Phenyl-substituted dihydropyrazines with DNA strand-breakage activity

Article abstract of DOI:10.1248/cpb.58.825

Monophenyl-substituted dihydropyrazines (Ph-DHP-1 to 4) of 2,3-dihydro-5,6-dimethylpyrazine (Me-DHP-1), which have the inductive effects of apoptosis and mutagenesis, were synthesized and their biological effect was investigated in terms of DNA strand-breakage. Differences between the phenyl- and methyl-substituted dihydropyrazines were examined.

Full text of DOI:10.1248/cpb.58.825

June 2010
Regular Article
Chem. Pharm. Bull. 58(6) 825—828 (2010)
825
Phenyl-Substituted Dihydropyrazines with DNA Strand-Breakage Activity
Shigeru ITO,a Shinji TAKECHI,b Kazuhide NAKAHARA,b Nobuhiro KASHIGE,c and
,b
*
Tadatoshi YAMAGUCHI
a Institute of Biomaterials and Bioengineering, Tokyo Medical and Dental University; Tokyo 101–0062, Japan: b Faculty of
Pharmaceutical Sciences, Sojo University; Kumamoto 860–0082, Japan: and c Faculty of Pharmaceutical Sciences,
Fukuoka University; Fukuoka 814–0180, Japan.
Received January 18, 2010; accepted March 15, 2010; published online March 17, 2010
Monophenyl-substituted dihydropyrazines (Ph-DHP-1 to 4) of 2,3-dihydro-5,6-dimethylpyrazine (Me-DHP-
1), which have the inductive effects of apoptosis and mutagenesis, were synthesized and their biological effect was
investigated in terms of DNA strand-breakage. Differences between the phenyl- and methyl-substituted dihy-
dropyrazines were examined.
Key words dihydropyrazine; radical generation; DNA breakage; biological effect; superoxide anion
In previous papers,1,2) we have showed the relationship be-
tween the chemical structures of dihydropyrazine (DHP) de-
rivatives and their DNA strand-breakage activity. For exam-
ple, we have found that the DNA strand-breakage activity is
remarkably enhanced upon addition of Cu2ꢀ. Additionally,
we observed an increase in the strand-breakage activity when
electron-donating methyl groups were added to the DHP
ring, and found that the ionization potential (IP) value was an
effective measure for the estimated activity. The use of the IP
value in this way offers the ability to actively design com-
pounds with the potential for DNA strand-breakage activity.
The phenyl-substituted dihydropyrazine, 2,3-dihydro-5-
methyl-6-phenylpyrazine (Ph-DHP-1), should have a much
higher DNA strand-breakage activity and produce a stronger
width 0.14 mT, time constant 0.3 s, amplitude 7ꢂ100, microwave power
10 mW, sweep time 2.0 min, and microwave frequency 9.427 GHz. The spec-
tra were recorded 30 min after mixing.
Synthesis of Dihydropyrazine Derivatives 2,3-Dihydro-5-methyl-6-
phenylpyrazine (Ph-DHP-1): To a solution of ethylenediamine (2.0 mmol)
in ether (5.0 ml) mechanically stirred on an ice bath, a solution of 1-phenyl-
propane-1,2-dione (2.0 mmol) in ether (5.0 ml) was slowly added dropwise.
The reaction mixture was stirred at room temperature until it became clear.
The mixture was then refluxed for 30 min. Potassium hydroxide was added
to the mixture to remove water and after filtration the filtrate was evaporated
under reduced pressure in nitrogen to give a solid product, which was puri-
fied by recrystallization from n-hexane. (Yellowish crystals. Yield 71.0%.
mp 36—37 °C. IR (Nujol) cmꢃ1: 1480 (CꢄN). 1H-NMR (500 MHz; CDCl3)
d: 2.10 (3H, t, Jꢄ1.46 Hz, CH3), 3.45—3.50 (2H, m, CH2), 3.54—3.60 (2H,
m, CH2), 7.40—7.60 (5H, m, aromatic-H). 13C-NMR (125 MHz; CDCl3) d:
24.5 (CH3), 45.0 (CH2), 45.5 (CH2), 127.5, 128.4, 129.6, 137.7 (aromatic-
C), 158.9 (CꢄN), 162.0 (CꢄN). FAB-MS (m/z): 173 (Mꢀꢀ1). HR-MS
intensity of radical species than the methyl-substituted dihy- Calcd for C11H13N2 (MꢀꢀH): 173.1079. Found: 173.1062.)
A mixture of (2R*,2S*)-2,3-dihydro-2,5-dimethyl-6-phenylpyrazine (Ph-
DHP-2ꢀ) and (3R*,3S*)-2,3-dihydro-3,5-dimethyl-6-phenylpyrazine (Ph-
dropyrazine, 2,3-dihydro-5,6-dimethylpyrazine (Me-DHP-1).
Therefore, we aimed to synthesize phenyl-substituted dihy-
DHP-2) was prepared in a similar manner from 1-phenylpropane-1,2-dione
dropyrazines (Ph-DHPs). It has already been shown3) that
and propane-1,2-diamine. (Yellowish oil. Yield 73.0%. IR (Nujol) cmꢃ1
:
1
Ph-DHPs cause cytotoxic and genotoxic damage that is dif-
ferent to that observed for Me-DHPs. In this report, the syn-
theses of four Ph-DHP derivatives and their DNA strand-
breakage activities are compared.
(CꢄN). H-NMR (500 MHz; CDCl3) d: 1.34 (6H, CH3), 2.07 (6H, CH3),
2.94—3.00 (2H, CH), 3.29—3.36 (2H, CH), 3.71—3.75 (1H, m, CH),
3.84—3.88 (1H, m, CH), 7.39—7.45 (10H, m, aromatic-H). 13C-NMR
(125 MHz; CDCl3) d: 19.05 (CH3), 19.18 (CH3), 24.44 (CH3), 24.67 (CH3),
49.79 (C2-CH), 50.58 (CH), 51.46 (CH2), 52.04 (CH2), 127.55, 127.63,
128.52, 129.70, 137.60, 137.81 (aromatic-C), 158.07 (CꢄN), 158.91
(CꢄN), 161.10 (CꢄN), 162.03 (CꢄN). FAB-MS (m/z): 187.1 (Mꢀꢀ1).
Experimental
The dihydropyrazine derivatives (Chart 1) employed were synthesized HR-MS Calcd for C12H15N2 (MꢀꢀH): 187.1235. Found: 187.1220.)
by the condensation of diketones and diamines. 2,3-Dihydro-5,6-dimethyl-
A mixture of 2,3-dihydro-2,2,5-trimethyl-6-phenylpyrazine (Ph-DHP-3ꢀ)
and 2,3-dihydro-3,3,5-trimethyl-6-phenylpyrazine (Ph-DHP-3) was pre-
pyrazine (Me-DHP-1) was synthesized by the method of Yamaguchi et al.1)
Similar methods were also used to generate the phenyl derivative, 2,3-dihy- pared in a similar manner from 1-phenylpropane-1,2-dione and 2-methyl-
dro-5-methyl-6-phenylpyrazine. The starting materials, 1-phenylpropane- propane-1,2-diamine. (Yellowish oil. Yield 70.4%. IR (Nujol) cmꢃ1: 1480
1
1,2-dione as the ketone and ethylenediamine, propane-1,2-diamine, 2- (CꢄN). H-NMR (500 MHz; CDCl3) d: 1.19 (6H, s, C2-CH3), 2.04 (3H, s,
methylpropane-1,2-diamine and trans-cyclohexane-1,2-diamine as diamines, C3-CH3), 3.47 (0.24H, d, Jꢄ6.3 Hz, C3-methylene), 3.55 (2H, s, C2-methyl-
were purchased from Wako Pure Chemical Ind., Ltd.
Assay of DNA Strand-Breakage Activity The method used to assess
ene), 7.37—7.44 (5H, m, aromatic-H). 13C-NMR (125 MHz; CDCl3) d: 24.8
(C3-CH3), 25.6, 25.8 (C2-CH3), 51.3 (C2), 56.8 (C1), 127.4, 127.6, 128.5,
the DNA strand-breakage activity of DHPs utilizes a covalently closed circu- 129.4, 129.6, 137.9 (aromatic-C), 156.2 (CꢄN), 161.8 (CꢄN). FAB-MS
lar duplex DNA (ccc-DNA) of plasmid pBR322 and has been described pre- (m/z): 201.1 (Mꢀꢀ1). HR-MS Calcd for C13H17N2 (MꢀꢀH): 201.1392.
viously.4)
Found: 201.1368. Signal assignments were confirmed by distortionless en-
Semi-empirical MO Calculations The ionization potential (IP) based hancement by polarization transfer (DEPT), heteronuclear multiple bond
on Koopman’s theorem, determined from PM3 molecular orbital (MO) cal- correlation (HMBC), heteronuclear multiple quantum correlation (HMQC),
culations, was used as a measure of the DNA strand-breakage activity.1) The
calculations were performed using MacSpartan Pro software.
correlation spectroscopy (COSY) spectra.)
trans-2-Methyl-3-phenyl-,5,6,7,8,9,10-hexahydroquinoxaline (Ph-DHP-
ESR Spectroscopy of Dihydropyrazines The ESR spectra were 4) was prepared in a similar manner from 1-phenylpropane-1,2-dione and
recorded on a JES-FA200 spectrometer (JEOL Co., Tokyo, Japan) using a trans-cyclohexane-1,2-diamine. (Yellowish oil. Yield 84.1%. IR (Nujol)
Mn2ꢀ marker as an external standard, and an ES-LC12 flat cell (JEOL Co.). cmꢃ1: 1445 (CꢄN). 1H-NMR (500 MHz; CDCl3) d: 1.39 (2H, m, C7H,
The spectra were measured in a 50 mM Tris–HCl buffer (pH 7.1) using 5,5-
dimethy-1-pyrroline N-oxide (DMPO) as a spin trapping agent, according to
a previous paper.4,5) The instrumental conditions used were: field center
C8H), 1.50 (2H, m, C7H, C8H), 1.86 (2H, s, C6H2), 2.10 (3H, s, C2-CH3),
2.39 (2H, s, C5H2), 2.68 (2H, s, C4aH, C8aH), 7.39—7.45 (5H, m, aro-
matic-H). 13C-NMR (125 MHz; CDCl3) d: 24.39 (C2-CH3), 25.44, 25.61
335.9 mT, scan width ꢁ5 mT, modulation frequency 100 kHz, modulation (C6H2, C7H2), 33.42, 33.57 (C5H2, C8H2), 58.75, 59.57 (C4aH, C8aH),
© 2010 Pharmaceutical Society of Japan
∗ To whom correspondence should be addressed. e-mail: tadayama@ph.sojo-u.ac.jp

826
Vol. 58, No. 6
Chart 1. The Synthesis of Phenyl-Substituted Dihydropyrazines
127.67, 128.56, 129.61, 137.65 (aromatic-C), 158.48, 161.60 (CꢄN). FAB-
MS (m/z): 227.1 (Mꢀꢀ1). HR-MS Calcd for C15H19N2 (MꢀꢀH): 227.1548.
Found: 227.1526.)
Results and Discussion
The dihydropyrazine derivatives (Fig. 1) employed were
synthesized by the condensation of a diketone as butane-2,3-
dione or 1-phenylpropane-1,2-dione, and a diamine as an
ethylenediamine derivative or cyclohexyldiamine. Diketone-
1 (butane-2,3-dione) reacted with diamine-1 (ethylenedi-
amine) to give Me-DHP-1 (2,3-dihydro-5,6-dimethyl-
pyrazine). In this case, no isomers were expected. However,
when the asymmetric species diketone-2 and diamine-2 or
diamine-3 were used as starting materials, several isomers
were produced. In the reaction of diketone-2 with diamine-
2, four isomers of Ph-DHP-2 [mixture of (R) and (S)] and
Ph-DHP-2ꢀ [mixture of (R) and (S)] were expected due to
the asymmetric carbon atom. In the reaction of diketone-2
with diamine-3, the formation of two isomers of Ph-DHP-3
ꢀ
and Ph-DHP-3 was predicted. The product ratios of these
Fig. 1. Influence of Added DMSO on Assay of DNA Strand-Breakage
Activity
isomers were estimated using the peak intensities from NMR
spectra of the products. For example, the product ratio of Ph-
ꢀ
DHP-2 and Ph-DHP-2 is 1 : 1 based on the signal intensi-
ties at d: 3.71—3.75 (1H, m, CH2-H) and d: 3.84—3.88 (1H, DHP-3ꢀ were obtained as product mixtures. The IP values of
m, CH2-H), which can be unambiguously assigned to one the mixture of Ph-DHP-2 and Ph-DHP-2ꢀ (Ph-DHP2, -2ꢀ),
proton of the methylene signal of Ph-DHP-2 and Ph-DHP- and the mixture of Ph-DHP-3 and Ph-DHP-3ꢀ (Ph-DHP-3,
2ꢀ, respectively. Similarly, the product ratio of Ph-DHP-3 -3ꢀ) can be roughly estimated to be ꢆ9.62 and ꢆ9.59, re-
and Ph-DHP-3ꢀ is 8 : 1 based on the signal intensities at d: spectively, based on the product ratio of the isomers. Thus,
3.55 (2H, s, C2-methylene) and d: 3.47 (0.24H, d, Jꢄ6.3 Hz, the order of IP values for the Ph-DHPs follows: R-DHP-3
C3-methylene); long-range coupling is observed between the (Ph-DHP-3, -3ꢀ)ꢅR-DHP-4ꢅR-DHP-2 (Ph-DHP-2, -2ꢀ)
C5-methyl and the C3-methylene for Ph-DHP-3ꢀ.
ꢅR-DHP-1, which is similar to that observed for the Me-
As shown in a previous paper,1) the PM3 method is the DHPs. However, the order suggested by the IP values was
most suitable for estimating the ionization potential (IP) later found to disagree with the actual order of DNA strand-
when using this value as a measure of the relative activity for breakage activity as described below, the reason for which is
DNA scission. The IP values obtained on the basis of the not known at this stage.
PM3 calculation are summarized in Table 1. The order of
The solubility of the Ph-DHPs in water decreases from
DNA strand-breakage activity of the Me-DHPs observed ex- Ph-DHP-4 to Ph-DHP-1, with Ph-DHP-1 being barely solu-
perimentally is R-DHP-3ꢅR-DHP-4ꢅR-DHP-2ꢅR-DHP-1, ble. As such, dimethyl sulfoxide (DMSO) may be added to
which agrees closely with the order expected based on the IP increase solubility and so its influence must be considered in
values.1)
any bioassay. For the assay of DNA strand-breakage activity
Ph-DHP-2 and Ph-DHP-2ꢀ as well as Ph-DHP-3 and Ph- as shown in Fig. 1, the activity of Me-DHP-1 employed as a

June 2010
827
Table 1. Ionization Potential (IP) Values Determined on the Basis of the PM3 Calculation
R-DHP-1
R-DHP-2
R-DHP-3
R-DHP-4
RꢄPh
9.6233
9.802
9.602
9.6237
9.580
9.599
9.596
9.770
(1 : 1)a)
9.776
(8 : 1)a)
9.721
DHP-2
DHP-2ꢀ
DHP-3
DHP-3ꢀ
RꢄMeb)
a) Product ratio of isomers. b) Obtained from a previous paper.1)
Table 2. DNA Strand-Breakage Activity of Ph-DHPs in the Absence or
Presence of Cu2ꢀ
Remaining ccc-DNA (%)
Conc.
(mM)
Test compound
Without Cu2ꢀ With Cu2ꢀ
for 3 h
for 1 h
0.1
100
95
1.0
99
89
47
21
10.0
0.1
100
95
Fig. 2. ESR Spectrum Obtained from the Reaction of Ph-DHP-1 (1.5 mM)
with CuCl2 (0.01 mM) in the Presence of DMPO (100 mM) at pH 7.1 (50 mM
Tris–HCl Buffer Containing 10% DMSO)
1.0
99
89
67
5
10.0
0.1
1.0
98
100
94
90
53
1
rized in Table 2.
The order of the activity in the presence of Cu2ꢀ is Ph-
DHP-2, 2ꢀꢅPh-DHP-1ꢅPh-DHP-3, 3ꢀꢅMe-DHP-1ꢅPh-
DHP-4. Although this order disagrees with that expected
based on the IP values, it is consistent with the activity based
on the radical signal peak heights for radical species in ESR
spectra, which were measured under reaction conditions of
DNA strand breakage. As shown in Fig. 2, all the compounds
(Me-DHP-1 and Ph-DHPs) revealed approximately similar
signal patterns for the DMPO-adducts,4) although with vary-
ing levels of intensity.
10.0
0.1
1.0
100
100
91
59
10.0
100
15
0.1
1.0
100
100
100
94
78
32
The spectrum was assigned to DMPO adducts of hydroxyl
(b: quartet of lines) and carbon-centered (a: doublet of sixtet
lines) radicals. The order of signal intensity for the carbon-
centered radical and the hydroxyl radical do not necessarily
parallel the order of DNA strand-breakage activity in some
cases (not described here). Full details of the ESR data for all
DHPs containing Me-DHPs and other compounds will be
published in another report, currently in preparation. Further-
10.0
Plasmid pBR322 ccc-DNA was incubated with various concentrations of DHP in
50 mM Tris–HCl bufer (pH 7.1) containing 10% DMSO at 37 °C for 1 h with 1 mM
CuCl2 or for 3 h without CuCl2. A small amount of remaining ccc-DNA indicates
strong breakage activity.
control is hardly affected by the amount of DMSO added more, as described in the our previous paper,3) the relative
(0—10%). However, for both Me-DHP-1 and Ph-DHP-1, lethal effect of Ph-DHPs on five strains (wild type, recA,
the activity was remarkably affected by the amount added uvrB, katE katG, sodA sodB) of Escherichia coli does not
(ca. 50%), with the phenyl-substituted DHP the more correspond with their radical signal intensity. Overall, the
strongly influenced. This result suggests that DMSO exerts lethal effect of the Ph-DHPs increased in the presence of
an influence as a scavenger of ·OH, indicating the activity of Cu2ꢀ. Interestingly, the Ph-DHPs, in contrast to the Me-
Ph-DHP-1 is mainly due to the generation of hydroxyl radi- DHPs, were toxic to the sodA sodB strain regardless of the
cals.2) Therefore, it is essential to reduce the amount of presence of Cu2ꢀ. We have previously predicted3) that Ph-
DMSO added, while maintaining the Ph-DHP solubility, in DHPs would affect cells differently than Me-DHPs because
order to limit its influence on the assay. Thus, the strand- of their greater solubility in fats, which should improve their
breakage activities of Ph-DHPs were observed in 50 mM ability to penetrate the cell membrane, as shown in Fig. 3.
Tris–HCl buffer (pH 7.1) containing 10% DMSO, as summa- The toxicity in the absence of Cu2ꢀ was Ph-DHP-4ꢅPh-

828
Vol. 58, No. 6
tails will be published in another report), the order of form-
ing 8-OHdG upon addition of Cu2ꢀ is Ph-DHP-3, 3 ꢅMe-
ꢀ
DHP-3 (2,3-dihydro-2,2,5,6-tetramethylpyrazine)ꢅPh-DHP-
ꢀ
2, 2 ꢅMe-DHP-2 (2,3-dihydro-2,5,6-trimethylpyrazine)ꢅ
Ph-DHP-4ꢅPh-DHP-1ꢅMe-DHP-1. This order agrees
with the order based on ESR signal intensity for the hydroxyl
and carbon-centered radicals.
In summary, the biological effect of DHPs was found to
depend on the amount of radical species generated from
DHPs and the sensitivity of bacterial cells to these species.
The chemical reactivity of DHPs with the component in vivo
was considered to also be a factor in their bacterial toxicity,
as shown with reactions of DHPs with ethylenediamine,6—8)
ketene9) or thiourea.10)
References
1) Yamaguchi T., Ito S., Kashige N., Nakahara K., Harano K., Chem.
Pharm. Bull., 55, 532—536 (2007).
2) Kashige N., Takeuchi T., Matsumoto S., Takechi S., Miake F., Yama-
guchi T., Biol. Pharm. Bull., 28, 419—423 (2005).
3) Takeda O., Takechi S., Ito S., Omori H., Katoh T., Yamaguchi T., Biol.
Pharm. Bull., 30, 1663—1667 (2007).
4) Yamaguchi T., Kashige N., Mishiro N., Miake F., Watanabe K., Biol.
Pharm. Bull., 19, 1261—1265 (1996).
5) Yamaguchi T., Matsumoto S., Watanabe K., Tetrahedron Lett., 39,
8311—8312 (1998).
6) Yamaguchi T., Ito S., Iwase Y., Watanabe K., Harano K., Heterocycles,
51, 2305—2309 (1999).
7) Yamaguchi T., Ito S., Iwase Y., Watanabe K., Harano K., Heterocycles,
53, 1677—1680 (2000).
8) Yamaguchi T., Eto M., Harano K., Kashige N., Watanabe K., Ito S.,
Teterahedron, 55, 675—686 (1999).
9) Nakahara K., Yamaguchi K., Yoshitake Y., Yamaguchi T., Harano K.,
Chem. Pharm. Bull., 57, 846—852 (2009).
10) Nakahara K., Takechi S., Yamaguchi K., Harano K., Yamaguchi T.,
Chem. Pharm. Bull., 57, 1371—1375 (2009).
Fig. 3. The Action of DHPs in Bacterial Toxicity towards E. coli
ꢀ
ꢀ
DHP-1ꢇPh-DHP-2, 2 ꢅPh-DHP-3, 3 , whereas in the
presence of Cu2ꢀ, it was Ph-DHP-3, 3ꢀꢅPh-DHP-2,
2ꢀꢅPh-DHP-1ꢅPh-DHP-4. In both the presence or absence
of Cu2ꢀ, since the sodA sodB strain is sensitive to Ph-DHPs,3)
the superoxide anion might be a reason for the bacterial toxi-
city. As with the DNA-strand breakage results, the order of
the ESR signal intensity for the hydroxyl and carbon-cen-
tered radicals did not appear to agree with the order of toxic-
ity for the Ph-DHPs for E. coli. The existence of the superox-
ide anion was recognized by the water-soluble tetrazolium
salt (WST)-1 assay.3) As described in our previous paper,2)
Ph-DHP-1 generated greater amounts of 8-OHdG than Me-
DHP-1 upon addition of Cu2ꢀ. At the present time (the de-