Study of chemical stability of antivirally active 5-azacytosine acyclic nucleoside phosphonates using NMR spectroscopy

Article abstract of DOI:10.1016/j.bmc.2008.05.058

Hydrolytic decomposition of four 5-azacytosine acyclic nucleoside phosphonates was studied. Products of the decomposition are carbamoylguanidine derivatives. Stability and decomposition products of HPMP-5-azaC (a 5-azacytosine derivative with strong antiviral activity) differ from the other derivatives. The reaction pathway of HPMP-5-azaC involves a formyl derivative formed by intramolecular transformylation reaction.

Full text of DOI:10.1016/j.bmc.2008.05.058

Bioorganic & Medicinal Chemistry 16 (2008) 6778–6782
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Study of chemical stability of antivirally active 5-azacytosine acyclic nucleoside
phosphonates using NMR spectroscopy
*
ˇ
´
ˇ
´
Martin Dracínsky , Marcela Krecmerová, Antonín Holy
Gilead Sciences & IOCB Research Centre, Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic, v.v.i., Flemingovo nám. 2,
CZ-166 10 Prague 6, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
Hydrolytic decomposition of four 5-azacytosine acyclic nucleoside phosphonates was studied. Products
of the decomposition are carbamoylguanidine derivatives. Stability and decomposition products of
HPMP-5-azaC (a 5-azacytosine derivative with strong antiviral activity) differ from the other derivatives.
The reaction pathway of HPMP-5-azaC involves a formyl derivative formed by intramolecular trans-
formylation reaction.
Received 17 March 2008
Revised 19 May 2008
Accepted 28 May 2008
Available online 12 June 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
5-Azacytosine
Triazine ring opening
NMR spectroscopy
Reaction kinetics
1. Introduction
teristic property of tumor cells that causes silencing of tumor sup-
pressor genes, and hence tumor progression.8–11
Acyclic nucleoside phosphonates (ANPs), a special class of
nucleotide analogues, are compounds of great interest due to their
broad-spectrum of biological activities (e.g., antiviral, cytostatic,
antiparasitic, immunomodulatory) resulting in their utilization in
clinical practice (antivirals: cidofovir, adefovir and tenofovir).1 Pre-
clinical and/or clinical investigation of several other ones is cur-
rently under way. Recently, we have synthesized a series of
acyclic nucleoside phosphonates with 5-azacytosine base moiety.2
This work follows up on a long-term investigation of 5-azacytosine
compounds in our Institute, especially nucleosides with antileuke-
mic activity. At present time, two of them are used as therapeutics:
5-azacytidine3 (azacytidine, VidazaTM) and 20-deoxy-5-azacytidine4
(decitabine, DacogenTM). Both compounds were approved for the
therapy of all subtypes of myelodysplastic syndrome (MDS), disor-
ders belonging to a group of bone marrow stem cell hyperplasias
and dysplasias that result in ineffective hematopoiesis.5 Myelodys-
plastic disorders and transformed leukemia have poor prognosis
and minimal response to chemotherapy. The great attention is also
paid to clinical investigation of azacytidine and decitabine for the
treatment of solid tumors, especially metastatic lung cancer6 and
hormone-independent prostate cancer.7 The efficacy of 5-azacyto-
sine nucleosides is bound to their inhibitory activity toward DNA
methylations: hypermethylation of DNA was found to be a charac-
Regarding these facts, introduction of a 5-azacytosine compo-
nent to the chemistry of ANPs resulted as consequent and very
promising step in our search for new biologically active com-
pounds. Three types of phosphonomethoxyalkyl side chains were
selected for attachment to N-1 position of 5-azacytosine: 2-(phos-
phonomethoxy)ethyl (PME-5-azaC, 1), 3-hydroxy-2-(phosphono-
methoxy)propyl (HPMP-5-azaC, 2), and 1-[(R)-2-(phosphono-
methoxy)propyl]-5-azacytosine (PMP-5-azaC). Contrary to expec-
tation, none of the compounds exhibited cytostatic efficacy
in vitro; therapeutic potential of 5-azacytosine acyclic nucleoside
phosphonates was shown to be bound to their antiviral effects.
While activities of PME derivative 1, (R) enantiomer of HPMP-5-
azaC and 5,6-dihydro form of (S)-HPMP-5-azaC were marginal only
and (R)-PMP derivative completely inactive, compound 2, 1-(S)-[3-
hydroxy-2-(phosphonomethoxy)propyl]-5-azacytosine
exerted
exceptionally strong activity against practically all types of DNA
viruses: adenoviruses, poxviruses (vaccinia virus, cowpox virus,
orf virus), herpes simplex (type 1 and 2) virus, varicella-zoster
virus (VZV), and human cytomegalovirus (HCMV). From the struc-
tural point of view, compound
2 represents a sym-triazine
analogue of cidofovir, that is, 1-(S)-[3-hydroxy-2-(phospho-
nomethoxy)propyl]cytosine, a broad-spectrum antiviral agent ap-
proved for intravenous treatment of cytomegalovirus retinitis in
AIDS patients (and topically used also for various other indications
caused by DNA viruses). The drawback of cidofovir (and all other
ANPs) is its nephrotoxicity and low bioavailability disallowing its
oral application. Comparison of cidofovir and its 5-aza analogue
* Corresponding author. Tel.: +420 220183139; fax: +420 220183123.
ˇ
´
E-mail addresses: dracinsky@uochb.cas.cz (M. Dracínsky), marcela@uochb.cas.cz
ˇ
´
(M. Krecmerová), holy@uochb.cas.cz (A. Holy).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.05.058

ˇ
´
6779
M. Dracínsky et al. / Bioorg. Med. Chem. 16 (2008) 6778–6782
2 showed that antiviral activity data of 2 [based on 50% effective
concentration (EC50 values)] are similar or in some cases higher,
and the antiviral selectivity index [ratio of 50% cytotoxic concen-
tration (CC50) to EC50] was 2- to 16-fold higher than that of cidofo-
vir.2 To improve bioavailability of 2, several structurally diverse
ester prodrugs were synthesised.12 Esterification was carried out
on the level of its cyclic form 3 using reaction of its tetrabutylam-
monium salt with corresponding alkyl bromides or acyloxymethyl
chloride. All these compounds exerted an extraordinary activity
against all types of DNA viruses; the most active compound was
found to be hexadecyloxyethyl ester with. EC50 values in the range
a new liquid chromatography/tandem mass spectrometry quantifi-
cation method. This study revealed that these previously unknown
compounds are partially an isomer of decitabine and partially two
various hydrated open-ring form isomers, that is, 20-deoxy-N-
(formylamidino)-N0-b- -ribofuranosylureas.19
D
The aim of this work was investigation of hydrolytic destruction
of both new 5-azacytosine antivirals (HPMP-5-azaC and its cyclic
form) targeted to identification of decomposition products for
the purpose of their antiviral and toxicity studies and investigation
of kinetic parameters and mechanism of these reactions.
of 0.003–0.008
60.00014 to 60.00038
HHV-6, and 0.037 g/ml for vaccinia virus. The activity of ester
l
g/ml for HSV, 60.0008 to 60.0014
l
g/ml for VZV,
2. Results and discussion
l
g/ml (HCMV), 0.008 to 0.037
lg/ml for
l
To approximate conditions in plasma or amino acids and/or
phosphate-containing buffers generally used as media for bio-
chemical assays we monitored the course of decomposition of
our compounds in 1 M glycine or PBS buffer (both as D2O solu-
tions) by NMR methodology. This method was selected as the most
suitable for identification and characterization of all reaction prod-
ucts in any moment of the decomposition process. Thus, quantifi-
cation of products is not bound to their UV absorption or other
physiochemical properties which can differ for single products or
intermediates. Glycine was selected due to its single structure
not disturbing NMR spectrum of reaction mixture. Its 1 M solution
(pH 5.75) enables to study decomposition process in acidic med-
ium and eventually possible interactions with amino acids from
plasma; PBS buffer (pH 7.4, a phosphate buffer with NaCl) was
used to study this process under slightly basic (physiological) con-
ditions. The following compounds were submitted to the stability
study: 1-[2-(phosphonomethoxy)ethyl]-5-azacytosine (1), 1-(S)-
[3-hydroxy-2-(phosphonomethoxy)propyl]-5-azacytosine (HPMP-
5-azaC, 2), its cyclic form, that is, 1-{[(5S)-2-hydroxy-2-oxido-
1,4,2-dioxaphosphinan-5-yl]methyl}-5-azacytosine (3), and (2S)-
[(2,3-dihydroxy)propyl]-5-azacytosine (5); the latter one to com-
pare decomposition process of 2 with an analogous compound
lacking phosphonomethyl group.
prodrugs was decreasing in the order: 2-(hexadecyl-
oxy)ethyl > pivaloyloxymethyl ꢀ octadecyl > erucyl (compounds
4a–4d, Fig. 1).12 Regarding exceptional activities of 2 and its ester
prodrugs, detailed studies directed to efficacy against particular
viruses (camelpox virus,13 polyomavirus14) in different cell culture
models as well as some in vivo studies15 were carried out. Progres-
sion in investigation of HPMP-5-azaC and its derivatives is still un-
der way.
In contrast to cidofovir, HPMP-5-azaC has more complicated
metabolic profile due to its chemical (and also enzymatic) instabil-
ity. Concerning chemical instability, in aqueous solutions ring
opening between C-6 and N-1 of the triazine moiety occurs and
HPMP-5-azaC is successively degraded to 2-{[(2S)-3-hydroxy-2-
(phosphonomethoxy)propyl]carbamoylguanidine (9). This ring-
opening reaction is a general property of all 5-azacytosine nucleo-
sides (e.g., riboside, 20-deoxyriboside, arabinoside) as well as other
derivatives with hydroxyl containing side-chain. The chemical sta-
bility of 5-azacytidine was first studied by Pithova et al. who
proved that triazine ring in 5-azacytosine opens between C-6 and
N-1 to form N-(formylamidino)-N0-b-
reversible reaction followed by irreversible loss of a formyl group
to form 1-b-
-ribofuranosyl-3-guanylurea.16,17 The first stage of
D-ribofuranosylurea in a
D
this hydrolysis consists in nucleophilic attack of hydroxyl ion in
position 6 of 5-azacytosine which electron density is much lower
compared to that of cytosine which results from quantum chemical
calculations.16 Similar course of alkaline hydrolysis was described
for 5-aza-20-deoxycytidine (decitabine) while in neutral conditions
this compound was decomposed not only to expected 20-deoxy-N-
Decomposition products and reaction courses were similar for
compounds 1, 3 and 5. In all these three cases a carbamoylguani-
dine derivative (6a–6c, Fig. 2) and formic acid were the only ob-
served products of the decomposition in both 1 M glycine
solution and PBS buffer. The decomposition of all three compounds
(1, 3, 5) is a first order reaction. This can be clearly seen from the
dependence of logarithm of reactant concentration on time
(Fig. 3). The rate constants and half-life times are given in Table
1. The reaction is in all cases slower in PBS buffer than in 1 M gly-
cine solution. The structure of all products was determined by 1H,
(formylamidino)-N0-b-
D-ribofuranosylurea but also to small
amount of several other products as resulted from HPLC analysis.18
The complete characterization of decomposition products of deci-
tabine in water and human plasma was carried out recently using
NH2
NH2
NH2
NH2
N
NH2
N
N
N
N
N
N
N
N
N
O
N
O
N
O
N
O
N
O
N
OH
O
P
O
P
OH
P(O)(OH)2
O
OH
O
OR
OH
O
O
O
O
P(O)(OH)2
4a, R = C18H37
5
2
3
1
(CH2)12-CH=CH -(CH2)7CH3
4b, R = cis
O
O
CH2
4c, R =
4d, R =
CH2CH2 -O -(CH2)15CH3
Figure 1. Structures of 5-azacytosine acyclic nucleoside phosphonates and other analogues: PME-5-azaC (1), HPMP-5-azaC (2), cyclic HPMP-5-azaC (3), ester prodrugs of
cyclic HPMP-5-azaC (4a–4d) and (2S)-(2,3-dihydroxy)propyl]-5-azacytosine (5).

ˇ
´
6780
M. Dracínsky et al. / Bioorg. Med. Chem. 16 (2008) 6778–6782
NH2
N
NH2
N
N
NH2
O
N
R
O
NH
R
R =
1
6a
O
P(O)(OH)2
O
O
OH
R =
R =
3
5
6b
O
P
OH
6c
Figure 4. Plot of concentration of HPMP-5-azaC (2) and its decomposition products
versus time.
OH
Figure 2. Decomposition products of compounds 1, 3 and 5.
Table 2
Rate constants and half-life times of decomposition reaction of compound
different buffers
5 in
Buffer
pH
4.0
5.8
7.4
k [dꢁ1
]
k [sꢁ1 ꢂ 10ꢁ7
4.51
5.09
4.17
7.75
]
t1/2 [d]
Phosphate
1 M glycine
PBS
0.039
0.044
0.036
0.067
17.8
15.8
19.6
10.3
Phosphate
10.0
tion path could be caused by protonation of triazine ring of com-
pound 2. Triazine derivative can form zwitterion with
2
a
negative charge on the phosphonate oxygens and protonated tri-
azine ring. To confirm or disconfirm this possibility we observed
the decomposition of compound 5 also in two 50 mM phosphate
buffers with pH 4 and 10. In the acidic buffer we expected proton-
ation of the triazine moiety but the reaction rate (see Table 2) dif-
fered only slightly from the PBS buffer, and no formyl derivative
was observed in the reaction mixture. In the basic buffer the reac-
tion was faster by a factor of two when compared with the PBS buf-
fer. The hydrolytic decomposition is probably catalyzed by both
acids and bases, but no qualitative change in the mechanism was
observed for the protonated triazine derivative.
Figure 3. Plot of logarithm of concentration of compounds 1, 3 and 5 versus time.
13C and 2D-H,C-HSQC, 2D-H,C-HMBC and 2D-H,H-COSY NMR spec-
tra. The biggest change in NMR shifts are experienced by the sig-
nals of hydrogens in position 10. In the triazine derivatives their
resonance is near 4 ppm while in the carbamoylguanidine products
their resonance is shifted upfield to 3.4 ppm.
The decomposition of compound 2 differs from the other reac-
tions. The reaction rate is higher (approximately four times) than
the decomposition rates of the other compounds, and intermediate
formyl derivative 8 is formed. The formyl derivative 8 is then
hydrolyzed to form carbamoylguanidine derivative 9 and formic
acid (Fig. 4). Compound 5 differs only slightly from HPMP deriva-
tive 2; it does not have the phosphonometylene group, which is
quite far from the triazine ring of compound 2. However, quite sur-
prisingly, no intermediate formyl derivative was observed during
its decomposition.
We considered two possible reasons for different behavior of
compounds 2 and 5 in aqueous solutions. First, the different reac-
Table 1
Rate constants and half-life times of decomposition reaction of compounds 1–3 and 5
1 M glycine
PBS
k [dꢁ1
]
k [sꢁ1 ꢂ 10ꢁ7
]
t1/2 [d]
k [dꢁ1
]
k [sꢁ1 ꢂ 10ꢁ7
]
t1/2 [d]
1 ? 6a
2 ? 8
3 ? 6b
5 ? 6c
0.027
0.149
0.048
0.044
3.13
17.2
5.56
5.09
26.2
4.75
14.7
16.1
0.018
0.142
0.023
0.036
2.08
16.4
2.70
4.17
39.3
4.98
29.7
19.6
Figure 5. Global minimum energy conformation of compound 2 (hydrogens are
omitted for clarity).

ˇ
´
6781
M. Dracínsky et al. / Bioorg. Med. Chem. 16 (2008) 6778–6782
based on the inhibition of cell growth (CC50 values, that is, the
compound concentration required to reduce cell growth by 50%)
or determination of MCC values, that is, the minimum cytotoxic
concentration of the compound that causes a microscopically
detectable alteration of cell morphology. Syntheses of pure formy-
lamido and carbamoylguanidine derivatives (7 and 9) were de-
scribed previously.2 Compound 6b was obtained from the cyclic
HPMP-5azaC after its 60 days lasting treatment with 1 M glycine.
None of these decomposition products exhibited cytotoxicity
in vitro: in all cases CC50 and MCC values were >100 lg/ml.
4. Experimental
Figure 6. Global minimum energy conformation of compound 5 (hydrogens are
NMR spectra were measured on a Bruker Avance-500 and/or
Bruker Avance-600 instruments and referenced to internal stan-
dard dioxane (d 3.75 and 67.19, respectively). The kinetics of
decomposition was studied at 22 °C. The samples (30 mg) were
dissolved in 1 mM solution of glycine in D2O or in a buffer. The
PBS buffer and the 50 mM phosphate buffers (pH 4.0 and 10.0)
were prepared in H2O, then lyophilized and dissolved in the same
volume of D2O. 1H NMR spectra were measured periodically and
signals from the low-field region were chosen for integration and
kinetic measurements. A singlet of hydrogen H-6 of the starting
azacytosine derivative, singlet of CHO of intermediate formiate 8,
and singlet of formic acid (which is the final product of the reac-
tions) appear in this region of 1H NMR spectra. NMR data in the fol-
lowing list are taken from the spectra in 1 M glycine.
All calculations were carried out using the Gaussian 03 software
package.20 DFT calculations were performed using the Becke3-
LYP21,22 functional with 6-31G(d) basis set. Full geometry optimi-
zations were carried out using polarizable continuum solvent
model23 (PCM).
Description of synthesis and full characterization of studied
compounds (elemental analyses, spectral data, physical constants)
are available in our previous papers: see Ref. 2 for compounds 1, 2,
5, 7, and 9, or Ref. 12 for compound 3. Decomposition products of
biologically inactive compounds 6a, 6b, and 6c were characterized
by NMR spectra only, their isolation from buffer solutions was not
performed.
omitted for clarity).
Second possible explanation of the different decomposition of
compound 2 was that its steric arrangement differs significantly
from that of compound 5. To answer the question whether the con-
formation of the studied triazine derivatives could play a role in
different reaction paths of 2 and 5 we performed a global energy
minimum conformational search using a DFT method with polariz-
able continuum model of water solvation. We systematically chan-
ged torsion angles C(2)–N(1)–C(10)–C(20), N(1)–C(10)–C(20)–C(30),
and C(10)–C(20)–O–C(40). The optimization started from 18 differ-
ent conformations of compound 2 and 6 different conformations
of compound 5. We found out that in the global energy minimum
conformation of compound 2 (see Fig. 5), the free hydroxy group in
position 30 is very close (3.3 Å) to carbon 6 of the triazine ring. The
conformation is stabilized by formation of intramolecular hydro-
gen bond between the phosphonate hydroxy group and carbonyl
oxygen of the azacytosine part of the molecule. The minimum en-
ergy conformation of compound 5 (see Fig. 6) differs significantly.
Here, the distance between triazine C-6 and the hydroxyl group in
position 30 is much bigger (5.6 Å).
3. Conclusions
We have shown that 5-azacytosine acyclic nucleoside phospho-
nates are instable and they are decomposed to carbamoylguani-
dine derivatives. Decomposition of HPMP-5-azaC (2) involves
formation of formiate 8 which was not observed in the decompo-
sition reaction of the other derivatives. The explanation for this dif-
ferent reaction pathway (supported by ab initio calculations) is
that intermediate formylamido derivative (7), which was previ-
ously shown to be intermediate in the 5-azacytidine decomposi-
tion,16 can react in a transformylation reaction with a spatially
close hydroxy group to form the formiate 8 (see Scheme 1). As a re-
sult, we can say that triazine can serve as a formylating agent when
a free hydroxyl group is in a suitable spatial arrangement.
4.1. 1-[2-(Phosphonomethoxy)ethyl]-5-azacytosine (1)
13C NMR: 165.91 (C-4); 161.41 (C-6); 156.02 (C-2); 70.04 d,
J(20,P) = 12.0 (C-20); 67.59 d, J(CH2P,P) = 156.8 (CH2P); 47.78 (C-
1
10). H NMR: 8.33 s, 1H (H-6); 4.04 t, 2H, J(10,20) = 4.9 (H-10); 3.81
t, 2H, J(20,10) = 4.9 (H-20); 3.65 d, 2H, J(CH2P,P) = 8.9 (CH2P).
4.2. 1-(S)-[3-Hydroxy-2-(phosphonomethoxy)propyl]-5-
azacytosine (2)
Decomposition products of both antivirals HPMP-5azaC and its
cyclic form were subjected to cytotoxicity assays performed in hu-
man embryonic lung (HEL) cells. Cytotoxicity measurements were
13C NMR: 166.19 (C-4); 161.48 (C-6); 156.50 (C-2); 79.71
d, J(20,P) = 12.1 (C-20); 66.67 d, J(CH2P,P) = 157.3 (CH2P); 60.57
NH2
NH2
NH2
NH2
OH
NH2
NH2
N
N
N
NH
CHO
N
N
CHO
O
NH
O
O
NH
O
NH
OH
O
N
OH
O
P(O)(OH)2
O
P(O)(OH)2
O
P(O)(OH)2
O
P(O)(OH)2
2
7
8
9
Scheme 1. Hydrolytic decomposition of HPMP-5-azaC (2).

ˇ
´
M. Dracínsky et al. / Bioorg. Med. Chem. 16 (2008) 6778–6782
6782
(C-30); 48.79 (C-10). 1H NMR: 8.30 s, 1H (H-6); 4.14 dd, 1H,
Jgem = 14.4, J(10a,20) = 3.2 (H-10a); 3.84 dd, 1H, Jgem = 12.6,
J(30a,20) = 3.8 (H-30a); 3.83 dd, 1H, Jgem = 14.3, J(10b,20) = 8.4 (H-
10b); 3.77 dd, 1H, Jgem = 13.0, J(CH2Pa,P) = 9.2 (CH2Pa); 3.73 m, 1H
(H-20); 3.62 dd, 1H, Jgem = 12.6, J(30b,20) = 4.0 (H-30b); 3.55 dd, 1H,
Jgem = 13.0, J(CH2Pb,P) = 9.8 (CH2Pb).
J(30b,20) = 4.9 (H-30b); 3.87 m, 1H (H-20); 3.80 m, 1H (CH2Pa);
3.71 m, 1H (CH2Pb); 3.51 dd, 1H, Jgem = 14.4, J(10a,20) = 4.3 (H-
10a); 3.37 dd, 1H, Jgem = 14.4, J(10b,20) = 6.9 (H-10b).
4.9. 2-[(2S)-3-Hydroxy-2-
(phosphonomethoxy)propyl]carbamoylguanidine (9)
4.3. 1-{[(5S)-2-Hydroxy-2-oxido-1,4,2-dioxaphosphinan-5-
yl]methyl}-5-azacytosine (3)
13C NMR: 155.81 (C-4); 155.07 (C-2); 81.14 d, J(20,P) = 11.3 (C-
20); 66.66 d, J(CH2P,P) = 156.9 (CH2P); 61.31 (C-30); 40.37 (C-10).
1H NMR: 3.78 dd, 1H, Jgem = 13.1, J(CH2Pa,P) = 9.2 (CH2Pa); 3.74
dd, 1H, Jgem = 12.0, J(30a,20) = 3.7 (H-30a); 3.72 dd, 1H, Jgem = 13.1,
J(CH2Pb,P) = 9.5 (CH2Pb); 3.63 m, 1H (H-20); 3.61 dd, 1H, Jgem = 11.8,
J(30b,20) = 5.3 (H-30b); 3.47 dd, 1H, Jgem = 14.3, J(10a,20) = 4.0 (H-10a);
3.31 dd, 1H, Jgem = 14.3, J(10b,20) = 6.6 (H-10b).
13C NMR: 166.04 (C-4); 161.35 (C-6); 156.11 (C-2); 74.01 d,
J(20,P) = 3.5 (C-20); 70.29 d, J(30,P) = 6.4 (C-30); 65.82 d,
J(CH2P,P) = 143.3 (CH2P); 47.51 (C-10). 1H NMR: 8.27 s, 1H (H-6);
4.22 m, 2H (H-30); 4.07 dd, 1 H, Jgem = 14.5, J(10a,20) = 3.0 (H-10a);
3.99 m, 1H (H-20); 3.95 dd, 1H, Jgem = 14.0, J(CH2Pa,P) = 8.9 (CH2Pa);
3.76 dd, 1H, Jgem = 14.6, J(10b,20) = 8.3 (H-10b); 3.72 dd, 1H,
Jgem = 14.1, J(CH2Pb,P) = 1.9 (CH2Pb).
Acknowledgments
This work is a part of the research project of the Institute
Z40550506. It was supported by the Centre of New Antivirals
and Antineoplastics 1M0508 by the Ministry of Education, Youth
and Sports of the Czech Republic, the Program of targeted projects
of Academy of Sciences of the Czech Republic 1QS40055050, by the
NIH Grant 1UC1AI062540-01 and by Gilead Sciences and IOCB Re-
search Centre. The authors’ thanks are due to Professor Jan Balza-
rini and his group in the Rega Institute for Medical Research,
Catholic University Leuven (Belgium) for cytotoxicity assays of de-
scribed compounds.
4.4. (2S)-[(2,3-Dihydroxy)propyl]-5-azacytosine (5)
13C NMR: 166.79 (C-4); 161.16 (C-6); 157.25 (C-2); 69.38 (C-20);
63.56 (C-30); 50.83 (C-10). 1H NMR: 8.23 s, 1H (H-6); 4.09 dd, 1H,
Jgem = 14.1, J(10a,20) = 3.3 (H-10a); 3.98 m, 1H (H-20); 3.68 dd, 1H,
Jgem = 11.9, J(30a,20) = 4.4 (H-30a); 3.66 dd, 1H, Jgem = 14.1,
J(10b,20) = 8.9 (H-10b); 3.59 dd, 1H, Jgem = 11.9, J(30b,20) = 5.9 (H-30b).
4.5. 2-[2-(Phosphonomethoxy)ethyl]carbamoylguanidine (6a)
13C NMR: 155.81 (C-4); 155.04 (C-2); 71.57 d, J(20,P) = 11.4 (C-
20); 67.42 d, J(CH2P,P) = 156.5 (CH2P); 39.77 (C-10). 1H NMR: 3.69
t, 2H, J(20,10) = 5.2 (H-20); 3.67 d, 2H, J(CH2P,P) = 9.7 (CH2P); 3.41
t, 2H, J(10,20) = 5.2 (H-10).
References and notes
´
1. Holy, A. Curr. Pharm. Des. 2003, 9, 2567–2592.
ˇ
´
2. Krecmerová, M.; Holy, A.; Pískala, A.; Masojídková, M.; Balzarini, J.; Andrei, G.;
Snoeck, R.; Naesens, L.; Neyts, J.; De Clercq, E. J. Med. Chem. 2007, 50, 1069–
1077.
4.6. 2-{[(5S)-2-Hydroxy-2-oxido-1,4,2-dioxaphosphinan-5-
yl]methyl}carbamoylguanidine (6b)
3. Pískala, A.; Šorm, F. Collect. Czech. Chem. Commun. 1964, 29, 2060–2076.
4. Pliml, J.; Šorm, F. Collect. Czech. Chem. Commun. 1964, 29, 2576–2578.
5. Kuykendall, J. R. Ann. Pharmacother. 2005, 39, 1700–1708.
6. Momparler, R. L.; Bouffard, D. Y.; Momparler, L. F.; Dionne, J.; Belanger, K.;
Ayoub, J. Anticancer Drugs 1997, 8, 358–368.
7. Thibault, A.; Figg, W. D.; Bergan, R. C.; Lush, R. M.; Myers, C. E.; Tompkins, A.;
Reed, E.; Samid, D. Tumori 1998, 84, 87–89.
8. Jones, P. A. Cell 1985, 40, 485–486.
9. (a) Momparler, R. L.; Cote, S.; Eliopoulos, N. Leukemia 1997, 11, S1–S6; (b)
Momparler, R. L.; Cote, S.; Eliopoulos, N. Leukemia 1997, 11, 175–180.
10. Bender, C. M.; Pao, M. M.; Jones, P. A. Cancer Res. 1998, 58, 95–101.
11. Jones, P. A.; Baylin, S. B. Nat. Rev. Genet. 2002, 3, 415–428.
13C NMR: 155.79 (C-4); 155.05 (C-2); 75.33 d, J(20,P) = 3.8 (C-20);
70.59 d, J(30,P) = 6.5 (C-30); 65.65 d, J(CH2P,P) = 143.2 (CH2P); 39.84
1
(C-10). H NMR: 4.18 m, 2H (H-30); 3.96 dd, 1H, Jgem = 14.1,
J(CH2Pa,P) = 8.9 (CH2Pa); 3.86 m, 1H (H-20); 3.77 dd, 1H, Jgem = 14.0,
J(CH2Pb,P) = 2.3 (CH2Pb); 3.36 dd, 1H, Jgem = 14.6, J(10a,20) = 4.5 (H-
10a); 3.31 dd, 1H, Jgem = 14.6, J(10b,20) = 7.9 (H-10a).
ˇ
´
12. Krecmerová, M.; Holy, A.; Pohl, R.; Masojídková, M.; Andrei, G.; Naesens, L.;
Neyts, J.; Balzarini, J.; De Clercq, E.; Snoeck, R. J. Med. Chem. 2007, 50, 5765–
5772.
4.7. 2-[(2S)-2,3-dihydroxypropyl]carbamoylguanidine (6c)
13C NMR: 156.18 (C-4); 155.84 (C-2); 70.84 (C-20); 63.72 (C-30);
42.53 (C-10). 1H NMR: 8.45 s, 1H (H-6); 3.82 m, 1H (H-20); 3.62 dd,
1H, Jgem = 11.8, J(30a,20) = 4.3 (H-30a); 3.54 dd, 1H, Jgem = 11.8,
J(30b,20) = 6.2 (H-30b); 3.37 dd, 1H, Jgem = 14.2, J(10a,20) = 4.5 (H-
10a); 3.24 dd, 1H, Jgem = 14.2, J(10b,20) = 7.4 (H-10b).
ˇ
´
13. Duraffour, S.; Snoeck, R.; Krecmerová, M.; van Den Oord, J.; De Vos, R.; Holy, A.;
Crance, J.-M.; Garin, D.; De Clercq, E.; Andrei, G. Antimicrob. Agents Chemother.
2007, 51, 4410–4419.
ˇ
´
14. Lebeau, I.; Andrei, G.; Krecmerová, M.; De Clercq, E.; Holy, A.; Snoeck, R.
Antimicrob. Agents Chemother. 2007, 51, 2268–2273.
15. Andrei, G.; Krecmerová, M.; Holy, A.; Naesens, L.; Neyts, J.; Balzarini, J.; De
Clercq, E.; Snoeck, R. Antiviral Res. 2007, 74, A33–A34.
ˇ
´
16. Pithová, P.; Pískala, A.; Pitha, J.; Šorm, F. Collect. Czech. Chem. Commun. 1965, 30,
2801–2811.
17. Beisler, J. A. J. Med. Chem. 1978, 21, 204–208.
18. Lin, T.-S.; Momparler, R. L.; Rivard, G. E. J. Pharm. Sci. 1981, 70, 1228–1232.
19. Liu, Z.; Marcucci, G.; Byrd, J. C.; Grever, M.; Xiao, J.; Chan, K. K. Rapid Commun.
Mass Spectrom. 2006, 20, 1117–1126.
20. Frisch, M. J. et al Gaussian 03; Gaussian, Inc.: Wallingford, CT, 2004.
21. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B Condens. Mater. Phys. 1988, 37, 785.
22. Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
4.8. 2-[(2S)-3-Formyloxy-2-
(phosphonomethoxy)propyl]carbamoylguanidine (8)
13C NMR: 164.28 (CO-formiate); 155.81 (C-4); 155.07 (C-2);
78.32 d, J(20,P) = 11.1 (C-20); 66.70 d, J(CH2P,P) = 157.1 (CH2P);
63.53 (C-30); 40.42 (C-10). 1H NMR: 8.19 s, 1H (formiate); 4.40 dd,
1H, Jgem = 12.1, J(30a,2) = 4.1 (H-30a); 4.26 dd, 1H, Jgem = 12.1,
23. Mennucci, B.; Cancès, E.; Tomasi, J. J. Phys. Chem. B 1997, 101, 10506.