Antitumor studies. Part 1: Design, synthesis, antitumor activity, and AutoDock study of 2-deoxo-2-phenyl-5-deazaflavins and 2-deoxo-2-phenylflavin-5-oxides as a new class of antitumor agents

Article abstract of DOI:10.1016/j.bmc.2006.09.063

Novel 2-deoxo-2-phenyl-5-deazaflavins and 2-deoxo-2-phenylflavin-5-oxides were prepared as a new class of antitumor agents and showed significant antitumor activities against NCI-H 460, HCT 116, A 431, CCRF-HSB-2, andKB cell lines. In vivo investigation,

Full text of DOI:10.1016/j.bmc.2006.09.063

Bioorganic & Medicinal Chemistry 15 (2007) 242–256
Antitumor studies. Part 1: Design, synthesis, antitumor activity, and
AutoDock study of 2-deoxo-2-phenyl-5-deazaflavins and 2-deoxo-2-
phenylflavin-5-oxides as a new class of antitumor agents
Hamed I. Ali,a Keiichiro Tomita,a Eiichi Akaho,b Hiroto Kambara,b Shinji Miura,c
Hiroyuki Hayakawa,c Noriyuki Ashida,c Yutaka Kawashima,d Takehiro Yamagishi,d
Hisao Ikeya,d Fumio Yonedae and Tomohisa Nagamatsua,*
aDivision of Pharmaceutical Sciences, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences,
Okayama University 1-1-1, Tsushima-naka, Okayama 700-8530, Japan
bFaculty of Pharmaceutical Sciences, High Technology Research Center (Life Science Center),
and Center for Area Research and Development, Kobe Gakuin University, 518 Arise,
Ikawadani-cho, Nishi-ku, Kobe 651-2180, Japan
cBiology Laboratory, Research and Development Division, Yamasa Shoyu Co., Choshi, Chiba 288-0056, Japan
dMedicinal Research Laboratories, Taisho Pharmaceutical Co., LTD, 1-403, Yoshino-cho, Ohmiya-shi, Saitama 303-8530, Japan
eFaculty of Pharmaceutical Sciences, Kyoto University, Yoshida, Sakyo-Ku, Kyoto 606-8501, Japan
Received 15 August 2006; revised 23 September 2006; accepted 28 September 2006
Available online 30 September 2006
Abstract—Novel 2-deoxo-2-phenyl-5-deazaflavins and 2-deoxo-2-phenylflavin-5-oxides were prepared as a new class of antitumor
agents and showed significant antitumor activities against NCI-H 460, HCT 116, A 431, CCRF-HSB-2, andKB cell lines. In vivo
investigation, 2-deoxo-10-methyl-2-phenyl-5-deazaflavin exhibited the effective antitumor activity against A 431 human adenocar-
cinoma cells transplanted subcutaneously into nude mouse. Furthermore, AutoDock study has been done by binding of the flavin
analogs into PTK pp60c-src, where a good correlation between their IC50 and AutoDock binding free energy was exhibited. In
particular, 2-deoxo-2-phenylflavin-5-oxides exhibited the highest potential binding affinity within the binding pocket of PTK.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
During the last three decades there has been consider-
able interest in synthesis, functional elucidation, and
biological evaluation of 5-deazaflavins {5-deazaisoallox-
azines, pyrimido[4,5-b]quinoline-2,4(3H,10H)-diones}.
Actually, 5-deazaflavins have attracted great interest be-
cause of the first synthesis as potential flavin antagonists
of flavin models1 and of the discovery that they serve as
co-factors for several flavin catalyzed reactions2
(Scheme 1). They also have the potent broad-spectrum
activity against coccidiosis.3 Recently, we have prelimi-
narily reported the selective protein kinase C (PKC)
inhibitory activities of 5-deazaflavins and 2-deoxo-2-
phenyl-5-deazaflavins {2-phenylpyrimido[4,5-b]quino-
Scheme 1.
Keywords: Antitumor activity; Flavin analog; AutoDock; Protein
tyrosine kinase.
lin-4(10H)-ones}, and their effective growth inhibition
against cancer cell lines such as A 431 cells and HT
1080 cells.4 Antitumor activities of nitro-5-deazaflavin-
*
Corresponding author. Tel.: +81 86 251 7931; fax: +81 86 251
7926; e-mail: nagamatsu@pheasant.pharm.okayama-u.ac.jp
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.09.063

H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
243
pyrrolecarboxamide hybrid molecules5 and 5-amino-5-
deazaflavin derivatives revealed potential inhibitory
activity against L 1210 or KB cells.6 In addition, cyto-
toxicities of the nitro-5-deazaflavins were evaluated in vi-
tro for hypoxic and oxic Chinese hamster cells (V79).
The 6- and 8-nitro derivatives were generally more toxic
against hypoxic cells than oxic cells and showing
marked hypoxic selectivity.7
200 ꢁC under nitrogen atmosphere for 0.5–5 h according
to the known procedure13 as shown in Scheme 2. The
commercially unavailable N-monomethylated anilines
were synthesized from appropriate aniline derivatives
in 2 steps, by reaction of anilines with ethyl formate un-
der reflux for 15 h to get the N-arylformamides followed
by reduction using LiAlH4 in THF to afford the corre-
sponding N-monomethylated anilines.14
Computer-aided drug design tools can generate many
useful and powerful models that explain structure–activ-
ity relationship (SAR) observations in a quantitative
manner. Using a model with a given structure, medicinal
chemists can compute the activity of a molecule.8 Very
recently, the impressive technological advances in areas
such as structural characterization of biomacromole-
cules, computer sciences, and molecular biology have
made rational drug design feasible.9 The explosion of
genomic, proteomic, and structural information has
provided hundreds of new targets and opportunities
for future discovery process of lead drug.10 Many new
protein structures of pharmaceutical interest have been
elucidated, and a substantial number of those proteins
contain an embedded inhibitor. Meanwhile, several
drugs that were designed by intensive use of computa-
tional methods are currently under investigation
through clinical trials.11 AutoDock is one of the most
widely used docking programs in computational binding
studies. It is said to offer a reasonable result in compar-
ison with other methods that specialize either in fast
database searches or in computationally intensive
calculations.12
The intended 10-alkylated 2-deoxo-2-phenyl-5-deazaf-
lavins (2a–g) were prepared by reaction of 1a, b, d–h
with Vilsmeier reagent (N,N-dimethylformamide–phos-
phoryl chloride) at 90 ꢁC for 1–4 h, and the 10-unsubsti-
tuted derivatives (2h–j) were prepared from 6-anilino-2-
phenylpyrimidin-4(3H)-ones15 in the same conditions of
Vilsmeier reaction. This synthetic method is a useful
preparation for 5-deazaflavins from 6-(N-monoalkylan-
ilino) uracils.16 The 10-methyl-2-phenylpyrimido[4,5-
b]quinolin-4(10H)-one (2a) was also prepared by meth-
ylation of 2-phenylpyrimido[4,5-b]quinolin-4(10H)-one
(2h) with excess methyl iodide in the presence of anhy-
drous potassium carbonate in DMF at room tempera-
ture in 80% yield. On the other hand, 2-deoxo-2-
phenylflavin-5-oxides (3a–i) were prepared by nitrosa-
tive cyclization of 6-N-monoalkylanilino-2-phenylpyr-
imidin-4(3H)-ones (1a–h, j) with excess NaNO2 in
AcOH at 10–15 ꢁC for 2–5 h. However, such cyclization
of 1i was difficult due to the peri steric hindrance be-
tween the ethyl at the N-ethyl and the ortho methyl of
aniline moiety.
The structures of the products 1a–j were confirmed in
particular by the presence of an equivalent proton reso-
nance at the 5-position as a singlet signal at dH 4.8–5.3 in
1H NMR spectra and by the presence of the C5 carbon
signal at dC 86–87 in 13C NMR spectra (Table 1). The
cyclized 2-deoxo-2-phenyl-5-deazaflavins (2a–j) showed
a characteristic singlet signal in the lower field at dH
9.2–9.4 due to the proton at the 5-position of 2a–j and
a signal at dC 137–141 due to the carbon at the 5-posi-
tion. It is implying that the 5-position of 2a–j is the most
electron-deficient methine and very reactive to nucleo-
philes. Whereas, the 2-deoxo-2-phenylflavin-5-oxides
(3a–i) showed no such signals at the 5-position. On the
other hand, the UV absorption spectra of the anilino-
pyrimidones (1a–j) showed one absorption maximum
at 253–257 nm together with an absorption maximum
as shoulder at 272–274 nm. Whereas, the 10-alkylated
deazaflavins (2a–g) showed four absorption maxima at
283–297, 336–355, 414–431, and 433–456 nm together
with an absorption maximum as shoulder at 262–
288 nm, and the 10-unsubstituted deazaflavins (2h–i)
also showed three absorption maxima at 279–290,
364–370, and 374–385 nm, which were observed with
hypsochromic shift in the region of the longest wave-
length in comparison with those of 2a–g together with
an absorption maximum as shoulder at 254–256 nm.
In contrast with the above UV absorption spectra, the
2-deoxo-2-phenylflavin-5-oxides (3a–i) exhibited the
absorptions in the region of longer wavelength, that is,
four absorption maxima at 303–312, 362–373, 470–
496, and 500–528 nm together with an absorption
maximum as shoulder at 292–302 nm. All compounds of
The main goal of this study is the development of poten-
tial inhibitors synthesized newly, for example, 5-deazaf-
lavins, flavins, flavin-5-oxides, and their analogs as
antitumor agents, based on molecular modeling with
the investigation of SAR between new inhibitors and
protein tyrosine kinase (PTK). Thereupon, 2-deoxo-2-
phenylflavin-5-oxide derivatives were designed and syn-
thesized for the aim to get stronger binding affinity and
more hydrogen bonds to the binding pocket of the PTK
due to the polar part of the N-oxide moiety. In the pres-
ent paper, we describe the preparation of 2-deoxo-2-
phenyl-5-deazaflavins and 2-deoxo-2-phenylflavin-5-ox-
ides and their antitumor activities, and also the Auto-
Dock study for new inhibitors docked into the binding
pocket of PTK. In this study, compounds having poten-
tial antitumor and PTK inhibitory activities, for exam-
ple, IC50 values less than 25 lM, were docked into the
binding pocket of PTK using AutoDock 3.05 program.
2. Results and discussion
2.1. Chemistry
The required 6-N-monoalkylanilino-2-phenylpyrimidin-
4(3H)-ones (1a–j), which were used as precursors for
synthesis of 2-deoxo-2-phenyl-5-deazaflavins (2a–g)
and 2-deoxo-2-phenylflavin-5-oxides (3a–i), were syn-
thesized by fusion of 6-chloro-2-phenylpyrimidin-
4(3H)-one with appropriate N-alkyl anilines at 180–

244
H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
Scheme 2. General method for the preparation of 6-(N-alkylanilino)-2-phenylpyrimidin-4(3H)-ones(1a–j), 2-deoxo-2-phenyl-5-deazaflavins (2a–j),
and 2-deoxo-2-phenylflavin-5-oxides (3a–i). Reagents and conditions: (a) fusion at 180–200 ꢁC, 0.5–5 h, N2; (b) Vilsmmeier reagent (DMF–POCl3),
90 ꢁC, 1–4 h; (c) NaNO2, AcOH, 10–15 ꢁC, 2–5 h. The compounds 3j and 4a–j were computationally designed compounds.
3a–i showed red color owing to the presence of absorp-
tion maximum at ca. 515 nm in the longest wavelength.
they showed less antitumor activities in comparison with
adriamycin. Though it is also inferior to the antitumor
activity of Ara-C (IC50: 0.15 lM), many compounds
showed 0.5–2.0 lM potential growth inhibitory activi-
ties against CCRF-HSB-2 cell line. In case against KB
cell line, many compounds showed 0.5–5.0 lM potential
growth inhibitory activities and the compounds 2e and
2g (IC50: 0.5 lM) exhibited more potent antitumor
activities than that of Ara-C (IC50: 0.70 lM).
2.2. In vitro antitumor activities of deazaflavin and flavin
analogs against human tumor cell lines
The synthesized compounds 2a–j and 3a–i were tested
in vitro for growth inhibitory activities against various
cultured tumor cell lines. Five human tumor cell lines
including human lung cancer cell line (NCI-H 460), hu-
man colon carcinoma cell line (HCT 116), human ade-
nocarcinoma cell line (A 431), human T-cell acute
lymphoblastoid leukemia cell line (CCRF-HSB-2), and
human oral epidermoid carcinoma cell line (KB) were
used, and the antitumor agents of Ara-C, cisplatin,
and adriamycin were also used as positive controls in
this study. As can be seen from Table 2, many com-
pounds of flavin analogs such as 2-deoxo-2-phenyl-5-
deazaflavins (2) and 2-deoxo-2-phenylflavin-5-oxides
(3) have been found to show fairly good antitumor
activities. Although the tested flavin analogs showed
slightly poorer antitumor activities than cisplatin against
NCI-H 460 cell line, their activities for the compounds
2a, 2c, 3a, and 3f against HCT 116 cell line were higher
than cisplatin. Especially, compound 3f (IC50: 0.72 lM)
exhibited 3-fold more potent antitumor activity than cis-
platin (IC50: 2.23 lM). Further, against A 431 cell line
2.3. In vivo antitumor activity against subcutaneously
transplanted human adenocarcinoma A 431 xenograft
In vivo research has an advantage to discover drugs as
the nature and properties of a chemical tool cannot be
considered independently. Therefore, the antitumor
activities of compounds 2a and 2b against subcutane-
ously transplanted human adenocarcinoma A 431 xeno-
graft in mice were measured, and the tumor volume and
body weights of mice were evaluated. Thus the human
adenocarcinoma A 431 was subcutaneously transplant-
ed into BALB/c nude mice, and the administration of
test compounds was started on the day (day 1) when
the mice show tumor growth in the 100–300 mm3 range.
Since the test compounds prepared here showed poor
solubility in water, the administrations were done by
intratumoral (it), intraperitoneal (ip) or oral (po) meth-

H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
245
Table 1. UV and and 13C NMR spectroscopic data for the compounds (la–j, 2a–j, and 3a–i)
Compound
k
max/nm (loge/dm3molꢀ1 cmꢀ1
)
dC (75 MHz)a
1a
256 (4.62), 273 (4.53)
256 (4.68), 274 (4.54)
253 (4.73), 274 (4.42)
257 (4.68), 274 (4.54)
254 (4.73), 273 (4.45)
257 (4.67), 274 (4.49)
257 (4.52), 272 (4.42)
257 (4.66), 274 (4.49)
254 (4.73), 274 (4.43)
257 (4.65), 274 (4.51)
38.84 (6-NCH3), 86.66 (C5), 127.19 ðC040Þ, 127.43 ðC020 and C600Þ,
128.09 ðC030 and C050Þ, 129.14 ðC30 and C05Þ, 130.04 ðC20 and C06Þ,
132.06 ðC04Þ, 132.83 ðC10 Þ, 145.05 ðC100Þ, 156.20 (C6), 163.50 (C2),
165.99 (C@O)
1b
1c
1d
1e
1f
14.09 (6-NCH2–CH3), 45.79 (6-NCH2–CH3) 86.52 (C5), 127.61
ðC400Þ, 128.03 ðC020 and C600Þ, 128.68 ðC300 and C050Þ, 129.14 ðC30 and C05Þ,
130.21 ðC02 and C60 Þ, 131.99 ðC40 Þ, 132.96 ðC10 Þ, 143.47 ðC010Þ, 156.17
(C6), 163.24 (C2), 165.85 (C@O)
21.93 (6-NCH(CH3)2), 47.64 (6-NCH(CH3)2), 87.02 (C5), 127.99
ðC400Þ, 128.32 ðC020 and C600Þ, 129.16 ðC300 and C500Þ, 130.0 ðC30 and C05Þ,
131.12 ðC02 and C60 Þ, 131.95 ðC40 Þ, 133.06 ðC10 Þ, 139.46 ðC010Þ, 155.91
(C6), 163.75 (C2), 165.65 (C@O)
21.45 (400-CH3), 38.87 (6-NCH3), 86.58 (C5), 127.19 ðC200 and C060Þ,
128.09 ðC300 and C050Þ, 129.13 ðC400Þ, 130.66 ðC03 and C50 Þ, 132.00 ðC02
and C60 Þ, 132.91 ðC04Þ, 137.04 ðC10 Þ, 142.44 ðC100Þ, 156.15 (C6), 163.38
(C2), 165.88 (C@O)
17.85 (200-CH3), 37.59 (6-NCH3), 85.81 (C5), 127.36 ðC020Þ, 128.09
ðC03 and C05Þ, 128.39 ðC400Þ, 129.16 ðC04 , C300 and C600Þ , 129.16 ðC500Þ,
132.05 ðC02Þ, 132.92 ðC60 Þ, 136.50 ðC10 Þ, 142.44 ðC100Þ, 156.18 (C6),
163.25 (C2), 165.93 (C@O)
39.00 (6-NCH3), 55.90 (400-OCH3), 86.36 (C5), 115.30 ðC200 and C060Þ ,
128.05 ðC030 and C050Þ , 128.65 ðC30 and C05Þ, 129.15 ðC20 and C06Þ,
132.02 ðC04Þ, 132.89 ðC10 Þ, 137.87 ðC100Þ, 156.12 ðC400Þ, 158.63 (C6),
163.88 (C2), 165.82 (C@O)
1g
1h
1i
38.79 (6-NCH3), 55.76 (400-OCH3), 86.84 (C5), 112.98 ðC020Þ, 113.10
ðC040Þ, 119.62 ðC060Þ, 128.04 ðC30 and C50 Þ, 129.17 ðC20 and C06Þ, 130.73
ðC500Þ, 132.07 ðC40 Þ, 132.82 ðC10 Þ, 146.14 ðC100Þ, 156.16 (C6), 161.0
ðC030Þ, 163.46 (C2), 165.83 (C@O)
14.10 (6-NCH2–CH3), 21.48 (400-CH3), 45.77 (6-NCH2–CH3), 86.36
(C5), 127.99 ðC200 and C600Þ, 128.43 ðC300 and C500Þ, 129.14 ðC400Þ, 130.83
ðC03 and C05Þ, 131.95 ðC20 and C60 Þ, 133.01 ðC40 Þ, 137.47 ðC01Þ, 140.77
ðC010Þ, 156.07 (C6), 163.38 (C2), 165.70 (C@O)
14.09 (6-NCH2–CH3), 18.0 (20-CH3), 45.08 (6-NCH2–CH3), 85.99
(C5), 127.82 ðC00Þ, 128.02 ðC30 and C50 Þ, 128.31 ðC400Þ, 129.18 ðC030 and
C060Þ, 129.49 ðC0520Þ, 131.97 ðC20 and C60 Þ, 133.12 ðC40 Þ, 137.01 ðC01Þ,
141.81 ðC010Þ, 156.19 (C6), 163.01 (C2), 165.77 (C@O)
1j
38.92 (6-NCH3), 56.55 (300- and 500-OCH3), 61.35 (400-OCH3), 86.72
(C5), 104.81 ðC200 and C600Þ, 128.06 ðC02 and ðC60 Þ, 129.20 ðC30 and C05Þ,
132.17 ðC04Þ, 132.81 ðC01Þ, 137.40 ðC400Þ, 140.72 ðC100Þ, 154.37 ðC030 and
C050Þ, 156.28 (C6), 163.73 (C2), 165.89 (C@O)
2a
2b
2c
2d
2e
2f
265 (4.42), 284 (4.51), 340 (3.95),
425 (3.88), 453 (3.88)
33.28 (N10–CH3), 116.63 (C7), 118.28 (C9), 123.99 (C6), 125.71(C8),
128.42 ðC03 and C50 Þ, 130.08 ðC20 and C60 Þ, 132.32 ðC04Þ, 132.44 (C5a),
136.12 ðC01Þ, 137.86 (C5), 141.05 (C4a), 145.30 (C9a), 157.21 (C10a),
171.41 (C2), 172.01 (C@O)
266 (4.44), 286 (4.55), 338 (4.11),
425 (3.98), 453 (3.99)
13.48 (N10–CH2–CH3), 41.61 (N10–CH2–CH3), 116.37 (C7), 118.32
(C9), 124.45 (C6), 125.55 (C8), 128.49 ðC03 and C50 Þ , 130.24 ðC20 and
C06Þ , 132.48 ðC04Þ, 132.62 (C5a), 136.08 ðC10 Þ, 138.22 (C5), 140.11
(C4a), 145.29 (C9a), 156.93 (C10a), 171.76 (C2), 172.13 (C@O)
21.19 (7-CH3), 33.26 (N10–CH3), 116.45 (C7), 118.19 (C9), 124.15
(C6), 128.43 ðC03 and C50 Þ, 130.10 ðC20 and C60 Þ, 131.22 (C8), 132.30
ðC04Þ, 136.07 (C5a), 138.16 ðC10 Þ, 138.16 (C5), 139.38 (C4a), 144.78
(C9a), 151.52 (C10a), 171.20 (C2), 172.08 (C@O)
25.18 (9-CH3), 40.12 (N10–CH3), 117.64 (C9), 125.71 (C7), 125.82
(C6), 127.55 (C8), 128.41 ðC03 and C50 Þ, 130.09 ðC20 and C06Þ, 130.89
ðC04Þ, 132.41 (C5a), 138.04 ðC10 Þ, 139.32 (C5), 140.52 (C4a), 146.10
(C9a), 159.26 (C10a), 171.24 (C2), 172.11 (C@O)
30.05 (N10–CH3), 56.31 (7-OCH3), 99.74 (C9), 116.28 (C6), 117.99
(C8), 120.59 (C5a), 128.45 ðC03 and C50 Þ, 130.23 ðC20 and C06Þ, 132.25
ðC04Þ, 134.58 (C7), 136.51 ðC10 Þ, 140.99 (C5), 142.38 (C4a), 144.03
(C9a), 154.42 (C10a), 171.96 (C2), 172.97 (C@O)
33.96 (N10–CH3), 57.49 (8-OCH3), 99.46 (C7), 115.51 (C9), 117.78
(C6), 119.86 (C5a), 129.14 ðC03 and C50 Þ, 129.86 ðC20 and C06Þ, 132.49
ðC04Þ, 134.65 ðC10 Þ, 139.11 (C5), 144.16 (C4a), 144.78 (C9a), 157.74
(C10a), 166.63 (C8), 171.33 (C2), 172.12 (C@O)
288 (4.60), 297 (4.65), 340 (3.28),
431 (4.22), 456 (4.26)
288 (4.45), 297 (4.46), 344 (4.09),
431 (3.94), 455 (3.93)
268 (4.46), 292 (4.54), 342 (4.10),
428 (4.02), 454 (4.02)
262 (4.42), 283 (4.60), 355 (4.02),
414 (4.24), 433 (4.37)
(continued on next page)

246
H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
Table 1 (continued)
Compound
2g
k
max/nm (loge/dm3molꢀ1 cmꢀ1
)
dC (75 MHz)a
287 (4.72), 297 (4.74), 336 (4.37),
430 (4.35), 455 (4.37)
13.50 (N10–CH2–CH3), 21.14 (7-CH3), 41.55 (N10–CH2–CH3),
116.20 (C7), 118.35 (C9), 124.64 (C6), 128.44 ðC03 and C50 Þ, 130.22
ðC02 and C06Þ, 131.47(C8), 132.28 ðC40 Þ, 135.82 (C5a), 138.03 C01Þ,
138.43 (C5), 138.46 (C4a), 144.68 (C9a), 156.53 (C10a), 171.56 (C2),
172.11 (C@O)
2h
2i
256 (4.35), 286 (4.47), 364 (3.91),
382 (4.14)
116.41 (C7), 127.21 (C8), 127.96 (C6), 128.95 ðC03 and C50 Þ, 129.01
(C9), 129.44 ðC20 and C60 Þ, 130.27 ðC40 Þ, 132.84 (C5a), 133.58 ðC01Þ,
138.89 (C5), 141.90 (C4a), 144.72 (C9a), 155.29 (C10a), 163.95 (C2),
168.15 (C@O)
254 (4.71), 290 (4.86), 365 (4.24),
385 (4.09)
18.40 (9-CH3), 115.97 (C9), 126.73 (C7), 126.92 (C6), 127.92 (C8),
128.80 ðC03 and C50 Þ, 129.18 ðC20 and C60 Þ, 132.49 ðC04Þ, 132.91 (C5a),
133.45 ðC01Þ, 136.81 (C5), 138.77 (C4a), 145.35 (C9a), 156.15 (C10a),
163.78 (C2), 168.21 (C@O)
2j
255 (4.63), 279 (4.88), 370 (4.52), 374 (4.48)
56.44 (8-OCH3), 106.95 (C7), 120.68 (C9), 122.59 (C6), 128.77 ðC30
and C05Þ, 129.26 ðC20 and C60 Þ, 131.39 ðC40 Þ, 132.56 (C5a), 133.50
ðC01Þ, 137.93 (C5), 139.54 (C4a), 143.68 (C9a), 154.22 (C8), 156.42
(C10a), 163.67 (C2), 168.47 (C@O)
3a
3b
3c
3d
3e
3f
292 (4.35), 303 (4.37), 366 (4.28),
472 (4.20), 500 (4.14)
33.40 (N10–CH3), 118.89 (C7), 121.43 (C9), 127.54 (C6), 129.04 ðC30
and C50 Þ, 129.84 ðC02 and C06Þ, 130.21 (C8), 132.63 ðC40 Þ, 135.88 C01Þ,
136.29 (C5a), 137.34 (C9a), 137.97 (C4a), 154.25 (C10a), 164.38 (C2),
166.68 (C@O)
293 (4.40), 303 (4.42), 365 (4.35),
472 (4.27), 500 (4.23)
13.17 (N10–CH2–CH3), 41.32 (N10–CH2–CH3), 118.37 (C9), 121.59
(C7), 127.32 (C8), 128.94 ðC0 and C50 Þ, 129.66 ðC20 and C06Þ, 130.04
(C6), 132.47 ðC40 Þ, 134.73 C103Þ, 136.24 (C5a), 137.97 (C9a), 145.92
(C4a), 153.45 (C10a), 164.30 (C2), 166.21 (C@O)
20.27 (N10–CH(CH3)2), 51.86 (N10–CH(CH3)2), 119.07 (C7),
121.71 (C9), 127.06 (C6), 129.00 ðC03 and C50 Þ, 129.59 ðC20 and C60 Þ,
130.82 (C8), 132.46 ðC40 Þ, 134.93 ðC10 Þ, 135.59 (C5a), 137.76 (C9a),
139.53 (C4a), 153.80 (C10a), 164.74 (C2), 166.55 (C@O)
21.06 (7-CH3), 33.21 (N10–CH3), 118.55 (C7), 120.15 (C9), 128.78
ðC30 and C50 Þ, 129.54 ðC20 and C60 Þ, 129.94 (C6), 131.94 (C8), 132.29
ðC04Þ, 137.01 ðC10 Þ, 137.66 (C5a), 137.88 (C9a), 138.03 (C4a), 153.55
(C10a), 164.53 (C2), 166.31 (C@O)
24.13 (9-CH3), 40.12 (N10–CH3), 119.53 (C9), 127.29 (C7), 128.06
(C6), 129.04 ðC0 and C50 Þ, 129.76 ðC20 and C60 Þ, 130.11 (C8), 132.66
ðC40 Þ, 133.06 C103Þ, 137.88 (C5a), 139.93 (C9a), 140.41 (C4a), 156.25
(C10a), 164.33 (C2), 166.76 (C@O).
33.92 (N10–CH3), 56.63 (7-OCH3), 113.37 (C6), 115.21 (C8), 118.01
(C9), 128.81 ðC30 and C50 Þ, 129.90 ðC20 and C60 Þ, 132.57 ðC04Þ, 134.04
C01Þ, 137.26 (C5a), 141.15 (C9a), 147.07 (C4a), 152.14 (C7), 153.81
(C10a), 165.01 (C2), 167.94 (C@O)
293 (4.41), 304 (4.21), 365 (4.08),
470 (4.02), 500 (3.93)
293 (4.46), 303 (4.47), 366 (4.34),
478 (4.27), 507 (4.26)
294 (4.33), 307 (4.38), 368 (4.23),
477 (4.11), 508 (4.06)
302 (4.47), 312 (4.48), 362 (4.20),
496 (4.21), 528 (4.22)
3g
3h
293 (4.39), 305 (4.32), 363 (4.02),
493 (4.32), 522 (4.27)
33.65 (N10–CH3), 58.78 (8-OCH3), 100.07 (C7), 116.33 (C9), 118.38
(C6), 128.85 ðC30 and C50 Þ, 129.89 ðC20 and C60 Þ, 132.43 ðC04Þ, 134.64
C01Þ, 138.55 (C5a), 143.38 (C9a), 146.04 (C4a), 152.24 (C8), 157.71
(C10a), 164.04 (C2), 166.37 (C@O)
13.18 (N10–CH2–CH3), 21.06 (7-CH3), 41.61 (N10–CH2–CH3),
118.19 (C7), 120.43 (C9), 122.93 (C6), 128.82 ðC03 and C50 Þ, 129.53
ðC20 and C60 Þ, 132.27 (C8), 132.83 ðC40 Þ, 137.11 ðC01Þ, 137.78 (C5a),
137.95 (C9a), 147.53 (C4a), 156.49 (C10a), 164.70 (C2), 166.42
(C@O)
294 (4.48), 306 (4.49), 366 (4.38),
479 (4.31), 510 (4.29)
3i
301 (4.43), 309 (4.41), 373 (4.00),
472 (4.21), 500 (4.18)
33.92 (N10–CH3), 58.01 (6- and 8-OCH3), 61.80 (7-OCH3), 93.57
(C9), 100.02 (C7), 128.71 ðC03 and C50 Þ, 129.74 ðC20 and C06Þ, 131.89
ðC04Þ, 132.28 ðC10 Þ, 135.98 (C5a), 137.12 (C9a), 138.11 (C4a), 149.42
(C10a), 151.65 (C6), 151.84 (C8), 165.21 (C2), 168.27 (C@O)
a Compounds la–j, 2a–e, and 2g were measured in CDC13, while compounds 2f, 2h–2j, and 3a–i were measured in (CD3)2SO.
ods. Table 3 shows the antitumor activity of 10-methyl
derivative (2a) and 10-ethyl derivative (2b) of 2-deoxo-
2-phenyl-5-deazaflavin against A 431 human adenocar-
cinoma cells. After the compound (2a, 25 mg/kg, it)
was given into the tumor cells by intratumoral adminis-
tration, the tumor growth was inhibited significantly as
can be seen in the T/C %, namely, the tumor volume
was reduced to 57.2% after 17 days of the administra-
tion. But the similar intratumoral administration of
compound (2b) did not inhibit the tumor growth. The
antitumor activity of compound (2a, 100 mg/kg, ip)
was further investigated by intraperitoneal administra-
tion. After the compound (2a) was given into the tumor
cells by ip, the tumor volume was reduced to 53.4% after
10 days of the administration, whereas staurosporine
(1 mg/kg, ip) similarly reduced to 84.9% the tumor vol-

H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
247
Table 2. Growth inhibitory activities against various tumor cell lines of 2-deoxo-2-phenyl-5-deazaflavins (2a–j) and 2-deoxo-2-phenylflavin-5-oxides
(3a–i)
Compound
Inhibitory activity against tumor cell lines [IC50 (lM)]
NCI-H 460
HCT 116
A 431
CCRF-HSB-2
KB
2a
2b
2.23
ND
2.08
ND
3.10
ND
ND
ND
ND
ND
2.72
ND
6.05
4.39
ND
1.33
ND
ND
3.72
0.82
ND
ND
2.11
ND
1.68
ND
2.91
ND
ND
ND
ND
ND
1.80
ND
6.58
3.25
ND
0.72
ND
ND
3.76
2.23
ND
ND
1.44
0.76
2.45
ND
ND
ND
ND
3.88
ND
ND
1.31
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.05
11.83
1.79
1.49
0.70
1.23
2.14
0.73
3.83
299.32
24.41
0.97
0.96
0.98
0.50
1.29
ND
2.33
2.62
1.86
3.98
0.50
5.04
0.51
5.12
>350.00
62.69
1.08
2.38
2.18
1.75
3.28
ND
2c
2d
2e
2f
2g
2h
2i
2j
3a
3b
3c
3d
3e
3f
3g
ND
ND
3h
3i
1.71
ND
2.97
ND
Cisplatin
Ara-C
Adriamycin
ND
ND
0.15
ND
0.70
ND
ND, not done.
Table 3. Antitumor activity of 5-deazaflavin analogs (2a and 2b)
against human adenocarcinoma A 431 xenograft modela
Compound
Dose (mg/kg) Tumor volume Weight change
(g)
T/C (%)
Intratumoral administration (it after 17 days)
Control
2a
100
57.2
109
ꢀ0.8
+0.1
ꢀ1.6
ꢀ0.6
25
25
1
2b
Staurosporine
b
Intraperitoneal administration (ip after 10 days)
Control
2a
Staurosporine
100
53.4
84.9
ꢀ0.4
ꢀ1.7
ꢀ1.0
100
1
Oral administration (po after 18 days)
100
97.6
Control
2a
ꢀ0.4
ꢀ0.9
ꢀ0.4
ꢀ2.0
75
150
1
2a
Staurosporine
51.8
72.9
a A 431 cells were implanted subcutaneously into nude mice and the
administration (it, ip, and po) was started on the day when the mice
show tumor growth in the range of 100–300 mm3. The tumor volume
(T/C %) was measured on the 17 days (it), 10 days (ip), and 18 days
(po) after the treatment of chemicals.
b Because the necrosis occurred, the reduced tumor volume could not
be evaluated.
Figure 1. A 431 tumor volume implanted into nude mouse treated with
chemicals (ip) after several days. A 431 cells were implanted subcu-
taneously into nude mice and the administration was started on the
day when the mice show tumor growth in the 100–300 mm3 range.
ume. The staurosporine was used only 1 mg/kg as a po-
sitive control because it was too toxic. Figure 1 shows A
431 tumor volume implanted into nude mouse treated
with chemicals (ip) after several days. Depending on
the gradual deduction of the tumor volume with the pro-
gress of time after the initial administration, the more
effectual antitumor activity of 2-deoxo-10-methyl-2-
phenyl-5-deazaflavin (2a, 100 mg/kg, ip) than that of
staurosporine was observed. Moreover, the antitumor
activity of compound (2a, 75 and 100 mg/kg, po) was
investigated by oral administration as shown in Table
3. After the compound (2a) was given by po, the tumor
volume was reduced to 97.6% (dose: 75 mg/kg) and
51.8% (dose: 150 mg/kg) after 18 days of the administra-
tion, whereas staurosporine reduced to 72.9% the tumor
volume. Figure 2 also shows A 431 tumor volume

248
H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
Table 4. AutoDock binding free energies (DGb) and inhibition
constants (Ki) for 2-deoxo-2-phenyl-5-deazaflavins (2a–j), 2-deoxo-2-
phenylflavin-5-oxides (3a–j), and 2-deoxo-2-phenylflavins (4a–j)
docked within protein tyrosine kinase
Compounda
Kib
DGb (kcal/mol)
2a
2b
2c
2d
2e
2f
3.78E-04
0.01
0.09
ꢀ4.67
ꢀ2.50
ꢀ1.44
ꢀ1.53
ꢀ2.68
ꢀ2.45
ꢀ1.42
ꢀ8.03
ꢀ4.18
ꢀ4.93
ꢀ7.23
ꢀ7.30
ꢀ6.23
ꢀ6.92
ꢀ7.44
ꢀ6.36
ꢀ7.21
ꢀ7.27
ꢀ5.95
ꢀ7.84
ꢀ4.99
ꢀ2.27
+0.22
ꢀ2.61
ꢀ1.89
ꢀ2.41
ꢀ2.05
ꢀ2.02
ꢀ1.38
ꢀ7.79
0.08
0.01
0.02
0.09
2g
2h
2i
1.29E-06
0.00
2j
3.97E-04
4.99E-06
4.45E-06
2.70E-05
8.45E-06
3.55E-06
2.19E-05
5.21E-06
4.68E-06
4.35E-05
1.71E-06
2.20E-04
0.02
3a
3b
3c
3d
3e
3f
3g
3h
3i
Figure 2. A 431 tumor volume implanted into nude mouse treated with
chemicals (po) after several days. A 431 cells were implanted
3jc
4a
4b
4c
4d
4e
4f
subcutaneously into nude mice and the administration was started
on the day when the mice show tumor growth in the 100–300 mm3
range.
d
—
0.01
implanted into nude mouse treated with chemicals (po)
after several days. Depending on the gradual deduction
of the tumor volume with the progress of time after the
initial administration, the more effectual antitumor
activity of 2-deoxo-10-methyl-2-phenyl-5-deazaflavin
(2a, 150 mg/kg, po) than that of staurosporine (1 mg/
kg, po) was observed, but the dose of 75 mg/kg was
insufficient antitumor activity.
0.04
0.02
4g
4h
4i
0.03
0.03
0.01
1.95E-06
4j
a The 2-deoxo-2-phenyflavins (4a–j), which are the N-deoxidated
derivatives of 2-deoxo-2-phenylflavin-5-oxides (3a–j), are computa-
tionally designed compounds.
b Inhibition constant.
c This is a computationally designed compound.
d This cannot be calculated because DGb is a positive value.
Accordingly, the antitumor activity of compound 2a
in vivo was shown to be more effective than compound
2b though compound 2a exhibited weaker activity than
that of compound 2b against A 431 tumor cells in vitro
as shown in Table 2. There is no fixed rule that the com-
pound, which has biological activity in vitro, must
exhibit the same effect in vivo. For example, the nonse-
lective protein kinase inhibitor, staurosporine, did not
exhibit any antitumor activity in vivo though it has
much greater antiproliferative activity than UCN-01
(7-hydroxy-staurosporine) in vitro.17
highest potential binding affinity into the binding site
of the 3D macromolecule (PTK, 1skj). The 5-deazaf-
lavins (2a–g) and the computationally designed N-
deoxidated derivatives, 2-deoxo-2-phenylflavins (4a–j),
showed less binding affinity than the N-oxides (3a–j).
The higher affinity is presumably attributed to the for-
mation of more and/or tighter hydrogen bonds be-
tween the C4-oxo of flavin-N-oxides and several
amino acids at the binding site owing to the increased
electronegativity of the oxo at the 4-posision due to
the N-oxide moiety. Therefore, the flavin-5-oxides
(3a–j) were docked deeply within the groove of the
binding pocket of PTK forming more hydrogen bonds
with Arg 12, Arg 32, Glu 35, and Lys 60. Considering
the fitting of the 5-deazaflavins (2a–g) and flavin ana-
logs (4a–i) into the PTK binding pocket, it was dem-
onstrated that their binding energies were higher in
comparison with the flavin-N-oxides (3a–j). Hence,
they were partially superimposed on the original
docked ur2 ligand (4-[3-carboxymethyl-3-(4-phospho-
nooxybenzyl)ureido]-4-[(3-cyclohexylpropyl)-methylcar-
bamoyl]butyric acid), namely, they are docked slightly
2.4. Molecular docking study
The computer simulated automated docking studies
were performed using the widely distributed molecular
docking software, AutoDock 3.05.18 The AutoDock
study of 2-deoxo-2-phenyl-5-deazaflavins (2a–j), 2-deo-
xo-2-phenylflavin-5-oxides (3a–i), and the computa-
tionally designed 3j and 2-deoxo-2-phenylflavins (4a–
j) was carried out, and they were docked within the
protein tyrosine kinase (PTK pp60c-src, PDB code:
1skj). As shown in Table 4, their AutoDock binding
free energies (DGb, kcal/mol) and inhibition constants
(Ki) were obtained. Among them, all flavin-5-oxides
(3a–j) exhibited the lowest free energy between ꢀ7.84
and ꢀ5.95 kcal/mol. In other words, they possess the

H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
249
away from the groove of the binding site. It is note-
worthy that the unsubstituted derivatives (2h, 3j, and
4j) exhibited especially stronger binding affinities to
possess the lower binding free energy ca. ꢀ8 kcal/
mol and also the derivatives (2a, 2i, 2j, and 4a) having
a methyl group at the 10-position or a methyl or
methoxy group on the benzene of the benzopteridine
ring showed the binding free energy between ꢀ4.99
and ꢀ4.18 kcal/mol. It was clarified that the deriva-
tives possessing less binding free energy than
ꢀ4.2 kcal/mol are allowed to fit well into the groove
of the binding site.
The overall good correlation between the growth
inhibitory activities (IC50,lM) of the flavin analogs
against tumor cells and the binding affinities (lower
binding free energy) predicted by AutoDock was
made clear as indicated in Tables 2 and 4. It was
found that the stronger binding affinity, the more
potent inhibitory activity against tumor cells. Espe-
cially, the correlation between the binding free ener-
gy (DGb) and IC50 (lM) values for compounds 2a–f
and 3a–d against CCRF-HSB-2 and KB tumor cells
was fairly good and showed almost same correlation
coefficients (R2) of 0.64 as shown in Figures 3
and 4.
Figure 4. Correlation between the binding free energy (DGb) and IC50
of 2-deoxo-2-phenyl-5-deazaflavins (2a, b, d, f) and 2-deoxo-2-phenyl-
flavin-5-oxides (3a–d) against human oral epithelial carcinoma (KB).
Figure 5 shows the differential docking mode of three
flavin analogs. Interestingly, the flavin-5-oxide (3b),
whose DGb was ꢀ7.30 kcal/mol, got deeply embedded
within the groove of the binding pocket. This is high-
ly correlated to its potent inhibition of the prolifera-
tion of human T-cell acute lymphoblastic leukemia
cells (CCRF-HSB-2, IC50: 0.96 lM). Whereas the 2-
deoxo-5-deazaflavin (2b), whose DGb and IC50 was
ꢀ2.50 kcal/mol and 1.79 lM against CCRF-SB2,
respectively, was docked at the same position as the
flavin isomer (4b), whose DGb was ꢀ2.27 kcal/mol.
Both flavin (4b) and 5-deazaflavin (2b) appeared to
be superimposed onto each other. Moreover, the fla-
vin-5-oxide (3b) exhibited three true hydrogen bonds
with Arg 12, Arg 32, and Glu 35 of bond angles
being 143.0ꢁ, 175.9ꢁ, and 152.3ꢁ, respectively. Another
˚
the false hydrogen bond (1.95 A) with Cys 42 was
omitted because of the angle through a hydrogen be-
tween C4–O of 3b and N of Cys 42 was 86ꢁ. While
compounds 2b and 4b did not exhibit any hydrogen
bond. As cited in the literature,19 the compound with
the higher number of hydrogen bonds, with the low-
est binding energy, and with the smallest RMSD
(root mean square deviation) is generally said to be
a reasonable candidate for potent inhibitors against
the enzyme. Actually the docking results showed that
there was no significant difference between flavins and
2-deoxo-5-deazaflavins with respect to docking affini-
ties into the binding site as shown in Table 4 and
Figure 3. Correlation between the binding free energy (DGb) and IC50
(lM) of 2-deoxo-2-phenyl-5-deazaflavins (2b, c, e, f) and 2-deoxo-2-
phenylflavin-5-oxides (3a–d) against human T-cell acute lymphoblastic
leukemia (CCRF-HSB-2).

250
H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
Figure 5. The differential docking mode of flavin analogs (2b, 3b, and 4b) and the binding pocket shown in transparent solid surface with labeled
amino acids, and the ur2 ligand is shown as a red line.
Figure 5. This may be explained as CH and N are
equivalent bioisosteres with similar size and compara-
ble electronic properties.
bonds involving Arg 32 and Glu 35, which are also
involved in ur2 ligand, and Arg 12 and Lys 60.
Also these inhibitors were involved in a hydropho-
bic interaction with only three amino acids, that
is, Tyr 59, Ile 71, and Asp 92. Therefore, the
docked inhibitors exhibited reasonable RMSD values
as shown in Figure 6, where compound 3e was
bound, nearly superimposed on the ur2 ligand
exhibiting DGb being ꢀ7.44 kcal/mol, three true
hydrogen bonds with Arg 12, Arg 32, and Glu 35
2.5. Mode of interaction of inhibitors within the binding
pocket
Most docked inhibitors interacted by the same mode
of the co-crystallized ur2 ligand within the PTK
binding site. They exhibited up to three hydrogen
Figure 6. AutoDock binding affinity of 2-deoxo-9,10-dimethyl-2-phenylflavin-5-oxide (3e) shown as ball and stick colored by element. The binding
pocket of the pp60c-src shown as solid backbone ribbon with the ur2 ligand (blue stick).

H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
251
Figure 7. Interaction of compound 3i (ball and stick, colored by element), docked into the binding pocket of PTK whose amino acids are shown as
labeled lines.
of angles being 138.6ꢁ, 173.9ꢁ, 145.8ꢁ, respectively,
and RMSD being 6.94 A.
460, HCT 116, A 431, CCRF-HSB-2, and KB cells
showed significant potential antitumor activities. Inter-
estingly, the activities of some compounds 2a, 2c, 3a,
and 3f against HCT 116 cells were higher than that of
cisplatin. Especially, compound 3f (IC50: 0.72 lM)
exhibited 3-fold more potent antitumor activity than cis-
platin (IC50: 2.33 lM). Approximately similar inhibitory
activities (IC50: 0.5–2.0 lM) to Ara-C (IC50: 0.15 lM)
against CCRF-HSB-2 cells were observed in many com-
pounds. Additional results of promising antitumor
activity were obtained in compounds showing potential
activities of 0.5–5.0 lM (IC50) against KB cells. Further-
more, in vivo antitumor activity of 2-deoxo-10-methyl-
2-phenyl-5-deazaflavin (2a) against human adenocarci-
noma A 431 cells implanted subcutaneously into mice
was undertaken. As the result, the tumor volume was re-
duced to 57.2% after 17 days of the intratumoral admin-
istration (25 mg/kg), to 53.4% after 10 days of the
intraperitoneal administration (100 mg/kg) and to
51.8% after 18 days of the oral administration
(150 mg/kg).
˚
Figure 7 illustrates the hydrogen bond and van der
Waals contacts similarly to the ur2 ligand with the Src
SH2 domain. Instead of the binding mode for the phenyl
phosphate group of ur2 within the pTyr (pY) pocket, the
C2-phenyl moiety of 3i assumes the same position but
has a different relative orientation, when compared with
the phosphate of other Src SH2 ligand complexes. The
benzene ring of 3i is parallel to the phenyl ring of Tyr
˚
59 within a distance of 6.65 A and angle of 94.4ꢁ, while
the imidazole ring of His 58 is oriented perpendicularly
to the planar benzopteridine ring within a distance of
˚
3.64 A and angle of 28.7ꢁ. Exceptionally, the same pY
pocket side chains are involved in the hydrogen-bonding
scheme of the C4-oxo group with Lys 60. Similarly to
ur2, Arg 12 does not make any hydrogen bond with
3i, while Arg 12 does in other structures 3b and 3e dis-
cussed previously in Figures 5 and 6, respectively. Also
3i interacted hydrophobically with Tyr 59, Ile 71, and
Asp 92.
The AutoDock study of 2-deoxo-2-phenyl-5-deazaflav-
ins (2a–j), 2-deoxo-2-phenylflavin-5-oxides (3a–j), and
2-deoxo-2-phenylflavins (4a–j) was carried out, and they
were docked within the protein tyrosine kinase (PTK
pp60c-src, PDB code: 1skj). All flavin-5-oxides (3a–j)
exhibited the lowest binding free energies between
ꢀ7.84 and ꢀ5.95 kcal/mol. Namely, they possessed the
highest potential binding affinity into the target site.
3. Conclusion
In this study, ten 2-deoxo-5-deazaflavins (2a–j) and nine
flavin-5-oxides (3a–i) were synthesized to investigate
their biological activities. In vitro growth inhibitory
activities of compounds 2a–j and 3a–i against NCI-H

252
H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
The correlation between the binding free energy (DGb,
kcal/mol) and IC50 (lM) values against CCRF-HSB-2
and KB tumor cells showed fairly good correlation coef-
ficients. 2-Deoxo-10-ethyl-2-phenylflavin-5-oxide (3b),
whose DGb was ꢀ7.30 kcal/mol, got deeply embedded
within the groove of the binding pocket, and is highly
correlated to its potent inhibition of the proliferation
of human T-cell acute lymphoblastic leukemia cells
(CCRF-HSB-2, IC50: 0.96 lM).
(2H, m, 2–Ph-oH), 12.60 (1 H, br s, 3-NH); IR (cmꢀ1
3240 (NH), 1650 (C@O).
)
4.1.3.
6-(N-Ethylanilino)-2-phenylpyrimidin-4(3H)-one
(1b)13. Yield, 4.01 g (81%); mp 268 ꢁC (lit.13 mp
1
265 ꢁC); H NMR (CDCl3), d 1.27 (3H, t, J = 7.2 Hz,
6-NCH2–CH3), 4.06 (2H, q, J = 7.2 Hz, 6-NCH2–
CH3), 5.10 (1H, s, 5-H), 7.24–7.27 (2H, m, N–Ph–oH),
7.31–7.37 (1H, m, N–Ph–pH), 7.42–7.45 (1H, m, N–
Ph–mH), 7.48–7.50 (3H, m, 2-Ph–m,pH), 8.16–8.19
Thus, SAR studies by computational design, chemical
synthesis, and biological investigation revealed new
structural requirements for the modified flavins that
would enhance the binding characteristics within the
protein tyrosine kinase. The most effective molecules
for flavin and 5-deazaflavins as antitumor activities were
elucidated that a phenyl substituent at the 2-position is
necessary.
(2H, m, 2-Ph–oH), 12.37 (1H, br s, 3-NH); IR (cmꢀ1
3210 (NH), 1635 (C@O).
)
4.1.4. 6-(N-Isopropylanilino)-2-phenylpyrimidin-4(3H)-
one (1c). Yield, 2.12 g (41%); mp > 300 ꢁC (MeOH);
1H NMR (CDCl3), d 1.18 (6H, d, J = 6.6 Hz, 6-N–
CH–(CH3)2), 4.78 (1H, s, 5-H), 5.40 (1H, quintet,
J = 6.6 Hz, 6-N–CH–(CH3)2), 7.13–7.18 (2H, m, N–
Ph–oH), 7.36–7.54 (6H, m, 2-Ph–m,pH and N–Ph–
m,pH), 8.10–8.14 (2H, m, 2-Ph–oH), 11.59 (1H, br s,
3-NH); IR (cmꢀ1) 3200 (NH), 1635 (C@O); Anal.
C19H19N3O requires C, 74.73; H, 6.27; N, 13.76. Found:
C, 74.69; H, 6.39; N, 13.60.
4. Experimental
4.1. Chemistry
Mps were obtained on a Yanagimoto micro melting
point apparatus and are uncorrected. Microanalyses
were measured by Yanaco CHN Corder MT-5 appara-
tus. IR spectra were recorded on a JASCO FT/IR-200
4.1.5. 6-(N-Methyl-4-methylanilino)-2-phenylpyrimidin-
4(3H)-one (1d). Yield, 4.75 g (96%); mp 265–267 ꢁC
(dioxane); 1H NMR (CDCl3), d 2.40 (3H, s, pMe),
3.54 (3H, s, 6-NMe), 5.23 (1H, s, 5-H), 7.16 (2H, dd,
1
spectrophotometer as Nujol mulls. H and 13C NMR
J2 ;3 ¼ 9:0 Hz; J3 ;5 ¼ 2:1 Hz, N-Ar-oH), 7.24 (2H,
00 00
00 00
00 00
00 00
spectra were obtained using a Varian VXR 300 and
75 MHz spectrophotometer, respectively, and chemical
shift values were expressed in d values (ppm) relative
to tetramethylsilane (TMS) as internal standard. Cou-
pling constants are given in Hz. UV spectra were mea-
sured in absolute EtOH using Hitachi U-2001 and
Beckman DU-68S UV spectrophotometers and absorp-
tion values in italic refer to wavelengths at which shoul-
ders or inflexions occur in the absorption. All reagents
were of commercial quality and were used without fur-
ther purification. Organic solvents were dried in the
presence of an appropriate drying agent and were stored
over suitable molecular sieves. Reaction progress was
monitored by analytical thin layer chromatography
dd, J2 ;3 ¼ 8:7 Hz; J2 ;6 ¼ 2:1 Hz, N-Ar-mH), 7.47–7.52
(3H, m, 2-Ph–m,pH), 8.16–8.20 (2H, m, 2-Ph–oH),
12.49 (1H, br s, 3-NH); IR (cmꢀ1) 3210 (NH), 1630
(C@O); Anal. C18H17N3O requires C, 74.20; H, 5.88;
N, 14.42. Found: C, 74.00; H, 5.91; N, 14.36.
4.1.6. 6-(N-Methyl-2-methylanilino)-2-phenylpyrimidin-
4(3H)-one (1e). Yield, 3.07 g (62%); mp 254–256 ꢁC
1
(EtOH–H2O); H NMR (CDCl3), d 2.20 (3H, s, oMe),
3.51 (3H, s, 6-NMe), 4.84 (1H, s, 5-H), 7.15–7.21 (1H,
m, N-Ar-oH), 7.27–7.45 (3H, m, N-Ar-m,pH), 7.46–
7.54 (3H, m, 2-Ph-m,pH), 8.15–8.30 (2H, m, 2-Ph–oH),
12.36 (1H, br s, 3NH); IR (cmꢀ1) 3210 (NH), 1640
(C@O); Anal. C18H17N3O requires C, 74.20; H, 5.88;
N, 14.42. Found: C, 73.83; H, 5.94; N, 14.42.
(TLC) on pre-coated glass plates (silica gel 60F254
-
plate-Merck) and the products were visualized by UV
light.
4.1.7. 6-(N-Methyl-4-methoxyanilino)-2-phenylpyrimidin-
4(3H)-one (1f). Yield, 4.86 g (93%); mp 273–275 ꢁC
(EtOH); 1H NMR (CDCl3), d 3.52 (3H, s, 6-NMe),
3.85 (3H, s, OMe), 5.18 (1H, s, 5-H), 6.96 (2H, dd,
4.1.1. General procedure for the preparation of 6-(N-
alkylanilino)-2-phenylpyrimidin-4(3H)-ones (1a–j).
A
mixture of 6-chloro-2-phenylpyrimidin-4(3H)-one13
(3.5 g, 17 mmol) and N-alkylaniline (0.051–0.085 mol)
was heated under nitrogen atmosphere at 180–200 ꢁC
for 0.5–5 h. After cooling, the resulting solid was
crushed with diethyl ether to get powdery crystals,
which were filtered off, washed with water, dried, and
recrystallized from an appropriate solvent with charcoal
to afford colorless needles.
J2 ;3 ¼ 9:0 Hz; J3 ;5 ¼ 2:1 Hz, N-Ar-oH), 7.19 (2H, d,
00 00
00 00
00 00
00 00
J2 ,3 = 9.0 Hz, J2 ;6 ¼ 2:1 Hz, N-Ar-mH), 7.46–7.52
(3H, m, 2-Ph–m,pH), 8.15–8.20 (2H, m, 2-Ph–oH),
12.29 (1H, br s, 3-NH); IR (cmꢀ1) 3210 (NH), 1630
(C@O); Anal. C18H17N3O2 requires C, 70.34; H, 5.58;
N, 13.67. Found: C, 70.10; H, 5.66; N, 13.28.
4.1.8. 6-(N-Methyl-3-methoxyanilino)-2-phenylpyrimidin-
4(3H)-one (1g). Yield, 3.87 g (74%); mp 235–237 ꢁC
(EtOH); 1H NMR (CDCl3), d 3.55 (3H, s, 6-NMe),
3.83 (3H, s, OMe), 5.29 (1H, s, 5-H), 6.82–6.90 (3H,
m, N-Ar-o,pH), 7.32–7.38 (1H, m, N-Ar-mH), 7.48–
7.54 (3H, m, 2-Ph–m,pH), 8.14–8.19 (2H, m, 2-Ph-oH),
12.18 (1H, br s, 3-NH); IR (cmꢀ1) 3200 (NH), 1645
4.1.2. 6-(N-Methylanilino)-2-phenylpyrimidin-4(3H)-one
(1a). Yield, 4.47 g (95%); mp 248 ꢁC (EtOH) (lit.13 mp
248 ꢁC); 1H NMR (CDCl3), d 3.56 (3H, s, 6-NMe),
5.27 (1H, s, 5-H), 7.28–7.33 (3H, m, N–Ph–o,pH),
7.42–7.54 (5 H, m, N–Ph–mH, 2-Ph–m,pH), 8.17–8.21

H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
253
(C@O); Anal. C18H17N3O2 requires C, 70.34; H, 5.58;
N, 13.67. Found: C, 70.09; H, 5.69; N, 13.62.
J = 7.2 Hz, 10-CH2–CH3), 7.42–7.54 (3H, m, 6-, 7-,
and 9-H), 7.57–7.65 (1H, m, 8-H), 7.87–7.92 (1H, m,
Ph–pH), 7.96–8.07 (2H, m, Ph–mH), 8.63–8.68 (2H, m,
Ph–oH), 9.34 (1H, s, 5-H); IR (cmꢀ1) 1640 (C@O).
4.1.9.
6-(N-Ethyl-4-methylanilino)-2-phenylpyrimidin-
4(3H)-one (1h). Yield, 4.77 g (92%); mp 295–297 ꢁC (de-
comp., EtOH); 1H NMR (CDCl3), d 1.25 (3H, t,
J = 7.0 Hz, 6-NCH2–CH3), 2.40 (3H, s, pMe), 4.04
(2H, q, J = 7.0 Hz, 6-NCH2–CH3), 5.07 (1H, s, 5-H),
7.11 (2H, br d, J = 8.4 Hz, N-Ar-oH), 7.24 (2H, br d,
J = 8.4 Hz, N-Ar-mH), 7.48–7.54 (3H, m, 2-Ph–m,pH),
8.15–8.19 (2H, m, 2-Ph–oH), 12.25 (1H, br s, 3-NH);
IR (cmꢀ1) 3200 (NH), 1635 (C@O); Anal. C19H19N3O
requires C, 74.73; H, 6.27; N, 13.76. Found: C, 74.66;
H, 6.36; N, 13.94.
4.1.15. 7,10-Dimethyl-2-phenylpyrimido[4,5-b]quinolin-
4(10H)-one (2c). Yield, 10.25 g (68%); mp 295–297 ꢁC
1
(decomp., DMF); H NMR [(CD3)3SO], d 2.44 (3H, s,
7-Me), 4.46 (3H, s, 10-Me), 7.43–7.66 (3H, m, 6-, 8-,
and 9-H), 7.88–8.22 (3H, m, Ph–m,pH), 8.43–8.61 (2H,
m, Ph–oH), 9.29 (1H, s, 5-H); IR (cmꢀ1) 1650 (C@O);
Anal. C19H15N3O requires C, 75.73; H, 5.02; N, 13.94.
Found: C, 75.31; H, 5.18; N, 13.57.
4.1.16. 9,10-Dimethyl-2-phenylpyrimido[4,5-b]quinolin-
4(10H)-one (2d). Yield, 9.19 g (61%); mp 262–264 ꢁC
(DMF); 1H NMR [(CD3)3SO], d 2.98 (3H,s, 9-Me),
4.57 (3H, s, 10-Me), 7.53–7.61 (4H, m, 7-H and Ph–
m,pH), 7.92 (1H, br d, J7,8 = 7.8 Hz, 8-H), 8.23 (1H, d,
4.1.10. 6-(N-Ethyl-2-methylanilino)-2-phenylpyrimidin-
4(3H)-one (1i). Yield, 3.01 g (58%); mp 249–251 ꢁC
1
(EtOH); H NMR (CDCl3), d 1.25 (3H, t, J = 7.2 Hz,
6-NCH2–CH3), 2.19 (3H, s, oMe), 4.27 (2H, br s, 6-
NCH2–CH3), 4.80 (1H, s, 5-H), 7.13–719 (2H, m, N-
Ar-300 and 600H), 7.28–7.34 (2H, m, N-Ar-400 and 500-H),
7.46–7.54 (3H, m, 2-Ph–m,pH), 8.16–8.22 (2H, m,
2-Ph–oH), 11.91 (1H, br s, 3-NH); IR (cmꢀ1) 3200
(NH), 1640 (C@O); Anal. C19H19N3OÆ1/10H2O requires
C, 74.29; H, 6.30; N, 13.68. Found: C, 74.13; H, 6.33; N,
13.46.
J
6,7 = 7.8 Hz, Ph–6H), 8.49–8.54 (2H, m, Ph–oH),
9.36 (1H, s, 5-H); IR (cmꢀ1) 1640 (C@O); Anal.
C19H15N3OÆ1/5 H2O requires C, 74.83; H, 5.09; N,
13.78. Found: C, 74.80; H, 5.21; N, 13.74.
4.1.17. 7-Methoxy-10-methyl-2-phenylpyrimido[4,5-b]-
quinolin-4(10H)-one (2e). Yield, 15.08 g (95%); mp
1
279–281 ꢁC (decomp., DMF); H NMR [(CD3)3SO], d
4.1.11. 6-(N-Methyl-3,4,5-trimethoxyanilino)-2-phenyl-
pyrimidin-4(3H)-one (1j). Yield, 5.93 g (95%); mp
294 ꢁC (decomp., EtOH); 1H NMR (CDCl3), d 3.55
(3H, s, 6-NMe), 3.85 (6H, s, N-Ar-300,500-diOMe), 3.89
(3H, s, N-Ar-400-OMe), 5.23 (1H, s, 5-H), 6.50 (2H, s,
N-Ar-200H and 600H), 7.45–7.60 (3H, m, 2-Ph–m,pH),
8.17–8.24 (2H, m, 2-Ph–oH), 12.54 (1H, br s, 3-NH) ;
IR (cmꢀ1) 3210 (NH), 1635 (C@O); Anal. C20H21N3O4
requires C, 65.38; H, 5.76; N, 11.44. Found: C, 64.98;
H, 5.79; N, 11.06.
3.95 (3H, s, 7-OMe), 4.51 (3H, s, 10-Me), 7.50–7.60
(3H, m, Ph-m,pH), 7.78 (1H, dd, J6,8 = 2.4 Hz,
J8,9 = 9.0 Hz, 6-H), 7.92 (1H, d, J6,8 = 2.4 Hz, 8-H),
8.20 (1H, d, J8,9 = 9.0 Hz, 9-H), 8.52–8.57 (2H, m, Ph–
oH), 9.36 (1H, s, 5-H); IR (cmꢀ1) 1620 (C@O); Anal.
C19H15N3O2Æ1/10 H2O requires C, 71.51; H, 4.80; N,
13.17. Found: C, 71.27; H, 4.97; N, 13.42.
4.1.18. 8-Methoxy-10-methyl-2-phenylpyrimido[4,5-
b]quinolin-4(10H)-one (2f). Yield, 14.60 g (92%); mp
1
265–267 ꢁC (decomp., EtOH); H NMR [(CD3)3SO], d
4.1.12. General procedure for the preparation of 2-deoxo-
2-phenyl-5-deazaflavin derivatives (2a–j). A mixture of 6-
(N-alkylanilino)- or 6-anilino-2-phenylpyrimidin-4(3H)-
ones15 (1, 0.05 mol) and phosphoryl chloride (77 g,
0.5 mol) in N,N-dimethylformamide (100 mL) was heat-
ed under stirring at 90 ꢁC for 1–4 h. Then, the reaction
mixture was poured onto ice and neutralized with aque-
ous ammonia (pH 7). The yellow crystals, which sepa-
rated, were filtered off, washed with water, dried, and
recrystallized from an appropriate solvent. All com-
pounds were obtained as yellow (2a–g) or pale yellow
needles (2h–j).
4.10 (3H, s, 8-OMe), 4.45 (3H, s, 10-Me), 7.35 (1H,
dd, J6,7 = 9.0 Hz, J7,9 = 2.1 Hz, 7-H), 7.44 (1H, d,
J7,9 = 2.1 Hz, 9-H), 7.25–7.62 (3H, m, Ph–m,pH), 8.31
(1H, d, J6,7 = 9.0 Hz, 6-H), 8.50–8.58 (2H, m, Ph–oH),
9.31 (1H, s, 5-H); IR (cmꢀ1) 1625 (C@O); Anal.
C19H15N3O2Æ3/10 H2O requires C, 70.71; H, 4.87; N,
13.02. Found: C, 70.30; H, 4.89; N, 12.88.
4.1.19. 10-Ethyl-7-methyl-2-phenylpyrimido[4,5-b]quino-
lin-4(10H)-one (2g). Yield, 11.8 g (75%); mp 291–
1
293 ꢁC (decomp., DMF); H NMR [(CD3)3SO], d 1.51
(3H, t, J = 7.0 Hz, 10-CH2–CH3), 2.54 (3H, s, 7-Me),
5.18 (2H, q, J = 7.0 Hz, 10-CH2–CH3), 7.50–7.60 (3H,
m, 6-H, 8-H, and Ph–pH), 7.98 (1H, d, J8,9 = 9.3 Hz,
9-H), 8.17–8.23 (2H, m, Ph–mH), 8.50–8.58 (2H, m,
Ph–oH), 9.33 (1H, s, 5-H); IR (cmꢀ1) 1640 (C@O); Anal.
C20H17N3O requires C, 76.17; H, 5.43; N, 13.32. Found:
C, 75.82; H, 5.65; N, 13.22.
4.1.13.
10-Methyl-2-phenylpyrimido[4,5-b]quinolin-
4(10H)-one (2a). Yield, 13.5 g (94%); mp 293–295 ꢁC
1
(EtOH) (lit.13 mp 290 ꢁC); H NMR (CDCl3), d 4.45
(3H, s, 10-Me), 7.34–7.49 (3H, m, 6-, 7-, and 9-H),
7.54–7.61 (1H, m, 8-H), 7.84–7.89 (1H, m, Ph–pH),
7.95–8.01 (2H, m, Ph–mH), 8.53–8.59 (2H, m, Ph–oH),
9.27 (1H, s, 5-H); IR (cmꢀ1) 1645 (C@O).
4.1.20. 2-Phenylpyrimido[4,5-b]quinolin-4(10H)-one (2h).
Yield, 11.75 g (86%); mp > 300 ꢁC (DMF); 1H NMR
[(CD3)3SO], d 7.5–7.76 (4H, m, 6-H,7-H, and Ph–mH),
4.1.14. 10-Ethyl-2-phenylpyrimido[4,5-b]quinolin-4(10H)-
one (2b)13. Yield, 12.21 g (81%); mp > 300 ꢁC (decomp.,
EtOH) (lit.13 mp 303 ꢁC); 1H NMR (CDCl3), d
1.64 (3H, t, J = 7.2 Hz, 10-CH2–CH3), 5.16 (2H, q,
7.94
(1H,
ddd,
J6,8 = 1.5 Hz,
J7,8 = 8.7 Hz,
J8,9 = 8.7 Hz, 8-H), 8.09 (1H, dd, J7,9 = 1.5 Hz,
J8,9 = 8.7 Hz, 9-H), 8.24–8.32 (3H, m, Ph–o,pH), 9.30

254
H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
(1H, s, 5-H), 12.64 (1H, br s, 10-NH) ; IR (cmꢀ1) 3170
(NH), 1680 (C@O); Anal. C17H11N3O requires C,
74.71; H, 4.06; N, 15.38. Found: C, 74.36; H, 4.31; N,
15.24.
8.47–8.51 (2H, m, Ph–oH); IR (cmꢀ1), 1540
(N–O), 1650 (C@O); Anal. C17H12N4O2 requires C,
67.10; H, 3.97; N, 18.41. Found: C, 67.11; H, 4.20; N,
18.62.
4.1.21. 9-Methyl-2-phenylpyrimido[4,5-b]quinolin-4(10H)-
one (2i). Yield, 9.77 g (68%); mp > 300 ꢁC (DMF); H
4.1.26. 10-Ethyl-4,10-dihydro-4-oxo-2-phenylbenzo[g]-
pteridine-5-oxide (3b). Yield, 2.42 g (76%); mp 262–
264 ꢁC (decomp., EtOH); H NMR [(CD3)3SO], d 1.52
1
1
NMR [(CD3)3SO], d 2.78 (3H, s, 9-Me), 7.45–7.70
(4H, m, 6-H, 7-H, and Ph-mH), 7.75–7.82 (1H, m, 8-
H), 8.10–8.19 (1H, m, Ph–pH), 8.25–8.30 (2H, m, Ph–
oH), 9.24 (1H, s, 5-H), 12.63 (1H, br s, 10-NH) ; IR
(cmꢀ1) 3160 (NH), 1670 (C@O); Anal. C18H13N3O re-
quires C, 75.25; H, 4.56; N, 14.63. Found: C, 74.85;
H, 4.82; N, 14.78.
(3H, t, J = 6.9 Hz, 10-CH2–CH3), 4.98 (2H, q,
J = 6.9 Hz, 10-CH2-CH3), 7.49–7.61 (3H, m, Ph–
m,pH), 7.72 (1H, ddd, J6,7 = 7.8 Hz, J7,8 = 7.8 Hz,
J7,9 = 1.5 Hz, 7-H), 8.07 (1 H, ddd, J6,8 = 1.5 Hz,
J7,8 = 7.8 Hz, J8,9 = 7.8 Hz, 8 H), 8.23 (1H, dd,
J7,9 = 1.5 Hz, J8,9 = 7.8 Hz, 9-H), 8.44–8.50 (2H, m,
Ph–oH), 8.56 (1H, dd, J6,7 = 7.8 Hz, J6,8 = 1.5 Hz, 6-
H); IR (cmꢀ1), 1540 (N–O), 1640 (C@O); Anal.
C18H14N4O2 requires C, 67.91; H, 4.43; N, 17.60.
Found: C, 67.94; H, 4.62; N, 17.58.
4.1.22. 8-Methoxy-2-phenylpyrimido[4,5-b]quinolin-4(10H)-
1
one (2j). Yield, 11.68 g (77%); mp > 300 ꢁC (DMF); H
NMR [(CD3)3SO], d 3.99 (3H, s, 8-OMe), 7.31 (1H,
dd, J6,7 = 7.5 Hz, J7,9 = 2.1 Hz, 7-H), 7.42 (1H, d,
J7,9 = 2.1 Hz, 9-H), 7.55–7.72 (3H, m, Ph–m,pH), 8.14
(1H, d, J = 7.8 Hz, 6-H), 8.24–8.29 (1H, m, Ph–oH),
4.1.27. 4,10-Dihydro-10-isopropyl-4-oxo-2-phenylbenzo[g]-
pteridine-5-oxide (3c). Yield, 2.66 g (80%); mp 267–
9.16 (1H, s 5-H), 12.57 (1H, br s, 10-NH) ; IR (cmꢀ1
)
269 ꢁC (MeOH); H NMR [(CD3)3SO], d 1.85 (6H, d,
1
3160 (NH), 1680 (C@O); Anal. C18H13N3O2Æ1/10 H2O
requires C, 71.28; H, 4.32; N, 13.85. Found: C, 71.11;
H, 4.54; N, 13.54.
J = 6.9Hz, 10-CH–(CH3)2), 6.19 (1 H, br s, 10-CH–
(CH3)2), 7.56–7.64 (3H, m, Ph-m,pH), 7.71 (1H, ddd,
J6,7 = 7.8 Hz, J7,8 = 7.8 Hz, J7,9 = 1.5 Hz, 7-H), 8.04
(1H, ddd, J6,8 = 1.5 Hz, J7,8 = 7.8 Hz, J8,9 = 7.8 Hz, 8-
H), 8.37 (1H, dd, J7,9 = 1.5 Hz, J8,9 = 7.8 Hz, 9-H),
8.44–8.50 (3H, m, 6, Ph–oH) ; IR (cmꢀ1), 1540
(N–O), 1640 (C@O); Anal. C19H16N4O2 requires C,
68.66; H, 4.85; N, 16.86. Found: C, 68.45; H, 4.94; N,
16.69.
4.1.23. Methylation of 2-phenylpyrimido[4,5-b]quinolin-
4(10H)-one (2h). A mixture of 2-phenylpyrimido[4,5-
b]quinolin-4(10H)-one (2h, 0.5 g, 1.83 mmol), methyl
iodide (1.30 g, 9.16 mmol), and anhydrous potassium
carbonate (1.01 g, 7.31 mmol) in DMF (50 mL) was
stirred at room temperature for 10 h. After the reac-
tion was complete, the precipitated potassium carbon-
ate was filtered off. The residue obtained by
concentration in vacuo was subjected to column chro-
matography on silica gel using a mixture of n-hexane
and ethyl acetate (1:1) as eluting solvent to isolate the
corresponding 10-methyl-2-phenylpyrimido[4,5-b]quin-
olin-4(10H)-one (2a, 0.425 g, 80%), which was identi-
cal to the product (2a) prepared by the reaction of
1a with Vilsmeier reagent in all spectral data as de-
scribed above.
4.1.28. 4,10-Dihydro-7,10-dimethyl-4-oxo-2-phenylbenzo[g]-
pteridine-5-oxide (3d). Yield, 2.16 g (68%); mp > 300 ꢁC
1
(decomp., DMF); H NMR [(CD3)3SO], d 2.51 (3H, s,
7-Me), 4.27 (3H, s, 10-Me), 7.49-7.61 (3H, m, Ph–
m,pH), 7.92 (1H, dd, J6,8 = 1.5 Hz, J8,9 = 8.7 Hz, 8-H),
8.12 (1H, d, J8,9 = 8.7 Hz, 9-H), 8.25 (1H, d,
J6,8 = 1.5 Hz, 6-H), 8.45–8.48 (2H, m, Ph–oH); IR
(cmꢀ1), 1540 (N–O), 1640 (C@O); Anal. C18H14N4O2
requires C, 67.91; H, 4.43; N, 17.60. Found: C, 68.12;
H, 4.56; N, 17.51.
4.1.24. General procedure for the preparation of 2-deoxo-
2-phenylflavin-5-oxide derivatives (3a–i). To a stirring
4.1.29. 4,10-Dihydro-9,10-dimethyl-4-oxo-2-phenylbenzo[g]-
pteridine-5-oxide (3e). Yield, 2.61 g (82%); mp > 300 ꢁC
1
solution
of
6-(N-alkylanilino)-2-phenylpyrimidin-
(DMF); H NMR [(CD3)3SO], d 1.91 (3H, s, 9-Me),
4(3H)-ones (1, 10 mmol) in acetic acid (5–15 mL) at
10–15 ꢁC was added sodium nitrite (20–40 mmol) by
portions, and the mixture was stirred for 2–5 h at
room temperature. The solid deposited was collected
by suction filtration and washed with water. Then,
the solid dried was recrystallized from an appropri-
ate solvent. All compounds were obtained as red
needles.
4.52 (3H, s, 10-Me), 7.63–7.84 (4H, m, 7-H and Ph–
m,pH), 8.08 (1H, dd, J6,8 = 1.5 Hz, J7,8 = 7.5 Hz, 8-H),
8.41–8.48 (3H, m, 6-H and Ph–oH); IR (cmꢀ1), 1540
(N–O), 1655 (C@O); Anal. C18H14N4O2 requires C,
67.91; H, 4.43; N, 17.60. Found: C, 68.16; H, 4.68; N,
17.63.
4.1.30. 4,10-Dihydro-7-methoxy-10-methyl-4-oxo-2-phenyl-
benzo[g]pteridine-5-oxide (3f). Yield, 2.67 g (80%); mp >
1
4.1.25. 4,10-Dihydro-10-methyl-4-oxo-2-phenylbenzo[g]-
pteridine-5-oxide (3a). Yield, 2.19 g (72%); mp > 300 ꢁC
300 ꢁC (DMF); H NMR [(CD3)3SO], d 4.07 (3H, s, 7-
OMe), 4.55 (3H, s, 10-Me), 7.65–7.73 (2H, m, Ph–
mH), 7.75–7.80 (1H, m, Ph–pH), 7.92 (1H, d,
J6,8 = 1.5 Hz, 6-H), 8.01 (1H, d, J = 9.0 Hz, 9-H),
8.41–8.47 (2H, m, Ph–oH), 8.50 (1H, dd, J6,8 = 1.5 Hz,
J8,9 = 9.0 Hz, 8-H); IR (cmꢀ1), 1539 (N–O), 1645
(C@O); Anal. C18H14N4O3 requires C, 64.66; H, 4.22;
N, 16.76. Found: C, 64.69; H, 4.47; N, 16.85.
1
(DMF); H NMR [(CD3)3SO], d 4.29 (3H, s, 10-Me),
7.50–7.60 (3H, m, Ph–m,pH), 7.73 (1H, ddd,
J6,7 = 7.8 Hz, J7,8 = 7.8 Hz, J7,9 = 1.5 Hz, 7-H), 8.09
(1H, ddd, J6,8 = 1.5 Hz, J7,8 = 7.8 Hz, J8,9 = 7.8 Hz,
8-H), 8.21 (1H, dd, J7,9 = 1.5 Hz, J8,9 = 7.8 Hz, 9-H),
8.45 (1H, dd, J6,7 = 8.7 Hz, J6,8 = 1.5 Hz, 6-H),

H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
255
4.1.31.
4,10-Dihydro-8-methoxy-10-methyl-4-oxo-2-
MTT-formazan that had formed. The optical density
of each well was measured at 570 nm (OD570) with a
microplate reader and the inhibition of cell growth (%)
was calculated as (1 ꢀ T/C) · 100, where C is the mean
OD570 of the control group and T is that of the treated
group. The IC50 was determined from the dose–response
curve.
phenylbenzo[g]pteridine-5-oxide (3g). Yield, 2.71 g
(81%); mp > 300 ꢁC (DMF); 1H NMR [(CD3)3SO], d
4.17 (3H, s, 8-OMe), 4.48 (3H, s, 10-Me), 7.57–7.70
(4H, m, 9-H and Ph–m,pH), 7.77 (1H, d, J6,7 = 7.5 Hz,
6-H), 8.42–8.47 (2H, m, Ph–oH), 8.52 (1H, dd,
J6,7 = 8.7 Hz, J7,9 = 1.5 Hz, 7-H); IR (cmꢀ1), 1540 (N–
O), 1640 (C@O); Anal. C18H14N4O3 requires C, 64.66;
H, 4.22; N, 16.76. Found: C, 64.68; H, 4.56; N, 16.82.
4.3. In vivo antitumor activities of 2-deoxo-2-phenyl-5-
deazaflavins (2a and 2b)
4.1.32. 10-Ethyl-4,10-dihydro-7-methyl-4-oxo-2-phenyl-
benzo[g]pteridine-5-oxide (3h). Yield, 2.56 g (77%); mp
268–269 ꢁC (MeOH); 1H NMR [(CD3)3SO], d 1.48
(3H, t, J = 7.2 Hz, 10-CH2–CH3), 2.49 (3H, s, 7-Me),
4.97 (2H, q, J = 7.2 Hz, 10-CH2–CH3), 7.51–7.59 (3H,
m, Ph–m,pH), 7.92 (1H, dd, J6,8 = 9.0 Hz, J8,9 = 1.5
Hz, 8-H), 8.17 (1 H, d, J = 9.0 Hz, 9-H), 8.28 (1H, d,
J6,8 = 1.5 Hz, 6-H), 8.44–8.49 (2H, m, Ph–oH) ; IR
(cmꢀ1), 1540 (N–O), 1650 (C@O); Anal. C19H16N4O2
requires C, 68.66; H, 4.85; N, 16.86. Found: C, 68.54;
H, 4.96; N, 16.78.
Antitumor activity of 2-deoxo-2-phenyl-5-deazaflavins
(2a and 2b) against subcutaneously transplanted hu-
man adenocarcinoma A 431 was tested according to
the modified method developed by Inaba et al.22 The
A 431 human adenocarcinoma was subcutaneously
transplanted into mice skin of neonate (BALB/c nude
mice), and the administration of test compounds was
started on the day (day 1) when the mice show tumor
growth in the 100–300 mm3 range. Staurosporine as a
positive control was used at 1 mg/kg and compounds
2a and 2b were used at 25 mg/kg for intratumoral
administration (it), at 100 mg/kg for intra-abdominal
administration (ip), and at 75 and 150 mg/kg for oral
administration (po). Since the test compounds showed
poor solubility in water, the administrations of these
suspended in 0.5% gum arabic were done into intra-tu-
mor (it), intraperitoneally (ip) or orally (po). The
administration schedule for it was five consecutive dai-
ly injections per week for 3 weeks (days 1–5, 8–12, and
15–19). Administration schedules for ip and po were
consecutive daily administration for 10 days (days 1–
10). The tumor volume (T/C %) was measured on the
17 days (it), 10 days (ip), and 18 days (po) after the
treatment of chemicals. The length (a) and width (b)
of the tumor mass were measured twice or three times
a week, and the tumor volume was calculated accord-
ing to the following formula: tumor volume
(mm3) = (a · b2)/2. Relative tumor volume (T/C %)
was calculated by the Battele Columbus formula, T/C
% = (Tn/T1) · 100/(Cn/C1), in which Cn = tumor volume
of control animals on day n and Tn = tumor volume of
actively treated animals on day n of the treatment.
4.1.33. 4,10-Dihydro-6,7,8-trimethoxy-10-methyl-4-oxo-
2-phenylbenzo[g]pteridine-5-oxide (3i). Yield, 3.19 g
1
(81%); mp > 300 ꢁC (DMF); H NMR [(CD3)3SO], d
3.94 (3H, s, 7-OMe), 4.18 (3H, s, 8-OMe), 4.21 (3H, s,
6-OMe), 4.43 (3H, s, 10-Me), 7.25 (1H, s, 9-H), 7.45–
7.60 (3H, m, Ph–m,pH), 8.53–8.56 (2H, m, Ph–oH); IR
(cmꢀ1),
1550
(N–O),
1645
(C@O);
Anal.
C20H18N4O5Æ7/10H2O requires C, 59.02; H, 4.80; N,
13.77. Found: C, 59.15; H, 4.67; N, 13.40.
4.2. Growth inhibitory activities of 2-deoxo-2-phenyl-5-
deazaflavins (2) and 2-deoxo-2-phenylflavin-5-oxides (3)
against human tumor cell lines
AraC was a commercial product of Yamasa Corpora-
tion (Choshi, Japan) and 3-(3,4-dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide (MTT) was purchased
from Sigma (St. Louis, MO). Five human tumor cell
lines of NCI-H 460 (human lung cancer), HCT 116 (hu-
man colon carcinoma), A 431 (human adenocarcinoma),
CCRF-HSB-2 (human T-cell acute lymphoblastoid leu-
kemia), and KB (human oral epidermoid carcinoma)
were used in this study. Cell lines were maintained in
RPMI 1640 medium (Life Technologies, Rockville,
MD) supplemented with 10% heat-inactivated fetal bo-
vine serum (Moregate Laboratories, Melbourne, Aus-
tralia) and 60 lg/mL of kanamycin. The MTT assay
developed by Mosmann20 was modified21 and used to
determine the inhibitory effects of test compounds on
cell growth in vitro. Briefly, medium containing 5 · 103
cells (NCI-H 460, HCT 116, A 431, CCRF-HSB-2 or
KB cells) was seeded into each well of a 96-well micro-
plate, with the compound solution added simultaneous-
ly to triplicate wells, before the final volume was made
up to 100 lL. The plate was incubated at 37 ꢁC for
72 h in a humidified atmosphere of 5% CO2. After the
incubation, 10 lL of an MTT solution (5 mg/mL in
phosphate-buffered saline lacking calcium and magne-
sium) was added to each well. After a further incubation
for 4 h at 37 ꢁC, 100 lL of 0.02 N HCl/50% N,N-dimeth-
ylformamide/20% SDS was added to solubilize any
4.4. Docking study
AutoDock 3.05,18 a grid-based docking program, was
used for docking study. The crystal structure of protein
tyrosine kinase pp60c-src (PDB code: 1skj),23 which is a
c-src tyrosine kinase of Rous sarcoma virus, was used
in this docking study. The 3D structure of 1skj was
reported by Plummer et al.23 using X-ray diffraction
˚
technique with a resolution of 2.10 A. The 1skj was re-
trieved from the Brookhaven protein database:
URL*http://www.rcsb.org/pdb/Wecome.do accessed on
January 10, 2006, as a complex bound with the inhibitor
ur2: 4-[3-carboxymethyl-3-(4-phosphonooxybenzyl)ure-
ido]-4-[(3-cyclohexylpropyl)-methylcarbamoyl]butyric
acid. For the docking study, water molecules and ur2 li-
gand were removed from the protein, then the polar
hydrogens and united atom Kollman charges were as-
signed for the protein. Then, various inhibitors were
docked within the prepared protein (1skj). The mode

256
H. I. Ali et al. / Bioorg. Med. Chem. 15 (2007) 242–256
of interaction of ur2 ligand against 1skj tyrosine was
used a standard docked model, the one used for the cal-
culation of the root mean square deviation (RMSD) of
the docked inhibitors.
are grateful to each organization for their generous sup-
port for our current work.
Supplementary data
4.4.1. Docking parameters. Prior to the AutoDock,
AutoGrid was carried out for the preparation of the grid
map using a grid box with a npts (number of points in
xyz) of 60–60–60 A box and 120–120–120 A box, which
encloses the original ligand ur2. The box spacing was
˚
0.3 A and grid center was designated at dimensions (x,
y, z): ꢀ5.143, 62.487 and ꢀ8.776. A scoring grid was cal-
culated from the original ligand structure (ur2) to mini-
mize the computation time. Finally AutoDock was run
using maximum number of retries and generations of
10,000 and 27,000, respectively. The genetic algorithm
with local search (GALS) was used for calculation of
the docking possibilities. The complexes obtained by
AutoDock were minimized using a maximum 300 itera-
tions and the hybrid GALS runs with max of 250 cycles
using different random number seeds to obtain score
convergence.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmc.2006.09.063.
˚
˚
References and notes
1. O’Brien, D. E.; Weinstock, L. T.; Cheng, C. C. J.
Heterocycl. Chem. 1970, 7, 99.
2. Hemmerich, P.; Massey, V.; Fenner, H. FEBS Lett. 1972,
84, 5.
3. Graham, D. W.; Brown, J. E.; Ashton, W. T.; Brown, R.
D.; Rogers, E. F. Experientia 1977, 33, 1274.
4. Nagamatsu, T.; Yoneda, F.; Kawashima,Y.; Yamagishi,
T.; Ikeya, H. The 6th Annual meeting of Division of
medicinal chemistry, 1987, Tsukuba, Japan, Book
Abstract, pp.148.
5. Kanaoka, Y.; Ikeuchi, Y.; Kawamoto, T.; Bessho, K.;
Akimoto, N.; Mikata, Y.; Nishida, M.; Yano, S.; Sasaki,
T.; Yoneda, F. Bioorg. Med. Chem. 1998, 6, 301.
6. Kimachi, T.; Yoneda, F.; Sasaki, T. J. Heterocycl. Chem.
1992, 29, 763.
7. Kawamoto, T.; Ikeuchi, Y.; Hiraki, J.; Eikyu, Y.; Shimizu,
K.; Tomishima, M.; Bessho, K.; Yoneda, F.; Mikata, Y.;
Nishida, M.; Ikehara, K. Bioorg. Med. Chem. Lett. 1995,
5, 2115.
8. Lewis, R. A. J. Med. Chem. 2005, 48, 1638.
9. Ooms, F. Curr. Med. Chem. 2000, 7, 141.
10. Anderson, A. C. Chem. Biol. 2003, 10, 787.
11. Jackson, R. C. Curr. Opin. Biotechnol. 1995, 6, 646.
12. Hu, Z.; Wang, S.; Southerland, W.M. Dock odysseys II.
ACS division of chemical information final program,
The 228th ACS national meeting, August 22–26, 2004,
Philadelphia, PA.
13. Yoneda, F.; Mori, K.; Ono, M.; Kadokawa, Y.; Nagao,
E.; Yamaguchi, H. Chem. Pharm. Bull. 1980, 28, 3514.
14. Coppola, G. M.; Schuster, H. F. J. Heterocycl. Chem.
1989, 26, 957.
15. Goldner, H.; Dietz, G.; Carstens, E. Ann. 1966, 694, 142.
16. Yoneda, F.; Sakuma, Y.; Mizumoto, S.; Ito, R. J. Chem.
Soc. Perkin Trans. 1 1976, 1, 1805.
17. Akinaga, S.; Gomi, K.; Morimoto, M.; Tamaoki, T.;
Okabe, M. J. Biol. Chem. 1991, 265, 3632.
18. Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.;
Hart, W. E.; Belew, R. K.; Olson, A. J. J. Comput. Chem.
1998, 19, 1639.
19. Akaho, E.; Fujikawa, C.; Runion, H. I.; Hill, C. R.;
Nakaho, H. J. Chem. Software 1999, 5, 147.
20. Mosmann, T. J. Immunol. Methods 1983, 65, 55.
21. Miura, S.; Yoshimura, Y.; Endo, M.; Machida, H.;
Matsuda, A.; Tanaka, M.; Sasaki, T. Cancer Lett. 1998,
129, 103.
4.4.2. Preparation of small molecules. ChemDraw ultra
8.0 software [Chemical Structure Drawing Standard;
Cambridge Soft corporation, USA (2003)] was used
for construction of compounds which were converted
to 3D structures using Chem3D ultra 8.0 software and
the constructed 3D structures were energetically mini-
mized by using MOPAC (semi-empirical quantum
mechanics) with AM1 mozyme geometry, 100 iterations
and minimum RMS gradient of 0.10.
4.4.3. Evaluation of docked results. Accelrys DS model-
ing 1.5 software [DS modeling 1.5; Accelrys Inc., San
Diego, CA (2005)] was utilized for molecular modeling
for the evaluation of hydrogen bonds in ligand–recep-
tor interaction. The correct hydrogen bonds were con-
sidered if their geometry angle within 110–180ꢁ
according to Murray-Rust et al.24 While the false
hydrogen bonds between N-oxide moiety and Cys42
were omitted because the bond angles were less than
90ꢁ which are not matched with the considered
parameters. Also DS modeling was used for measure-
ment of RMSD, which was computed and expressed
˚
in A as a structural comparison of two molecules in
terms of distance. That is, it was measured as distance
between the centroid of the docked inhibitor and the
original ur2 ligand.
Acknowledgments
This work was supported by Research Grant of Taisho
Pharmaceutical Co., LTD, Saitama, Japan. This was
also supported by Collaborative Research Grant A of
Kobe Gakuin University, in Japan. The authors are
indebted to the SC-NMR Laboratory of Okayama Uni-
versity for the NMR spectral measurements and the
Okayama University Information Technology Center
for using Accelrys Discovery Studio 1.5. The authors
22. Inaba, M.; Tashiro, T.; Kobayashi, T.; Fujimoto, S.;
Sakurai, Y.; Maruo, K.; Ohnishi, Y.; Ueyama, Y.;
Nomura, T. Jpn. J. Cancer Res. 1986, 77, 190.
23. Plummer, M. S.; Holland, D. R.; Shahripour, A.; Lunney,
E. A.; Fergus, J. H.; Marks, J. S.; McConnell, P.; Mueller,
W. T.; Sawyer, T. K. J. Med. Chem. 1997, 40, 3719.
24. Murray-Rust, P.; Glusker, J. P. J. Am. Chem. Soc. 1984,
106, 1018.