Phosphonium betaines derived from hexafluoro-1,4-naphthoquinone: Synthesis and cytotoxic and antioxidant activities

Article abstract of DOI:10.1016/j.jfluchem.2016.10.014

Fluorinated derivatives of 1,4-naphthoquinones are highly potent inhibitors of Cdc25A and Cdc25B phosphatases; they suppress the growth of tumor cells. Four derivatives of phosphonium betaines derived from hexafluoro-1,4-naphthoquinone: (triphenyl[5,6,7,8-tetrafluoro-1-oxido-4-oxo-3-(phenylimino)-3,4-dihydronaphthalen-2-yl]phosphonium) (4), ((3,5-difluorophenyl)(methyl)phenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-dihydronaphthalen-2-yl)phosphonium) (5), ((2,5-difluorophenyl)(methyl)phenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-dihydronaphthalen-2-yl)phosphonium) (6) and ((3,5-difluorophenyl)diphenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-dihydronaphthalen-2-yl)phosphonium) (7) were synthesized for the first time. Their cytotoxicity toward human mammary adenocarcinoma, human myeloma, hamster and murine and fibroblasts as well as their antioxidant and mutagenic effects on a Salmonella tester strain were analyzed. All four substances showed comparable IC₅₀values in terms of suppression of tumor cell growth, which were from two- to ninefold lower comparing with those of fibroblasts. To identify the features of spatial orientation and noncovalent interactions of the new phosphonium betaines in the binding site of Cdc25B, a molecular docking analysis was carried out. It showed that the interactions of the analyzed compounds with a Cdc25B model binding site are characterized by the presence of a large number of acceptors (fluorine and oxygen atoms, forming halogen and hydrogen bonds) and by participation of pi-systems and phosphorus in specific electrostatic interactions that may result in inhibition of enzymes of the Cdc25 family. In addition, compounds 5 and 6 (especially the latter) were found to be effective antioxidants protecting bacterial cells from H₂O₂-induced and spontaneous and mutagenesis at significantly lower concentrations (IC₅₀?=?0.09 to 1.8?μM) than those of derivatives 4 and 7 (86–92?μM). Taking into account these data (together with the good cytotoxic effect on cancer cells comparing with normal mammalian cells) we can propose compounds 5 and 6 as possible useful inhibitors of tumor cell growth and antioxidants.

Full text of DOI:10.1016/j.jfluchem.2016.10.014

Accepted Manuscript
Title: Phosphonium betaines derived from
hexafluoro-1,4-naphthoquinone: Synthesis and cytotoxic and
antioxidant activities
Author: Svetlana I. Zhivetyeva Olga D. Zakharova Ludmila P.
Ovchinnikova Dmitry S. Baev Irina Yu. Bagryanskaya Vitalij
D. Shteingarts Tatiana G. Tolstikova Georgy A. Nevinsky
Evgeny V. Tretyakov
PII:
S0022-1139(16)30301-3
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.jfluchem.2016.10.014
FLUOR 8879
To appear in:
FLUOR
Received date:
Revised date:
Accepted date:
9-9-2016
21-10-2016
21-10-2016
Please cite this article as: Svetlana I.Zhivetyeva, Olga D.Zakharova,
Ludmila P.Ovchinnikova, Dmitry S.Baev, Irina Yu.Bagryanskaya, Vitalij
D.Shteingarts, Tatiana G.Tolstikova, Georgy A.Nevinsky, Evgeny V.Tretyakov,
Phosphonium betaines derived from hexafluoro-1,4-naphthoquinone: Synthesis
and cytotoxic and antioxidant activities, Journal of Fluorine Chemistry
http://dx.doi.org/10.1016/j.jfluchem.2016.10.014
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Phosphonium
Betaines
Derived
from
Hexafluoro-1,4-
naphthoquinone: Synthesis and Cytotoxic and Antioxidant
Activities
Svetlana I. Zhivetyeva a, Olga D. Zakharova b, Ludmila P. Ovchinnikova c, Dmitry. S. Baev a, Irina
Yu. Bagryanskaya a, d, Vitalij. D. Shteingarts a**, Tatiana G. Tolstikova a, Georgy A. Nevinsky b, d, *
and Evgeny. V. Tretyakov a, d, *
,
a
N.N. Vorozhtsov Institute of Organic Chemistry, Russian Academy of Sciences, Novosibirsk
630090, Russia
b Institute of Chemical Biology and Fundamental Medicine, Novosibirsk 630090, Russia
c
Institute of Cytology and Genetics, Siberian Division of Russian Academy of Sciences, 10
Lavrentiev Ave., Novosibirsk 630090, Russia
d National Research University – Novosibirsk State University, Novosibirsk 630090, Russia

Graphical abstract
Abstract: Fluorinated derivatives of 1,4-naphthoquinones are highly potent inhibitors of Cdc25A and
Cdc25B phosphatases; they suppress the growth of tumor cells. Four derivatives of phosphonium
betaines derived from hexafluoro-1,4-naphthoquinone: (triphenyl[5,6,7,8-tetrafluoro-1-oxido-4-oxo-3-
(phenylimino)-3,4-dihydronaphthalen-2-yl]phosphonium)
(4),
((3,5-
difluorophenyl)(methyl)phenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-dihydronaphthalen-2-
yl)phosphonium) (5), ((2,5-difluorophenyl)(methyl)phenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-
1,4-dihydronaphthalen-2-yl)phosphonium)
(6)
and
((3,5-difluorophenyl)diphenyl(5,6,7,8-
tetrafluoro-3-oxido-1,4-dioxo-1,4-dihydronaphthalen-2-yl)phosphonium) (7) were synthesized for
the first time. Their cytotoxicity toward human mammary adenocarcinoma, human myeloma, hamster
and murine and fibroblasts as well as their antioxidant and mutagenic effects on a Salmonella tester
strain were analyzed. All four substances showed comparable IC50 values in terms of suppression of
tumor cell growth, which were from two- to ninefold lower comparing with those of fibroblasts. To
identify the features of spatial orientation and noncovalent interactions of the new phosphonium
betaines in the binding site of Cdc25B, a molecular docking analysis was carried out. It showed that
the interactions of the analyzed compounds with a Cdc25B model binding site are characterized by
the presence of a large number of acceptors (fluorine and oxygen atoms, forming halogen and
hydrogen bonds) and by participation of pi-systems and phosphorus in specific electrostatic
interactions that may result in inhibition of enzymes of the Cdc25 family. In addition, compounds 5
and 6 (especially the latter) were found to be effective antioxidants protecting bacterial cells from
H2O2-induced and spontaneous and mutagenesis at significantly lower concentrations (IC50 = 0.09 to
1.8 μM) than those of derivatives 4 and 7 (8692 μM). Taking into account these data (together with
the good cytotoxic effect on cancer cells comparing with normal mammalian cells) we can propose
compounds 5 and 6 as possible useful inhibitors of tumor cell growth and antioxidants.

Keywords: Fluorinated 1,4-naphthoquinones, triphenylphosphanes, phosphonium betaines,
cytotoxicity.
1. Introduction
Mammalian cell-to-cell communication and division are regulated by phosphorylation of
proteins, which is the result of a dynamic equilibrium between the activity of protein kinases and
phosphatases. The subfamily of dual-specificity phosphatases of proteins retains some of the
structural attributes of post-translational modifications of proteins, but these enzymes have a
unique ability to hydrolyze both phospho-serine/-threonine and phosphotyrosine residues on the
same protein substrate. Key members of this family of dual-specificity phosphatases are Cdc25
phosphatases controlling the cell cycle. Phosphatases of the Cdc25 subfamily remove phosphate
groups from cyclin-dependent kinases (CDKs) and thereby activate complexes cyclin-CDK
regulating the cell cycle [1,2]. Cdc25A and Cdc25B are strongly expressed in many tumor cells,
frequently showing correlations with a higher grade or aggressiveness of tumors and unfavorable
prognosis [3]. The hypothetical involvement of Cdc25 phosphatases in tumor cells growth makes
these cell cycle-regulatory proteins possible therapeutic targets in cancer [4,5].
Small-molecule inhibitors have served as effective tools for elucidation of the function of
kinases and phosphatases involved in certain cellular signaling pathways because the effects of
such inhibitors are usually nonquantal and reversible, and these molecules easily penetrate the
plasma membrane. Some of the many tested substances have been shown to suppress activity of
the enzyme of Cdc25 subfamily [6,7], in particular, NSC 95397 (2,3-bis[(2-hydroxyethyl)thio]-
1,4-naphthoquinone (National Cancer Institute Library) was demonstrated to be the strongest
inhibitor of Cdc25 [8]. 7-Aminoquinoline-5,8-quinones and p-naphthoquinones represent the
backbones for the synthesis of possible Cdc25 inhibitors [8, 9], including NSC 663284 [10]. For
instance, Cpd5, or 2-(2-mercaptoethanol)-3-methyl-1,4-naphthoquinone, was found to be an
effective inhibitor of Cdc25 phosphatases [11].
Most quinones have been shown to suppress Cdc25 activity via sulfhydryl arylation in the
quinone moiety because of redox properties, but they can also cause formation of toxic species of
oxygen [12], which may be harmful to healthy tissues and thereby reduce the therapeutic utility of
quinones. One possible way to overcome the inherent cytotoxicity of quinones is to use derivative
of compounds that are stable when reduced and therefore pose a lower risk of the formation of
radicals and resultant nonspecific damage to cells. Fluorinated derivatives of 1,4-naphthoquinones
are less likely to cause formation of reactive oxygen species (ROS); among several candidates,
fluorinated Cpd5 – 5,6,7,8-tetrafluoro-2-(2-mercaptoethanol)-3-methyl-[1,4]-naphthoquinone (F-
Cpd5) was found to be the strongest inhibitor of Cdc25A and Cdc25B phosphatases [13-14,15].

Calculations predicted F-Cpd5 to have much more potent reductive properties comparing with
Cpd5 without generation of ROS [13]. F-Cpd5 is threefold more effective than Cpd5 in
suppressing the growth of human hepatoma Hep3B cells [13-15] and inhibits mitogen-stimulated
DNA replication 12-fold less strongly in healthy rat hepatocytes than in Hep3B cells [14].
The rationale aim in the research of Cdc25 phosphatases inhibitors is the search for better
inhibitors of tumor growth that have no mutagenic, carcinogenic, and cytotoxic effects. Because
all naphthoquinones are inhibitors of Cdc25 phosphatases [8-15], it is only logical to hypothesize
that some new polyfluorinated naphthoquinones should also inhibit these phosphatases.
Accordingly, we recently synthesized several novel polyfluorinated phenylamino-, n-alkylamino-,
and alkylthioderivatives of 1,4-naphthoquinone and characterized their antioxidant and mutagenic
properties by means of a special type of Salmonella cells as well as their toxicity toward primary
mouse fibroblasts comparing with murine and human tumor cells [16=17,18]. It is noteworthy that
for all these compounds, the effectiveness of inhibition of cancer cell growth at significantly lower
concentrations in comparison with healthy cells as well as the mutagenic and antioxidant activities
in the above-mentioned bacterial system substantially depend on the structure of functional
groups; only three or four of these compounds are promising, i.e., have all of these important
properties [16-18].
In the present study, we carried out experiments to synthesize four novel phosphonium betaine
derivatives of polyfluorinated 1,4-naphthoquinones including one with imino-functionality and
several compounds with the fluorinated phenyl-substituent attached to a phosphorous atom. We
compared these compounds in terms of specific toxicity toward cancer cells and prevention of
spontaneous and H2O2-induced mutagenesis in bacterial cells.
2. Results and Discussion
2.1. Reactions of hexafluoro-1,4-naphthoquinone with PPh3 and fluorophenylphosphanes
In our first report [19] we proposed that the reaction of hexafluoro-1,4-naphthoquinone (1) with
PPh3 in MeOH initially yields phosphonium salt A, in which a triphenylphosphonium group
effectively activates the neighboring position 3 for a nucleophilic attack, whereupon the rapid F3
substitution by methanol takes place to produce (after demethylation) phosphonium betaine 3 with a
~90% yield. To prove or to refute this idea, we conducted a reaction of hexafluoro-1,4-
naphthoquinone 1 with PPh3 (1:1) in C6H6 instead of MeOH (Scheme 1). In this case, the reaction
also produced phosphonium betaine 3 with a slightly lower (75%) yield comparing with the reaction
in MeOH. The second product that we isolated was also Ph3PO (~25%); its formation indicated that
the triphenylphosphanodefluorination reaction competed with reduction of the quinone moiety.

Because under normal conditions, the products of reaction of quinone 1 with PPh3 in C6H6 and
in MeOH were identical, for detection of the reaction intermediates, we designed a series of specific
experiments, which were carried out in thoroughly dried benzene in an atmosphere of dry argon.
19
31
The results were obtained rapidly (Scheme 1); in the F and P{1H} NMR spectra of the
solution immediately after mixing of quinone 1 with PPh3, we detected signals not only of betaine
3
but
also
(3,3,5,6,7,8-hexafluoro-1-oxido-4-oxo-3,4-dihydronaphthalen-2-
19
yl)triphenylphosphonium (2) and Ph3PF2 (38%, 54%, and 8%, respectively). In the NMR F
spectrum, the unusual in its structure betaine
2:1:1:1:1 respectively at δF –77.6 ppm (br. s, 2F, CF2), –138.8 and –139.2 ppm (ddd, 1F,
3JFF 20, 4JFF 10, and 5JFF  13 Hz; 5,8), –144.9, and –151.0 ppm (ddd, 1F, 3JFF 20, 4JFF
10 Hz;
6,7). In the NMR 31P{1H} spectrum of the solution, the singlet at δP 16.8 ppm
belonged to compound 2. As for the known PPh3F2 [20], its signals were readily identified in the
reaction mixture: a doublet at δF –39.1 ppm with JFP 665 Hz and a triplet at δP –54.5 ppm
JPF
665 Hz were observed in the NMR 19F and 31P{1H} spectra correspondingly
Thus, these experiments revealed two important products in the reaction of 1 with PPh3: betaine
2 showed five multiples in the intensity ratio


F


F


.
2 and PPh3F2. In structure 2, the >CF2 fragment is present, from which we can hardly expect high
reactivity. Therefore, it can be assumed that betaine 2 is in rapid equilibrium with minor
intermediate A, which, in turn, easily undergoes hydrolysis even in the presence of traces of water.
The formation of PPh3F2 indicates a parallel reaction of reduction of quinone 1 by PPh3 [21]. In
general, in all these experiments, the process of phosphanodefluorination mostly dominated over
reduction [ratio (2+3)/PPh3F2  12:1], that did not allow us to register the corresponding reduction
intermediate of type B. The formation of last was detected in case of chloranil where the reduction
process dominates [22].
To confirm the intended A2 equilibrium, the reaction of quinone
C6H6 was carried out followed by treatment with aniline (Scheme 1). According to 19F and
31P{1H} NMR data (~30%),† the reaction also resulted in
along with omnipresent betaine
1 with PPh3 in dried
,
3
the formation of previously unknown triphenyl[5,6,7,8-tetrafluoro-1-oxido-4-oxo-3-(phenylimino)-
3,4-dihydronaphthalen-2-yl]phosphonium (4, ~50%, isolated yield 36%), whose structure was
unambiguously solved by means of NMR and XRD data. The X-ray molecular structure and
selected bond lengths are depicted in Figure 1. The bond lengths and bond angles are the same as
the statistical means [23]. In solid 4, the benzoquinone ring has an envelopelike conformation
†
The formation of betaine 3 is result of hydrolysis of intermediate A by atmospheric moisture during working-up
procedure.

characterized by a 0.289(4) Å deviation of the C3 atom from the plane of the other atoms (with
standard deviations 0.030 Å). In the crystal, the C–F..., C=O..., and C–H... interactions lead to
formation of 3D networks. The atom-to-plane distances are 3.23 Å and 3.18 Å for interactions
between atoms F1, O2, and the fluorinated ring, and 2.89 Å for interactions between atom H5 and
phenyl ring C11–C16 (Fig. 2).
Previously, it was found that betaine 3 and its derivatives can effectively inhibit growth of
cancer cell (for RPMI 6228 cells, IC50 reached 3.4  0.3 μg/ml) [24]. At the same time, they show
substantial antioxidant activities, which make this type of compounds good candidates for use in
anticancer treatments [19]. In the present work, we tried to extend the range of type 3 compounds
by means of fluorinated phosphane PR1R2Ph (Scheme 2). Experiments revealed that in reactions of
1 with PR1R2Ph, the corresponding betaines formed successfully but with a low yield owing to the
greater predisposition of the analyzed phosphanes to oxidation. Accordingly, the reaction of
quinone 1 with (3,5-difluorophenyl)(methyl)phenylphosphane PMe(3,5-F2C6H3)Ph, along with the
desired betaine (3,5-difluorophenyl)(methyl)phenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-
dihydronaphthalen-2-yl)phosphonium (5) (~60%, isolated yield 38%), produced a substantial
amount of the corresponding phosphane oxide (~40%). Fortunately, it was found that the use of
DMSO as a solvent noticeably increased the phosphanodefluorination reaction rate and,
consequently, raised the yields of betaines. For example, the reaction of quinone 1 with (2,5-
difluorophenyl)(methyl)phenylphosphane or (3,5-difluorophenyl)diphenylphosphane resulted in
reaction mixtures containing betaines (2,5-difluorophenyl)(methyl)phenyl(5,6,7,8-tetrafluoro-3-
oxido-1,4-dioxo-1,4-dihydronaphthalen-2-yl)phosphonium
(6)
or
(3,5-
difluorophenyl)diphenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-dihydronaphthalen-2-
yl)phosphonium (7) (with 90% NMR yields), from which they were isolated with the yield of 67%
and 57%, respectively (Scheme 2).
2.2. Biological experiments
Naphthoquinone betaines 47 were analyzed for their ability to inhibit the growth of three
mammalian cell lines: human myeloma (RPMI 8226), tumor cell lines from human mammary
adenocarcinoma (MCF-7), and a hepatocellular carcinoma epithelial tumor (HepG2), as well as
normal Chinese hamster Ag 17 cells (AG) and healthy murine fibroblasts (LMTK).
At the first step of biological assessment of compounds 47, we study their ability to suppress
the growth of three tumor cell lines rather than mouse fibroblasts. Figure 3 shows representative

results on these compounds in the case of HEP, RPMI, and AG cells. The data obtained for
compounds 47 with all types of cells are shown in Table 1.
One can see that compounds 47 showed more or less comparable IC50 values in the case of
MCF-7 (4.7–8.5 M), RPMI (3.6–5.5 M), and HEP (4.5–10.6 M) cancer cells. Antitumor drugs
are regarded as potentially more effective when they are stronger growth suppressors of tumor cells
comparing with healthy mammalian cells. Consequently, we compared the actions of these
compounds on cancer cell lines and normal mouse (LMTK) and hamster (Ag-17) fibroblasts. All four
compounds inhibited the growth of healthy cells at 1.0- to 9.1-fold higher concentrations relative to
tumor cells. The biggest difference between the inhibition of growth of tumorous and healthy cells
was observed for compounds 5 and 6 (Table 1).
Recently, we compared the growth-inhibitory effects of F-Cpd5 with those of 19 novel
containing no phosphorus atoms polyfluoro-1,4-naphthoquinone derivatives [16-18]. Fifteen of them
suppressed the growth of tumor cells at a 2- to 72-fold lower concentration in comparison with F-
Cpd5 (IC50 = 14.8 ± 0.9 μM for RPMI cells and 173.0 ± 21.0 μM for MCF cells) [16-18]. IC50
values of compounds 47 are also ~1.9- to 2.7-fold and 30- to 38-fold less than those of F-Cpd5 in
the case of RPMI and MCF cells, respectively (Table 1).
To clarify the details of spatial orientation and noncovalent interactions of the novel
phosphonium betaines derived from hexafluoro-1,4-naphthoquinone in the binding site of Cdc25B,
a molecular docking analysis was conducted. Previously, several studies have revealed two catalytic
binding sites in Cdc25B structure that are available for interaction with small molecules [25,26].
The first one (His472 through Arg479) has a relatively compact cavity and can bind only small
anions. The second binding site is composed of residues 379–380, 397–399, 471, 484–485, 488–
489, 492, 500, 503, 505, and 507 and has a binding cavity suitable for polycyclic molecules. The
conserved structure of this region in the group of dual-specificity phosphatases, the anion binding,
and the reactive sulfhydryl group of Cys484 indicate that this region performs some important
function [25]. Predictions of binding modes of the phosphonium betaines under study are presented
in Figure 4.
The predicted spatial orientation of molecules analyzed in the binding site depends on structure
of the substituents. Compounds 4, 5, and 6 have the orientation of the naphthoquinone core toward
the outer hydrophilic area of the binding site, but compound 4 has an inverted conformation of the
naphthoquinone core. We observed a difference in orientations of the substituents in the inner
hydrophobic region of the binding site. For compound 4, this region is occupied by the

iminophenylic group of the naphthoquinone core, and the triphenylphosphonium part of the
molecule is not involved in the interactions. As for compounds 5 and 6, the inner region is occupied
by one of substituents of phosphonium: phenyl (5) and 2,5-difluorophenyl (6). Compound 7 is
inserted into the inner portion of the binding site by the naphthoquinone core. Fluorine atoms
forming halogen and hydrogen bonds with amino acid residues (Glu377, Arg485, Arg488, Arg492,
Asp397, Tyr382, Cys484, and others) of the binding site are important for binding of this type of
compounds to the Cdc25B model. Compounds 5 and 6 show participation of naphthoquinone in the
formation of hydrogen bonds and electrostatic interactions of the positively charged phosphorus
atom. It should be mentioned that -systems of the analyzed compounds participate in electrostatic
interactions with Glu377, Arg488, and Arg492. Therefore, the characteristics of the predicted
interactions of the analyzed compounds with the binding site of the Cdc25B model are determined
by the presence of a large number of acceptors: fluorine and oxygen atoms, forming halogen and
hydrogen bonds and by participation of -systems and phosphorus in electrostatic interactions. The
reactive sulfhydryl group of Cys484 (involved in interactions with 4, 6 and 7) may lead to blockage
of disulfide bond formation, which may be responsible for binding to another domain of the full-
length Cdc25 proteins, to cyclin/kinases, or other proteins [25].
It is known that some compounds interacting with many cell targets at the same time can be
polyfunctional and have cytoprotective properties, or conversely, may have mutagenic or
carcinogenic effects. Obviously, drugs are more successful when they are not mutagenic at
therapeutic concentrations. The mutagenic activity of compounds 47 was estimated by the Ames
test [27] based on S. typhimurium TA102 as described by Kemeleva et al. [28]. The mutagenicity in
the Ames assay is assessed by computing the frequency of reversion from histidine auxotrophy to
prototrophy under the influence of the substance being tested [27,28]. Figure 5 shows the
representative data on quinones 47, and all results are summarized in Table 2. One can see that
although all four compounds have comparable IC50 values for the inhibition of cancer cell growth,
they have different effectiveness at suppressing the spontaneous and Н2O2-induced mutagenesis.
Compound 6 showed the highest potency (IC50 = 0.26 M) of suppression of the spontaneous
formation of mutants, whereas this activity for compound 5 (1.8 M) was ~6.9-fold lower.
Compounds 4 and 7 inhibited the spontaneous mutagenesis very weakly. At a relatively high
concentration (50 μg/ml), compound 7 suppressed spontaneous mutagenesis only by 15.5% ± 1.2%
(50 μg/ml = 92.20 M), while compound 4 by 44.4% ± 3.5% (86 M; Table 2).
At low concentrations (0.04–0.09 M), all four quinones effectively decreased the H2O2-
dependent formation of mutants: from 137.7% to 100% of revertants (the number of revertants
observed in controls without H2O2 was set to 100%; Table 2). At the same time, the total inhibition
of peroxide-induced and spontaneous mutagenesis by these compounds was quite different.

Compound 6 (IC50 = 0.09 M) showed the best protection of the cells from peroxide-induced and
spontaneous mutagenesis (a decrease in formation of mutants from 100% to 50% in the presence of
H2O2), while compounds 5 (0.62 M) and 4 (86.0 M) were 6.9- and 956-fold less efficacious,
respectively. Compound 7 was found to offer the worst protection of cells either from spontaneous
or from total mutagenesis; in the presence of H2O2 at high concentration (50 μg/ml = 92.2 M), it
decreased formation of mutants from 100% to 50% only for 40.0% ± 3.0%. The efficiency of
protection of cells from the harmful effects of H2O2 can ranked as follows: 6 > 5 > 4 > 7 (descending
order; Table 2).
For many previously analyzed polyfluorinated alkylthio-, phenylamino-, and n-
alkylaminoderivatives of 1,4-naphthoquinone [16-18], most often (with infrequent exceptions), the
suppression of mutant formation in the presence of H2O2 was observed at similar or even greater
concentrations than those for the suppression of spontaneous mutagenesis. Figure 5 shows that
compounds 47 suppress the formation of mutants in the presence of H2O2 at concentrations lower
than those in the absence of this peroxide. Moreover, all four compounds highly effectively
decrease H2O2-driven mutagenesis (from 137.7% to 100%) at low and similar concentrations (0.04–
0.09 M) (Table 2). This finding suggests that all four compounds can effectively interact with
H2O2 thus efficiently reducing its concentration. The bigger decrease in mutagenesis in the presence
of H2O2 may be explained as follows: the products of H2O2-mediated oxidation of these compounds
are more efficacious traps of oxidative compounds. Nevertheless, the four initial compounds and
products of their H2O2-mediated modification differ substantially in the inhibition of spontaneous
mutagenesis. The differences observed here can be related to the structural features of these
compounds. The best antioxidants (6 and 5) contain the PPhMe moiety, which is absent in weaker
antioxidants 4 and 7.
These findings imply that quinones 47 are not mutagenic by themselves and inhibit
spontaneous mutagenesis as well as mutagenic effect of H2O2. Because compounds 6 and especially
5 did not enhance the spontaneous mutagenesis and effectively suppressed the mutagenic effect of
H2O2 at low concentrations, they can be considered efficient antioxidants. Our data indicate that
compounds 47 cause a more pronounced decrease in tumor cell growth comparing with the control
compound (F-Cpd5, a potent inhibitor of Cdc25 phosphatases) [13-15]. The structures of 47
resemble the structure of F-Cpd5 and of the above-mentioned previously described derivatives of
quinones [16-18]; accordingly, we may expect similar effects of these compounds on Cdc25
phosphatases and on tumor cells growth. It is obvious that all these compounds (including F-Cpd5)
perform a twofold function, acting both as inhibitors of cellular phosphatases and as antioxidants.
Compounds 5 and 6 (especially the latter) suppress spontaneous and H2O2-mediated mutagenesis
(IC50) on average at concentrations 5- and 41-fold lower, respectively, than concentrations at which

they inhibit the growth of tumor cells. Taking into account the efficacy of suppression of tumor cell
proliferation and antioxidant effects, compounds 6 and 5 (especially the latter) appear to be more
promising inhibitors of Cdc25 than 1,4-naphthoquinone.
3. Conclusions
We studied the reaction of hexafluoro-1,4-naphthoquinone with PPh3 in dry benzene and
identified two main fluorinated intermediates: the betaine with two germinal fluorine atoms and
Ph3PF2. Formation of these intermediates proves the existence of two parallel processes of
interaction of hexafluoro-1,4-naphthoquinone with PPh3. The first process is the
phosphanodefluorination yielding the intermediate betaine, and the second one is a redox
transformation. The intermediate betaine is highly reactive and in the presence of water or aniline
easily yields the corresponding stable phosphonium betaines. We showed that the fluorinated
phosphanes PR1R2Ph (R1 = Me, Ph; R2 = Ph, 3,5-, 2,5-F2C6H3) also effectively participate in the
reaction with hexafluoro-1,4-naphthoquinone and thus offer a variety of unknown polyfluorinated
betaines.
Fluorinated derivatives of 1,4-naphthoquinone are less active in terms of production of ROS
and may be promising inhibitors of Cdc25 as compared to 1,4-naphthoquinone. Our results suggest
that compounds 5 and 6 have the best potential: they show stronger cytotoxic effects on cancer cells
as compared to healthy mammalian cells and have the best antioxidant action on bacterial cells in
the presence and absence of H2O2.
4. Experimental section
4.1. Materials and methods
Benzene was dried by distillation over P2O5 and finally stored over molecular sieves.
PPh3 was crystallized from Et2O. Aniline was dried over CaH2 and redistilled at 0.09 kPa.
Reactions were conducted in Schlenk tubes.
19
The NMR spectra were recorded with a Bruker AV-300 (1H: 300.13 MHz, F: 282.36
MHz, 31P{1H}: 121.49 MHz) and AV-400 (1H: 400.13 MHz, 13C{1H}: 100.61 MHz)
spectrometers in the respective deuterated solvents relative to the residual proton chemical
1
13
shifts of chloroform (δH 7.25 ppm, δC 77.00 ppm) in H NMR spectra, Me4Si in C{1H}
19
31
NMR spectra, external C6F6 (δF = –162.9 ppm) in F NMR spectra, and H3PO4 in P{1H}
spectra. High-resolution mass spectra (HRMS) were measured with a DFS Thermo scientific
instrument (EI, 70 eV).

4.1.1. Reaction between quinone
betaine
PPh3 (0.493 g, 1.88 mmol) in anhydrous C6H6 (1 mL) was added slowly to a stirred
solution of quinone (0.500 g, 1.88 mmol) in anhydrous C6H6 (20 mL) at room temperature
1 and PPh3 in anhydrous C6H6 and fixing the formation of
2.
1
under dried argon. After 30 min, then the reaction mixture changed color from orange to
green, the 19F и 31P{1H} NMR spectra of solution were recorded under dried argon. 19F NMR
(282.36 MHz, C6D6 + C6H6): –77.6 (br. s, 2F, CF2), –138.8 and –139.2 [ddd, 1F, 3JFF
4JFF 10 Hz, 5JFF 5,8], –144.9 and –151.0 [ddd, 1F, 3JFF 20 Hz, 4JFF 10 Hz; F6,7
13 Hz;
ppm. 31P{1H} NMR (121.49 MHz, C6D6 + C6H6): 16.8 (
ppm
 20 Hz,


F


]
s
)
.
4.1.2.
Triphenyl[5,6,7,8-tetrafluoro-1-oxido-4-oxo-3-(phenylimino)-3,4-dihydronaphthalen-2-
yl]phosphonium (4).
PPh3 (0.493 g, 1.88 mmol) in anhydrous C6H6 (1 mL) was added slowly to a stirred
solution of quinone
1 (0.500 g, 1.88 mmol) in anhydrous C6H6 (20 mL) at room temperature
under dried argon. After 30 min, then the reaction mixture changed color from orange to
green, the 19F и 31P{1H} NMR spectra of solution were recorded under dried argon. Then
aniline (0.184 g, 1.98 mmol) was slowly added to benzene solution and after 30 min, the
NMR spectra of mixture were recorded under dried argon (Scheme 1). The solvent was
distilled off, the residue was purified by TLC (Sorbfil, chloroform–hexane, 6:1, Rf = 0.4, 1
time). After double TLC under the same conditions, betaine
4 was obtained as black
crystals, (0.394 g, 36%), m.p. 286–288 °C. Crystals suitable for XRD analysis were grown
1
from chloroform/hexane (1:4). H NMR (400.13 MHz, CDCl3): δ = 7.75–7.83 (m, 6H, C6H5),
3
7.56–7.63 (m, 3H, C6H5), 7.47–7.55 (m, 6H, C6H5), 7.1 [tm, 2H, JHH = 7.8 Hz, NC6H5], 6.9 [tm,
3
1H, 3JHH = 7.4 Hz, NC6H5], 6.1 [d, 2H, JHH = 8.3 Hz, NC6H5] ppm. 13C{1H} NMR (100.61 MHz,
2
3
2
CDCl3): δ = 180.2 [d, JCP = 10.2 Hz, C1], 174.8 [d, JCF = 6.4 Hz, C4], 154.1 [d, JCP = 3.5 Hz, C3],
150.7 (s, NC6H5), 146.2 [dm, 1JCF ~ 274 Hz, C5,8], 144.8 [dm, 1JCF ~ 262 Hz, 2JCF = 14.5 Hz, C6 or 7],
2
141.8 [dm, 1JCF ~ 260 Hz, JCF = 14.6 Hz, C6 or 7], 133.5 [d, 3JCP = 10.2 Hz, C6H5], 132.3 [d, 4JCP
=
2
1
2.8 Hz, C6H5], 128.8 [d, JCP = 12.8 Hz, C6H5], 128.4 (s, NC6H5), 123.9 [d, JCP = 93.2 Hz, C1’],
122.3 (s, NC6H5), 120.9 [dm, 2JCF = 13.7 Hz, C10F4O2], 118.9 (s, NC6H5), 79.2 [d, 1JCP = 106.2 Hz,
19
ortho
C2] ppm. F NMR (282.36 MHz, CDCl3): δ = –138.5, –138.8, –143.6, –149.0, [ddd, 1F,
J
FF
19.1÷20.5 Hz, metaJFF 7.6÷9.2 Hz, paraJFF 14.2÷14.3 Hz] ppm. 31P{1H} (121.49 MHz, CDCl3): δ = 15.6
(s) ppm. HRMS: calcd. for С34H20F4NO2P [M] 581.1162; found 581.1156; Anal. calcd. for
С34H20F4NO2P: C, 70.23; H, 3.47; N, 2.41; found: C, 70.47; H, 3.51; N, 2.49.

4.1.3.
(3,5-Difluorophenyl)(methyl)phenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-
dihydronaphthalen-2-yl)phosphonium (5).
A mixture of quinone 1 (0.158 g, 0.59 mmol), (3,5-difluorophenyl)(methyl)phenylphosphane
(0.140 g, 0.59 mmol) and methanol (2.6 mL) was stirred for 72 h at room temperature under
dried argon. A precipitate was centrifuged off, washed with methanol (2×0.5 mL) and dried on air
to give the title compound 5 (0.037 g, 13%) as bright yellow crystals. After evaporation of the
solvent, the dry residue was crystallized from methanol to yield an additional amount (0.070 g) of
the product, an overall yield of 5 being 0.107 g (38%), bright yellow crystals, m.p. 195–197 °C. 1H
NMR (300.13 MHz, CDCl3): δ = 7.63–7.79 (m, 3H, C6H5), 7.51–7.63 (m, 2H, C6H5), 7.17 [ddt,
3
3
4
3
4
2H, JHF = 14.0 Hz, JHP = 5.3 Hz, JHH = 1.6 Hz, C6F2H3], 7.0 [tt, 1H, JHF = 8.5 Hz, JHH = 2.0
2
Hz, C6F2H3], 2.54 [d, 3H, JHP = 13.9 Hz, CH3] ppm. 13C{1H} NMR (100.61 MHz, CDCl3): δ =
178.2 (m, C1,4), 174.0 [d, 2JCP = 3.7 Hz, C3], 163.2 [ddd, 1JCF = 254.9 Hz, 3JCF = 21.2 Hz, 3JCP = 11.6
1
2
1
2
Hz, C3’,5’], 147.5 [ddm, JCF ~ 272 Hz, JCF ~ 12 Hz, C5 or 8], 146.7 [ddm, JCF ~ 267 Hz, JCF ~ 11
Hz, C5 or 8], 145.3 [dddd, 1JCF ~ 264 Hz, 2JCF ~ 16 Hz, 2JCF ~ 12 Hz, 3JCF ~ 3 Hz, C6 or 7], 142.8 [dddd,
1JCF ~ 262 Hz, 2JCF ~ 16 Hz, 2JCF ~ 13 Hz, 3JCF ~ 3 Hz, C6 or 7], 133.8 [d, 4JCP = 3.0 Hz, C6H5], 132.1
3
2
1
[d, JCP = 11.0 Hz, C6H5], 129.6 [d, JCP = 13.2 Hz, C6H5], 128.9 [dt, JCP = 93.0 Hz, 3JCF = 8.0 Hz,
C1’], 121.9 [d, 1JCP = 92.7 Hz, C6H5], 119.0 [dm, 2JCF = 11.4 Hz, C10F4O2], 114.8 [dd, 2JCF = 27.7 Hz,
3’
2JCP = 11.6 Hz, C2’,6’], 114.8 [dd, 2JCF = 11.3 Hz, C4’], 108.8 [ddd, 2JCF
5’ = 24.9 Hz, 4JCP = 1.3
and CF
1
1
Hz, C4’], 89.1 [d, JCP = 99.5 Hz, C2], 12.8 [d, JCP = 60.2 Hz, CH3] ppm. 19F NMR (282.36 MHz,
CDCl3): δ = –105.8 (m, 2F, C6F2H3), –137.9, –140.6, –144.4, –149.1, [ddd, 1F, orthoJFF 19.4÷20.0 Hz,
meta
para
4
J
FF
9.4÷11.3 Hz,
J
FF
13.4÷13.5 Hz] ppm. 31P{1H} (121.49 MHz, CDCl3): δ = 14.6 [t, JPF 7.0
Hz] ppm. HRMS: calcd. for С23H11F6O3P [M] 480.0345; found 480.0346; Anal. calcd. for
С23H11F6O3P: C, 57.52; H, 2.31; F, 23.73; P, 6.45; found: C, 57.58; H, 2.52; F, 23.78; P, 6.30.
4.1.4.
(2,5-Difluorophenyl)(methyl)phenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-
dihydronaphthalen-2-yl)phosphonium (6)
A mixture of quinone 1 (0.088 g, 0.33 mmol), (2,5-difluorophenyl)(methyl)phenylphosphane
(0.078 g, 0.33 mmol) and DMSO (1.0 mL) was stirred for 7 h at room temperature under dried
31
argon and analyzed by 19F NMR and P{1H} NMR (Scheme 2). To solution was quickly added
(2,5-difluorophenyl)(methyl)phenylphosphane (0.016 g, 0.07 mmol) in DMSO (1.0 mL). The
resulting mixture was stirred for 1 h at room temperature under dried argon. Water (~8 mL)
was added; a precipitate was centrifuged off, washed with water (16 mL) and dried on air. The dry
residue was crystallized from methanol to yield the title compound 6 (0.107 g, 67%) as bright
yellow crystals, m.p.213–215 °C. 1H NMR (400.13 MHz, CDCl3): δ = 7.63–7.75 (m, 3H, C6H5),
7.52–7.61 (m, 2H, C6H5), 7.24–7.33 (m, 1H, C6F2H3), 7.12–7.23 (m, 1H, C6F2H3), 6.87–7.00 (m,

1H, C6F2H3), 2.62 [dd, 3H, 2JHP = 14.4 Hz, 5JHF = 1.8 Hz, CH3] ppm. 13C{1H} NMR (100.61 MHz,
1
3
CDCl3): δ = 178.1 (m, C1,4), 173.7 [d, 2JCP = 3.8 Hz, C3], 159.5 [dd, JCF = 247.4 Hz, JCP = 2.6 Hz,
C5’], 158.6 [ddd, 1JCF = 247.6 Hz, 2JCP = 16.0 Hz, 4JCF = 2.2 Hz, C2’], 147.5 [dd, 1JCF ~ 268 Hz, 2JCF =
10.1 Hz, C5 or 8], 146.7 [dd, 1JCF ~ 268 Hz, 2JCF = 10.7 Hz, C5 or 8], 144.8 [dddd, 1JCF ~ 264 Hz, 2JCF ~ 17
Hz, 2JCF ~ 13 Hz, 3JCF ~ 3 Hz C6 or 7], 142.8 [dddd, 1JCF ~ 262 Hz, 2JCF ~ 16 Hz, 2JCF ~ 13 Hz, 3JCF ~ 3
Hz C6 or 7], 133.7 [d, 4JCP = 3.1 Hz, C6H5], 131.8 [d, 3JCP = 11.2 Hz, C6H5], 129.6 [d, 2JCP = 13.4 Hz,
2
3
4
1
C6H5], 122.5 [ddd, JCF = 24.2 Hz, JCF = 9.2 Hz, JCP = 2.5 Hz, C4’], 121.8 [d, JCP = 94.1 Hz, C1’],
2
3
2
121.0 [ddd, JCF = 26.6 Hz, JCF = 8.3 Hz, 3JCP = 3.5 Hz, C3’], 119.1 [dm, JCF = 11.5 Hz, C10F4O2],
118.2 [dt, 2JCF = 25.3 Hz, 3JCF ~ 2JCP = 7.5 Hz, C6’], 116.5 (s, C3), 113.9 [ddd, 1JCP = 91.2 Hz, 2JCF
=
3
1
1
19.6 Hz, JCF = 6.8 Hz, C1’], 88.6 [d, JCP = 101.5 Hz, C2], 11.6 [dd, JCP = 61.0 Hz, 4JCF = 4.2 Hz,
CH3] ppm. 19F NMR (282.36 MHz, CDCl3): δ = –108.7 (m, C6F2H3), –116.7 (m, C6F2H3), –138.6, –
ortho
meta
para
141.4, –145.2, –149.9, [ddd, 1F,
J
FF
19.5÷20.0 Hz,
J
FF
9.3÷11.6 Hz,
J
FF
13.1÷13.4 Hz]
ppm. 31P{1H} (121.49 MHz, CDCl3): δ = 11.2 [d, 3JPF = 3.6 Hz] ppm. HRMS: calcd. for С23H11F6O3P
[M] 480.0345; found 480.0353; Anal. calcd. for С23H11F6O3P: C, 57.52; H, 2.31; F, 23.73; P, 6.45;
found: C, 57.47; H, 2.57; F, 23.77; P, 6.38.
4.1.5. (3,5-Difluorophenyl)diphenyl(5,6,7,8-tetrafluoro-3-oxido-1,4-dioxo-1,4-dihydronaphthalen-
2-yl)phosphonium (7).
A mixture of quinone 1 (0.06 g, 0.23 mmol), (3,5-difluorophenyl)diphenylphosphane (0.081 g,
0.27 mmol) and DMSO (1.6 mL) was stirred for 1 h at room temperature under dried argon and
19
31
analyzed by F NMR and P{1H} NMR (Scheme 2). Water (~8 mL) was added; a precipitate was
centrifuged off, washed with water (16 mL) and dried on air. The dry residue was crystallized from
methanol to yield the title compound 7 (0.07 g, 57%) as bright yellow crystals, m.p.231–233 °C. 1H
NMR (300.13 MHz, CDCl3): δ = 7.62–7.73 (m, 6H, C6H5), 7.50–7.59 (m, 4H, C6H5), 7.15 [ddt,
2H, 3JHF = 13.6 Hz, 3JHP = 5.4 Hz, 4JHH = 1.8 Hz, C6F2H3], 7.02 [tt, 1H, 3JHF = 8.5 Hz, 4JHH = 2.1
Hz, C6F2H3] ppm. 13C{1H} NMR (100.61 MHz, CDCl3): δ = 178.3 (m, C1,4), 173.7 [d, 2JCP = 3.2 Hz,
C3], 162.8 [ddd, 1JCF = 254.3 Hz, 3JCF = 21.0 Hz, 3JCP = 11.6 Hz, C3’,5’], 147.4 [dddd, 1JCF ~ 271 Hz,
3
4
1
2
3
2JCF ~ 11 Hz, JCF ~ 4 Hz, JCF ~ 2 Hz, C5 or 8], 146.5 [dddd, JCF ~ 270 Hz, JCF ~ 11 Hz, JCF ~ 3
Hz, 4JCF ~ 2 Hz, C5 or 8], 145.3 [dddd, 1JCF ~ 230 Hz, 2JCF ~ 16 Hz, 2JCF ~ 12 Hz, 3JCF ~ 3 Hz, C6 or 7],
142.7 [dddd, 1JCF ~ 230 Hz, 2JCF ~ 16 Hz, 2JCF ~ 12 Hz, 3JCF ~ 3 Hz, C6 or 7], 133.5 [d, 4JCP = 3.4 Hz,
3
2
1
C6H5], 133.5 [d, JCP = 11.0 Hz, C6H5], 129.4 [d, JCP = 13.2 Hz, C6H5], 128.0 [dt, JCP = 93.1 Hz,
1
2
3JCF = 8.1 Hz, C1’], 121.1 [d, JCP = 93.4 Hz, C1’], 119.3 [dm, JCF = 11.0 Hz, C10F4O2], 116.3 [dd,
3’
2JCF = 28.0 Hz, 2JCP = 11.1 Hz, C2’,6’], 108.6 [ddd, 2JCF
5’ = 25.0 Hz, 4JCP = 1.7 Hz, C4’], 88.6 [d,
and CF
1JCP = 100.1 Hz, C2] ppm. 19F NMR (282.36 MHz, CDCl3): δ = –106.8 (m, 2F, C6F2H3), –138.4, –
ortho
meta
para
140.7, –145.0, –149.8, [ddd, 1F,
J
FF
19.4÷20.1 Hz,
J
FF
9.1÷11.2 Hz,
J
FF
13.6÷13.7 Hz]

ppm. 31P{1H} (121.49 MHz, CDCl3): δ = 15.4 [t, 4JPF = 7.0 Hz] ppm. HRMS: calcd. for С28H13F6O3P
[M] 542.0501; found 542.0494; Anal. calcd. for С28H13F6O3P: C, 62.01; H, 2.42; F, 21.02; P, 5.71;
found: C, 62.36; H, 2.68; F, 20.83; P, 5.75.
4.2. Crystallographic Study
XRD data for 4 were obtained on a Bruker Kappa Apex II CCD diffractometer using φ, ω scans
of narrow (0.5°) frames with Mo Kα radiation (λ = 0.71073 Å) and a graphite monochromator. The
structures were solved by direct methods and refined by full-matrix least-squares method against all
F2 in anisotropic approximation using the SHELX-97 programs set [29]. The H atoms positions
were calculated with the riding model. Absorption corrections were applied empirically using
SADABS programs. Crystals of compound 4 are triclinic, space group P-1, a = 10.463(1), b =
10.519(1), c = 13.224(1) A,  = 82.623(5),  = 69.266(5), γ = 79.776(5) , V = 1336.1(3) A3, Z = 2,
C34H20F4NO2P, Dc = 1.445 г/см3,  = 0.165 mm-1, F(0 0 0) = 596, independent reflections 4758, for
1722 reflections with I > 2σ: R = 0.0534, wR2 = 0.1431, S = 1.06 (for all reflections: R = 0.0715,
wR2 = 0.1743). The obtained crystal structures were analyzed for short contacts between non-
bonded atoms using the PLATON program [30]. Tables listing detailed crystallographic data, atomic
positional parameters, and bond lengths and angles are available as CCDC 1452176 from the
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
4.3. Biological activity
4.3.1. Determination of mutagenisity of compounds
In the Ames test, the histidine-dependent strain of S. typhimurium TA102 having a mutation at
the histidine operon was used. [27]. The mutagenic activities of compounds 47 were analyzed by
the standard method (without metabolic activation) [27] according to [28]. TA102 (a liquid culture)
was prepared by growth of cells (from a frozen stock) during 16-h at 37 °C in LB medium
containing penicillin. The cells obtained were plated on minimal glucose agar containing histidine
and antibiotics at the density sufficient to obtain isolated colonies. A separate bacterial colony was
inoculated into 5 mL LB medium supplemented with tetracycline (2 g/mL) and ampicillin (50
g/mL); the growth of cells was performed with shaking (130 rpm) for 15 h at 37°C. The Ames test
was performed using the double-layer method according to [27]. To the overnight culture of
bacteria (100 l) one of the tested compounds was added (in different concentrations). In some
experiments 0.1 ml of 63 mM H2O2 (final concentration 3 mM), were mixed with 2 mL of liquid
0.6% top agar at 42° C. This mixture was poured onto plates containing a minimal medium
supplemented with 3% agar and 0.2% glucose; it was uniformly distributed on the surface of the
solid agar. The plates were incubated at 37°C for 48 h, and the revertants were counted. The cells

incubated with H2O2 in the absence of any compounds were served as positive controls, while cells
incubated in the absence of antioxidants and H2O2 was used as negative controls for mutation
induction. The data of experiments are expressed as mean  S.E. corresponding to at least 3
independent experiments.
4.3.2. Cytotoxicity assays
Tumor cell lines from human mammary adenocarcinoma MCF-7, human myeloma RPMI
8226, hepatocellular carcinoma epithelial tumor cells HepG2, as well as normal Chinese hamster
Ag 17 cells (AG) fibroblast cell line, and normal mouse fibroblasts (LMTK) were used. The cells
(~2x103 cells per well) were incubated at 37ºC for 24 h using RPMI 1640 or IMDM medium (5%
CO2) and then compounds 47 were added as in [31]. The relative amount of live cells (after 72 h
of cell incubation) was estimated using 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium
bromide according to a standard MTT-colorimetric test [31]. The compound concentration causing
50% cell growth inhibition (IC50) was determined. The results were presented as mean  S.E. of
three independent experiments.
4.4. Docking studies
The docking analysis of molecules was carried out using Autodock Vina [32]. Ligand
molecules were sketched in 3D format using OpenBabel module of PyRx GUI. Merck molecular
force field minimization algorithm [33] was used to produce low-energy conformers. Cdc25B
model (PDB ID 1QB0, 1.91 Å resolution) and phosphonium betaines were prepared by adding
hydrogens, adding missing residues, and converting to PDBQT format. All crystallographic water
molecules and other chemical components were omitted, the right bond orders as well as charges
and atom types were assigned. The grid coordinates for molecular docking was chosen according to
Cdc25B catalytic pockets study [25]. Molecules were docked using Vina with exhaustiveness grade
8, with up to 9 poses saved per molecule. The docking procedure was carried out for the unchanged
conformation of the receptor and flexible ligand molecules. The lowest energy modes were selected
and the ligands interactions with binding site were determined. The UCSF Chimera 1.10 and
Accelrys Discovery Studio Visualizer 2016 were utilized for docking and interaction visualization.
Acknowledgements
This work was financially supported by the Siberian Branch of RAS (III.2P.1; The project
1.7.15), Foundation for Basic Research (No. 16-33-00005 mol_a). Authors would like to
acknowledge the Multi-Access Chemical Service Center SB RAS for spectral and analytical
measurements.

References
[1] J. Pines, Four-dimensional control of the cell cycle, Nat. Cell. Biol. 1 (1999) E73–E79.
[2] P.R. Mueller, T.R. Coleman, A. Kumagai, W.G. Dunphy, Myt1: a membrane-associated
inhibitory kinase that phosphorylates Cdc2 on both threonine-14 and tyrosine-15, Science 6
(1995) 86–90.
[3] K. Kristjansdottir, J.K. Rudolph, Cdc25 phosphatases and cancer, J. Chem. Biol. 11 (2004)
1043–1051.
[4] R. Boutros, C. Dosier, B. Ducommun, The when and wheres of CDC25 phosphatases, Curr.
Opin. Cell. Biol. 18 (2006) 185–191.
[5] R. Boutros, V. Lobjois, B. Ducommun, CDC25 phosphatases in cancer cells: key players? Good
targets? Nat. Rev. Cancer 7 (2007) 495–507.
[6] J.W. Eckstein, Cdc25 as a potential target of anticancer agents, Invest. New. Drugs 18 (2000)
149–156.
[7] K.E. Pestell, A.P. Ducruet, P. Wipf, J.S. Lazo, Small molecule inhibitors of dual specificity
protein phosphatases, Oncogene 19 (2000) 6607–6612.
[8] J.S. Lazo, K. Nemoto, K.E. Pestell, K. Cooley, E.C. Southwick, D.A. Mitchell, W. Furey, R.
Gussio, D.W. Zaharevitz, B. Joo, P. Wipf, Identification of a potent and selective pharmacophore
for Cdc25 dual specificity phosphatase inhibitors, Mol. Pharmacol. 61 (2002) 720–728.
[9] P. Wipf, B. Joo, T. Nguyen, J.S. Lazo, Synthesis and biological evaluation of caulibugulones A-
E, Org. Biomol. Chem. 2 (2004) 2173–2174.
[10] J.S. Lazo, D.C. Aslan, E.C. Southwick, K.A. Cooley, A.P. Ducruet, B. Joo, A. Vogt, P. Wipf,
Discovery and biological evaluation of a new family of potent inhibitors of the dual specificity
protein phosphatase Cdc25, J. Med. Chem. 44 (2001) 4042–4049.
[11] Y. Nishikawa, B.I. Carr, M. Wang, S. Kar, F. Finn, P. Dowd, Z.B. Zheng, J. Kerns, S.
Naganathan, Growth inhibition of hepatoma cells induced by vitamin K and its analogs, J. Biol.
Chem. 270 (1995) 28304–28310.
[12] P. Wardman, Electron transfer and oxidative stress as key factors in the design of drugs
selectively active in hypoxia, Curr. Med. Chem. 8 (2001) 739–761.
[13] S.W. Ham, J.I. Choe, M.F. Wang, V. Peyregne, B.I. Carr, Fluorinated quinoid inhibitor:
possible "pure" arylator predicted by the simple theoretical calculation, Bioorg. Med. Chem.
Lett. 14 (2004) 4103–4105.
[14] S. Kar, M. Wang, S.W. Ham, B.I. Carr, Fluorinated Cpd 5, a pure arylating K-vitamin
derivative, inhibits human hepatoma cell growth by inhibiting Cdc25 and activating MAPK,
Biochem. Pharmacol. 72 (2006) 1217–1227.

[15] H. Park, B.I. Carr, M. Li, S.W. Ham, Fluorinated NSC as a Cdc25 inhibitor, Bioorg. Med.
Chem. Lett. 17 (2007) 2351–2354.
[16] O.A. Zakharova, L.I. Goryunov, N.M. Troshkova, L.P. Ovchinnikova, V.D. Shteingarts, G.A.
Nevinsky, Cytotoxicity of new n-butylamino and sulfur-containing derivatives of polyfluorinated
1,4-naphthoquinone, Eur. J. Med. Chem. 45 (2010) 270–274.
[17] O.D. Zakharova, L.P. Ovchinnikova, L.I. Goryunov, N.M. Troshkova, V.D. Shteingarts, G.A.
Nevinsky, Cytotoxicity of new alkylamino- and phenylamino-containing polyfluorinated
derivatives of 1,4-naphthoquinone, Eur. J. Med. Chem. 45 (2010) 2321–2326.
[18] L.I. Goryunov, N.M. Troshkova, G.A. Nevinsky, V.D. Shteingarts, Synthesis of 2-
aminopentafluoro-1,4-naphthoquinone derivatives, Russ. J. Org. Chem. 45 (2009) 835–841.
[19] L.I. Goryunov, S.I. Zhivetyeva, G.A. Nevinsky, V.D. Shteingarts, Synthesis of
diphenyl(X)phosphonium betaines (X = CH3, C6H5, 2,5-F2C6H3) from hexafluoro-1,4-
naphthoquinone, ARKIVOC (2011) 185–191.
[20] (a) G.S. Reddy, R. Schmutzler, 31P and 19F nuclear magnetic resonance studies of phosphorus-
fluorine compounds, Z. Naturforsch. 25b (1970) 1199–1214.
(b) V.V. Grushin, Thermal stability, decomposition paths, and Ph/Ph exchange reactions of
[(Ph3P)2Pd(Ph)X] (X = I, Br, Cl, F, and HF2), Organometallics 19 (2000) 1888–1900.
[21] S.I. Zhivetyeva, G.A. Selivanova, L.I. Goryunov, I.Yu. Bagryanskaya, V.D. Shteingarts,
Triphenylphosphanodefluorination of fluoranil and its derivatives, J. Fluorine Chem. 180 (2015)
21–32.
[22] F. Ramirez, S. Dershowitz, The structure of quinone-donor adducts. I. The action of
triphenylphosphine on p-benzoquinone, 2,5-dichloro-p-benzoquinone and chloranil, J. Am.
Chem. Soc. 78 (1956) 5614–5622.
[23] F.H. Allen, O. Kennard, D.G. Watson, L. Brammer, A.G. Orpen, R.Taylor, Tables of bond
lengths determined by X-ray and neutron diffraction. Part 1. Bond lengths in organic
compounds, J. Chem. Soc., Perkin Trans. II (1987) S1–S19.
[24] G.A. Nevinsky, O.D. Zakharova, L.I. Goryunov, V.D. Shteingarts, S.I. Zhivetyeva,
Phosphorus-containing fluorinated 1,4-naphthoquinone derivatives, possessing cytotoxic
naphthoquinone derivatives, possessing cytotoxic activity with respect to human cancer cells in
culture, RU Patent 2535676, 2014.
[25] R.A. Reynolds, A.W. Yem, C.L. Wolfe, M.R. Deibel Jr, C.G. Chidester, K.D. Watenpaugh,
Crystal structure of the catalytic subunit of Cdc25B required for G2/M phase transition of the
cell cycle, J. Mol. Biol. 293 (1999) 559–568.
[26] A. Lavecchia, C. Di Giovanni, A. Pesapane, N. Montuori, P. Ragno, N.M. Martucci, M.
Masullo, E. De Vendittis, E. Novellino, Discovery of new inhibitors of Cdc25B dual specificity
phosphatases by structure-based virtual screening, J. Med. Chem. 55 (2012) 4142–4158.
[27] D.M. Maron, B.N. Ames, Revised methods for the Salmonella mutagenicity test, Mutat. Res.
113 (1983) 173–215.
[28] E.A. Kemeleva, E.A. Vasunina, O.I. Sinitsyna, A.S. Khomchenko, M.A. Gross, Kandalintseva
N.B. A.E. Prosenko, G.A. Nevinskii, New promising antioxidants based on 2,6-dimethylphenol,
Bioorg. Khim. (Moscow) 34 (2008) 558–569.

[29] G.M. Sheldrick, SHELX-97 – Programs for Crystal Structure Analysis (Release 97-2),
University of Göttingen, Germany. – 1997.
[30] A.L. Spek, Single-crystal structure validation with the program PLATON, J. Appl. Crystallogr.
36 (2003) 7–13.
[31] T. Mosmann, Rapid colorimetric assay for cellular growth and survival: application to
proliferation and cytotoxicity assays, J. Immunol. Methods 65 (1983) 55–63.
[32] O. Trott, A.J. Olson, AutoDock Vina: improving the speed and accuracy of docking with a new
scoring function, efficient optimization and multithreading, J. Comput. Chem. 31 (2010) 455–
461.
[33] T.A. Halgren, Merck molecular force field. I. Basis, form, scope, parameterization, and
performance of MMFF94, J. Comput. Chem. 17 (1996) 490–519.

Fig. 1. The X-ray molecular structure of compound 4. Selected bond lengths (Å): P1C1 1.767(3),
N1C2 1.294(4), O1–C3 1.203(4), O2–C6 1.231(4), C1–C6 1.420(4), C1–C2 1.436(4), C2–C3
1.504(4), C3–C4 1.488(4), C4–C5 1.408(4), and C5–C6 1.519(4).
Fig. 2. A part of packing and atom... interactions in the crystal structure of 4 [Centroid...F1
3.579(3) and 3.814(3), Centroid....O2 3.619(4), and Centroid...H5 2.89 A].

RA, %
100
5 (AG)
80
60
40
20
0
6 (HEP)
7 (RPMI)
4 (RPMI)
0
10
20
30
[Compound], M
Fig. 3. Effects of fluorinated compounds 47 (relative activity, RA) on growth of various cells.
4 (–7.9 kcal/mol)
5 (–8.5 kcal/mol)
6 (–8.5 kcal/mol)
7 (–8.4 kcal/mol)
Fig. 4. The best binding modes of phosphonium betaines predicted by means of the AutoDock Vina
scoring function. The noncovalent interactions are indicated by dashed lines: green, hydrogen
bonds; blue, fluorine bonds; orange, electrostatic interactions. The hydrophobic intensity of the
binding site ranges from −3.00 (the least hydrophobic region: blue shading) to 3.00 (a highly
hydrophobic region: brown shading).

A
B
RA, %
RA, %
160
160
140
120
100
80
140
120
100
80
- H2O2
+ H2O2
60
60
- H2O2
40
40
20
20
+ H2O2
2
0
0
0
10 20 30 40 50 60 70 80
0
4
6
8
10 12 14 16 18 20
[Compound 11], M
[Compound 13], M
C
D
RA, %
160
RA, %
160
140
120
100
80
140
120
100
80
- H2O2
+ H2O2
60
60
40
40
- H2O2
20
20
+ H2O2
0
0
0
2
4
6
0
10 20 30 40 50 60 70 80 90 100
[Compound 18], M
[Compound 17], M
Fig. 5. Analysis of antioxidant and mutagenic effects of compounds 4 (A), 5 (B), 6 (C), and 7 (D)
by the standard Ames test using the S. typhimurium strain TA102 in the absence (□) or presence (■)
of 3 mM H2O2 (A–D). The number of revertants without H2O2 was set to 100%.

Scheme 1. Reaction of hexafluoro-1,4-naphthoquinone 1 with PPh3 in dried C6H6 followed by
hydrolysis or treatment with aniline.
Scheme 2. Reactions of hexafluoro-1,4-naphthoquinone 1 with fluorinated PR1R2Ph.

Table 1
Cytotoxicity (IC50) of fluorinated derivatives of phosphonium betaines 47.
IC50, M
Average ratio
of 1-3 and 4-5
Compound
MCF-7
RPMI 6228
HEP
LMTK
AG-17
1
2
3
4
5
5.2±0.4
4.6±0.3
5.5±0.4
5.7±0.3
5.8±0.3
6.4±0.4
7.9±0.5
5.5±0.4
4.5±0.3
7.1±0.5
7.9±0.4
7.6±0.9
7.6±0.6
7.7±0.5
10.4±0.7
7.6±0.6
19.8±0.7
23.0±1.6
50.0±3.0
19.8±1.1
4
5
6
7
1.34.4
1.15.0
1.39.1
1.03.6
*Mean ± SD from three independent experiments.
Table 2
IC50 characterizing suppression of spontaneous and H2O2-induced mutations by polyfluorinated
derivatives of 1,4-naphthoquinone.
IC50, M (% of inhibition at 50 μg/ml)*
Suppression of H2O2-induced and
spontaneous mutagenesis, %
Suppression
spontaneous
of
Compound
mutagenesis
(from 100 to 50%)
From 137.7
to 100%
From 100
to 50%
(44.4 ±3.5 %)**
1.8±0.09
0.09±0.006
0.057±0.004
0.04±0.003
0.05±0.004
86.0±7.0
4
5
6
7
0.62±0.05
0.09±0.005
(40.0 ±3.0 %)
0.26±0.03
(15.5 ±1.2 %)
*Mean ± SD from four independent experiments.
**The percentage of inhibition at 50 μg/ml; compound 4: 86 M, compound 7: 92.2 M.