in intensity. Compound 12 is a decomposition product of
both 7 and 9a (See Section 3.1.2), so the lack of change in its signals
is a further indication that 7 reacts only with 8a and that 9a is sta-
ble under the reaction conditions. Reactions of 7 with methacrylo-
nitrile (2-methylpropenenitrile), 8b, and crotononitrile (a mixture
of E- and Z-2-butenenitrile), 8c, produced results very similar to
that of 7 with 8a but differing in the rate of the reactions (Appendix
1.1 and 1.2). The relative rate of appearance of the Ene products
were 9a > 9b > 9c judged from the relative times of appearance
of 9 intermediates in Fig. 2, A1, and A2 and the relative integrations
of the terminal methyl groups of 9a, 9b and 9c. (Fig. A3) All cyano
products studied, 9a–c, are stable in chloroform-d solution. Similar
results were obtained when reaction were run in acetone-d6.
first observed after 15 min of reaction and were intense enough to
be characterized after 2 h. These first new intermediate spectra
match those from products from the slower reaction of 7 and 8a
in chloroform-d or acetone-d6 in the absence of water (See Sec-
tion 3.1.1 and Tables A1–A5). After 4 h reaction in acetone-d6/
H2O a second new set of signals, 10a, was detected and character-
izable after 40 h (Fig. 3 and Table A4). Intermediate 10a is formed
by the loss of the trimethylsilyl group from the nitrogen of 9a1&2
.
From 7 to 40 h the reactant signals continued to decrease; interme-
diate signals from 9a1&2 began to decrease, while signals from 12
increased and those for 14a were detected. (Fig. 3 and Table A5)
After 106 h of reaction the reactant 1HNMR spectra and those of
9a1&2 were no longer detectable. After 210 h the spectrum of 10a
was no longer detectable. So the sequence of appearance and
disappearance of 1HNMR signals supports the reaction sequence
summarized in Fig. 3, steps 7 + 8a ? 9a(1&2) ? 10a ? 12 + 14a.
Apparently the increased availability of acidic protons and/or more
polar solvent facilitated both the formation and decomposition of
9a1&2 which are stable under reaction conditions in chloroform-
d. (See Section 3.1.1).
3.2. Reactions of a,b unsaturated carbonyl substrates
3.2.1. Methyl vinyl ketone
As shown in Fig. 4, 1HNMR monitoring of 7 reacting with
methyl vinyl ketone(3-butene-2-one) (8d) in chloroform-d first
showed formation of 9d (detected at 5 min. and characterized at
24 min.) followed by the simultaneous increase of 12 and appear-
ance of 13d1 (detected at 24 min. and characterized at 50 min.),
then 13d2 (detected at 50 min. and characterized at 105 min.)
and finally appearance of 14d (detected at 4 h and characterized
at 198 h).
3.1.2. In acetone-d6/H2O solution
When water is present in the reaction of 7 with 8a in acetone-d6
(Fig. 3), the first detectable intermediate signals, 9a1 and 9a2, were

R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
61
Fig. 3. Reaction of 1,4-di(trimethylsilyl)-1,4-dihydropryridine, 7, with acrylonitrile, 8a, in acetone-d6/H2O: starting material 7 and 8d and irremovable impurities 12 and (I)
from 7 are labeled in the first time point. Intermediates and products are labeled at times when they were first characterizable. Chemical shifts: 7: H(a&e) 5.66d d, H(b&d)
4.25d dd, Hc 2.11d t; 8a: H2 5.84d dd, H3cis 6.12d dd, H3trans 6.23d dd, I: 1.24d s; water: 2.90d s; 9a1: Ha 3.90d dd, Hb 5.53d d, Hd 5.35d d, He 6.17d d, H2 2.74d dq, H3 1.11d d;
9a2: Ha 4.08d dd, Hb 5.53d d, Hd 5.38d m, He 6.18d d, H2 2.88d m, H3 1.20d d; 10a: Ha 4.08d dd, Hb 5.37d d, Hd 4.76d d, He 6.17d d, H2 2.73d dq, H3 1.19d d; 12: H(a&e) 8.45d
dd, H(b&d) 7.40d dd; 14a H2 2.37d q, H3 1.17d t. (See Tables A1-A5 for coupling constants.)
Intermediates 13d(1&2) are formed as a mixture of E–Z-isomers
with distinct 1HNMR spectra as indicated (Fig. 4). The E–Z-isomers
were formed in a ratio of approximately 13d1:13d2 = 4:1 based on
the relative integrations of the 1.71 d and 1.68 d signals. The forma-
tion of 13d1 and 13d2 from 9d may occur in one-step because of
the possibility of a six-atom cyclic reaction step that concertedly
transfers the trimethylsilyl group from the ring nitrogen to the
substituent oxygen. (Scheme A1) Alternatively, conversion of 9d
to 13d + 12 may be a stepwise process with 10d as an intermediate
present at a steady state concentration too low to be observed by
1HNMR. (Scheme A2) Intermediates 13d(1&2) progress slowly to
14d in chloroform-d solution. Intermediate 9d was no longer
detectable by 1HNMR after 4 h of reaction, and intermediate 13d
was still the major product when reaction monitoring was termi-
nated at 198 h. The sequence of appearance of intermediates (9d,
then 12 and 13d1, followed by 13d2 and then 14d) support the
reaction sequence 7 + 8d ? 9d ? 12 + 13d ? 14d illustrated in
Fig. 4.
The reaction of 7 with 8d in acetone-d6 without added water
(Fig. A4 and Tables A5–A7) progressed very similarly but a bit more
slowly than the reaction in chloroform-d: 9d signals detected with-
in 40 min, those for 12 and 13d(1&2) appeared after 2 h, 14d signals
detected after 10 h, however, the chemical shifts and coupling con-
stants for individual peaks were slightly different (Tables A4–A7,
acetone-d6). As with the reaction in chloroform-d, formation of
14d in acetone-d6 was sufficiently slow that 13d(1&2) were still
the major species present after 139 h of reaction. So the reaction
path for 8d with 7 in acetone-d6 is the same as that in chloro-
form-d. (Fig. 4: 7 + 8d ? 9d ? 12 + 13d ? 14d).
As with reactions of cyano substrates, addition of water to the
reaction of 8d with 7 in acetone-d6 led to a significantly faster pro-
cess compared to their reactions in chloroform-d or acetone-d6
alone (Fig. 4, A4 and A5). After 15 min. of reaction in the presence
of water, 9d was barely detectable from its two major signals at 2.0
and 0.89 d and within 40 min of mixing, signals for 13d(1&2), 12 (in-
crease) and 14d were detected simultaneously. (Fig. A5 and Tables
A5–A7) For the first 48 h of the reaction, the signals for 13d(1&2) in-
creased slowly as the signals for 12 and 14d increased more rapidly
and those for 7, 8d and H2O decreased. After 48 h of reaction the
signals for 13d(1&2) began to decrease in intensity as signals for
12 and 14d continued to increase. While intermediate spectra were
less intense than for reactions in the absence of water, all aspects
for the reaction of 8d with 7 in the presence of added water are
consistent with the mechanism observed in chloroform-d and ace-
tone-d6 in the absence of water, 7 + 8d ? 9d ? 12 + 13d ? 14d
(Fig. 4 and Tables A5–A7). The lower intensities of NMR spectra
of 9d and 13d(1&2) and the earlier appearance of the product,
14d, suggest that the rates of decomposition of 9d and 13d(1&2)
are more rapid in the presence of water than in its absence
(Figs. A4 and A5).
3.2.2. Methyl acrylate
The first detectable intermediate in the reaction of methyl acry-
late(methyl propenoate) 8e with 7 in acetone-d6/H2O (Fig. 5) is 9e,
detected after 1 h of reaction. As with 9a, two diastereomers of
9e(1&2) were characterized. Three hours after mixing, the intensi-
ties of the 9e(1&2) signals had increased, signals from methyl pro-
panoate, 14e, were detected and the signals for 12 had begun to

62
R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
Fig. 4. Reaction of 1,4-ditrimethylsilyl-1,4-dihydropryridine, 7, with methyl vinyl ketone, 8d, in chloroform-d: starting material 7 and 8d and irremovable impurities 12 and
(I) from 7 are labeled in the first time point. Intermediates and products are labeled at times when they were first characterizable. Chemical shifts: 7, 12 and I (See Fig. 2) 8d:
H1 2.24 d s, H2 5.85 d dd, H3cis 6.30d dd, H3trans 6.14d dd; 9d: Ha 3.12d dd, Hb 5.32d d, Hc 3.92d d, Hd 6.13d d, H1 2.05d d, H2 2.91d dq, H3 0.90d d; 13d1 H1 1.45d dq, H2 4.45d
qq, H3 1.71d dq; 13d2 H1 1.47d dq, H2 4.63d qq, H3 1.68d dq; 14d H1 2.07d s, H2 2.45d q, H3 0.90d t. (See Tables A1–A3 and A5–A7 for coupling constants.)
increase. The intensities of the 9e(1&2) signals reached a maximum
after 36 h of reaction and then began to decrease. The intensities of
the signals from 14e increased until the signals from 7, 8e and
9e(1&2) were no longer detectable (172 h of reaction). The reaction
of 8e with 7 showed no 1HNMR evidence for a 13e intermediate
which would be expected to exhibit signals in the region of 1–2
d and 4–5 d as did 13d. (Fig. 4) So the reaction either follows the
mechanistic sequence 9e(1&2) ? 12 + 14e as illustrated in Fig. 5 or
the sequence 9e(1&2) ? 12 + 13e ? 14e where decomposition of
13e is fast enough that it does not accumulate sufficiently to be de-
tected by 1HNMR.
Rates of decomposition of 9 intermediates under similar condi-
tions seem to follow the order 9d > 9e > 9a (Relative times of dis-
appearance of 9 intermediates in Figs. 3 and 5 and A5, and
progress curves of formation of final products in Fig. A6). Regard-
less of the pathway for decomposition of 9, at some point the
CAC bond between the acceptor substrate and pyridine ring must
break. The order of reactivity of the intermediates in this study,
increasing with decreasing pKa value of the product
gests that the CAC bond breaks heterolytically leaving extra elec-
tron density on the -atom of the product. Consequently, a
a-atom, sug-
a
greater ability of the acceptor substrate to accommodate the ex-
cess electron density should increase the rate of decomposition
of 9. If the atom attached to the pyridine ring in 9 was oxygen,
as in many NADH dehydrogenases studied, the rate of decomposi-
tion of 9 should be very fast, making the steady-state concentra-
tion of 9 very low. Thus, it is not surprising that, even if they do
form, analogous Ene intermediates have not been observed in bio-
logical systems before now. The data presented here suggest that if
such intermediates do form in enzymatic reactions their concen-
trations should be very low.
4. Discussion
All of our data are consistent with the general Ene mechanism
presented in Fig. 6. All reactions of 7 with various acceptors, 8a–
e (Fig. 1) involve an initial Ene reaction (Step A in Fig. 6) forming
Ene adducts, 9a–e. With varying substrate structure or reaction
conditions, there are three observable decomposition pathways
for 9, Fig. 6 paths B + E (See Fig. 3), C + F (see Fig. 4), or D (see
Fig. 5).
The relative rates of formation of all 9 intermediates seem to in-
crease as solvent polarity is increased with addition of water to the
solvent. (see Sections 3.1.2 and 3.2.1) Recent theoretical studies of
Retro-Ene reactions of allyl ethers and thioethers, the reverse of
Fig. 6, Step A, indicate that they employ an asynchronous concerted
mechanism [8]. The transition states of these reactions have some
separation of charges. So increases in solvent polarity should have
a larger energy lowering effect on the transition states than on the
neutral reactants leading to a rate increase.
The chemical shifts in the 1HNMR spectra of the ring protons in
the first observable intermediates, 9a–e, in the reactions of 8a–e
with 7, vary only slightly with changes in substrate structure
(Tables A3 and 7A). Our spectra for 9a are consistent with the
1HNMR spectrum reported by Sulzbach and Iqbal for 9a (5.9 d dou-
blet, 5.7 d doublet, 5.4 d doublet, 3.6 d doublet of doublets, 2.6 d
multiplet) [9]. It is clear that all of our reactions form an initial
Ene adduct. However, subsequent reactions of 9a,d,e vary with
changes in substrate structure and reaction conditions. Our ability
to observe intermediate 9 indicates that step A is initially faster
than any of the modes of decomposition of 9, otherwise spectra

R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
63
Fig. 5. Reaction of methyl acrylate, 8e, with 1,4-ditrimethylsilyl-1,4-dihydropryridine, 7, in acetone-d6/water: Starting material 7 and 8e and irremovable impurities 12 and
(I) from 7 are labeled in the first time point. Intermediates and products are labeled at times when they were first characterizable. Chemical shifts: 7: H(a&e) 5.66d d, H(b&d)
4.25d dd, Hc 2.11d t; water: 3.25 d s, 8e: H1 3.65 d s, H2 5.80d dd, H3cis 6.29d dd, H3trans 6.07d dd; 9e1: Ha 3.92d dd, Hb 5.38d d, Hd 5.33d dd, He 6.23d d, H1 3.63d s, H2 2.84d
dq, H3 1.04d d; 9e2: Ha 3.89d dd, Hb 5.39d d, Hd 5.34d dd, He 6.20d d, H1 3.64d s, H2 2.7–2.9d m, H3 1.02d d; 12: H(a&e) 8.45d dd, H(b&d) 7.40d dd; I: 1.24d s; 14e H1 3.60d s,
H2 2.31d q, H3 1.09d t. (See Tables A1, A2, A5 and A6 for coupling constants.)
Fig. 6. General mechanistic scheme for the reaction of 1,4-di(trimethylsilyl)-1,4-dihydropyridine, 7, with acceptor substrates 8a–e. 8a: R1 = R2 = H, Q = CN; 8b: R1 = CH3
R2 = H, Q = CN; 8c: R1 = H, R2 = CH3, Q = CN; 8d: R1 = R2 = H, Q = COCH3; 8e: R1 = R2 = H, Q = COOCH3.
of 9 would not have been detectable. So all of our reactions involve
a covalent intermediate in the dihydropyridine reduction process.
Previous NADH redox model studies have provided examples
singlet electron transfer (SET) and apparently concerted (no
detectable intermediate) mechanisms [2,8,10–14]. Our identifica-
tion of covalent intermediates, 9a,d,e, completes the spectrum of
observed mechanisms with NADH model reactions. Results from
reactions of dihydropyridines and a wide range of model com-
pounds seem to involve one of these mechanisms. We suggest that
a general initial Ene-shaped transition state structure could accom-
modate all three mechanisms for transferring a hydride equivalent.
Scheme 1 describes the general shape of a transition state that is
consistent with the model reactions, theoretical calculations and
biological data.

64
R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
Scheme 1. Ene-shaped dehydrogenase transition state structure.
The cyclic transition state shape provides a six atom delocalized
From their ab initio calculations on the reaction between
methyleniminium ion and dihydropyridine, Wu and Houk pro-
posed a transition state model that involved interaction of the p⁄
structure which is aromatic. The aromaticity lowers the transition
state energies for these reactions compared with transition states
that lack the cyclic structure. We suggest that, depending on the
structures of reactants and any catalyst, transition states would
vary in the distances labeled a, b and c in Scheme 1. Variations in
these distances could result in any of the three observed reaction
paths from a common transition framework.
From a theoretical investigation of reactions of dihydropyri-
dines, Inagaki and Hirabayashi [10] concluded that the key orbital
interactions in hydrogen transfer reactions between dihydropyri-
orbitals of the C3AC4 bond of NAD+ and the
p
orbital of the accep-
tor molecule [12]. They treated the hydrogen as a hydride ion and
described its interactions with the bonds between C4AC3 on
p
NAD+ and CAN of the iminium ion. Although they did not mention
involvement of the orbital associated with C2 or C6 of NAD+, the
geometry of their lowest energy syn transition state places the
reacting atoms in a cyclic orientation similar to that suggested in
Scheme 1 with a relatively long distance represented by c. More re-
cent ab initio work on the transition state for hydride transfer be-
tween dihydropyridine and pyridinium ion also yielded a syn
type transition state in which orbital interactions between the
two rings were cited as factors that lower the transition state en-
ergy [13]. Schiott, Zheng and Bruice did calculations using a contin-
uum solvation model for the hydride transfer from formate ion to
NAD+ as a model for the reaction catalyzed by formate dehydroge-
nase. Their calculated 15N isotope effects agreed with experimental
values, however, they reported that the transition states for addi-
tion of formate ion to the 2, 4 or 6 position of the NAD+ ring had
lower energies than that for the hydride transfer reaction [14].
They concluded that the enzyme must orient the substrate to pre-
vent this competing reaction. Addition of formate ion to the 2 or 6
position of the ring is the first step in the Ene mechanism (Eq. (2)).
There have been many studies of both primary and secondary
isotope effects on nicotinamide dependent enzymatic reactions
[15]. The lack of a secondary deuterium isotope effect for 2-deute-
ro-NAD+ with alcohol dehydrogenase suggests that the 2 position
of the NAD ring does not change its bonding significantly in the
reaction [16]. In our general Ene-type transition state, the sub-
strate oxygen would have to be interacting with C6 of the nicotin-
amide ring as in intermediate 6 of Eq. (2). Thus, for the cyclic
transition state to be involved in enzymatic reactions, the oxygen
atom of the alcohol substrate must be positioned so that it is able
to interact with C6 of the NAD ring. Recent X-ray crystallographic
data on alcohol dehydrogenase ternary complexes provides some
dines and acceptor
p-bonds are those between the p orbital of
the C2AC3 (or C5AC6) bond in the dihydropyridine and the p⁄
orbital of the acceptor molecule. Their model indicates that as
the acceptor p⁄ orbital energy is lowered there is a more favorable
interaction with the dihydropyridine
p orbital; such an interaction
would favor an Ene mechanism. They further suggested that suffi-
cient lowering of the acceptor p⁄ orbital energy could lead to a con-
certed two-electron transfer between the two
p systems in a single
step. Although they did not consider the effects of changes in the
p
orbital of the dihydropyridine, it would follow that raising the en-
ergy of that orbital should favor the formation of an Ene adduct.
Since the two trimethylsilyl groups in 7 are more electron donating
to the dihydropyridine ring than the corresponding substituents in
NADH, they should raise the energy of the
p orbitals in 7 relative to
those in NADH favoring the formation of Ene adducts. Our results
prove that, under appropriate conditions, the dihydropyridine ring
can support the type of Ene reaction chemistry suggested by
Hamilton.
Yasui and Ohno have suggested that a solution to the dichotomy
of concerted vs. multi-step reaction sequences for these reactions
might be to consider a range of reaction paths with varying ener-
gies of potential intermediates and transition states [11]. Our pro-
posal provides a structural shape of an initial transition state that
meets the variability described conceptually by Inagaki and Hira-
bayashi and the mechanistic flexibility proposed by Yasui and
Ohno [10,11].

R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
65
insight [17]. With a series of substrate analogs, all of the structures
of ternary complexes of enzyme, nicotinamide and substrate ana-
logs that place the alcohol carbon atom within 4.0 Å of C4 of the
nicotinamide ring also place the substrate analog oxygen over
the ring and within 4.0 Å C6 of the ring.2 Very little movement
would be required for the alcohol oxygen to be in position to bond
or at least have significant orbital interaction with C6 of NAD+. Crys-
tal structures of complexes of lactate and malate dehydrogenases
[18] also place substrate analog oxygen atoms over the NAD ring.
With isocitrate dehydrogenase one structure shows the substrate
analog oxygen oriented away from the NADP+ ring, while another
places the oxygen closer to C6 of the NADP+ than is the hydrogen
to be transferred to C4 [19]. (See Appendix A2 for additional crystal
data.) It is clear that the crystal structures of ternary complexes are
just pictures of relatively stable orientations of substrates on the en-
zyme surfaces and do not necessarily represent transition state
structures, however the majority of the available data does not rule
out the possibility of our proposed initial, general, cyclic Ene-shaped
transition state structure in these nicotinamide dehydrogenases.
Recently there has been much work using molecular dynamics
in conjunction with semi-empirical and ab initio approaches to ex-
plore pathways of enzymatic catalysis and determine transition
state structures on enzyme surfaces [20]. These studies use exper-
imentally determined enzyme complex structures as starting
points for their calculations of reaction pathways and transition
state structures. Most studies have focused on the distances be-
tween the hydrogen nucleus being transferred and its adjacent car-
bon atoms as well as the CAHAC bond angles in the transition
states. Consequently, little information has been reported on the
distance between the substrate oxygen and either C2 or C6 of the
NAD ring. However, the two-dimensional diagrams of transition
states reported appear to be consistent with cyclic Ene-shaped
transition states (Scheme 1) in which the relative distances for
the interactions at a, b and c vary.
documentation of a covalent intermediate in a 1,4-dihydropyridine
redox reaction and provide an example of the Ene mechanism for
nicotinamide reactions first proposed by Hamilton. We do not sus-
pect that all nicotinamide redox reactions involve an Ene interme-
diate, but we do suggest that this demonstration of the ability of
dihydropyridines to support Ene reactions provides a more general
model for a versatile initial Ene-shaped transition state structure
(Scheme 1) which could account for the variety of observed mech-
anistic paths in nicotinamide model reactions and is consistent
with the vast majority of the available experimental data and the-
oretical studies on enzymatic systems. The aromatic character of
this cyclic transition state structure also provides a theoretically
sound explanation for the high rates of the chemical steps of many
nicotinamide dehydrogenases. Our proposal changes the concept
of the shape of the dehydrogenase transition state and provides a
new model for design of potential tight binding inhibitors for
NAD dehydrogenases.
Acknowledgment
We thank professor Carl Salter, Moravian College Chemistry
Department for many important suggestions in the preparation
of this manuscript.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bioorg.2011.10.002.
References
[1] F.H. Westheimer, H.F. Fisher, E.E. Conn, B. Vennesland, Enzymatic transfer of
hydrogen to DPN, J. Am. Chem. Soc. 73 (1951) 2403.
[2] (a) F.H. Westheimer, Mechanism of action of the pyridine nucleotides, in: D.
Dolphin, O. Avramovic, R. Poulson (Eds.), Pyridine Nucleotide Coenzymes:
Chemical, Biochemical, and Medical Aspects, Part A, John Wiley & Sons, New
York, 1987, pp. 253–322;
Finally, Bull et al. reported that Finasteride, an a,b-unsaturated
amide, which is a mechanism based slow binding inhibitor of ste-
roid 5a-reductase, forms an adduct with NADH as a consequence of
its inhibition of the enzymatic reaction [21]. They isolated the ad-
duct and proposed that it consists of the inhibitor molecule cova-
lently bonded from it’s a-carbon to C4 of NADH. They supported
their structural characterization with mass spectral data showing
the presence of the masses of both NADH and Finasteride in the ad-
duct, but their assignment of the position of attachment is based
on the similarity of its kmax, 326 nm, to that of the C4 cyanide ad-
duct of NADH, reported as ‘‘in the region of 340 ml’’ [22]. How-
ever, with the correspondence of the Finasteride adduct kmax to
that of 9a, 326 nm (see Section 2.7), their data seem equally consis-
tent with attachment of the Finasteride a-carbon to the 6-position
(b) N.J. Oppenheimer, A.L. Handlon, third ed., in: D. Sigman (Ed.), Mechanism
of NAD-dependent enzymes, The Enzymes, vol. 20, Academic Press, Inc., San
Diego, CA, 1992, pp. 453–505;
(c) S. Yasui, M. Okamura, M. Fujii, Multistep mechanism for ‘‘hydride’’ transfers
in the redox reactions induced by NAD(P)H-and NAD(P)+-models, Rev. Het.
Chem. 20 (1999) 145–165.
[3] G.A. Hamilton, Proton in biological redox reactions, Prog. Bioorg. Chem. 1
(1971) 83–157.
[4] (a) L.J. Rodriguez, A.B. Vidal, R.E. Izquierdo, J.R. Fermin, R. Anez, Activation
energy calculation for retro-ene elimination reaction of propylene from diallyl
ether, Theochem 769 (2006) 211–216;
(b) A. Viola, J.J. Collins, N. Filipp, Intramolecular pericyclic reaction of
acetylenes. Part 9. Intramolecular pericyclic reactions of acetylene
compounds, Tetrahedron 37 (1981) 3765–3811. and references therein;
(c) T. Ibuki, Y. Takezaki, Unimolecular decomposition of chemically activated
methyl allyl ether, Int. J. Chem. Kinect. 9 (1977) 201–213;
of the NADH. Thus, the isolated adduct could be a dead end Ene ad-
duct formed in an Ene reaction of NADH and Finasteride. Our data
suggest that the rate of decomposition of the Ene-adducts, 9a,d,e,
vary inversely with the pKa of the product (see Section 4); it is
not surprising that an Ene adduct between NAD and Finasteride
would form and decompose very slowly. Binding to the enzyme
might also stabilize the adduct producing an effective tight binding
inhibitor of the enzyme.
(d) P. Vitins, K.W. Egger, Thermochemical kinetics of retro-ene reactions of
molecules with the general structure (allyl)XYH in the gas phase. IX.
Unimolecular thermal decomposition of diallyl ether, J. Chem. Soc. Perkin II
11 (1974) 1292–1293;
(e) K. W Egger, P. Vitins, Thermochemical kinetics of retro ene reactions of
molecules with the general structure (allyl)XYH in the gas phase. IX. Thermal
unimolecular decomposition of ethyl allyl ether in the gas phase, Int. J. Chem.
Kinect. 6 (1974) 429–435.
[5] K. Alder, F. Pascher, A. Schmitz, Substituting additions. I. Addition of maleic
anhydride and azodicarboxylic esters to singly unsaturated hydrocarbons.
Substitution processes in the allyl position, Chem. Ber. 76 (1943) 27–53.
[6] L.F. Fieser, M. Fieser, Reagents for Organic Synthesis, vol. 1, John Wiley & Sons,
Inc., 1967. pp. 1140.
5. Conclusions
[7] R.A. Sulzbach, Synthesis and properties or 1,4-bis(trimethylsilyl)-1,4-
dihydropyridine, J. Organomet. Chem. 24 (1970) 307–314.
[8] (a) M. Izadyar, A.H. Jahangir, M.R. Gholami, DFT calculations on retro-ene
reactions part I: allyl n-butyl sulfide pyrolysis in the gas phase, J. Chem. Res. 9
(2004) 585–588;
Although our 1,4-ditrimethylsilyl-1,4-dihydropyridine model
differs significantly from NADH, it contains 1,4-dihydropyridine
ring that reduces organic substrates. Our results produce the first
(b) M. Izadyar, M.R. Gholami, M. Haghgu, DFT calculations on retro-ene
reactions, part II: allyl n-propyl sulfide pyrolysis in the gas phase, Theochem
686 (2004) 37–42.
2
Distances determined from coordinates in files 1AXE.PDB, 1QV6.PDB, 1QV7.PDB
[9] R.A. Sulzbach, A.F.M. Iqbal, Addition of acrylonitrile to derivatives on 1,4-
dihydropyridine, Angew. Chem. Int. Ed. Eng. 10 (1971) 733.
and 1MGO.PDB.

66
R.D. Libby, R.A. Mehl / Bioorganic Chemistry 40 (2012) 57–66
[10] S. Inagaki, Y. Hirabayashi, Mechanism of NAD(P)H reduction reactions, Bull.
Chem. Soc. Jpn. 50 (1977) 3360–3364.
alcohol dehydrogenases as determined by NMR and X-ray crystallographic,
Biochemistry 41 (2002) 15770–15779.
[11] S. Yasui, A. Ohno, Model studies with nicotinamide derivatives, Bioorg. Chem.
14 (1986) 70–96.
[18] (a) D.B. Wigley, S.J. Gamblin, J.P. Turkenburg, E.J. Dodson, Piontek, H.
Muirhead, J.J. Holbrook, Structure of
a ternary complex of an allosteric
[12] T.-D. Wu, K.N. Houk, Theoretical transition structures for hydride transfer to
methyleneiminium ion from methylamine and dihydropyridine. On the
nonlinearity of hydride transfers, J. Am. Chem. Soc. 109 (1987) 2226–2227.
[13] T.-D. Wu, D.K.W. Lai, K.N. Houk, Transition structures of hydride transfer
reactions of protonated pyridinium ion with 1,4-dihydropyridine and
protonated nicotinamide with 1,4-dihydronicotinamide, J. Am. Chem. Soc.
117 (1995) 4100–4108.
[14] B. Schiott, Y.-J. Zheng, T.C. Bruice, Theoretical investigation of the hydride
transfer from formate to NAD+ and the implications for the catalytic
mechanism of formate dehydrogenase, J. Am. Chem. Soc. 120 (1998) 7192–
7200.
[15] P.F. Cook, B.L. Bertagnolli, Kinetics of pyridine nucleotide utilizing enzymes, in:
D. Dolphin, O. Avramovic, R. Poulson (Eds.), Pyridine Nucleotide Coenzymes:
Chemical, Biochemical, and Medical Aspects, Part A, John Wiley & Sons, New
York, 1987, pp. 405–447.
[16] (a) P.F. Cook, N.J. Oppenheimer, W.W. Cleland, Secondary deuterium and
nitrogen-15 isotope effects in enzyme-catalyzed reactions. Chemical
mechanism of liver alcohol dehydrogenase, Biochemistry 20 (1981) 1817–
1825;
lactate dehydrogenase from Bacillus stearothermophilus at 2.5 Å resolution, J.
Mol. Biol. 223 (1992) 317–335;
(b) M.D. Hall, D.G. Levitt, L.J. Banaszak, Crystal structure of Escherichia coli
malate dehydrogenase. A complex of the apoenzyme and citrate at 1.87 A
resolution, J. Mol. Biol. 226 (1992) 867–882;
(c) A.D. Chapman, A. Cortes, T.R. Dafforn, A.R. Clarke, R.L. Brady, Structural
basis of substrate specificity in malate dehydrogenases: crystal structure of a
ternary complex of porcine cytoplasmic malate dehydrogenase, alpha-
ketomalonate and tetrahydoNAD, J. Mol. Biol. 285 (1999) 703–712.
[19] (a) J.M. Bolduc, D.H. Dyer, W.G. Scott, P. Singer, R.M. Sweet, D.E. Koshland Jr.,
B.L. Stoddard, Mutagenesis and Laue structures of enzyme intermediates:
isocitrate dehydrogenase, Science 268 (1995) 1312–1318;
(b) B.L. Stoddard, A. Dean, D.E. Koshland Jr., Structure of isocitrate
dehydrogenase with isocitrate, nicotinamide adenine dinucleotide
phosphate, and calcium at 2.5-A resolution:
complex, Biochemistry 32 (1993) 9310–9316.
a pseudo-Mechaelis ternary
[20] (a) T.C. Bruice, A view at the millennium: the efficiency of enzymatic catalysis,
Acc. Chem. Res. 35 (2002) 139–148;
(b) S. Hammes-Schiffer, S. Billeter, Hybrid approach for the dynamical
simulation of proton and hydride transfer in solution and proteins, Int. Rev.
Phys. Chem. 20 (2001) 591–616;
(c) J. Villa, A. Warshel, Energetics and dynamics of enzymatic reactions, J. Phys.
Chem. B 105 (2001) 7887–7907.
(b) J.W. Burgner II, N. J Oppenheimer, W.J. Ray, Remote nitrogen-15 isotope
effects on addition of cyanide to NAD, Biochemistry 26 (1987) 91–96;
(c) N. Rotberg, W.W. Cleland, Secondary nitrogen-15 isotope effects on the
reactions catalyzed by alcohol dehydrogenases, Biochemistry 30 (1991) 4068–
4071.
[21] H.G. Bull, M. Barcia-Calvo, S. Andersson, W.F. Baginsky, H.K. Chan, D.E.
Ellsworth, R.R. Miller, R.A. Sterns, R.K. Bakshi, G.H. Rasmusson, R.L. Tolman,
R.W. Myers, J.W. Kozarich, G.S. Harris, Mechanism-based inhibition of human
steroid 5a-reductase by Finasteride: enzyme-catalyzed formation of NADP-
dihydrofinasteride, a potent bisubstrate analog inhibitor, J. Am. Chem. Soc. 118
(1996) 2359–2365.
[17] (a) S. Ramaswamy, H. Eklund, B.V. Plapp, Structure of horse liver alcohol
dehydrogenase complexed with NAD+ and substituted benzyl alcohols,
Biochemistry 33 (1994) 5230–5237;
(b) L.A. LeBrun, D-H. Park, D.-H.S. Ramaswamy, B.V. Plapp, Participation of
histidine-51 in catalysis by horse liver alcohol dehydrogenase, Biochemistry
43 (2004) 3014–3026;
[22] S.P. Colowick, N.O. Kaplan, M.M. Ciotti, The reaction of pyridine nucleotide
with cyanide and its analytical use, J. Biol. Chem. 191 (1951) 447–459.
(c) J.K. Rubach, B.V. Plapp, Mobility of fluorobenzyl alcohols bound to liver