Synthesis, Hydrolytic Stability, and Antileukemic Activity of Azacytidine Nucleoside Analogs

Article abstract of DOI:10.1007/s11094-016-1375-4

New azacytidine nucleoside analogs with modified carbohydrate moieties were synthesized. Screening identified a highly active 2′-fluoro-containing azacytidine analog that could potentially be of interest as an agent for treating acute myelogenous leukemia and myelodysplastic syndrome.

Full text of DOI:10.1007/s11094-016-1375-4

DOI 10.1007/s11094-016-1375-4
Pharmaceutical Chemistry Journal, Vol. 49, No. 12, March, 2016 (Russian Original Vol. 49, No. 12, December, 2015)
SEARCH FOR NEW DRUGS
SYNTHESIS, HYDROLYTIC STABILITY, AND ANTILEUKEMIC
ACTIVITY OF AZACYTIDINE NUCLEOSIDE ANALOGS
T. S. Bozhok,1 E. N. Kalinichenko,1,* B. B. Kuz’mitskii,1 and M. B. Golubeva1
Translated from Khimiko-Farmatsevticheskii Zhurnal, Vol. 49, No. 12, pp. 7 – 11, December, 2015.
Original article submitted October 28, 2014.
New azacytidine nucleoside analogs with modified carbohydrate moieties were synthesized. Screening identi-
fied a highly active 2¢-fluoro-containing azacytidine analog that could potentially be of interest as an agent for
treating acute myelogenous leukemia and myelodysplastic syndrome.
Keywords: myelodysplastic syndrome, azacytidine, glycosylation, fluoro-deoxynucleosides, hydrolytic sta-
bility, antileukemic activity.
Epigenetic therapy occupies
a
special place in
modified carbohydrates that possessed higher antileukemic
activity than Aza-C, which is used in tumor chemotherapy.
Fluorine atoms in biologically active molecules are
known to affect considerably their physicochemical and bio-
logical properties [8]. The increased biological activity can
be explained by several factors. The affinity for lipophilic
and also aqueous phases increases the permeability through
cell membranes. The special biological properties of fluori-
nated nucleosides are also due to the ability to form F••H
H-bonds, the electrostatic interaction, and the effects on the
stereochemistry and conformation of the compounds. It is
noteworthy that the F atom acts as a bioisostere of groups
such as methyl, amine, and hydroxyl. Bioisosteric replace-
ment is one approach used in medicinal chemistry to design
more efficacious and safe drugs [9].
myelodysplastic syndrome (MDS) therapy. The term “epige-
netic” is understood to mean inherited mutations in cells that
are not related to damage of gene nucleotide sequences and
are reversible, e.g., DNA methylation [1]. One approach to
MDS epigenetic therapy is the use of hypomethylating
agents [2] such as azacytidine (5-azacitidine, Aza-C, Vidaza)
and decitabine (2¢-deoxy-5-azacytidine, Dacogen). These
antimetabolites exert their hypomethylating effect through
the N atom, which replaces the cytosine C-5 atom, makes ad-
dition of a methyl to this part of the cytidine analog impossi-
ble, and inhibits DNA methyltransferases. As a result,
onco-suppressor gene functions and cell-cycle regulatory
mechanisms in the hematopoietic myeloid lineage are re-
stored [3, 4]. Currently, Aza-C is an effective therapeutic for
treating MDS, acute myelogenous leukemia (AML), and
chronic myelomonocytic leukemia without MDS symptoms.
Furthermore, the clinical efficacy of the drug in combination
with other treatments was proved against not only MDS but
also solid tumors [5]. Broad use of Aza-C in medical practice
is limited by the high general toxicity and hydrolytic degra-
dation at physiological pH values [6, 7]. The goal of the pres-
ent investigation was to find Aza-C nucleoside analogs with
5¢-Deoxy derivatives of pyrimidine nucleosides, which
have a broad spectrum of antitumor activity and can accumu-
late selectively in tumor tissue (through the action of tumor
angiogenesis factor thymidine phosphorylase), are especially
interesting. This minimizes the systemic activity of the drug
on healthy tissues [10].
We synthesized Aza-C and its nucleoside analogs that in-
cluded modifications at the carbohydrate 2¢-, 3¢-, or 5¢-posi-
tions. Condensation of 2,4-bis(trimethylsilyl)-5-azacytosine
(II), which was synthesized from 5-azacytosine, with the
sugars 1-O-acetyl-2,3,5-tri-O-benzoyl-b -D-ribofuranoside
(Ia); 1,2,3-tri-O-acetyl-5-deoxy-b -D-ribofuranoside (Ib);
1-O-acetyl-2,5-di-O-benzoyl-3-deoxy-3-fluoro-a /b -D-ribofu
1
Institute of Bioorganic Chemistry, National Academy of Sciences of
Belarus, 5/2 Akad. V. F. Kuprevicha St., Minsk, 220141 Republic of
Belarus.
*
e-mail: kalinichenko@iboch.bas-net.by.
804
0091-150X/15/4912-0804 © 2015 Springer Science+Business Media New York

Synthesis, Hydrolytic Stability, and Antileukemic Activity
805
Scheme 1
Ia – c
1
IIIa – c
1
IVa – c
1
2
3
2
3
3
2
1
2
3
Ia: R = R = R = OBz, b -isomer; Ib: R = R = OAc, R = H, b -isomer; Ic: R = R = OBz, R = F, a /b -isomers; IIIa: R = R = R = OBz;
IIIb: R = R = OAc, R = H; IIIc: R = R = OBz, R = F; IVa: R = R = R = OH; IVb: R = R = OH, R = H; IVc: R = R = OH, R = F.
1
2
3
1
3
2
1
2
3
1
2
3
1
3
2
Scheme 2
V
VI
VIIa, c
VIIIa, c
VIIa, VIIIa: b -isomer; VIIb, VIIIb: a -isomer.
vacuo to dryness. The residue was evaporated again with an-
hydrous toluene (10 mL). Solid II was treated with a solution
of the appropriate acetate Ia-c (1.24 mmol) in anhydrous
CH Cl (6 mL), cooled to 0°C, treated with trimethylsilyltri-
ranoside (Ic); or 3,5-di-O-benzoyl-2-deoxy-2-fluoro-a -D-
arabinofuranosyl bromide (VI) gave benzoyl derivatives
IIIa-c and VIIa and -b. These were isolated pure via column
chromatography over silica gel. Debenzoylation of interme-
diate nucleosides IIIa-c and VIIa and -b by ammonia in
MeOH produced Aza-C (IVa) and its 5¢-deoxyribo-,
3¢-fluorodeoxyribo-, and 2¢-fluorodeoxyarabino-analogs IVb
and -c and VIIIa and -b, respectively (Schemes 1 and 2).
The chemical structures of the synthesized nucleosides were
elucidated using PMR and 19F NMR data (Table 1) and UV
and mass spectrometry. CD spectroscopy established the
stereochemical features of the a /b -fluoronucleoside Aza-C
analogs [11].
2
2
fluoromethanesulfonate (0.26 mL, 1.438 mmol), stirred un-
der Ar for 4 h at room temperature, diluted with CH Cl , and
2
2
poured with stirring into cold saturated aqueous NaHCO .
3
The organic layer and aqueous phase were extracted with
CH Cl . The organic extract was dried and evaporated in
2
2
vacuo to dryness. The solid was chromatographed over silica
gel using first a mixture of EtOAc:hexane (1:3) and then
EtOAc:hexane:MeOH (30:90:7). Fractions containing reac-
tion products IIIa-c were combined and evaporated to dry-
ness.
4-Amino-1-(2,3,5-tri-O-benzoyl-b -D-ribofuranosyl)-
1,3,5-triazin-2(1H)-one (IIIa). Yield 65%, white stable
EXPERIMENTAL CHEMICAL PART
foam. Mass spectrum, m/z (I ): 579.1 [M + Na]+ (100),
rel
445.1 (65), 595.1 [M + K]+ (58).
UV spectra in the range 200 – 300 nm were taken on a
Varian Cary 100 spectrophotometer (USA). NMR spectra
were obtained on a Bruker Avance-500 spectrometer (USA)
at operating frequency 500 MHz for PMR and 470 MHz for
19F NMR. Mass spectra were obtained in a LC-MS system
using an Accela HPLC and LCQ Fleet triple quadrupole
ion-trap mass detector (Thermo Electron) with electrospray
ionization. CD spectra were taken on a Jasco J-20
spectropolarimeter (Japan).
4-Amino-1-(2,3-di-O-acetyl-5-deoxy-b -D-ribofuranosyl)-
1,3,5-triazin-2(1H)-one (IIIb). Yield 80%, white stable
foam. Mass spectrum, m/z (I ): 335.1 [M + Na]+ (100),
rel
201.1 (45), 351.1 [M + K]+ (27).
4-Amino-1-(2,5-di-O-benzoyl-3-deoxy-3-fluoro-b -D-
ribofuranosyl)-1,3,5-triazin-2(1H)-one (IIIc). Yield 67%,
white stable foam. Mass spectrum, m/z (I ): 477.1
rel
[M + Na]+ (100), 493.1 [M + K]+ (58), 343.1 (53), 150.1
General method for glycosylation of 5-azacytosine by
1-O-acyloxysugars Ia-c. A suspension of 5-azacytosine
(0.16 g, 1.428 mmol) and ammonium sulfate (0.01 g) in
hexamethyldisilazane (HMDS, 6 mL) was refluxed for 10 h.
The resulting homogeneous solution was evaporated in
(15).
Glycosylation of 5-azacytosine by bromide VI. A sus-
pension of 5-azacytosine (0.215 g, 1.92 mmol) and ammo-
nium sulfate (0.01 g) in HMDS (8 mL) was refluxed for
10 h. The homogeneous solution was evaporated in vacuo to

806
T. S. Bozhok et al.
dryness. The residue was evaporated again with anhydrous
toluene (10 mL). The residue of II was treated with a solution
of bromide VI [prepared from 1,3,5-tri-O-benzoyl-2-deoxy-
2-fluoro-a -D-arabinofuranoside (V, 0.7 g, 1.51 mmol)] in
anhydrous 1,2-dichloroethane (7.5 mL), refluxed for 20 h
[M + Na]+ (100), 455.1 [M + H]+ (64), 493.1 [M + K]+ (43),
343.1 (30), 150.1 (29).
4-Amino-1-(3,5-di-O-benzoyl-2-deoxy-2-fluoro-a -D-a
rabinofuranosyl)-1,3,5-triazin-2(1H)-one (VIIb). Yield
5%, white amorphous powder. Mass spectrum, m/z (I ):
rel
477.1 [M + Na]+ (100), 455.1 [M + H]+ (47), 493.1 [M + K]+
under Ar, cooled, diluted with CHCl , and poured with stir-
3
(38), 343.1 (37), 150.1 (31).
ring into ice water. The organic layer and aqueous phase
General procedure for debenzoylation of nucleosides
IIIa-c and VIIa and -b. A suspension of the appropriate
nucleoside IIIa-c or VIIa or -b (0.318 mmol) in anhydrous
MeOH (5 mL) was treated with MeOH (8 mL) saturated
were extracted with CHCl . The organic extract was dried
3
and evaporated in vacuo to dryness until a thick stable foam
formed. The residue was chromatographed over silica gel us-
ing first EtOAc and then CHCl :petroleum-ether:MeOH
with dry NH at 0°C, stirred for 7 h at room temperature,
3
3
(15:7:1). Fractions containing reaction products VIIa and -b
were combined and evaporated to dryness.
evaporated to dryness, and evaporated again with anhydrous
MeOH. The solid was chromatographed over silica gel using
4-Amino-1-(3,5-di-O-benzoyl-2-deoxy-2-fluoro-b -D-a
rabinofuranosyl)-1,3,5-triazin-2(1H)-one (VIIa). Yield
a linear gradient of MeOH (0® 30%, v/v) in CHCl . Frac-
tions containing reaction product were combined and evapo-
3
82%, white stable foam. Mass spectrum, m/z (I ): 477.1
rated to dryness. The solid was crystallized from Et O.
rel
2
19
TABLE 1. PMR and F NMR Spectra of III-IVa-c and VII-VIIIa and -b
PMR spectrum,* d , ppm, J, Hz
19F NMR spectrum,* d , ppm
Compound
8.18 (c, 1H, H6), 8.08 – 7.33 (m, 15H, arom), 6.79 (br.s, 1H, NHa), 6.07 (d, 1H, J 4.0, H1¢),
5.99 (t, 1H, H2¢), 5.91 (m, 1H, J 4.0, J 5.9, H3¢), 5.78 (br.s, 1H, NHa), 4.81 (dd, 1H, J 3.0, J
12.0, H5¢), 4.77 (m, 1H, H4¢), 4.71 (dd, 1H, J 4.7, H5¢¢)
IIIa
IIIb
–
–
8.37 (c, 1H, H6), 7.72 (d, 2H, J 22.0, NH2), 5.60 (d, 1H, J 3.5, H1¢), 5.54 (m, 1H, J 6.3,
H2¢), 5.17 (t, 1H, H3¢), 4.06 (m, 1H, H4¢), 2.06 (d, 6H, J 2.3, OAc,), 1.31 (d, 3H, J 6.2,
CH3)
8.08 (c, 1H, H6), 8.08 – 7.44 (m, 10H, arom), 7.02 (br.s, 1H, NHa), 5.96 (d, 1H, J 4.5, H1¢),
5.78 (br.s, 1H, NHá), 5.77 (m, 1H, J 10.1, H2¢), 5.67 (dt, 1H, J 53.1, H3¢), 4.73 (dd, 1H, J
3.5, H5¢), 4.69 (dm, 1H, J 20.8, H4¢), 4.63 (dd, 1H, J 4.8, J 12.0, H5¢¢)
– 202.38 (m, FC3¢)
IIIc
VIIa
VIIb
IVa
8.30 (d, 1H, J 1.6, H6), 8.08 – 7.44 (m, 10H, arom), 6.53 (br.s, 1H, NHa), 6.37 (dd, 1H, J
2.4, J 21.5, H1¢), 5.87 (br.s, 1H, NHb), 5.62 (dd, 1H, J 16.3, H3¢), 5.42 (dd, 1H, J 49.7, H2¢),
4.81 (dd, 1H, J 5.3, J 12.0, H5¢¢), 4.76 (dd, 1H, J 3.6, H5¢), 4.55 (m, 1H, H4¢)
– 201.27 (m, FC2¢)
8.46 (s, 1H, H6), 8.02 – 7.51 (m, 12H, arom, NH2), 6.11 (d, 1H, J 16.4 Hz, H1¢), 5.79 – 5.67
(m, 2H, J 50.8, J 18.0, H2¢, H3¢), 5.22 (m, 1H, H4¢), 4.61 (dd, 1H, J 3.9, J 12.0, H5¢), 4.55
(dd, 1H, J 5.4, H5¢¢)
– 186.58 (m, FC2¢)
–
–
8.58 (s, 1H, H6), 7.53 (d, 2H, J 11.0, NH2), 5.66 (d, 1H, J 3.6, H1¢), 5.43 (d, 1H, J 5.0,
2¢OH), 5.12 (t, 1H, J 4.9, 5¢OH), 5.04 (d, 1H, J 5.8, 3¢OH), 4.08 – 3.98 (m, 2H, H2¢, H3¢),
3.84 (m, 1H, H4¢), 3.68 (m, 1H, H5¢), 3.54 (m, 1H, H5¢¢)
IVb
8.25 (s, 1H, H6), 7.57 (d, 2H, J 4.8, NH2), 5.56 (d, 1H, J 3.7, H1¢), 5.34 (d, 1H, J 4.7,
2¢OH), 5.02 (d, 1H, J 5.9, 3¢OH), 4.15 (m, 1H, J 5.2, H2¢), 3.84 (m, 1H, H3¢), 3.74 (m, 1H,
H4¢), 1.27 (d, 3H, J 6.3, ÑH3)
IVc
8.49 (s, 1H, H6), 7.66 (d, 2H, J 11.9, NH2), 5.83 (d, 1H, J 6.0, 2¢OH), 5.79 (d, 1H, J 7.2,
H1¢), 5.25 (t, 1H, J 5.1, 5¢OH), 4.99 (dd, 1H, J 54.3, H3¢), 4.45 (dt, 1H, J 23.4, H2¢), 4.18
(dm, 1H, J 25.3, H4¢), 3.61 – 3.60 (m, 2H, H5¢, H5¢¢)
– 199.43 (m, FC3¢)
– 197.43 (m, FC2¢)
– 186.58 (m, FC2¢)
VIIIa
VIIIb
8.31 (s, 1H, H6), 7.63 (d, 2H, J 3.2, NH2,), 6.04 (dd, 1H, J 3.7, J 17.5, H1¢), 5.86 (d, 1H, J
4.8, 3¢OH), 5.05 (t, 1H, J 5.8, 5¢OH), 4.97 (dm, 1H, J 2.3, J 52.1, H2¢), 4.18 (dm, 1H, J 14.4,
H3¢), 3.79 (m, 1H, H4¢), 3.58 (m, 1H, H5¢), 3.51 (m, 1H, J 12.0, H5¢¢)
8.23 (s, 1H, H6), 7.61 (d, 2H, J 10.0, NH2,), 5.84 (dd, 1H, J 1.5, J 15.8, H1¢), 5.77 (d, 1H, J
4.1, 3¢OH), 5.18 (dt, 1H, J 50.9, H2¢), 5.01 (t, 1H, J 5.6, 5¢OH), 4.33 (m, 1H, H4¢), 4.28
(dm, 1H, J 19.9, H3¢), 3.49 – 3.47 m (2H, H5¢, H5¢¢)
*
IIIa and -c and VIIa, CDCl ; IVa-c, IIIb, VIIb, and VIIIa and -b, DMSO-d .
3
6

Synthesis, Hydrolytic Stability, and Antileukemic Activity
807
chromatograph with a Waters 996 spectrophotometric detec-
tor. Chromatography used an EC 250/4.6 Nucleodur 100-5
4-Amino-1-b -D-ribofuranosyl-1,3,5-triazin-2(1H)-one
(IVa). Yield 60%, white amorphous powder, mp
C
ec column (Macherey-Nagel AG) packed with octade-
228 – 230°C (lit. 231 – 233°C [12]). UV spectrum (H O),
18
2
cylsilyl silica gel for chromatography P (5 m m) and mobile
phase of KH PO (2.72 g/L) in MeOH adjusted to pH 6.8 us-
l
, nm (loge ): 241 (3.94). CD spectrum (H O), l , nm
2
max
([q × 10 – 3]): 215 (+ 2.04), 250 (+ 10.82), 285 (0). Mass spec-
2
4
trum, m/z (I ): 267.1 [M + Na]+ (100), 113.0 (81), 245.1
ing Et N (50:950, v/v) at flow rate 1 mL/min. Chromato-
3
rel
[M + H]+ (24).
grams were recorded at 241 nm. The injected sample volume
was 5 m L; chromatography time, 20 min; column tempera-
ture, 25°C. Peaks of impurities were identified by comparing
chromatograms of the test solution with those of dissolved
USP standards 5-azacytosine, Aza-C, and 1-b -D-ribofurano-
syl-3-guanylurea picrate. In addition, chromatograms of the
test solution and degradation products were recorded on a
LC-MS.
4-Amino-1-(5-deoxy-b -D-ribofuranosyl)-1,3,5-triazin-
2(1H)-one (IVb). Yield 92%, white amorphous powder, mp
212 – 214°C. UV spectrum (H O), l , nm (loge ): 242
2
max
(3.77). Mass spectrum, m/z (I ): 251.1 [M + Na]+ (100),
rel
113.0 (52), 229.1 [M + H]+ (23).
4-Amino-1-(3-deoxy-3-fluoro-b -D-ribofuranosyl)-1,3,5-
triazin-2(1H)-one (IVc). Yield 67%, white amorphous pow-
der, mp 237 – 239°C. UV spectrum (H O), l , nm (loge ):
241 (3.94). CD spectrum (MeOH), l , nm ([q × 10 – 3]): 215
2
max
EXPERIMENTAL BIOLOGICAL PART
(0), 250 (+ 5.59), 280 (0). Mass spectrum, m/z (I ): 269.1
rel
[M + Na]+ (100), 113.0 (70), 247.1 [M + H]+ (23), 285.0
[M + K]+ (11).
Preliminary in vitro screening for antileukemic activity
of synthesized IVa-c and VIIIa used standard study methods
[14, 15]. Human tumor cell lines from the cell-culture collec-
tion of the RAS Institute of Cytology, St. Petersburg were
used to study the specific cytotoxicity of the synthesized nu-
cleosides. These were KG-1 (acute myelogenous leukemia),
TPH-1 (acute monocytic leukemia), HL-60 (acute promyelo-
cytic leukemia), K-562 (chronic myelogenous leukemia),
and MOLT-3 (acute T-lymphoblastic leukemia). Cells were
cultured in RPMI 1640 medium with 20% added fetal bovine
serum for HL-60, KG-1, and MOLT-3 and 10%, for THP-1
and K-562.
4-Amino-1-(2-deoxy-2-fluoro-b -D-arabinofuranosyl)-
1,3,5-triazin-2(1H)-one (VIIIa). Yield 75%, white amor-
phous powder, mp 204 – 206°C. UV spectrum (H O), l
,
max
2
nm (loge ): 241 (3.94). CD spectrum (MeOH), l , nm
([q × 10 – 3]): 215 (+ 2.15), 245 (+ 7.54), 285 (0). Mass spec-
trum, m/z (I ): 269.1 [M + Na]+ (100), 113.0 (59), 247.1
rel
[M + H]+ (33), 285.0 [M + K]+ (27).
4-Amino-1-(2-deoxy-2-fluoro-a -D-arabinofuranosyl)-
1,3,5-triazin-2(1H)-one (VIIIb). Yield 78%, white amor-
phous powder. UV spectrum (H O), l , nm (loge ): 241
2
max
Cytotoxic activity was studied at concentrations
0.01 – 50.0 m M. The reference drug was Aza-C drug sub-
stance prepared by the developed method. Solutions (10 m L)
of test compounds were placed in triplicate into wells of a
96-well plate with leukemia cells (1.5 ´ 105 cells/mL) and
(3.94). CD spectrum (MeOH), l , nm ([q × 10 – 3]): 215
(– 3.45), 240 (– 4.63), 285 (0). Mass spectrum, m/z (I ):
rel
269.1 [M + Na]+ (100), 113.0 (59), 247.1 [M + H]+ (33),
285.0 [M + K]+ (27).
The hydrolytic stability was studied using an HPLC
method [13]. Solutions of IVa-c and VIIIa of concentration
0.082 ´ 10–4 M were prepared in water for injection (pH
6.7), phosphate buffer (pH 3.2), and borate buffer (pH 10.2).
incubated at 37°C in a 5% CO atmosphere for 48 and 72 h.
2
When the cultivation with the test nucleosides was finished,
HPLC chromatograms were obtained on
a
Waters
TABLE 3. Specific Cytotoxicity of New Aza-C Nucleoside Ana-
logs in Human Hemoblastosis Cell Culture
Mean effective concentration (EC50, m M)
of tested compounds
Leukemia
TABLE 2. Time Required to Reduce the Concentration of Aza-C
and Its Nucleoside Analogs VIIIa and IVb by 10, 20, and 30% in
Water for Injection at 21°C
cell line
VIIIa
IVc
IVb (5¢-deoxy-
Aza-C)
IVa (Aza-C)
(2¢F-ara-Aza-C) (3¢F-Aza-C)
Time, min
Residual
KG-1
> 1.0
> 10.0
> 50.0
> 50.0
> 50.0
i/a#
amount,
3.66 ± 0.10 0.35 ± 0.06*
4.25 ± 0.25 0.51 ± 0.01*
5.37 ± 0.07 0.58 ± 0.05*
VIIIa
IVc,
IVa
IVb (5¢-deoxy-
Aza-C)
%
(Aza-C)
(2¢F-ara-Aza-C) (3¢F-Aza-C)
TPH-1
HL-60
K-562
MOLT-3
> 10.0
> 10.0
> 50.0
i/a#
90
80
70
25 ± 1.2
80 ± 3.9
43 ± 1.7*
89 ± 4.2
12 ± 0.5*
30 ± 1.5
75 ± 3.4
65 ± 3.1*
175 ± 8.2
…
> 10.0
4.64 ± 0.15
i/a#
i/a#
350 ± 14.8
360 ± 15.1
*
*
Statistically significant difference vs. reference compound
Statistically significant difference vs. reference compound
(p < 0.05); # i/a, inactive.
(p < 0.05).

808
T. S. Bozhok et al.
each well was treated with MTT solution in phosphate buffer
and incubated for an additional 4 h. Then, cells were precipi-
tated by centrifugation. The medium was removed from the
wells. Each well was treated with DMSO (150 m L) to dis-
solve formazan crystals. The optical density of the dissolved
formazan in the multi-well plates was measured on a Vityaz?
reader (Rep. Belarus) at 570 nm.
statistically significant cytotoxicity against highly sensitive
KG-1 leukemia cells at concentrations of 1.0 and 10.0 m M,
respectively. However, the cytotoxicity in the cell cultures
did not strengthen but regressed as the amount of test com-
pound increased. The low in vitro cytotoxicity of 5¢-deoxy
analog IVb was probably due to the lack of a 5¢-OH in the
carbohydrate and different metabolism than Aza-C and its
fluorodeoxy analogs IVc and VIIIa.
The average of three parallel measurements was calcu-
lated for each concentration of test compound. The degree of
cell growth suppression was plotted as a function of the loga-
rithm of the concentration. The mean effective concentration
for 50% suppression of the viability of the leukemia cells
Table 2 also shows that nucleoside analog VIIIa with a
C-2¢ F atom in the arabino configuration was 1.7 times more
hydrolytically stable than Aza-C at neutral pH values. The
time required to degrade 10% of VIIIa was 43 min whereas
this limit for Aza-C was reached already after 25 min. This
was an important parameter because the elimination half-life
(EC ) was determined graphically for active compounds.
50
Table 3 presents the results for the specific in vivo cytotoxic
activity of the Aza-C nucleoside analogs. Results were pro-
cessed statistically using the Student t-criterion.
of Aza-C after s.c. injection was (41 ± 8) min. The C
max
(750 ± 403 ng/mL) was reached 0.5 h after injection. It is
noteworthy that the hydrolytic stability of 3¢F-Aza-C de-
creased noticeably compared with the isomeric arabino ana-
log. Obviously, conformational features caused by introduc-
ing the F atom in the carbohydrate part had a significant in-
fluence on the chemical stability of the F-containing Aza-C
analogs. Apparently, the F atom in 5-Aza-C analogs IVc and
VIIIa, like for 1-(3-deoxy-3-fluoro-b -D-ribofuranosyl)ade-
nine and 1-(3-deoxy-3-fluoro-b -D-xylofuranosyl)adenine,
shifted most effectively the S« N equilibrium toward one of
the conformers [16]. One of the principal structural differ-
ences of the S and N conformers was the distance between
RESULTS AND DISCUSSION
The studies showed that nucleoside analog VIIIa
(2¢F-ara-Aza-C) exhibited substantially greater inhibition of
the viability for three lines of human acute myelogenous leu-
kemia models than analogs IVc (3¢F-Aza-C) and IVb
(5¢-deoxy-Aza-C). The cytotoxicity of this nucleoside was
also significantly greater than that of the currently widely
used antileukemic drug Aza-C.
Nucleoside analog 2¢F-ara-Aza-C, like reference drug
Aza-C, suppressed selectively and dose-dependently the via-
bility of cell lines KG-1 acute myelogenous leukemia,
TPH-1 acute monocytic leukemia, and HL-60 acute
promyelocytic leukemia. Chronic myelogenous leukemia
K-562 cells exhibited low sensitivity to 2¢F-ara-Aza-C
whereas Aza-C in the K-562 model exhibited practically the
same cytotoxicity as against the three acute myelogenous
leukemia cell lines. Acute T-lymphoblastic leukemia
MOLT-3 cells were completely resistant to the new
nucleoside analog and Aza-C.
the hetero-base and exocyclic -5¢CH OH substituent. It was
2
substantially greater in the S conformer than in the N con-
former. Intramolecular interaction between the 5¢-OH and
the triazine ring C-6 probably facilitated hydrolysis with ring
opening. As a result, the degradation half-life was shorter for
the F-containing nucleosides than for the 5¢-deoxy analog. It
was shown earlier that intramolecular interaction between
the 5¢-OH and triazine ring of decitabine facilitated its hydro-
lysis with ring opening, as a result of which the degradation
half-life of the b -anomer was shorter than that of the
a -anomer [17].
It is noteworthy that nucleoside analog 2¢F-ara-Aza-C
exhibited pronounced cytotoxicity against acute myeloge-
nous leukemia cells at a concentration of 0.1 m M whereas the
threshold concentration for Aza-C was 1.0 m M. A saturation
effect was attained if the concentration of the test nucleoside
was increased to 1.0 m M. The saturating concentration of
Aza-C was an order of magnitude greater (10.0 m M) com-
Thus, the novel 2¢-F-containing Aza-C analog VIIIa was
most promising for further studies of the acute toxicity and in
vivo antileukemic activity.
REFERENCES
pared with that of 2¢F-ara-Aza-C. The EC values, which
1. A. D. Shirin, Klin. Onkogematol., 1, No. 1, 21 – 33 (2008).
2. G. S. Mack, J. Natl. Cancer Inst., 98, No. 20, 1443 – 1444
(2006).
50
represent the nucleoside concentrations that inhibit by 50%
compared with a control the viability of leukemia cells, were
0.35 m M (KG-1), 0.51 m M (TPH-1), 0.58 m M (HL-60) for
2¢F-ara-Aza-C and 3.66 m M (KG-1), 4.25 m M (TPH-1), and
5.37 m M (HL-60) for Aza-C. This indicated that the specific
cytotoxicity of the new Aza-C analog was 9 – 10 times
greater than that of the control.
3. J. K. Christman, Oncogene, 21(35), 5483 – 5495 (2002).
4. E. A. Griffiths and S. D. Gore, Semin. Hematol., 45(1), 23 – 30
(2008).
5. J. Bauman, C. Verschraegen, S. Belinsky, et al., Cancer
Chemother. Pharmacol., 69(2), 547 – 554 (2012).
6. K. Raj and G. J. Mufti, Ther. Clin. Risk Manage., 2(4),
377 – 388 (2006).
Nucleoside analogs 3¢F-Aza-C and 5¢-deoxy-Aza-C were
inactive at concentrations of 0.1 m M although they did show
7. J. A. Beisler, J. Med. Chem., 21(2), 204 – 208 (1978).

Synthesis, Hydrolytic Stability, and Antileukemic Activity
809
8. A. Strunecka, J. Patocka, and P. Connett, J. Appl. Biomed., 2(3),
141 – 150 (2004).
13. E. N. Kalinichenko, T. S. Bozhok, T. N. Buravskaya, et al., Tr.
Beloruss. Gos. Univ., 8(1), 235 – 240 (2013).
9. O. N. Zefirova and N. S. Zefirov, Vestn. Mosk. Univ., Ser. 2:
Khim., 43(4), 251 – 256 (2002).
14. A. N. Mironov and N. D. Bunyatyan (eds.), Handbook for Pre-
clinical Drug Trials [in Russian], Part 1, Grif i K, Moscow
(2012), pp. 640 – 654.
10. T. Ishikawa, Y. Fukase, T. Yamamoto, et al., Biol. Pharm. Bull.,
21(7), 713 – 717 (1998).
15. T. Mosmann, J. Immunol. Methods, 65(1 – 2), 55 – 63 (1983).
16. J. E. van den Boogaart, E. N. Kalinichenko, T. L. Podkopaeva,
et al., Eur. J. Biochem., 221(3), 759 – 768 (1994).
11. T. S. Bozhok and E. N. Kalinichenko, Izv. Nats. Akad. Nauk.
Belarusi, Ser. Khim. Nauk, No. 3, 60 – 65 (2014).
12. A. Piskala and F. Sorm, Collect. Czech. Chem. Commun., 29(9),
2060 – 2076 (1964).
17. D. K. Rogstad, J. L. Herring, J. A. Theruvathu, et al., Chem.
Res. Toxicol., 22(6), 1194 – 1204 (2009).