Dependence of the reactivity of acridine on its substituents: A computational and kinetic study

Article abstract of DOI:10.1002/ejoc.201101017

Aromatic nucleophilic substitution on the C9 carbon of acridine plays an important role in multiple biological and medicinal applications. The rate of the key reaction is strongly dependent on the environment and acridine substituents. In this study, the

Full text of DOI:10.1002/ejoc.201101017

FULL PAPER
DOI: 10.1002/ejoc.201101017
Dependence of the Reactivity of Acridine on Its Substituents: A Computational
and Kinetic Study
[a]
[a]
[a]
[a]
ˇ
ˇ
ˇ
ˇ
Zbigniew Zawada, Jaroslav Sebestík, Martin Safarík, and Petr Bour*
Keywords: Nitrogen heterocycles / Kinetics / Reaction mechanisms / Transition states / Density functional calculations
Aromatic nucleophilic substitution on the C9 carbon of acri-
dine plays an important role in multiple biological and me-
dicinal applications. The rate of the key reaction is strongly
dependent on the environment and acridine substituents. In
this study, the factors influencing the reaction mechanism
were studied theoretically and verified experimentally for
simplified systems. Density functional theory was used for
the computations. The activation energy of a model deriva-
tive was determined experimentally from a kinetic study.
Also, the relative reactivities of selected compounds were
verified by a competition experiment. The theoretical predic-
tions correlate well with the observations. The computed re-
action path confirms that the thiol group attacks the aromatic
core activated by the nitrogen heteroatom and replaces halo-
gen or amino substituents via a well-defined Meisenheimer
transition state. The reaction barrier is strongly influenced by
both the solvent and substituents on the aromatic system. A
multistep mechanism is proposed for the reaction of aminoac-
ridine in which the aqueous solvent participates in the reac-
tion. Strong correlations between reaction energies and geo-
metrical parameters were observed and can be used to ratio-
nalize the design of acridine drugs as well as to avoid lengthy
computations of the transition states.
matic arthritis, Lupus erythrematosus, malaria, tapeworm
infections, Chagas disease, and epilepsy.[4a,4b,6] Quinacrine
inhibits the generation of the toxic isoform of the prion
protein in an intoxicated cell culture.[4b] Bis-acridinylated
compounds have also been tested for anti-prion activity.[4c,7]
Finally, the 9-aminoacridine drug tacrine (9-amino-1,2,3,4-
tetrahydroacridine) has been used in the treatment of Alz-
heimer’s disease.[8]
Introduction
The tricyclic aromatic acridine system (Scheme 1) exhib-
its increased reactivity at the 9-position relative to the corre-
sponding all-carbon arene due to the presence of the nitro-
gen heteroatom.[1] Acridine and its derivatives, such as 9-
aminoacridines, are widely used in medicine and biochemis-
try as drugs and labeling agents.[2]
The reactivity of acridines can be modified by their sub-
stituents. For instance, 9-amino substitution has been pro-
posed to increase biological activity.[9] Acridines without a
substituent at C9 are potent anticancer drugs.[10] Neverthe-
less, there are only a limited number of rules for predicting
the reactivity of a particular derivative. The influence of
acridine substituents on its reactivity has not been sup-
ported by a systematic theoretical analysis.
Scheme 1. Acridine and the atomic numbering pattern.
Acridine-based compounds have been considered, for ex-
ample, for the treatment of protozoal infections, cancer, vi-
ral, and prion diseases. The covalent attachment of acri-
dines can modulate the biological activity of peptides and
proteins.[2a,2b,2d,3] Derivatives of 9-aminoacridine are used
as efficient drugs;[2b,4] their anticancer activity is increased
by their covalent binding to nucleic acids.[5] The quinacrine
derivative is used, for example, in the treatment of rheu-
In this work, to enhance the future design of other biolo-
gically functional derivatives and to understand the func-
tion of this molecule on the atomic level, we have employed
quantum chemical modeling. For a limited number of sys-
tems, the results have been verified experimentally. It ap-
pears that the reactivity of acridine can be modeled quite
reliably and modeling confirms that the most probable reac-
tion mechanism of the binding of acridine in organisms is
[a] Molecular Spectroscopy, Institute of Organic Chemistry and aromatic nucleophilic substitution.[11] The influence of sol-
Biochemistry,
vent was simulated by using the polarizable continuum di-
electric model.
Although there are many possible mechanisms for the
nucleophilic substitution of aromatic halides,[12] we have
found the formation of the Meisenheimer complex as a
Flemingovo námeˇstí 2, 16610 Prague, Czech Republic
Fax: +420-220183578
E-mail: bour@uochb.cas.cz
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201101017 or from
the author.
Eur. J. Org. Chem. 2011, 6989–6997
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6989

ˇ
ˇ
Z. Zawada, J. Sebestík, M. Safarˇík, P. Bourˇ
FULL PAPER
transition state to be the most reasonable pathway for our senheimer complex during the aromatic nucleophilic substi-
systems. For example, radical mechanisms have been ex- tution is also the rate-determining step in this case.
cluded due to the excess of mercaptans, which are efficient
scavengers of free radicals. Moreover, calculated energy pro-
files along the reaction coordinate indicate that Meisenhei-
mer’s complex is the transition state (TS) of the rate-de-
termining step rather than an intermediate.
Results and Discussion
Energetics and Transition Complex of a Model Reaction
Aromatic nucleophilic substitution is omnipresent in ac-
ridine chemistry, for example, in reactions with NH2OH,
PhNHNH2, or thiols.[6c,11,13] The reactions strongly depend
on temperature, pH, and the nucleophilicity of the attack-
ing agent.[13a,14] To the best of our knowledge, the reactions
have not been extensively investigated theoretically, except
for single molecule energetics.[11,15]
Because of the significance for prion biochemistry, we
have concentrated on the condensation reaction between ac-
ridine and the thiol group, which can take place on reduced
disulfide linkages within the protein molecule. The reactiv-
ity and structure of a prion can potentially be influenced
by the attachment of an acridine label. In particular, re-
duction and subsequent alkylation prevent the conversion
of a prion into a protease resistant form.[16] An unperturbed
disulfide linkage is important for a proper insertion of a
prion into a membrane.[16b,16c]
The validity of the methods was tested on the simplified
reaction (1). The activation barrier, reaction energy, and se-
lected geometry parameters, calculated at different levels of
approximation, are summarized in Table 1. We can see that
the reaction energies (ΔE) are relatively indifferent to the
approximation used, whereas larger differences are appar-
ent for the activation barriers (E*). The barriers are also
more influenced by the solvent environment. For water,
their magnitudes are in the range 81–192 kJ/mol, whereas
in vacuo, very low barriers were calculated by methods
comprising electron correlation (DFT, MP2). The BPW91/
6-31G** level could not be included because the computed
“barrier” is negative, which implies that the transition state
was not found.
First, activation barriers and reaction energies were cal-
culated for a model pyridine system in which a chlorine
atom was used as the leaving group. The computationally
simplest SH– nucleophile was used in most computations.
For a limited number of systems, we verified that the reac-
tion mechanism and relative energy ordering are the same
for MeS– and benzyl mercaptan (BM). Secondly, the opti-
mal computational method (B3LYP/6-31+G**/CPCM) was
then chosen for the whole acridine. This simplification en-
abled us to consistently compare more methods; months of
CPU time would be needed for a TS computation of a
larger system, even at a relatively low approximation level.
The computational model also provides trends of reactivity
that are in very good agreement with the kinetic control
and competition experiments, which, clearly, could not be
modeled in full.
Table 1. Calculated activation (E*) and reaction (ΔE) energies for
the model reaction (1) and selected distances (r) in the transition
state (Figure 1).
E*
ΔE
rC–Cl rC–S
[Å]
rS–H
[Å]
[kJ/mol] [kJ/mol] [Å]
Vacuum calculations
HF/6-31G**
MP2/6-31G**
B3PW91/6-31G**
B3LYP/6-31G**
B3PW91/6-31++G**
B3LYP/6-311++G**
91
1
0
4
12
16
–107
–105
–97
–97
–101
–99
1.837 2.251
1.776 2.535
1.824 2.444
1.876 2.386
1.829 2.376
1.874 2.340
1.331
1.336
1.348
1.351
1.348
1.349
The computation suggests a limited role of the aromatic
acridine side-rings in the underlying reaction, which is lo-
cated on the internal ring. The specific role of more distant
acridine substituents (e.g., MeO, Cl, NO2, and CF3) on the
CPCM (H2O)
reaction rate was investigated for a large set of acridine de- HF/6-31G**
192
101
106
101
105
122
81
–119
–117
–109
–116
–130
–137
–124
1.863 2.183
1.862 2.183
1.952 2.287
1.877 2.237
1.925 2.265
1.926 2.264
1.871 2.183
1.330
1.333
1.350
1.347
1.349
1.349
1.337
MP2/6-31G**
rivatives. The correlation between geometry and energy pa-
B3LYP/6-31G**
rameters provides a way to estimate reactivity without a
B3PW91/6-31++G**
laborious search for the transition state, perhaps also for
B3LYP/6-311++G**
more complicated acridine systems.
B3LYP/6-31+G**
MP2/6-311++G2dp
The activation energy was then estimated experimentally
for a model system. For a larger set of acridine derivatives,
relative reactivities obtained from a competition experiment
The CPCM results (lower part of Table 1) are more rel-
are qualitatively compared with calculated activation ener- evant to the usual experimental conditions and also more
gies. Finally, a multistep mechanism in an aqueous environ- consistent for different levels of calculation. The sensitivity
ment is proposed for the case in which chlorine at the 9- to the solvent is in accord with the observed dependence
position of acridine is replaced by an amine and verified of acridine reactivity on the environment.[1,13a,14a] The HF
by computations of the intrinsic reaction coordinate. The method provides the highest activation barrier for both the
computations suggest that the formation of the Mei- in vacuo and CPCM calculations, which indicates the im-
6990
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 6989–6997

The Reactivity of Acridines
portance of involving the correlation energy in the calcula- rameters are summarized in Table 2. The 9-chloroacridines
tions. The MP2 results are quite close to those obtained by are clearly predicted to react faster than the 4-chloropyr-
the DFT methods. The variation of the basis set causes a idine benchmark model. For the acridines, the activation
relatively minor change in energy, approximately within 10– energy is lowest for the protonated nitro analogue 1 (5 kJ/
20% for the basis sets examined. The B3LYP/6-31+G**/ mol) and highest for the partially saturated compound 18.
CPCM method was used for computation of the larger sys-
Table 2. Activation energies (E*) and selected bond lengths in the
tems involving acridine described below as a compromise
reactants (rC–Cl,react.) and transition states (rC–Cl,TS, rC–S, and rS–H
)
between speed and accuracy. It is reasonably fast and in-
volves diffuse functions, at least on heavy atoms, which is
presumably very important for a study of the reaction
mechanism.
for the acridine analogues summarized in Figure 2, as calculated at
the B3LYP/6-31+G**/CPCM(H2O) level.
System
E*
[kJ/mol]
rC–Cl,react. rC–Cl,TS
rC–S
[Å]
rS–H
[Å]
[Å]
[Å]
All the methods also predict a similar geometry for the
activated complex (Figure 1, Table 1). The H–S–C and S–
C–Cl angles are predicted to vary in the range of 83–95° by
all models, and thus are not listed. The sulfur atom ap-
proaches the C4 atom from a direction approximately per-
pendicular to the aromatic ring and the system adopts a
tetrahedral arrangement. Within the CPCM environment,
the C–S bond becomes significantly shorter than deter-
mined by in vacuo calculations. Note also that consider-
ation of the solvent leads to a reduction of the reaction
energy. In more polar solvents the reaction is thus supposed
to proceed more slowly than in vacuo, but with a higher
reaction yield.
4-Chloropyridine
1
2
3
4
5
6
7
122
5
1.757
1.726
1.733
1.732
1.739
1.749
1.926
1.732
1.753
1.755
1.766
1.789
–
1.801
1.819
1.820
1.821
1.823
1.829
1.833
1.830
1.846
1.851
1.869
1.959
2.264
3.102
2.748
2.717
2.623
2.451
2.583
2.380
2.311
2.318
2.358
2.305
2.293
2.287
2.294
2.259
2.249
2.233
2.261
1.349
1.352
1.351
1.351
1.350
1.350
1.350
1.350
1.349
1.350
1.349
1.350
1.349
1.349
1.348
1.349
1.349
1.349
1.350
18
20
32
52
59
62
69
71
72
74
76
79
81
86
88
96
116
1.751
1.752
1.753
1.750
1.754
1.754
1.755
1.752
1.757
1.757
1.759
1.764
8
9
10
11
12
13
14
15
16
17
18
3 (exp.)
35
There is a low variance in the total reaction energies
(from –118 kJ/mol for 18 to –136 kJ/mol for 1, see Table S2
in the Supporting Information), slightly lower than for the
pyridine model (–137 kJ/mol). The geometries of the transi-
tion states are also similar to those of pyridine.
Figure 1. Geometry of the activated complex (B3LYP/6-
311++G**/CPCM) in reaction (1).
Reactivity of Acridine Derivatives
When a methylamine group was attached to C9 instead
At the B3LYP/6-31+G**/CPCM level we investigated the of chlorine (compound 6), a similar transition state was cal-
reactivity of the aminoacridine 6 and selected 9-chloroacrid- culated in the rate-determining step. Acridine protonation
ines (Figure 2). Protonated species were considered sepa- (compounds 1–3 and 6) leads to lower activation energies
rately (e.g., 1 ϶ 5) as they are treated as individual mole- and thus speeds up the reaction in comparison with the
cules in the calculations. The reaction energies calculated corresponding unprotonated compounds. This is in agree-
with the hydrogen sulfide anion and selected geometry pa- ment with the known behavior of acridine compounds.
Figure 2. Acridine derivatives investigated computationally.
Eur. J. Org. Chem. 2011, 6989–6997
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6991

ˇ
ˇ
Z. Zawada, J. Sebestík, M. Safarˇík, P. Bourˇ
FULL PAPER
The H–S–C and S–C–Cl angles in the transition states
The dependence of ΔE on E* (Figure 3) again corre-
(Table S2) are quite uniform for all the derivatives, confined sponds to Hammond’s reaction rules and to the concept of
to 92–96° and 93–100°, respectively. However, a larger vari- product development control,[12] which relates very negative
ance is apparent for the bond lengths listed in Table 2. Their reaction energies to low barriers. In other words, the fast
correlation with the activation energies is discussed in detail kinetics parallel the relative stabilities of the products.
in the next section.
When a TS becomes similar to the product, the activation
energy increases.
Correlation of the Activation Energies with the Geometry
Parameters
Experimental Activation Energy for Reaction (2)
The plots of the calculated C–Cl (reactant and TS) and
C–S (TS) bond lengths and the reaction energies against the
activation energies are shown in Figure 3. A greater number
of compounds have been included in this figure with the
additional derivatives (E1–E31, see Table S1 in the Support-
ing Information) confirming the trends discussed above for
1–18.
Figure 4 shows the dependence of the experimental rate
constant for reaction (2) on temperature. Because of the ex-
cess of one reactant (BM), its concentration is roughly con-
stant and the reaction behaves like a pseudo-first-order sys-
tem. The exponential fit corresponding to the Arrhenius
formula provides an activation energy of 35 kJ/mol, which
is in reasonable agreement with the calculated values listed
in Table 2. For compound 3, derived from 16 by in situ
protonation (reaction carried out in AcOH), we calculated
a value of 20 kJ/mol, although for a different SH– reagent.
Computations with BM were slow and numerically un-
stable. Nevertheless, the correlation of the calculated results
with experiment indicates that the SH– model, in spite of
the approximations, is appropriate for qualitatively de-
termining the reaction mechanism and energetics. Note,
however, that the experimental value may depend on vari-
ous parameters that are difficult to take into account theo-
retically. For example, the relatively slow formation of the
acridone byproduct was neglected. The good agreement of
the temperature dependence of the rate constant with the
theoretical dependence (Arrhenius law, Figure 4) suggests
that pseudo-first-order kinetics is appropriate, at least at the
beginning of the reaction (up to about 20% conversion).
Figure 3. Calculated (B3LYP/6-31+G**/CPCM) dependencies of
the C–Cl (reactant and TS) and C–S (TS) bond lengths and the
reaction energy on the activation energy. Compounds 1–18 (Fig-
ure 2, without derivative 6) and E1–E31 (Table S1 in the Support-
ing Information) are included.
The strong correlation between E* and the C–Cl reactant
bond length (Figure 3, top) can be conveniently used to
provide a very quick estimation of the activation energy, as
the full computation of the transition state is much more
computationally demanding than optimization of an equi-
librium geometry.
Higher activation energies are clearly associated with
longer C–Cl bonds in the TS. A similar correlation was also
Figure 4. The dependence of the experimental rate constant for re-
action (2) on the temperature; an exponential fit is indicated.
observed for the relative change in this bond [(dTS
–
dreactant)/dreactant, not shown]. This corresponds very well
with Hammond’s postulate,[12,17] that is, E* is qualitatively
proportional to the change in geometry between the reac-
tants and TS.
An opposite correlation is found for the C–S bond
formed (Figure 3), that is, it becomes shorter for reactions
with higher E*. This is also in accordance with Hammond’s
postulate: There is no C–S bond in the reactants, the
shorter the C–S bond the greater the geometrical change
between the reactants and TS.
6992
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 6989–6997

The Reactivity of Acridines
Competition Experiment
pose that the rate-determining step (formation of a Mei-
senheimer complex) found for the chloro derivative is sim-
ilar also for quinacrine, although some specifics for the
amino substituent in the latter have to be taken into ac-
count: Quinacrine (pKa = 10.4)[2a] is more basic than chlo-
roacridine (16, pKa = 6.0).[21] Under physiological condi-
tions (pH = 7.4), quinacrine is completely protonated on
N10 and the reaction is accelerated (cf. the activation ener-
gies for 16 and 3 are 88 and 20 kJ/mol, respectively). Pro-
tonation of the quinacrine derivative is further enhanced by
a resonance with the iminium form (20, Figure 5).
The reactivities of the compounds decrease in the follow-
ing order: 12 Ͼ 14 Ͼ 15 Ͼ 16 ϾϾ 18. Reaction half-times
for 12, 14, and 16 were determined to be approximately 0.5,
1, and 7 h, respectively. Derivative 18 did not react within
the timescale of the HPLC monitoring. For 15, its HPLC
signal could not be completely distinguished from the ac-
ridone byproduct, nevertheless, we can estimate its reactiv-
ity to be higher than that of 16 and lower than that of 14.
The general behavior thus agrees well with the reactivities
predicted by the calculations summarized in Table 2. For
example, the lowest activation energy (76 kJ/mol) corre-
sponds to the highest reaction rate (t1/2 = 0.5 h) for 12,
whereas for 18 the highest activation energy (116 kJ/mol)
correlates with the slowest reaction (t1/2 ϾϾ 24 h).
We also noticed that the reaction yields during the syn-
thesis of compounds 12, 14, 15, and 16 (89, 90, 82, and
65%, respectively) also correlate reasonably well with the
theoretical predictions, that is, with increasing activation
energies and decreasing reactivities. The yields in real ex-
periments are clearly only weakly related to our abstract
theoretical models, nevertheless the correlation confirms the
decisive role of the activation energies and the activated
complex.
Activation of the Acridine Reactant
According to the computations, most of the acridine sub-
stituents, for example, the nitro group and chlorine atom,
reduce the activation energy. Only the methoxy group and
saturation of the aromatic ring (compound 18) cause an
increase. This is in accordance with described substituent
effects for aromatic nucleophilic substitution reactions.[18]
The nitro group, for example, is an electron-withdrawing
substituent that enhances the reaction. Owing to their nega-
tive inductive effects, the chlorine atom and the CF3 group
also increase the reactivity (9, 11–14, Figure 1, Table 2),
whereas electron-donating substituents (the methoxy group
and alkyl chains in partially saturated systems, 16, 17, 18)
inhibit the nucleophilic substitution. Note that when both
a methoxy group and a chlorine atom are present (16), the
activation energy is similar to that of bare 9-chloroacridine.
The effects of the two substituents cancel each other out.
Although the reaction mechanisms for 4-chloropyridine
and 9-chloroacridines are, according to the calculations, al-
most identical, it should be noted that their reactivities in
general are not equivalent. For example, the thermal sta-
bility of acridine is lower than that of pyridine.[19]
Figure 5. Reaction mechanism for the reaction of 4-methylami-
nopyridine discussed in the text. Relative energies along the reac-
tion path (bottom, in kJ/mol) of the intermediates were calculated
at the B3LYP/6-31+G**/CPCM(H2O) level of theory.
The reactivity of the quinacrine drug is too complex to
be modeled in detail by our computational methods, never-
theless, in Figure 5, we propose a reaction mechanism for
a simpler system, 4-methylaminopyridine. Similarly to the
chloropyridine, the formation of a Meisenheimer complex
(20TS) as a transition state is the rate-determining step with
an activation energy of 112 kJ/mol. The protonation of the
initial amine 19 is associated with a large drop of energy
(step 1, Figure 5, ΔE = –198 kJ/mol). A second protonation
of amine 21 (step 4) makes the reaction again more energet-
ically viable. CH3NH2 is eliminated from 22 with a low re-
action barrier (5 kJ/mol). 4-Thiopyridone (24) is the final,
most energetically-viable product.
Modification of the Reaction Mechanism for Quinacrine
Alternative reaction paths were investigated but not
Apart from 9-chloroacridine, many aminoacridine deriv- found to be reasonable, mostly due to high energy barriers.
atives attract attention because of their biological activity. These included (1) the reaction of SH– with unprotonated
For example, quinacrine is used as a lysosomotropic agent 4-methylaminopyridine (19), (2) an intramolecular proton
and cysteine protease inhibitor in biochemical applica- transfer in 21 (from SH to NH), and (3) simultaneous me-
tions.[20] On the basis of the calculations (Table 2), we sup- thylamine cleavage with thiol group deprotonation.
Eur. J. Org. Chem. 2011, 6989–6997
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6993

ˇ
ˇ
Z. Zawada, J. Sebestík, M. Safarˇík, P. Bourˇ
FULL PAPER
We are aware that the CPCM solvent model could intro- high activation energies. Also remarkable is the additivity
duce some errors into the estimated energies, especially of the fluorine substitution effect on the activation energy,
those of the proton considered as H3O+. Nevertheless, the not observed for the other groups.
calculations seem to provide a useful insight into the
mechanism. The proton transfer to basic nitrogen atoms
Remarks on the Role of the Substituents
(steps 1, 4, and 6, Figure 5) did not exhibit activation barri-
ers. This was verified by the potential energy surface scans
performed with the Gaussian suite of programs.[22] Note
that at physiological pH, almost all the substrates would
already be protonated.
The relative influence of substituent positions on the ac-
ridine reactivity is summarized in Table 4. In spite of pos-
sible computational errors, a more detailed analysis reveals
interesting properties of the substituted acridine systems.
For example, the OMe group increases the activation en-
ergy, but its positive mesomeric (+M) effect is not additive;
for the 3,4-dimethoxy derivative the activation energy is
lower than for either 3-methoxy or 4-methoxy derivatives.
We can explain this by the nonplanar arrangement of aro-
matic and OMe π-electron systems at the 4-position, which
only allows for the negative inductive (–I) effect. This is in
agreement with the activation energy of the 2,3-dimethoxy
derivative being greater than for either the 2-methoxy or
3-methoxy derivatives, in which the π conjugation is not
perturbed. The nitro group displays opposite behavior to
OMe, having a strong negative mesomeric effect.
Empirical Rules Derived from the Theoretical Results
We can thus conclude that the computations provide re-
alistic estimates of the relative reactivities of the acridines,
which can be used to plan aromatic nucleophilic experi-
ments. Some of the predicted strong correlations between
geometry and activation energy are summarized in Table 3.
As mentioned above (Figure 3 and Table 2), the dependence
of E* on the reactant geometry is a particularly valuable
tool, for example, for future drug design, as it avoids the
time-consuming and often laborious TS calculations.
Table 4. Calculated relative influence of substituent positions on
reactivity.
Table 3. Empirical relationships and approximate rules obtained
from the reaction path calculations (B3LYP/6-31+G**/CPCM/
H2O) of the acridine derivatives reacting with the HS– ion.[a]
Monosubstitution
NO2: 2 Ͼ 4 Ͼ 3 Ͼ 1
CF3: 2 Ͼ 4 Ͼ 3 Ͼ 1
Cl: 1 Ͼ 4 Ͼ 3 Ͼ 2
F: 1 Ͼ 4 Ͼ 3 Ͼ 2
Unprotonated substrate
E* = 3598rC–Cl,reactant – 6234, R2 = 0.98
rC–Cl,TS = 6.69rC–Cl,reactant – 9.90, R2 = 0.88
OMe: 1 Ͼ 3 Ͼ 4 Ͼ 2
rC–S,TS = 837.1rC–Cl,reactant – 2951.4rC–Cl,reactant + 2603.71, R2 = 0.94
2
Polysubstitution
Protonated substrate
F: 1,2,3,4 Ͼ 1,3,4 Ͼ 1,2,4 Ͼ 1,2,3 Ͼ 1,4 Ͼ 1,3 Ͼ 2,3,4 Ͼ 1,2 Ͼ
3,4 Ͼ 1 Ͼ 2,4 Ͼ 2,3 Ͼ 4 Ͼ 3 Ͼ 2
OMe: 1 Ͼ 3,4 Ͼ 3 Ͼ 4 Ͼ 2 Ͼ 2,3
E* = 2285rC–Cl,reactant – 3939, R2 = 0.95
rC–Cl,TS = 3.27rC–Cl,reactant – 3.91, R2 = 0.84
rC–S,TS = 1958.4rC–Cl,reactant – 6824.2rC–Cl, reactant + 5947.4, R2 = 0.91
2
Functional groups at a given position
Additivity effect of the fluorine substitutions on acridine
C-1: NO2 Ͼ Cl Ͼ F Ͼ CF3 Ͼ OMe Ͼ H
C-2: NO2 Ͼ CF3 Ͼ Cl Ͼ F Ͼ H Ͼ OMe
C-3: NO2 Ͼ CF3 Ͼ Cl Ͼ F Ͼ H Ͼ OMe
C-4: NO2 Ͼ CF3 Ͼ Cl Ͼ F Ͼ H Ͼ OMe
C-2, C-7: NO2, NO2 Ͼ NO2, SO2H Ͼ SO2H, SO2H Ͼ CHO,
CHO Ͼ H, H Ͼ NH2, NH2
E* = 86.1 – 11.9f1 – 1.4f2 – 5.7f3 – 6.7f4, fi = 1 for fluorine at posi-
tion i, otherwise fi = 0
[a] E* is in kJ/mol, distances in Å, R2 is the square of correlation
coefficient.
Also, if the TS computation is inevitable, the relation be-
tween the reactant and TS geometries can save an enor-
mous amount of computational time. In particular, unreal-
istic C–Cl and C–S bond lengths in the starting structures
typically lead to convergence problems during the reaction
path calculation.
The 1-methoxy derivative is more reactive than unsubsti-
tuted 9-chloroacridine or the 2-fluoro derivative. This can-
not be explained by the decrease in the +M effect of OMe
because the OMe group is coplanar with the aromatic sys-
tem. In this case, the chlorine and oxygen atoms repel each
other (Figure 6).
In Table 3, relationships for the protonated and unpro-
tonated compounds are given separately. Although the
trends are the same, the protonated compounds (E*,
rC–Cl,TS) are notably less influenced by the substituents. This
can be explained by the strong effect of the hydrogen
charge, which dominates other substituent effects in the
protonated species.
Substitution at the 1-position of acridine requires par-
ticular attention for steric reasons. Bulky substituents with
Figure 6. Distortion of the chlorine nonbonding orbital by the
a strong electron-withdrawing effect (NO2) usually require presence of the OMe group at the 1-position.
6994 Eur. J. Org. Chem. 2011, 6989–6997
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

The Reactivity of Acridines
The special character of the 1-position is also apparent tions, as did the relative reactivities determined by the com-
for the nitro group. It is not coplanar with the aromatic petition experiment.
system, the torsion angle is about 60° for this position, and
For the quinacrine drug model, a multistep reaction was
the –M effect is quite weakened. The activation energy of suggested that is compliant with physiological conditions.
1-NO2-acridine is the highest among the nitroacridines. The calculated reaction path indicates that the reaction pro-
Also, the 4-NO2-acridine is slightly distorted (ca. 30°).
ceeds through a similar rate-determining step as for the
We also see that the activation energy can be most influ- chloro derivative. In addition, unlike the chloro derivative,
enced by substituents with a strong mesomeric effect at C2. quinacrine is completely protonated at N10 at physiological
The activity can be further supported by an attachment of pH, which can at least partially explain the high biological
another substituent at C7, equivalent to C2. In this case, activity of this compound.
the effect is approximately additive.
Finally, we found simple relationships between geometry
Although the Mulliken charges calculated for the model and reactivity, which can speed up the computational
reaction of compound 15 (Figure 7) are not directly related screening of the acridine derivatives. In the future, we hope
to observable properties, they document the large electronic that the theoretical modeling will be useful in the rational
redistribution during the substitution and approximately design of acridine aromatic labels, particularly in medicinal
correlate with the positions of importance for the reactivity. chemistry.
For example, the charge on C3 changes only slightly (–0.04
Ǟ –0.08) and this carbon is relatively unimportant. The
largest change in charge is on C1, which is clearly very im- Experimental Section
portant (cf. Table 4), except for the cases complicated by
Calculations: In the calculations on reaction (1), acridine was sim-
steric effects. The change in charge is also larger on C4 than
plified to the pyridine ring. The chlorine atom was attached at C4
on C3, in agreement with their relative influence on reactiv-
of pyridine as a simple leaving group. The reaction path was deter-
mined by the QST3[23] method using the Gaussian suite of pro-
grams. Activation barriers and reaction energies were estimated at
the HF, MP2,[24] B3PW91,[25] and B3LYP[25] levels using the 6-
31G**, 6-31++G**, and 6-311++G** Gaussian standard Pople-
ity.
style basis sets. Activation energies were defined as E* = ETS – EA
–
EB, in which TS is the transition state, A is 4-chloropyridine, and
B is SH–. Similarly, the reaction energy ΔE = EC + ED – EA – EB,
in which C is pyridine-4-thiol and D is Cl–, was calculated. The
solvent environment was accounted for by the CPCM[26] correc-
tion. Geometries of the transition states and minima on the poten-
tial energy surface were verified by calculation of the harmonic
vibrational frequencies. The number of CPCM cavities changes
along the reaction coordinate, however, in a model computation,
we verified that the energy discontinuity is less than 1 kJ/mol.
Figure 7. Calculated Mulliken charges of reactants (top), TS
(middle), and products (bottom) for the substitution of Cl by SH–
in 15.
The computations of the reaction course were then repeated for a
broader range of acridine derivatives 1–18 (Figure 1) using the
B3LYP/6-31+G**/CPCM(H2O) level of approximation and, fi-
nally, completed with the extended set of compounds E1–E31 listed
in Table S1 in the Supporting Information.
Conclusions
Reactivity Studies: The reaction kinetics were studied for the reac-
tion involving 6,9-dichloro-2-methoxyacridine (16, Figure 2) and
benzyl mercaptan (BM). Acridine 16 (ca. 2.3 mg) was dissolved in
glacial acetic acid (1.6 mL). The mixture was briefly sonicated and
then stirred for 10 min to ensure temperature equilibration. BM
(ca. 200 μL) was then added. The amounts of all substances were
adjusted so the final concentrations were 5.00 mm of 16 and 1.00 m
of BM (e.g., 2.335 mg of 16, 198 μL of BM, and 1.48 mL of acetic
acid). After addition of the thiol, the mixture was rigorously stirred
for 20 s and then 3 μL of reaction mixture was injected into the
HPLC apparatus.
We tested the performance of the computational meth-
ods on a reduced pyridine system and roughly estimated
their reliability. The similarity in activation energies sug-
gested that the model reproduces the behavior of the more
complex acridine derivatives reasonably well. The B3LYP/
6-31+G**/CPCM(water) method was then chosen as a
compromise between speed and accuracy to investigate the
larger systems and to analyze the influence of substituents.
For all acridine derivatives, the energetics and geometry
of the activated complex remained qualitatively similar.
However, the reaction barrier and thus the actual reactivity
could be significantly reduced or increased by substitution
of the aromatic systems. In particular, the protonation of
the aromatic nitrogen was confirmed to accelerate the reac-
tion.
The reaction mixture was analyzed by HPLC at 381 nm with the
following solvent gradient: 0 min 40% B, 5 min 66% B, 5.1 min
100% B, 5.7 min 100% B, 5.8 min 40% B. HPLC was in continuous
mode (i.e., it was not stopped between analyses) and two subse-
quent injections were separated by at least 7.3 min, otherwise the
peak integrals were inaccurate. The cleanliness of the Hamilton
syringe was crucial; after each injection it was rinsed 15 times with
The experimentally determined activation energy for the
model reaction agreed well with these theoretical predic- THF and 6 times with tBuOMe (full volume of the syringe), the
Eur. J. Org. Chem. 2011, 6989–6997
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6995

ˇ
ˇ
Z. Zawada, J. Sebestík, M. Safarˇík, P. Bourˇ
FULL PAPER
syringe was dried with dry air for around 20 s, and the piston was
wiped dry. The temperature of the syringe and piston was allowed
to partially equilibrate with that of the reaction mixture. Before
injection, the Hamilton syringe was rinsed three times with the re-
action mixture. The time (t) dependence of the product areas (S)
was fitted by a first-order kinetic model, S(t) = S(0) + K[1 –
exp(–kt)]. By using the Arrhenius relation [k = A exp(–E*/RT)],
the activation energy E* was calculated from the rate constants (k)
obtained at six different temperatures (T) within the interval 293–
323 K; R is the universal gas constant and A is a constant indepen-
dent of temperature.
9.1, 0.8 Hz, 1 H), 8.36 (ddd, J = 8.8, 1.4, 0.7 Hz, 1 H), 8.16 (ddd,
J = 8.8, 1.2, 0.7 Hz, 1 H), 7.79 (ddd, J = 8.8, 6.6, 1.4 Hz, 1 H),
7.68 (dm, J = 9.1 Hz, 1 H), 7.63 (ddd, J = 8.8, 6.6, 1.2 Hz, 1
H) ppm. 13C NMR (101 MHz, CDCl3): δ = 150.0, 147.6, 141.6,
132.2 (q, J = 33.0 Hz), 131.5, 130.3, 128.3 (q, J = 4.7 Hz), 128.2,
126.5, 125.2, 125.2, 124.8, 123.9 (q, J = 272.6 Hz), 122.2 (q, J =
2.9 Hz) ppm. HRMS (EI): calcd. for C14H7ClF3N 281.0219 [M]+;
found 281.0226.
9-Chloro-1-(trifluoromethyl)acridine (14): Pale-green solid, yield
178 mg, 31%. HPLC: tR = 8.55 min; m.p. 78–81 °C. 1H NMR
(400 MHz, CDCl3): δ = 8.63 (ddd, J = 8.9, 1.3, 0.7 Hz, 1 H), 8.42
(ddq, J = 9.0, 1.2, 0.6 Hz, 1 H), 8.22 (ddd, J = 8.7, 1.2, 0.7 Hz, 1
H), 8.18 (dp, J = 7.3, 1.2 Hz, 1 H), 7.88 (ddd, J = 8.7, 6.6, 1.3 Hz,
1 H), 7.78 (ddq, J = 9.0, 7.3, 0.9 Hz, 1 H), 7.72 (ddd, J = 8.9, 6.6,
1.2 Hz, 1 H) ppm. 13C NMR (101 MHz, CDCl3): δ = 149.8, 148.6,
139.4, 136.3, 131.5, 129.7, 129.3 (q, J = 8.2 Hz), 128.4, 127.8, 126.3,
125.7, 125.1 (q, J = 31.5 Hz), 124.2 (q, J = 273.0 Hz), 121.5 ppm.
HRMS (EI): calcd. for C14H7ClF3N 281.0219 [M]+; found
281.0212.
To obtain information about the dependence of the reaction rate
on the substituents and to verify the theoretical predictions, a com-
petition experiment was set up to compare the reactivity of com-
pounds 12, 14, 15, 16, and 18 (Figure 1) with BM. Each of the five
compounds in 2.5 mm acetone solutions (100 μL) were mixed and
then BM (2 μL) and acetic acid (500 μL) were added. Pure acetic
acid was not used as a solvent for stock solutions because, unlike
acetone, it slowly reacts with 9-chloroacridines. After calibration,
5 μL samples were analyzed by HPLC over 24 h so that the relative
concentrations of the reactants could be monitored at 239, 247,
and 268 nm with the following gradient elution: 0 min 15% B,
1 min 18% B, 2 min 23% B, 3 min 30% B, 4 min 36% B, 12 min
75% B, 12.1 min 100% B. To exclude possible interference by light,
the experiment was performed in the dark, although we did not
notice any instability in our previous experiments in daylight. BM
acts as a radical scavenger and stabilizes potentially light-sensitive
acridine chromophores. The relative reactivities were determined
according to the time for consumption of the substrates. The reac-
tion was monitored over 24 h.
General Procedure for the Synthesis of 9-(Benzylsulfanyl)acridines:
9-Chloroacridine (0.4 mmol), BM (110 μL, 0.9 mmol), and NaOH
(30 mg, 0.8 mmol) were suspended in a mixture of methanol (1 mL)
and THF (0.5 mL). After approximately 2 min of sonication, water
(5 mL) and chloroform (5 mL) were added. The organic layer was
separated and the water fraction was extracted with chloroform
(3ϫ10 mL). The chloroform solution was dried with Na2SO4 and
the solvent was evaporated. The residue was dissolved in toluene
(10 mL) and the solvent was evaporated. This procedure was re-
peated three times. The crude product was purified by chromatog-
raphy (column or TLC).
Chromatographic Procedures: TLC silica gel plates (2 mm) from
Merck with F254 were employed for preparative TLC, silica gel 60
from Merck was used for column chromatography. HPLC was per-
9-(Benzylsulfanyl)-3-(trifluoromethyl)acridine: 9-Chloro-3-(trifluoro-
methyl)acridine (12) (112 mg, 0.4 mmol) was used. The crude prod-
uct was purified by preparative TLC (hexane/acetone, 90:10) to
yield 131 mg (89%) of a pale-yellow solid. HPLC: tR = 11.17 min;
m.p. 92–94 °C. 1H NMR (400 MHz, CDCl3): δ = 8.67 (dp, J = 9.1,
0.9 Hz, 1 H), 8.61 (ddd, J = 8.8, 1.4, 0.7 Hz, 1 H), 8.46 (m, J =
1.8, 0.9 Hz, 1 H), 8.16 (ddd, J = 8.8, 1.3, 0.7 Hz, 1 H), 7.74 (ddd,
J = 8.8, 6.6, 1.4 Hz, 1 H), 7.55 (ddq, J = 9.1, 1.8, 0.4 Hz, 1 H),
7.53 (ddd, J = 8.8, 6.6, 1.3 Hz, 1 H), 7.03–6.93 (m, 3 H), 6.79 (m,
2 H), 3.99 (s, 2 H) ppm. 13C NMR (101 MHz, CDCl3): δ = 149.7,
147.4, 142.7, 137.0, 131.64 (q, J = 32.7 Hz), 131.0, 130.5, 130.2,
130.1, 128.8, 128.6, 128.5, 128.4 (q, J = 4.7 Hz), 127.8, 127.6, 126.9,
124.0 (q, J = 272.5 Hz), 121.6 (q, J = 2.9 Hz), 42.1 ppm. IR
formed
with
a
ZORBAX
Poroshell
120
SB-C18
3ϫ50 mmϫ2.7 μm column with a flow of 1.00 mL/min at 25.0 °C
and gradients obtained from solutions A (0.05% TFA in water)
and B (100% ACN). HPLC (199, 220, 270, and 381 nm) was used
to determine the purity of compounds using the following gradient:
0 min 15% B, 1 min 18% B, 2 min 23% B, 3 min 30% B, 4 min
36% B, 12 min 75% B, 12.1 min 100% B; purity of all prepared
standards Ͼ97%.
9-Chloro-3-(trifluoromethyl)acridine (12) and 9-Chloro-1-(trifluoro-
methyl)acridine (14): The two chloroacridines were prepared ac-
cording to ref.[27] 3-{[3-(Trifluoromethyl)phenyl]amino}benzoic
acid (575 mg, 2 mmol) was dissolved in POCl3 (4 mL) and heated
at reflux for 2 h. Excess POCl3 was evaporated. The resulting very
viscous liquid was dissolved in ethanol (10 mL). The solution was
cooled (dry-ice/ethanol bath) and a concentrated ammonia/water
solution (3 mL) was added dropwise. Then the suspension was rig-
orously stirred for another 20 min in a cooling bath, water was
added (40 mL), and the cooling bath removed. The volume was
reduced by evaporation to around 10 mL and then extracted with
diethyl ether (4ϫ20 mL). The ether fractions were dried with
Na2SO4 and the solvent was evaporated. The mixture was dissolved
in acetone and the products were separated by preparative TLC
(DCM/toluene/ethyl acetate, 75:25:5). The products were extracted
from silica gel by ethyl acetate. The solvent was evaporated and the
residue was extracted with hexane. The hexane was evaporated and
the two isomers were obtained in a total yield of 82%.
(CHCl ): ν = 1328 (vs, CF ), 2931 (w, CH ) cm–1. HRMS (EI):
˜
3
3
2
calcd. for C21H14F3NS 369.0799 [M]+; found 369.0800.
9-(Benzylsulfanyl)-1-(trifluoromethyl)acridine: 9-Chloro-1-(trifluoro-
methyl)acridine (14) (112 mg, 0.4 mmol) was used. The crude prod-
uct was purified by preparative TLC (hexane/acetone, 90:10) to
yield 133 mg (90%) of a yellow solid. HPLC: tR = 9.30 min; m.p.
1
104–107 °C. H NMR (400 MHz, CDCl3): δ = 8.84 (ddd, J = 8.8,
1.3, 0.7 Hz, 1 H), 8.19 (ddq, J = 8.6, 1.3, 0.5 Hz, 1 H), 8.13 (ddd,
J = 8.7, 1.2, 0.7 Hz, 1 H), 7.91 (ddq, J = 7.2, 1.3, 0.7 Hz, 1 H),
7.76 (ddd, J = 8.7, 6.6, 1.3 Hz, 1 H), 7.61 (ddd, J = 8.8, 6.6, 1.2 Hz,
1 H), 7.59 (ddq, J = 8.6, 7.2, 0.9 Hz, 1 H), 6.88–6.77 (m, 3 H), 6.48
(m, 2 H), 3.69 (s, 2 H) ppm. 13C NMR (101 MHz, CDCl3): δ =
148.6, 148.6, 143.0, 136.4, 135.6, 131.0, 130.4, 130.0, 129.2 (q, J =
7.2 Hz),128.5, 128.0, 127.9, 127.8, 127.4, 127.1, 127.0, 126.5 (q, J
= 31.2 Hz), 124.7 (q, J = 272.8 Hz), 43.1 ppm. IR (CHCl ): ν =
˜
3
1288 (vs, CF3), 2930 (w, CH2) cm–1. HRMS (EI): calcd. for
C21H14F3NS 369.0799 [M]+; found 369.0798.
9-Chloro-3-(trifluoromethyl)acridine (12): Pale-yellow solid, yield
293 mg, 51%. HPLC: tR = 9.48 min, m.p. 118–120 °C. 1H NMR
9-(Benzylsulfanyl)acridine: 9-Chloroacridine (15) (89 mg, 0.4 mmol)
(400 MHz, CDCl3): δ = 8.47 (m, J = 0.8 Hz, 1 H), 8.46 (dp, J = was used. The crude product was purified by preparative TLC (hex-
6996 Eur. J. Org. Chem. 2011, 6989–6997
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

The Reactivity of Acridines
S. J. DeArmond, Proc. Natl. Acad. Sci. USA 2008, 105, 10595–
ane/ethyl acetate, 80:20) to yield 103 mg (82%) of a pale-yellow
solid. HPLC: tR = 5.55 min; m.p. 103–106 °C. 1H NMR (400 MHz,
CDCl3): δ = 8.59 (ddd, J = 8.8, 1.4, 0.7 Hz, 2 H), 8.14 (ddd, J =
8.7, 1.2, 0.7 Hz, 2 H), 7.68 (ddd, J = 8.7, 6.6, 1.4 Hz, 2 H), 7.45
(ddd, J = 8.8, 6.6, 1.2 Hz, 2 H), 7.03–6.94 (m, 3 H), 6.87–6.82 (m,
2 H), 3.98 (s, 2 H) ppm. 13C NMR (101 MHz, CDCl3): δ = 148.9,
142.1, 137.3, 130.2, 130.2, 129.3, 128.8, 128.43, 127.4, 126.9, 126.7,
41.9 ppm. HRMS (EI): calcd. for C20H15NS 301.0925 [M]+; found
301.0922.
10600.
[8] a) J. K. Marquis, Biochem. Pharmacol. 1990, 40, 1071–1076; b)
S. E. Freeman, R. M. Dawson, Prog. Neurobiol. 1991, 36, 257–
277.
[9] B. R. Brown, W. J. Firth, L. W. Yielding, Mutat. Res. 1980, 72,
373–388.
[10] D. J. A. Bridewell, G. J. Finlay, B. C. Baguley, Anti-Cancer
Drug Des. 2001, 16, 317–324.
ˇ
ˇ
[11]
[12]
[13]
J. Sebestík, M. Safarˇík, I. Stibor, J. Hlavácˇek, Biopolymers
2006, 84, 605–614.
9-(Benzylsulfanyl)-6-chloro-2-methoxyacridine: 6,9-Dichloro-2-meth-
oxyacridine (16) (114 mg, 0.4 mmol) was used. The crude product
was purified by column chromatography (first chloroform 100%,
then chloroform/methanol, 92:8) to yield 97 mg (65%) of an orange
solid. HPLC: tR = 9.61 min; m.p. 153–155 °C. 1H NMR (400 MHz,
CDCl3): δ = 8.47 (dd, J = 9.3, 0.4 Hz, 1 H), 8.10 (dd, J = 2.1,
0.4 Hz, 1 H), 7.98 (dd, J = 9.4, 0.4 Hz, 1 H), 7.66 (dd, J = 2.8,
0.4 Hz, 1 H), 7.36 (m, 2 H), 7.05–6.93 (m, 3 H), 6.78 (m, 2 H), 3.94
(s, 2 H), 3.83 (s, 3 H) ppm. 13C NMR (101 MHz, CDCl3): δ =
158.3, 147.1, 146.6, 139.0, 137.4, 135.0, 131.7, 130.8, 128.8, 128.5,
128.5, 128.0, 127.9, 127.5, 126.1, 102.3, 55.7, 41.4 ppm. HRMS
(EI): calcd. for C21H16NOSCl 365.0641; found 365.0643.
R. Bruckner, Advanced Organic Chemistry: Reaction Mecha-
nisms, Springer, New York, 2002.
a) A. Paul, S. Ladame, Org. Lett. 2009, 11, 4894–4897; b) M.
Weltrowski, A. Ledochowski, P. Sowinski, Pol. J. Chem. 1982,
56, 77–82.
[14]
[15]
a) A. Kunikowski, A. Ledochowski, Pol. J. Chem. 1981, 55,
1979–1984; b) C. M. Johnson, T. W. Linsky, W. D. Yoon, M. D.
Person, W. Fast, J. Am. Chem. Soc. 2011, 133, 1553–1562.
a) J. R. Goodell, B. Svensson, D. M. Ferguson, J. Chem. Inf.
Model. 2006, 46, 876–883; b) A. Wrooblewska, J. Meszko, K.
Krzyminsky, Y. Ebead, J. Blazejowski, Chem. Phys. 2004, 303,
301–308.
a) L. M. Herrmann, B. Caughey, Neuroreport 1998, 9, 2457–
2461; b) J. I. Shin, J. Y. Shin, J. S. Kim, Y. S. Yang, Y. K. Shin,
D. H. Kweon, Biochem. Biophys. Res. Commun. 2008, 377, 995–
1000; c) J. Y. Shin, J. I. Shin, J. S. Kim, Y. S. Yang, Y. K. Shin,
K. K. Kim, S. Lee, D. H. Kweon, Mol. Cells 2009, 27, 673–
680.
[16]
Supporting Information (see footnote on the first page of this arti-
cle): Experimental and computational details, compound charac-
terizations, computations for other acridine derivatives.
Acknowledgments
[17]
G. S. Hammond, J. Am. Chem. Soc. 1955, 77, 334–338.
[18] M. B. Smith, J. March, March’s Advanced Organic Chemistry,
5th ed., Wiley, New York, 2001.
[19] A. R. Katritzky, R. A. Barcock, M. Siskin, W. N. Olmstead,
Energy Fuels 1994, 8, 990–1001.
[20] K. Doh-Ura, T. Iwaki, B. Caughey, J. Virol. 2000, 74, 4894–
4897.
This work was supported by the Czech Science Foundation (203/
07/1517, P208/11/0105), the Czech Academy of Sciences
(M200550902), and the Ministry of Education (LH11033). We
thank Dr. Pavel Fiedler for the interpretation of IR spectra.
[21] M. G. Ferlin, C. Marzano, G. Chiarelotto, F. Baccichetti, F.
Bordin, Eur. J. Med. Chem. 2000, 35, 827–837.
[1] J. Chiron, J. P. Galy, SynLett 2003, 15, 2349–2350.
ˇ
[2] a) J. Sebestík, J. Hlavácˇek, I. Stibor, Curr. Protein Pept. Sci.
[22] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starov-
erov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M.
Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Ad-
amo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Mar-
tin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09, Re-
vision A.02, Gaussian, Inc., Wallingford, CT, 2009.
2007, 8, 471–483; b) M. Demeunynck, F. Charmantray, A.
Martelli, Curr. Pharm. Des. 2001, 7, 1703–1724; c) S. Claude,
J. M. Lehn, J. P. Vigneron, Tetrahedron Lett. 1989, 30, 941–944;
ˇ
ˇ
d) Z. Zawada, J. Sebestík, M. Safarˇík, A. Krejcˇírˇiková, A. Brˇez-
inová, J. Hlavácˇek, I. Stibor, K. Holada, P. Bourˇ, in: Peptides
2010 (Eds.: M. Lebl, M. Meldal, K. Jensen, T. Hoeg-Jensen),
Proceedings of the 31st European Peptide Symposium, Euro-
pean Peptide Society, Copenhagen, 2010, pp. 84–85.
ˇ
[3] J. Sebestík, I. Stibor, K. Hlavácˇek, J. Pept. Sci. 2006, 12, 472–
480.
[4] a) D. J. Wallace, Semin. Arthritis Rheum. 1989, 18, 282–297; b)
C. Korth, B. C. H. May, F. E. Cohen, S. B. Prusiner, Proc. Natl.
Acad. Sci. USA 2001, 98, 9836–9841; c) B. C. H. May, A. T.
Fafarman, S. B. Hong, M. Rogers, L. W. Deady, S. B. Prusiner,
F. E. Cohen, Proc. Natl. Acad. Sci. USA 2003, 100, 3416–3421.
[5] I. Lasnitzki, J. H. Wilkinson, Br. J. Cancer 1948, 2, 369–375.
[6] a) R. L. Krauth-Siegel, H. Bauer, H. Schirmer, Angew. Chem.
2005, 117, 698; Angew. Chem. Int. Ed. 2005, 44, 690–715; b) A.
Saravanamuthu, T. J. Vickers, C. S. Bond, M. R. Peterson,
W. N. Hunter, A. H. Fairlamb, J. Biol. Chem. 2004, 279, 29493–
[23] C. Peng, P. Y. Ayala, H. B. Schlegel, M. J. Frisch, J. Comput.
Chem. 1996, 17, 49–56.
[24] C. Møller, M. S. Plesset, Phys. Rev. 1934, 46, 618–622.
[25] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
29500; c) F. Wild, J. M. Young, J. Chem. Soc. 1965, 7261–7274; [26] a) A. Klamt, in: The Encyclopedia of Computational Chemistry,
d) J. C. Burnett, J. J. Schmidt, R. G. Stafford, R. G. Panchal,
T. L. Nguyen, A. R. Hermone, J. L. Vennerstrom, C. F.
McGrath, D. J. Lane, E. A. Sausville, D. W. Zaharevitz, R.
Gussio, S. Bavari, Biochem. Biophys. Res. Commun. 2003, 310,
84–93.
vol. 1 (Eds.: P. v. Ragué Schleyer, N. L. Allinger, T. Clark, J.
Gasteiger, P. A. Kollman, H. F. Schaefer III, P. R. Schreiner),
Wiley, Chichester, 1998, pp. 604–615; b) E. Cances, B. Men-
nucci, J. Tomasi, J. Chem. Phys. 1997, 107, 3032–3041.
[27] J. H. Wilkinson, I. L. Finar, J. Chem. Soc. 1948, 32–35.
Received: July 12, 2011
[7] P. Spilman, P. Lessard, M. Sattavat, C. Bush, T. Tousseyn, E. J.
Huang, K. Giles, T. Golde, P. Das, A. Fauq, S. B. Prusiner,
Published Online: October 17, 2011
Eur. J. Org. Chem. 2011, 6989–6997
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6997