Elemental step thermodynamics of various analogues of indazolium alkaloids to obtaining hydride in acetonitrile

Article abstract of DOI:10.1039/c5ob01715g

A series of analogues of indazolium alkaloids were designed and synthesized. The thermodynamic driving forces of the 6 elemental steps for the analogues of indazolium alkaloids to obtain hydride in acetonitrile were determined using an isothermal titration calorimeter (ITC) and electrochemical methods, respectively. The effects of molecular structure and substituents on the thermodynamic driving forces of the 6 steps were examined. Meanwhile, the oxidation mechanism of NADH coenzyme by indazolium alkaloids was examined using the chemical mimic method. The result shows that the oxidation of NADH coenzyme by indazolium alkaloids in vivo takes place by one-step concerted hydride transfer mechanism.

Full text of DOI:10.1039/c5ob01715g

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Elemental step thermodynamics of various
analogues of indazolium alkaloids to obtaining
hydride in acetonitrile†
Cite this: Org. Biomol. Chem., 2015,
13, 11472
Nan-Ping Lei,*‡ Yan-Hua Fu‡ and Xiao-Qing Zhu*
A series of analogues of indazolium alkaloids were designed and synthesized. The thermodynamic driving
forces of the 6 elemental steps for the analogues of indazolium alkaloids to obtain hydride in acetonitrile
were determined using an isothermal titration calorimeter (ITC) and electrochemical methods, respecti-
vely. The effects of molecular structure and substituents on the thermodynamic driving forces of the 6
steps were examined. Meanwhile, the oxidation mechanism of NADH coenzyme by indazolium alkaloids
was examined using the chemical mimic method. The result shows that the oxidation of NADH coenzyme
by indazolium alkaloids in vivo takes place by one-step concerted hydride transfer mechanism.
Received 14th August 2015,
Accepted 24th September 2015
DOI: 10.1039/c5ob01715g
www.rsc.org/obc
Introduction
Well known as a traditional medicine, the seeds of Nigella
sativa have been extensively studied for their antidiabetic
effect, but the actual bioactive compounds and the molecular
mechanism responsible for this activity had not yet been
well elaborated.1–4 Several indazolium alkaloids were success-
fully isolated from the seeds of Nigella sativa (Scheme 1) and
uncovered to show good antidiabetic effects, and an initiatory
study concerning the molecular mechanism of their antidia-
betic effects showed these indazolium alkaloids could increase
the consumption of glucose by liver hepatocytes through the
activation of AMP-activated protein kinase (AMPK).5 However,
the deeper molecular mechanism about how these alkaloids
activate AMPK was heretofore not investigated. It was revealed
that the activity of mammalian AMPK could be biochemically
regulated by the redox potential of NADH/NAD+, which is
directly related to the ratio of NADH/NAD+, and it was further
clarified that AMPK is activated by NAD+ in a dose-dependent
manner, whereas AMPK is inhibited by NADH.6 A question
might arise whether the activation of AMPK by these alkaloids
is due to the fact that these indazolium alkaloids would act
upon (consume) NADH through chemical reactions. As a
Scheme 1 The structures of several indazolium alkaloids derived from
the seeds of Nigella sativa with antidiabetic effects.5
result, it should be of significance to investigate thoroughly
the possibilities of reactions between these alkaloids and
NADH. It was known that NADH plays a vital role in biochemi-
cal reactions through donating hydride, hydrogen atoms, or
electrons to other substrates,7–12 whilst the cationic motifs of
these indazolium alkaloids tend to exhibit their good abilities
to obtain hydride, hydrogen atoms, electrons (Scheme 2),
etc.,13–18 thus the possible reactions between these alkaloids
and NADH should include hydride transfer, hydrogen atom
transfer and electron transfer, all of which could be assigned
to the possible elemental steps of hydride transfers between
these alkaloids and NADH, as illustrated in Scheme 3. Since
thermodynamics could offer intrinsic criteria in diagnosing
the possibilities of reactions, it might be efficient and compre-
hensive if we could elucidate the possibilities of all the
elemental steps of hydride transfer between these alkaloids
and NADH from the perspective of thermodynamics. In fact,
this issue might be addressed if we turn to the unique thermo-
dynamic tool of “Molecular ID Card” developed in our
group.19–22
The State Key Laboratory of Elemento-Organic Chemistry, Department of Chemistry,
Collaborative Innovation Center of Chemical Science and Engineering, Nankai
University, Tianjin 300071, China. E-mail: leinp@mail.nankai.edu.cn,
xqzhu@nankai.edu.cn
†Electronic supplementary information (ESI) available: Detailed synthetic routes
and preparation procedures of the analogues of indazolium alkaloids; the plots
of thermodynamic parameters ΔHHA(X+), ΔHHA(X•), ΔHPA(X•), E(X+/0) and E(XH+/0
against the Hammett substituent parameter σ. See DOI: 10.1039/c5ob01715g
‡These two authors contributed equally to this paper.
)
11472 | Org. Biomol. Chem., 2015, 13, 11472–11485
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
X+ to obtain hydride, hydrogen atoms and electrons, respecti-
vely; the hydrogen atom affinities [ΔHHA(X•)] and the proton
affinities [ΔHPA(X•)] for their neutral radicals (X•) to obtain the
hydrogen atom and proton, respectively; and the reduction
potentials [E(XH+/0)] for their hydrogen adducts (XH•+) to
obtain electrons. All these thermodynamic driving forces are
very important and desired parameters for chemists not only
to diagnose the reactivities of X+ and their various reaction
intermediates, but to thoroughly elucidate the mechanism of
hydride transfer reactions to X+ via “Molecular ID Cards”.19–22
Based on the structures designed in Scheme 4, we figured out
how these thermodynamic parameters were affected by struc-
tural isomerizations, variations of heteroatoms, and remote
substituents. The possibilities of reactions between ind-
azolium alkaloids and NADH were elucidated using the deter-
mined thermodynamic parameters of elemental steps.
Scheme 2 Possible elementary steps and corresponding driving forces
for the analogues of indazolium alkaloids (X+) to obtain hydride anion.
In addition, since most biochemical processes with the
indazolium alkaloids and NADH coenzyme as a hydride accep-
tor or donor in a living body all take place in the polar organic
regions constructed with enzyme proteins rather than with
pure water, the chemical information of the indazolium alka-
loids as a hydride acceptor or donor in the polar organic
regions constructed with enzyme proteins should be more
important and valuable than that in the pure aqueous solu-
tion. In order to derive the characteristic chemical information
of the indazolium alkaloids as a hydride acceptor in the polar
organic regions constructed with enzyme proteins, in this
work, acetonitrile was chosen as the solvent to imitate the
polar organic regions constructed with enzyme proteins,
because the polarity of acetonitrile (ε = 37.5) is quite close
to that of the peptide bond in proteins (ε = 37.0, 37.8 and 38.3
for HCONMe2, MeCONMe2 and N,N-dimethylbenzamide,
respectively).22
Scheme 3 Possible elemental steps of hydride transfer from NADH to
indazolium alkaloids.
Results
All the analogues of indazolium alkaloids (X+) in Scheme 4
were synthesized according to conventional procedures and
1
the target molecules were identified by H NMR and MS, and
the detailed synthetic routes and procedures are provided in
the ESI.† Hydride affinities of these analogues [ΔHH-A(X+)] in
this work were defined as the enthalpy changes of the reac-
tions of X+ with a free hydride anion in acetonitrile to form
their conjugated amines (XH) [eqn (1) and (2)] at 298 K in
acetonitrile. The determinations of the enthalpy changes of X+
to gain hydride in acetonitrile were performed according to
the following two strategies: (i) for 1+–7+, the hydride affinities
were obtained according to eqn (3)–(8), derived from the
hydride exchange reactions between their conjugated amines
(XH) and 4-acetamido-2,2,6,6-tetramethylpiperidine-1-oxo-
ammonium perchlorate (TEMPO+ClO4−) or trityl perchlorate
(Ph3C+ClO4−) in acetonitrile, respectively. In eqn (3)–(8), ΔHr
were the reaction enthalpy changes, which could be deter-
mined via titration calorimetry (Fig. 1); ΔHH-A(TEMPO+) was
the hydride affinity of TEMPO+ClO4−, which was determined
Scheme 4 Structures of analogues of indazolium alkaloids (X+) exam-
ined in this work (all of them are in the form of perchlorate salts).
In this work, a series of analogues of indazolium alkaloids
(X+) were designed and synthesized (Scheme 4) to mimic the
naturally occurring antidiabetic alkaloids in Scheme 1. The
thermodynamic driving forces of 6 possible elementary steps
for X+ to obtain hydride (Scheme 2) in acetonitrile were deter-
mined, namely, the hydride affinities [ΔHH-A(X+)], hydrogen
atom affinities [ΔHHA(X+)] and reduction potentials [E(X+/0)] for
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 11472–11485 | 11473

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Reaction enthalpy changes of indazolium alkaloid analogues
(X+), together with reduction potentials of X+ and reduction potentials
of their hydrogen adducts (XH•+) in acetonitrile
E(X+/0 a,c,d
)
E(XH+/0 a,c,d
)
Indazolium
analogues (X+)a
ΔHr
CV
OSWV
CV
OSWV
b
1+
2+
3+
4+
−51.9
−49.9
−48.0
−50.5
−1.867
−1.893
−1.813
−1.926
−1.859
−1.871
−1.797
−1.909
0.665
0.643
0.751
0.703
0.491
0.766
0.720
0.511
5+ (a–i)
p-MeO (a)
p-Me (b)
p-H (c)
−47.8
−47.5
−47.0
−46.2
−45.2
−46.5
−47.2
−45.7
−45.5
−55.0
−1.780
−1.778
−1.776
−1.727
−1.699
−1.755
−1.763
−1.720
−1.695
−1.201
−1.759
−1.743
−1.715
−1.675
−1.624
−1.694
−1.726
−1.653
−1.642
−1.162
−0.059
−0.056
−0.040
−0.017
0.008
−0.035
−0.051
−0.010
0.004
−0.089
−0.080
−0.065
−0.045
−0.020
−0.055
−0.072
−0.033
−0.028
0.198
Fig. 1 Isothermal titration calorimetry (ITC) graph of the reaction heat
of
1H
with
4-acetamido-2,2,6,6-tetramethylpiperidine-1-oxo-
ammonium perchlorate (TEMPO+) in acetonitrile at 298 K. Titration was
conducted by adding 10 μL of TEMPO+ (1.0 mM) every 300 s into the
acetonitrile solution containing 1H (ca. 12.0 mM).
p-Cl(d)
p-CF3 (e)
m-MeO (f)
m-Me (g)
m-Cl (h)
m-CF3 (i)
6+
0.244
7+ (a–e)
p-MeO (a)
p-Me (b)
p-H (c)
p-Cl(d)
p-CF3 (e)
8+
to be −105.6 kcal mol−1 23
affinity of Ph3C+ClO4−, which was previously determined to be
−104.6 kcal mol−1 24 (ii) For 8+–11+, since the enthalpy
changes [ΔHH-D (XH)] for their conjugated amines to release
hydride had already been determined by our previous work,
the enthalpy changes for 8+–11+ to gain hydride in acetonitrile,
could be derived by switching the sign of ΔHH-D(XH).25 The
enthalpy changes of these hydride exchange reactions are
listed in Table 1, while the hydride affinities of these alkaloid
analogues are summarized in Table 2.
,
and ΔHH-A(Ph3C+) was the hydride
−19.6
−19.0
−17.9
−16.3
−14.5
−1.073
−1.051
−1.024
−0.983
−0.938
−2.217
−2.055
−1.430
−1.141
−1.055
−1.035
−1.001
−0.955
−0.909
−2.173
−2.024
−1.400
−1.111
0.703
0.713
0.729
0.776
0.809
−0.145
−0.068
0.365
0.595
0.667
0.679
0.703
0.743
0.787
−0.179
−0.103
0.332
0.564
.
9+
10+
11+
a X+ stands for the analogues of indazolium alkaloids in this work,
while X• stands for their neutral radicals. b ΔHr were the reaction
enthalpy changes of eqn (3), (5), and (7) measured by titration
calorimetry in acetonitrile at 298 K, respectively. The data, given in
kcal mol−1, were average values of at least three independent runs. The
ΔHH‐AðXþÞ
298 K; in MeCN
X
þ þ Hꢀ ꢀꢀꢀꢀꢀꢀꢀꢀꢀ! XH
ð1Þ
ð2Þ
reproducibility was estimated to be 0.5 kcal mol−1 c Measured by CV
.
and OSWV methods in acetonitrile at 298 K, the unit in volt vs. Fc+/0
and reproducible to 5 mV or better. The reduction potentials [E(XH+/0)]
of 1H•+–7H•+ were determined by the oxidation potentials [E(XH0/+)] of
1H–7H, since they are equal in value.27,28 d E(X+/0) and E(XH+/0) of 8+,
9+, 10+, and 11+ and their intermediates were derived from our
previous work.25
ΔHH-AðXþÞ ¼ Hf ðXHÞ ꢀ ½Hf ðXþÞ þ HfðHꢀÞꢁ
ð3Þ
ΔHHAðX•Þ ¼ ΔHH-AðXþÞ ꢀ F½EðH0=ꢀÞ ꢀ EðXþ=0Þꢁ
ΔHPAðX•Þ ¼ ΔHHAðXþÞ ꢀ F½EðHþ=0Þ ꢀ EðXHþ=0Þꢁ
The hydrogen atom affinities of X+ were defined as the
enthalpy changes for X+ to gain a hydrogen atom [ΔHHA(X+)] in
acetonitrile at 298 K; the hydrogen atom affinities and proton
affinities of the neutral radical intermediates (X•) were also
defined as the enthalpy changes for X• to gain a hydrogen
atom [ΔHHA(X•)] and to gain a proton [ΔHPA(X•)] in acetonitrile
at 298 K, respectively. These parameters were indicators of
their hydrogen atom accepting abilities or proton accepting
abilities. To derive these parameters, three thermodynamic
ð10Þ
ð11Þ
ΔHH-AðXþÞ ¼ ΔHH-AðTEMPOþÞ ꢀ ΔHr
ð4Þ
ð5Þ
ΔHH-AðXþÞ ¼ ΔHH-AðTEMPOþÞ ꢀ ΔHr
ð6Þ
ð7Þ
ΔHH-AðXþÞ ¼ ΔHH-AðPh3CþÞ ꢀ ΔHr
ð8Þ cycles were constructed following the possible routes of hydride
transfer to these indazolium analogues (Scheme 5). Through
ΔHHAðXþÞ ¼ ΔHH-AðXþÞ ꢀ F½EðH0=ꢀÞ ꢀ EðXHþ=0Þꢁ
ð9Þ
these cycles, eqn (9)–(11) could be derived according to Hess’s
11474 | Org. Biomol. Chem., 2015, 13, 11472–11485
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 2 Enthalpy changes of X+ to accept hydride and to accept
hydrogen atom, as well as the enthalpy changes of X• to accept proton
and to accept hydrogen atom in acetonitrile (kcal mol−1 a,b
)
Indazolium
analogues (X+)
ΔHH-A(X+)c
ΔHHA(X+)d
ΔHPA(X•)d
ΔHHA(X•)d
1+
2+
3+
4+
−53.7
−55.7
−57.6
−55.1
−12.7
−12.2
−15.2
−17.6
−2.3
−2.1
−3.4
−8.4
−70.3
−72.6
−72.8
−72.9
5+ (a–i)
p-MeO (a)
p-Me (b)
p-H (c)
−57.8
−58.1
−58.6
−59.4
−60.4
−59.1
−58.4
−59.9
−60.1
−60.6
−33.6
−33.7
−33.9
−34.2
−34.6
−34.0
−33.8
−34.4
−34.5
−29.8
−21.0
−20.7
−20.2
−19.6
−18.9
−20.0
−20.5
−19.4
−19.1
−3.4
−72.1
−72.1
−71.9
−71.8
−71.6
−71.9
−72.0
−71.8
−71.7
−61.2
p-Cl(d)
p-CF3 (e)
m-MeO (f)
m-Me (g)
m-Cl (h)
m-CF3 (i)
6+
Fig. 2 Cyclic voltammetry (CV) and Osteryoung square wave voltam-
metry (OSWV) of 1+ in deaerated acetonitrile containing 0.1
M
n-Bu4NBF4 as the supporting electrolyte. The full line: CV graph (sweep
rate = 0.1 V s−1), the dashed line: OSWV graph.
7+ (a–e)
p-MeO (a)
p-Me (b)
p-H (c)
p-Cl(d)
p-CF3 (e)
8+ e
−84.7
−85.3
−86.4
−88.0
−89.8
−49.5
−54.1
−73.0
−91.2
−43.1
−43.4
−44.0
−44.6
−45.4
−27.2
−30.2
−39.1
−51.9
−14.2
−14.1
−13.9
−13.5
−13.2
−24.1
−23.7
−18.2
−24.3
−82.8
−83.0
−83.3
−83.8
−84.5
−73.4
−74.6
−79.1
−90.6
9+ e
10+ e
11+ e
a Relative uncertainties were estimated to be smaller than or close to
1.0 kcal mol−1 in each case. b X+ stands for the analogues of
indazolium alkaloids in this work, while X• stands for their neutral
radicals. c ΔHH-A(X+) were estimated from eqn (4), (6) and (8),
respectively. d ΔHHA(X+), ΔHPA(X•) and ΔHHA(X•) were estimated from
eqn (9)–(11), respectively, with a unit of kcal mol−1, taking E(H+/0) =
−2.307 (V vs. Fc+/0), E(H0/−) = −1.137 V (V vs. Fc+/0) (Fc = ferrocene),
and choosing the redox potentials of X+ and XH•+ measured by the
OSWV method (Table 1) as E(X+/0) and E(XH+/0), since the values by
OSWV were proved to be closer to corresponding standard redox
potentials than those by CV.25 e Derived from our previous work.25
Fig. 3 Cyclic voltammetry (CV) and Osteryoung square wave voltam-
metry (OSWV) of 1H in deaerated acetonitrile containing 0.1
M
n-Bu4NBF4 as the supporting electrolyte. The full line: CV graph (sweep
rate = 0.1 V s−1), the dashed line: OSWV graph.
law, where E(X+/0), E(XH+/0), E(H0/−) and E(H+/0) are the stan-
dard redox potentials of X+, XH•+, H− and H+, respectively.
Since E(H0/−) and E(H+/0) could be obtained from the litera-
ture,26 E(X+/0), E(XH+/0) could be determined through electro-
chemical methods (Table 1, Fig. 2 and 3), and ΔHH-A(X+) could
be obtained from the above studies, and the values of
ΔHHA(X+), ΔHHA(X•) and ΔHPA(X•) could easily be calculated
via eqn (9)–(11). These parameters, together with ΔHH-A(X+),
are summarized in Table 2.27,28
Discussion
Hydride affinities of the analogues of indazolium alkaloids
(X+) in acetonitrile
In this study, the hydride affinities of X+ were defined as the
standard state enthalpy changes for X+ to obtain hydride in
acetonitrile at 298 K. From the second column in Table 2, it is
clear that the scales of hydride affinities of 1+–11+ scale almost
Scheme 5 Three thermodynamic cycles constructed on the basis of
the reactions of the analogues of indazolium alkaloids (X+) with hydride
anion (H−).
41 kcal mol−1, ranging from −49.5 to −90.6 kcal mol−1
,
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 11472–11485 | 11475

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 4 Comparison of the hydride affinities of the analogues of indazolium alkaloids (X+).
offering us a library of analogues of indazolium alkaloids with consume energy, which would offset the energy released in the
diverse electrophiles to hydride. For an intuitive comparison, formation of the new C–H bond. In sharp contrast, if the
the hydride affinities of X+ were ranked in the decreasing hydride affinities of these indazolium analogues are compared
order: 8+ > 1+ > 9+ > 4+ > 2+ > 3+ > 5c+ > 6+ > 10+ > 7c+ > 11+, with those of imines (e.g. −40.8 kcal mol−1 for N-benzylidene-
corresponding to the increasing order of electrophilicities of aniline),19 it seems a paradox that the hydride affinities of
these alkaloid analogues to hydride: 8+ < 1+ < 9+ < 4+ < 2+ < 3+ these analogues are more negative than those of the imines by
< 5c+ < 6+ < 10+ < 7c+ < 11+, as illustrated in Fig. 4. To be over 12 kcal mol−1, since at first glance, the hydride transfer to
specific, indazolium ions (1+–5+), benzo[d]imidazolium ions an alkaloid analogue is accompanied by one more energy-
(8+, 9+), and benzo[d]isothiazolium ions (6+) belong to weak consuming loss of aromatization of a heterocycle compared to
electrophiles to hydride. When reducing them to their conju- the hydride transfer to an imine, apart from the common
gated amines, some strong inorganic hydrides like boron energy-releasing formation of one C–H bond and an energy-
hydrides or aluminum hydrides were recommended.16,18,29 In consuming cleavage of the CvN bond. Nevertheless, if
contrast, benzo[d]thiazolium ions (10+), benzo[d]isoxazolium the contribution from the molecular charge is considered, it
ions (7+) and benzo[d]oxazolium ions (11+) belong to the cate- may be concluded that it is more spontaneous for cationic
gory of strong electrophiles to hydride, some well-known alkaloid analogues to accept the hydride than for neutral
organic hydride donors such as dihydrobenzo[d]imidazoles imines to accept the hydride, which is different from the case
(like 8H and 9H) and Hantzsch esters are capable of reducing between indazolium alkaloid analogues and the benzyl
them to their conjugated amines.25,30 Structurally, 5c+ shows a carbon cations mentioned above. This result might suggest to
slightly stronger electrophilicity to the hydride than its iso- us that the cationic indazolium analogues of alkaloids more
meric 9+, which is consistent with the fact that 5c+ was able to readily obtain hydride than their neutral indazole analogues.
slowly grab the hydride from the conjugated amine of 9+ under
Hydrogen atom affinities of the analogues of indazolium
harsh conditions, illustrated as evidence for the reliability of
alkaloids (X+) in acetonitrile
our data.31 Variation of the heteroatom in the structures of
The hydrogen atom affinities of X+ were defined as the stan-
dard state enthalpy changes for X+ to obtain hydrogen atom in
acetonitrile at 298 K. As shown in the third column of Table 2
and Fig. 5, the hydrogen atom affinities of X+ decrease in the
order: 2+ > 1+ > 3+ > 4+ > 8+ > 6+ > 9+ > 5c+ > 10+ > 7c+ > 11+,
corresponding to the increasing order of the electrophilicities
of indazolium alkaloid analogues to the hydrogen atom. Gen-
erally, these analogues are poor electrophiles to hydrogen
atoms, since they are not able to be reduced by some strong
antioxidant reagents such as vitamin E (80.9 kcal mol−1 to
release hydrogen atom), the commercially available BHT (2,6-
di-tert-butyl-4-methylphenol, 81.6 kcal mol−1 to release hydro-
gen atom) and phenothiazine (79.7 kcal mol−1 to release
hydrogen atoms).33–35 When comparing their hydrogen atom
these analogues could lead to a change of the hydride affinity
up to around 37 kcal mol−1, such as for 9+ (−54.1 kcal mol−1
)
vs. 11+ (−91.2 kcal mol−1), indicating that the electrophilicities
of these analogues to hydride could be adjusted in a flexible
manner to meet diverse demands.
If the hydride affinities of these analogues are compared
with those of primary benzyl carbon cations (e.g. −106 kcal
+
mol−1 for 4-MeOPhCH2 ) in acetonitrile,32 it is found the
hydride affinities of these analogues are more positive than
those of the benzyl carbon cations by at least 14 kcal mol−1
.
Unlike the reduction of a benzyl carbon cation in which only
one new C–H bond formed, the reduction of an alkaloid ana-
logue is accompanied by the additional cleavage of one CvN
bond and the loss of aromatization of the heterocycle to
11476 | Org. Biomol. Chem., 2015, 13, 11472–11485
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 5 Comparison of the hydrogen atom affinities of the analogues of indazolium alkaloids (X+).
affinities with their hydride ion affinities (Table 2, column 3 their standard reduction potentials, 6+, 7+, 10+ and 11+ should
vs. column 2), it may be concluded that these analogues are belong to the category of good electrophiles to electrons, while
more readily reduced by the hydride ion than by the hydrogen 1+, 2+, 3+, 4+, 5+, 8 and 9+ are due to weak electrophiles to elec-
atom, thus the hydride transfer is probably not initiated by trons. In living bodies, it should be thermodynamically un-
hydrogen atom transfer. Notably, the hydrogen atom affinities favorable for these alkaloid analogues to be reduced by electrons
of 1+–4+ are much more positive than that of 5c+ by about over from naturally existing reducing reagents like NADH or vitamin
16 kcal mol−1, which is probably due to the fact that the de- C, since the standard oxidation potentials of both NADH (0.280 V
localization effect by the phenyl group at the C(3) position vs. Fc+/0 in neutral water) and of vitamin C (−0.276 V vs. Fc+/0
greatly stabilizes the yielded radical cation of 5c+ after hydro- in neutral water) are more positive than that of the alkaloid
gen atom transfer, which will be discussed vide infra.
analogues by over 1.281 V (equivalent to 29.5 kcal mol−1) and
0.725 V (equivalent to 16.7 kcal mol−1), respectively.19
Electron affinities of the analogues of indazolium alkaloids
As illustrated in Fig. 6, the standard reduction potentials of
these analogues are ranked in the increasing order: 8+ < 9+ < 4+
< 2+ < 1+ < 3+ < 5c+ < 10+ < 6+ < 11+ < 7c+, corresponding to the
increasing order of electron-withdrawing abilities: 8+ < 9+ < 4+
< 2+ < 1+ < 3+ < 5c+ < 10+ < 6+ < 11+ < 7c+. Considering the
effects of structural isomerization, the E(X+/0) of 1+, 5c+, 6+ and
7c+ is more positive than their corresponding isomers 8+, 9+,
10+ and 11+ by more than 0.1 V (equivalent to 2.3 kcal mol−1),
(X+) in acetonitrile
The standard reduction potentials of these analogues of ind-
azolium alkaloids were employed as the indicators of their
electron affinities. From column 4 in Table 1, it is found that
the one-electron reduction potentials of these analogues locate
in a scale ranging from −2.173 to −1.001 V (vs. Fc+/0), or an
energy scope of ca. 27 kcal mol−1. Judging from the values of
Fig. 6 Comparison of the one electron reduction potentials of the analogues of indazolium alkaloids (X+).
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 11472–11485 | 11477

View Article Online
Paper
Organic & Biomolecular Chemistry
implying the electrophilicities of the former ones to electrons 71 kcal mol−1, which means the antioxidant abilities of these
are stronger than their isomers. When examining the effect of analogues might be greatly strengthened after obtaining one
the variation of heteroatoms on the reduction potentials of electron. As a result, after accepting the electron, the ana-
these alkaloid analogues, it is found that altering heteroatoms logues 7+ and 11+ will be scavenged by antioxidant reagents
could result in a change up to 0.9 V (equivalent to 20.7 kcal like vitamin E, BHT, and phenothiazine. By comparing
mol−1), with a similar trend of N < S < O corresponding to the ΔHH-A(X+) with the corresponding ΔHPA(X•), it is evident that
potentials.
electron transfer could convert these analogues from Lewis
acids to bases. Notably, ΔHPA(5c•) is smaller than ΔHPA(1•) by
almost 20 kcal mol−1, which is close to the difference
(ca. 21 kcal mol−1) between ΔHHA(5c+) and ΔHHA(1+), again this
might be attributed to the fact that the yielded 5cH•+ is more
stable than 1H•+, which will be provided further evidence
vide infra. For an intuitive comparison, the hydrogen or proton
obtaining abilities of X• were ranged in order and are shown in
Fig. 7 and 8. Apparently, both the hydrogen atom accepting
abilities and the proton accepting abilities of the neutral
radicals of these alkaloid analogues could be regulated in
an energy scope of over 20 kcal mol−1 through structural
variations.
Hydrogen atom affinities and proton affinities of neutral
radical intermediates (X•) of the analogues of indazolium
alkaloids in acetonitrile
As stated above and shown in Scheme 2, if hydride transfer to
these analogues of indazolium alkaloids is initiated by single
electron transfer, neutral radical intermediates (X•) could be
formed as transient species, which could either grab a hydro-
gen atom to give their conjugated amines (XH), or accept a
proton to generate their corresponding radical cations (XH•+).
Consequently, the thermodynamic driving forces for the two
possible elementary steps of hydride transfer, namely hydro-
gen atom affinities [ΔHHA(X•)] for X• to obtain hydrogen atoms
and proton affinities [ΔHPA(X•)] for X• to obtain protons in
acetonitrile at 298 K, should be vital parameters not only in
Electron affinities of radical cation intermediates (XH•+) of the
analogues of indazolium alkaloids in acetonitrile
diagnosing the chemical activities of X• but in predicting Electron affinities of radical cation intermediates (XH•+) of
whether hydrogen atom transfer or proton transfer to X• is these alkaloid analogues were defined as the one electron
thermodynamically more favorable. From columns 4 and 5 in reduction potentials of XH•+. From column 6 of Table 1, it is
Table 2, it is found that, ΔHHA(X•) locates on a scale from found that reduction potentials of XH•+ range from 0.787 V to
−61.2 to −84.5 kcal mol−1, while ΔHPA(X•) ranges from −2.1 to −0.179 V (vs. Fc+/0), and the one electron reduction potential of
−20.5 kcal mol−1. For each neutral radical, ΔHHA(X•) is much XH•+ decreases in the order: 7cH•+ > 2H•+ > 3H•+ > 1H•+
>
more negative than ΔHPA(X•) by at least 49 kcal mol−1, imply- 11H•+ > 4H•+ > 10H•+ > 6H•+ > 5cH•+ > 9H•+ > 8H•+, corres-
ing that the attack on X• by the hydrogen atom should be ponding to the decreasing order of the electrophilicity of XH•+
thermodynamically much more favorable than that by proton, to electron (Fig. 9). For an intuitive comparison, the one elec-
that is, once the hydride transfer to an alkaloid analogue is tron reduction potentials of XH•+ were ranged in the order
initiated by the single electron transfer, the following step corresponding to their structures, as illustrated in Fig. 9. From
should be a hydrogen atom transfer rather than a proton trans- the perspective of effects of structural variations, it is evident
fer. When ΔHHA(X•) is compared with the corresponding that E(XH+/0) could be adjusted within a scale approximating
ΔHHA(X+) (Table 2, column 5 vs. column 3), it is found that 1.0 V (equivalent to 23.06 kcal mol−1) via isomerizations, vari-
ΔHHA(X•) is much more negative than ΔHHA(X+) by up to ations of heteroatoms or remote substituents. Notably, the one
Fig. 7 Comparison of the hydrogen atom affinities of neutral radical intermediates of indazolium alkaloid analogues (X•).
11478 | Org. Biomol. Chem., 2015, 13, 11472–11485
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 8 Comparison of the proton affinities of the neutral radical intermediates of indazolium alkaloid analogues (X•).
Fig. 9 Comparison of one electron reduction potentials of the radical cation intermediates of indazolium alkaloid analogues (XH•+).
electron reduction potential of 1H•+ (0.643 V) is much larger second (0.1 V s−1) to 1 V s−1, indicating 5cH•+ could exist in
than that of 5cH•+ (−0.065 V), implying a much smaller stabi- solution for a while. As a result, it may be concluded that
lity of 1H•+ than 5cH•+, which is coincident with our deduc- 5cH•+ is more stable than 1H•+, which could respond to the
tions in the preceding paragraphs. For further verification, differences of the other thermodynamic parameters men-
cyclic voltammetry (CV) technology was employed to test the tioned hereinbefore.
relative stabilities of radical cations.36,37 For example, the
radical cation 1H•+ would be generated during the sweeping of
Effects of remote substituents on the enthalpy changes and
reduction potentials
oxidant potential of 1H in anaerobic acetonitrile by CV. If the
newly generated transient species 1H•+ is stable within the
From Tables 1 and 2, it is clear that all the enthalpy change
values of X+ and X• and the reduction potentials of X+ and
sweeping time, a reversible CV graph will be obtained and vice
XH•+ are strongly dependent on the nature of substituents on
versa, so CV can be utilized to semi quantitatively compare the
the phenyl group at the position of C(3) in 5+ and 7+, as well as
relative stability of radical cations. To compare the stabilities
of 1H•+ and 5cH•+, CV sweepings were carried out for 1H and
their reaction intermediates. To elucidate the relationship of
5cH. As illustrated in Fig. 10, the CV graphs of 1H are found to
be irreversible even at a high sweeping rate of 1 V s−1, indicat-
ing 1H•+ could not stably exist in solution during the sweeping
time. In contrast, the CV graphs for 5cH are found to be nearly
reversible in a wide range of sweeping rate, from 0.1 volt per
the substituents with the enthalpy changes and the reduction
potentials, the effects of remote substituents at para- or meta-
position are examined on the ΔHH-A(X+) and ΔHHA(X+), on the
ΔHPA(X•) and ΔHHA(X•), as well as on the E(X+/0) and E(XH+/0),
respectively. The results in Fig. 11 and S1–S5 (ESI†) show that
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 11472–11485 | 11479

View Article Online
Paper
Organic & Biomolecular Chemistry
phenyl ring at the position of the C(3), it was not difficult to
estimate the values of ΔHH-A(X+), ΔHHA(X+), ΔHPA(X•),
ΔHHA(X•), E(X+/0) as well as E(XH+/0) according to eqn (12)–
(23), as long as the corresponding Hammett substituent para-
meters (σp or σm) are available and the standard deviation of
these estimations is less than 0.25 kcal mol−1 for enthalpies
or less than 25 mV for reduction potentials. Also,
these formulae might quantitatively guide us to select species
of the alkaloid analogues with suitable substituents for
special use.
Diagnosing possible reactions between indazolium alkaloids
with NADH via chemical mimics
As mentioned in the Introduction section, several indazolium
alkaloids isolated from the seeds of Nigella sativa (Scheme 1)
showed good effects against diabetes, but the molecular mech-
anism of their antidiabetic effects has not yet been well elabo-
rated to date. The latest study showed that their antidiabetic
effects originate from the activation of AMP-activated protein
kinase (AMPK) by the alkaloids, but the in depth mechanism
of the activation of AMPK by the alkaloids remains uninvesti-
gated. It was revealed that the activity of AMPK could be regu-
lated by the redox potential of NADH/NAD+, which is directly
related to the ratio of NADH/NAD+, and it was further clarified
that AMPK is activated by NAD+ in a dose-dependent manner,
whereas AMPK is inhibited by NADH. We then suspected that
the activation of AMPK by these alkaloids might be due to the
fact that these alkaloids would act upon (consume) NADH
through chemical reactions. Possible reactions between these
alkaloids and NADH should include hydride transfer, hydro-
gen atom transfer, and electron transfer, which could be
assigned to the possible elemental steps of hydride transfer
reactions between these alkaloids and NADH (Scheme 3).
Since thermodynamics could offer us intrinsic criteria in diag-
nosing the possibilities of reactions, it might be efficient and
comprehensive to elucidate the possibilities of all the elemen-
tal steps of hydride transfer between these alkaloids and
NADH from the perspective of thermodynamics. To tackle this
problem, a mimic reaction of hydride transfer is employed,
where BPH and 3+ are chosen as the model compounds of
NADH and of indazolium alkaloids, respectively.40 As verified
by our previous studies, a “Molecular ID Card” might be used
as a unique thermodynamic tool to efficiently diagnose the
possibilities of elemental steps of hydride transfer between 3+
(hydride acceptor) and BPH (hydride donor), provided that
their “Molecular ID Cards” are available. Fortunately, the
“Molecular ID Card” for 3+ to accept hydride could be obtained
by this work, whilst the “Molecular ID Card” for BPH to
release hydride could be derived based on our previous
Fig. 10 CV graphs at different sweeping rates in anaerobic MeCN con-
taining 0.1 M n-Bu4NBF4 as the supporting electrolyte. The upper graph
is the CV graph of 1H, while the lower one is the CV graph of 5cH.
Fig. 11 Plot of ΔHH-A(X+) against Hammett substituent parameter (σ).
all these values are linearly dependent on the Hammett substi- work.41 According to Hess’s Law, it is easy to access the
tuent parameters σp or σm with very good correlation coeffi- thermodynamic driving forces of all possible elemental steps
cients, implying that the Hammett linear free energy of hydride transfer from BPH to 3+, as shown in Scheme 6.
relationship38,39 holds in these chemical and electrochemical Based on the driving forces of six possible elemental
processes. From the slopes and the intercepts of these steps of hydride transfer from BPH to 3+, we can make the fol-
linear correlations, twelve mathematical formulae are derived lowing predictions: (i) hydride transfer (step c) from BPH to
[eqn (12)–(23)]. Evidently, for any para- or meta-position on the 3+ should be spontaneous, since this reaction is thermo-
11480 | Org. Biomol. Chem., 2015, 13, 11472–11485
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 6 Possible elementary steps of hydride transfer from BPH to 3+ and thermodynamic driving forces of each step derived from the “Mole-
cular ID Cards” of BPH and 3+.
ΔHH-AðXþÞ ¼ ꢀ3:30σ ꢀ 58:7 for 5þ
ð12Þ
dynamically favorable by 4.6 kcal mol−1, while both hydrogen
atom transfer (step b) and electron transfer (step a) from BPH
to 3+ are thermodynamically forbidden by more than 41.8 kcal
mol−1. (ii) Even if hydrogen atom transfer or electron transfer
is triggered, the yielding transient ion pairs of BP• and 3H•+ or
BPH•+ and 3H• are not likely to exist for long in solution, and
would react with each other following the possible routes (step
b–step f, step a–step e, or step a–step d–step f) of stepwise
hydride transfer, until they give the same pair of stable
products of BP+ and 3H as produced by the concerted hydride
transfer (step c). Therefore, it might be concluded that, when
indazolium alkaloids encounter NADH in vivo, the hydride
transfer reaction is likely to happen, rather than hydrogen
atom transfer or electron transfer reaction, to give the
corresponding indazoles and NAD+ as products, and the
route of hydride transfer tends to be concerted rather than
stepwise.
ΔHH-AðXþÞ ¼ ꢀ6:37σ ꢀ 86:4 for 7þ
ΔHHAðXþÞ ¼ ꢀ1:29σ ꢀ 33:9 for 5þ
ΔHHAðXþÞ ¼ ꢀ2:84σ ꢀ 43:9 for 7þ
ΔHHAðX•Þ ¼ 0:598σ ꢀ 72:0 for 5•
ΔHHAðX•Þ ¼ ꢀ2:10σ ꢀ 83:3 for 7•
ΔHPAðX•Þ ¼ 2:60σ ꢀ 20:3 for 5•
ΔHPAðX•Þ ¼ 1:28σ ꢀ 13:9 for 7•
ð13Þ
ð14Þ
ð15Þ
ð16Þ
ð17Þ
ð18Þ
ð19Þ
EðXþ=0Þ ¼ 0:1681σ ꢀ 1:7142 for 5þ
EðXþ=0Þ ¼ 0:1821σ ꢀ 1:0030 for 7þ
ð20Þ
ð21Þ
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 11472–11485 | 11481

View Article Online
Paper
Organic & Biomolecular Chemistry
change decreases in the order: 4+ (−2.1 kcal mol−1) > 2+
(−2.7 kcal mol−1) > 3+ (−4.6 kcal mol−1), corresponding to an
increasing driving force of hydride transfer with the change of
the size of bridge ring as follows: seven membered < five mem-
EðXHþ=0Þ ¼ 0:0866σ ꢀ 0:0654 for 5H•þ
EðXHþ=0Þ ¼ 0:1511σ þ 0:7058 for 7H•þ
ð22Þ
ð23Þ
On the basis of the information disclosed above, we pro- bered < six membered, suggesting that the natural preference
posed a molecular mechanism to understand the deeper mole- of the indazolium alkaloids to bear a six membered bridge
cular mechanism of the antidiabetic effects of indazolium ring happens to facilitate hydride transfer from NADH to
alkaloids as follows: when indazolium alkaloids encounter them, which might make a positive contribution to the anti-
NADH in vivo, indazolium alkaloids would oxidize NADH to diabetic effects of these alkaloids via a superior activation of
NAD+ through hydride transfer reactions, leading to a lower AMPK. Besides, when examining the substituents at the C(11)
concentration of NADH and a higher concentration of NAD+, position in the structures of these alkaloids, it was found that
both of which would contribute to the activation of AMPK, fol- either a phenyl group or H might occupy the C(11) position of
lowed by increasing consumption of glucose to cure hyper- these natural alkaloids (Scheme 1). Which substituent might
glycaemia. Additionally, since the driving forces of hydride be favourable for the antidiabetic effects of indazolium alka-
transfer reactions between these alkaloids and NADH are not loids? With 5c+ and 1+ as the corresponding model com-
far from zero, these reactions should be reversible, i.e., their pounds, it is demonstrated that the hydride transfer from BPH
adjustments for the concentrations of NADH, NAD+ and to 5c+ is thermodynamically more favorable than that to 1+ by
the activity of AMPK should not be violent, rendering their almost 5 kcal mol−1 (Table 3), implying that those alkaloids
regulations of the metabolism of glucose mild and not structurally bearing a phenyl group at the C(11) position are
likely to slop over the safe level to avoid the side effect of more ready to grab the hydride from NADH and might show
hypoglycaemia.
better antidiabetic activities than those bearing H at the C(11)
If our proposal about the deeper molecular mechanism of position, which might explain why the antidiabetic activities of
antidiabetic effects of indazolium alkaloids is tenable, it alkaloids C and D were disclosed to be stronger than those of
should also be of importance for us to examine the potential alkaloids A and B in Scheme 1.5 Obviously, these thermo-
impacts of the structures of indazolium alkaloids on their anti- dynamic implications might shed light on screening the suitable
diabetic effects. As illustrated in Scheme 1, the most eye-catch- indazolium alkaloids and their analogues for antidiabetic use.
ing feature about the structures of indazolium alkaloids is that
they all bear a six membered nonaromatic ring bridged to the
core of indazoles. Pharmaceutical chemists might ask whether
the existence of the bridge ring and its size would impact on
Conclusions
the antidiabetic effects of alkaloids by affecting hydride trans- In this study, a series of analogues of indazolium alkaloids
fer reactions between these alkaloids and NADH. To address (X+) were designed and synthesized. The thermodynamic
this issue, we examined the driving force of hydride transfer driving forces of 6 possible elementary steps for X+ to obtain
from BPH to 1+–4+, where 1+ was chosen as the model com- hydride in acetonitrile were determined. Based on these para-
pound of alkaloids whose structure bears no bridge ring but meters, we diagnosed the possible reactions between indazo-
two methyl substituents instead, and 2+, 3+ and 4+ were chosen lium alkaloids and NADH, and proposed
a molecular
as the model compounds structurally bearing a five, six, and mechanism to understand the root of the antidiabetic effects
seven membered bridge ring, respectively. The thermodynamic of these alkaloids. We further evaluated the potential struc-
driving forces of hydride transfer from BPH to these model tural impacts on the antidiabetic effects of indazolium alka-
compounds were calculated and listed in Table 3. As shown in loids. After detailed discussions, we arrived at the following
Table 3, the driving force of hydride transfer from BPH to 1+ is conclusions:
smaller than that from BPH to 2+, 3+ or 4+ by at least 2 kcal
(1) These analogues of indazolium alkaloids are weak to
mol−1, implying that the bridge ring in the structures of inda- strong electrophiles to hydride. 1+–5+, 8+, 9+ and 6+ belong to
zolium alkaloids would make the hydride transfer from NADH weak electrophiles to hydride, and when reducing them to con-
to the alkaloids more spontaneous. In addition, when the jugated amines, some strong inorganic hydrides like boron
enthalpy change of 3+ to accept hydride from BPH was com- hydrides or aluminum hydrides are recommended; 10+, 7+,
pared with that of 2+ and of 4+, it was found that the enthalpy and 11+ belong to strong electrophiles to hydride, and some
organic hydrides like Hantzsch ester and dihydrobenzo[d]imi-
dazoles are capable of reducing them to their conjugated
amines. The cationic indazolium analogues of alkaloids are
predicted to be more capable of obtaining hydride than their
neutral indazole analogues, mainly due to the contribution of
molecular charge.
Table 3 Thermodynamic driving forces (ΔHr(H–T)) of hydride transfer
from the model of NADH (BPH) to the models of indazolium alkaloids
(1+–5c+) in acetonitrile
(2) These analogues are poor electrophiles to hydrogen
atoms, and are not likely to be reduced even by strong antioxi-
dant reagents such as vitamin E, commercially available BHT
Models of alkaloids
1+
2+
3+
4+
5c+
ΔHr(H–T) (kcal mol−1
)
−0.7
−2.7
−4.6
−2.1
−5.6
11482 | Org. Biomol. Chem., 2015, 13, 11472–11485
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
and phenothiazine. Since the hydrogen atom affinities of these at the C(11) position, which might answer the question why
indazolium analogues are much more positive than their the antidiabetic activities of alkaloids C and D were found to
hydride affinities, these analogues are more readily reduced by be stronger than those of alkaloids A and B. Obviously, our
hydride ions than by hydrogen atoms, and hydride transfers results might provide vital implications on the root of anti-
are probably not initiated by hydrogen atom transfers.
diabetic activities of indazolium alkaloids and shed light on
(3) These analogues are weak to good electrophiles to screening suitable indazolium alkaloids or their analogues for
electrons. 6+, 7+, 10+ and 11+ should belong to the category of antidiabetic use.
good electrophiles to electrons, while 1+, 2+, 3+, 4+, 5+, 8+ and
9+ are due to weak electrophiles to electrons. In living bodies,
it should be thermodynamically unfavorable for these ana-
logues to be reduced by an electron from NADH or vitamin C.
Experimental section
Materials
(4) If hydride transfer to these analogues is initiated by
single electron transfer, neutral radical intermediates (X•)
All reagents were of commercial quality from freshly opened
containers or were purified before use. Reagent grade aceto-
could form, which could either grab a hydrogen atom or
accept a proton. For each X•, the attack on X• by a hydrogen
nitrile was refluxed over KMnO4 and K2CO3 for several hours
and was doubly distilled over P2O5 under argon before use.
atom is thermodynamically much more favorable than that by
The commercial tetrabutylammonium hexafluorophosphate
a proton, that is, if hydride attack on these indazolium ana-
(n-Bu4NPF6, Aldrich) was recrystallized from CH2Cl2/ether
and was vacuum-dried at 110 °C overnight before the prepa-
logues is initiated by single electron transfer, the following
step tends to be hydrogen atom transfer rather than proton
ration of the supporting electrolyte solution. 4-Acetamido-
transfer.
2,2,6,6-tetra-methylpiperidine-1-oxo-ammonium
perchlorate
(5) The reduction potentials of the radical cation inter-
mediates of these indazolium analogues range from 0.787 to
−0.179 V, implying these radical cations are unstable and
easily accept the electron. Electrochemical experiments
verified that the stability of 5cH•+ is much larger than that
of 1H•+, which could explain the related differences of the
thermodynamic parameters.
(6) All the thermodynamic parameters could be adjusted in
a flexible manner through variation of the heteroatoms or
structural isomerizations. The good Hammett linear free-
energy relationship held between these thermodynamic para-
meters and the nature of the remote substituents on the
phenyl ring at the C(3) position, and 12 mathematical for-
mulae were derived. These results might guide us to select the
analogues with suitable structures or remote substituents for
applications.
(7) On the basis of the thermodynamic parameters deter-
mined, the oxidation mechanism of NADH coenzyme by inda-
zolium alkaloids in vivo were predicted to take place by a one-
step concerted hydride transfer mechanism; a deeper mole-
cular mechanism of the antidiabetic effects of indazolium
alkaloids was then proposed as follows: when indazolium alka-
loids encounter NADH in vivo, they would oxidize NADH to
NAD+ through hydride transfer reactions, leading to a lower
concentration of NADH and a higher concentration of NAD+,
both of which would activate AMPK and then increase the con-
(TEMPO+ClO4−) and trityl perchlorate (Ph3C+ClO4−) were pre-
pared following operations described in the literature.42,43
Syntheses of the analogues of indazolium alkaloids (1+–7+)
The analogues of indazolium alkaloids 1+–4+,44,45 5+,46 and 6+–
7+ (ref. 47) were synthesized according to the literature,
respectively, and the detailed synthetic routes are provided in
the ESI.† The analogues of indazolium alkaloids (1+–7+) are all
known compounds.
Measurement of redox potentials
The electrochemical experiments were carried out by CV or
OSWV using a BAS-100B electrochemical apparatus in deaer-
ated acetonitrile under an argon atmosphere at 298 K as
described previously.48 0.1 M n-Bu4NPF6 in acetonitrile was
employed as the supporting electrolyte. A standard three-
electrode cell consists of a glassy carbon disk as the working
electrode, a platinum wire as the counter electrode, and 0.1 M
AgNO3/Ag (in 0.1 M n-Bu4NPF6–acetonitrile) as the reference
electrode. The ferrocenium/ferrocene redox couple (Fc+/Fc) was
taken as the internal standard. The reproducibilities of the
potentials were usually ≤5 mV for ionic species and ≤10 mV
for neutral species.
Isothermal titration calorimetry (ITC)
sumption of glucose to cure hyperglycaemia, and their regu- The titration experiments were performed on a CSC4200 iso-
lation of the metabolism of glucose should be mild and not thermal titration calorimeter in acetonitrile at 298 K as
likely to exceed the safe level to avoid hypoglycaemia. The described previously.49 The performance of the calorimeter
natural preference for indazolium alkaloids to bear a six mem- was checked by measuring the standard heat of neutralization
bered bridge ring happens to facilitate hydride transfer from of an aqueous solution of sodium hydroxide with a standard
NADH to them, which might make a positive contribution to aqueous HCl solution. Data points were collected every 2 s.
the antidiabetic effects of these alkaloids; these alkaloids The heat of reaction was determined following 10 automatic
structurally bearing a phenyl group at the C(11) position are injections from a 250 μL injection syringe containing a stan-
more ready to grab the hydride from NADH and accordingly dard solution (2 mM) into the reaction cell (1.30 mL) contain-
might show better antidiabetic activities than those bearing H ing 1 mL of other concentrated reactants (∼12 mM). Injection
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 11472–11485 | 11483

View Article Online
Paper
Organic & Biomolecular Chemistry
volumes (10 μL) were delivered in 0.5 s time intervals with 19 X.-Q. Zhu, Q.-Y. Liu, Q. Chen and L.-R. Mei, J. Org. Chem.,
300–450 s between every two injections. The reaction heat was 2010, 75, 789.
obtained by the integration of each peak except for the first 20 Y. Cao, S.-C. Zhang, M. Zhang, G.-B. Shen and X.-Q. Zhu,
one.
J. Org. Chem., 2013, 78, 7154.
21 S.-M. Si, Y.-H. Fu and X.-Q. Zhu, J. Phys. Chem. C, 2015,
119, 62.
22 K. Xia, G.-B. Shen and X.-Q. Zhu, Org. Biomol. Chem., 2015,
13, 6255.
Acknowledgements
Financial support from the National Natural Science Foun- 23 X.-Q. Zhu, X. Chen and L.-R. Mei, Org. Lett., 2011, 13,
dation of China (Grant No. 21472099, 21390400 and 21102074)
and the 111 Project (B06005) is gratefully acknowledged.
2456.
24 The hydride affinity of Ph3C+ cation in acetonitrile
(−104.6 kcal mol−1) is derived from the reaction heat of
Ph3C+ with BNAH in acetonitrile at 298 K (40.4 kcal mol−1
)
and the molar enthalpy change of BNAH to release hydride
anions in acetonitrile at 298 K (64.2 kcal mol−1). The
former is measured using ITC and the latter is a known
number (see ref. 25).
Notes and references
1 K. M. Fararh, Y. Atoji, Y. Shimizu, T. Shiina, H. Nikami and
T. Takewaki, Res. Vet. Sci., 2004, 77, 123.
2 H. Rchid, H. Chevassus, N. Rachid, C. Guiral, P. Petit, 25 X.-Q. Zhu, M.-T. Zhang, A. Yu, C.-H. Wang and J.-P. Cheng,
M. Chokairi and Y. Sauvaire, Fundam. Clin. Pharmacol.,
2004, 18, 525.
J. Am. Chem. Soc., 2008, 130, 2501.
26 D. D. M. Wayner and V. D. Parker, Acc. Chem. Res., 1993, 26,
3 B. Meddah, R. Ducroc, M. E. A. Faouzi, B. Eto, L. Mahraoui,
287.
A. Benhaddou-Andaloussi, L. C. Martineau, Y. Cherrah and 27 Eqn (9)–(11) were constructed from the thermodynamic
P. S. Haddad, J. Ethnopharmacol., 2009, 121, 419.
4 Y.-M. Liu, Y.-H. Jiang, Q.-H. Liu and B.-Q. Chen, Evid Based
Complement Alternat Med., 2011, 2011, 538671.
5 T. Yuan, P. Nahar, M. Sharma, K. Liu, A. Slitt, H. A. Asia
and N. P. Seeram, J. Nat. Prod., 2014, 77, 2316.
6 R. Rafaeloff-Phail, L.-Y. Ding, L. Conner, W.-K. Yeh,
D. McClure, H.-H. Guo, K. Emerson and H. Brooks, J. Biol.
Chem., 2004, 279, 52934, and related references cited
therein.
7 C. A. Coleman, J. G. Rose and C. J. Murray, J. Am. Chem.
Soc., 1992, 114, 9755.
8 J.-P. Cheng, Y. Lu, X.-Q. Zhu and L.-J. Mu, J. Org. Chem.,
1998, 63, 6108.
cycles (I), (II), and (III) in Scheme 5, respectively. Note that
we used the term free energy change ΔGet to replace the
enthalpy change ΔHet in eqn (9)–(11) for the electron trans-
fer process. The justification of this treatment was that
entropies associated with electron transfer were negligible,
as verified by Arnett et al.: E. M. Arnett, K. Amarnath,
N. G. Harvey and J.-P. Cheng, J. Am. Chem. Soc., 1990, 112,
344.
28 The energetic data herewith was suggested to be used in a
relative sense for the purpose of comparison within a
family. This was because the electrode potentials were
measured under irreversible conditions and some uncer-
tainties in the absolute data were involved (possibly by a
few tens of millivolts). For a method to correct an irrevers-
ible potential to a reversible one, please see ref. 26.
9 J. Yuasa, S. Yamada and S. Fukuzumi, J. Am. Chem. Soc.,
2006, 128, 14938.
10 S. Fukuzumi, H. Kotani, Y.-M. Lee and W. Nam, J. Am. 29 I. H. Lee, E. H. Jeoung and M. M. Kreevoy, J. Am. Chem.
Chem. Soc., 2008, 130, 15134. Soc., 1997, 119, 2722.
11 X.-Q. Zhu, Y.-Y. Mu and X.-T. Li, J. Phys. Chem. B, 2011, 115, 30 X.-Q. Zhu, F.-H. Deng, J.-D. Yang, X.-T. Li, Q. Chen,
14794.
N.-P. Lei, F.-K. Meng, X.-P. Zhao, S.-H. Han, E.-J. Hao and
12 J. A. Birrell and J. Hirst, Biochemistry, 2013, 52, 4048.
Y.-Y. Mu, Org. Biomol. Chem., 2013, 11, 6071.
13 The motifs of these alkaloids belong to iminium ions, 31 A. V. El’tsov and M. Z. Girshovich, Zh. Org. Khim., 1967, 3,
which are likely to accept hydride, hydrogen atoms and 1332.
electrons as supported by: X.-M. Zhang, J. W. Bruno and 32 J.-P. Cheng, K. L. Handoo and V. D. Parker, J. Am. Chem.
E. Enyinaya, J. Org. Chem., 1998, 63, 4671, and nucleophilic
addition reactions in ref. 14–18.
14 J. Elguero, E. G. Gutierrez-Puebla, A. Monge, C. Parco and
M. Ramos, Tetrahedron, 1993, 49, 11305.
Soc., 1993, 115, 2655.
33 F. G. Bordwell and W.-Z. Liu, J. Am. Chem. Soc., 1996, 118,
10819.
34 D. D. M. Wayner, E. Lusztyk, K. U. Ingold and P. Mulder,
J. Org. Chem., 1996, 61, 6430.
15 A. Gonzalez-Ortega, Eur. J. Org. Chem., 2005, 4663.
16 A. Schmidt, T. Habeck, B. Snovydovych and W. Eisfeld, Org. 35 F. G. Bordwell, X.-M. Zhang and J.-P. Cheng, J. Org. Chem.,
Lett., 2007, 9, 3515. 1993, 58, 6410.
17 A. M. Gonzalez-Nogal, M. Calle and P. Cuadrado, 36 B. Gollas and B. Speiser, Angew. Chem., Int. Ed. Engl., 1992,
Eur. J. Org. Chem., 2007, 6089. 31, 332.
18 A. Schmidt, A. Beutler and B. Snovydovych, Eur. J. Org. 37 Y. Li, K. C. Mondal, P. P. Samuel, H.-P. Zhu, C. M. Orben,
Chem., 2008, 4073.
S. Pannerselvam, B. Dittrich, B. Schwederski, W. Kaim,
11484 | Org. Biomol. Chem., 2015, 13, 11472–11485
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
T. Mondal, D. Koley and H. W. Roesky, Angew. Chem., Int. 43 H. Dauben, J. L. Honnen and K. J. Harmon, Org. Chem.,
Ed., 2014, 126, 4252.
1960, 25, 1442.
38 P. R. Wells, Chem. Rev., 1963, 63, 171.
39 C. D. Johnson, Chem. Rev., 1975, 75, 755.
40 BPH was chosen as the model compound of NADH, as BPH
44 A. Schmidt, B. Snovydovych, T. Habeck, P. Dröttboom,
M. Gjikaj and A. Adam, Eur. J. Org. Chem., 2007, 4909–
4916.
and NADH share the same core structure of 1,4-dihydro- 45 J. Yang and P. Gharagozloo, Synth. Commun., 2005, 35(3),
pyridine, and the hydride donating ability of BPH 409–415.
(53.0 kcal mol−1 to release hydride in acetonitrile)41 is 46 Z. N. Timofeeva, M. V. Petrova, M. Z. Girshovich and
close to that of NADH (53.6 kcal mol−1 to release hydride
A. V. El’tsov, Zh. Obshch. Khim., 1969, 39, 54–59.
in neutral water); 3+ was selected as the model compound 47 T. Kitamura, M. Todaka, I. Shin-machi and Y. Fujiwara,
of indazolium alkaloids, since 3+ is structurally quite close
to the indazolium alkaloids A and B in Scheme 1.
41 X.-Q. Zhu, Y. Tan and C.-T. Cao, J. Phys. Chem. B, 2010, 114,
2058.
Heterocycl. Commun., 1998, 4(3), 205–208.
48 X.-Q. Zhu, Q. Li, W.-F. Hao and J.-P. Cheng, J. Am. Chem.
Soc., 2002, 124, 9887.
49 X.-Q. Zhu, W.-F. Hao, C.-H. Tang and J.-P. Cheng, J. Am.
Chem. Soc., 2005, 127, 2696.
42 J. M. Bobbit, J. Org. Chem., 1998, 63, 9367.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 11472–11485 | 11485