Activation of p16 gene silenced by DNA methylation in cancer cells by phosphoramidate derivatives of 2′-deoxyzebularine

Article abstract of DOI:10.1021/jm8005965

We report herein the application of the phosphoramidate ProTide technology to improve the metabolism of the DNA methytransferase inhibitor, zebularine (Z). Zebularine is a riboside that must undergo a complex metabolic transformation before reaching the critical 2′-deoxyzebularine 5′-triphosphate (dZTP). Because 2′-deoxyzebularine (dZ) is not phosphorylated and therefore inactive, the ProTide strategy was employed to bypass the lack of phosphorylation of dZ and the inefficient reduction of zebularine 5′-diphosphate by ribonucleotide-diphosphate reductase required for zebularine. Several compounds were identified as more potent inhibitors of DNA methylation and stronger inducers of p16 tumor suppressor gene than zebularine. However, their activity was dependent on the administration of thymidine to overcome the potent inhibition of thymidylate synthase (TS) and deoxycytidine monophosphate (dCMP) deaminase by dZMP, which deprives cells of essential levels of thymidine. Intriguingly, the activity of the ProTides was cell line-dependent, and activation of p16 was manifest only in Cf-Pac-1 pancreatic ductal adenocarcinoma cells.

Full text of DOI:10.1021/jm8005965

J. Med. Chem. 2008, 51, 7593–7601
7593
Activation of p16 Gene Silenced by DNA Methylation in Cancer Cells by Phosphoramidate
Derivatives of 2′-Deoxyzebularine
Christine B. Yoo,† Rocco Valente,‡ Costantino Congiatu,‡ Federica Gavazza,‡ Annette Angel,‡ Maqbool A. Siddiqui,§
Peter A. Jones,*,† Christopher McGuigan,*,‡ and Victor E. Marquez*,§
Departments of Biochemistry and Molecular Biology, USC/Norris ComprehensiVe Cancer Center, Keck School of Medicine, UniVersity of
Southern California, Los Angeles, California 90033, Welsh School of Pharmacy, Cardiff UniVersity, King Edward VII AVenue, Cardiff CF10
3XF, U.K., and Laboratory of Medicinal Chemistry, Center for Cancer Research, National Cancer InstitutesFrederick, National Institutes of
Health, Frederick, Maryland 21702
ReceiVed May 20, 2008
We report herein the application of the phosphoramidate ProTide technology to improve the metabolism of
the DNA methytransferase inhibitor, zebularine (Z). Zebularine is a riboside that must undergo a complex
metabolic transformation before reaching the critical 2′-deoxyzebularine 5′-triphosphate (dZTP). Because
2′-deoxyzebularine (dZ) is not phosphorylated and therefore inactive, the ProTide strategy was employed to
bypass the lack of phosphorylation of dZ and the inefficient reduction of zebularine 5′-diphosphate by
ribonucleotide-diphosphate reductase required for zebularine. Several compounds were identified as more
potent inhibitors of DNA methylation and stronger inducers of p16 tumor suppressor gene than zebularine.
However, their activity was dependent on the administration of thymidine to overcome the potent inhibition
of thymidylate synthase (TS) and deoxycytidine monophosphate (dCMP) deaminase by dZMP, which deprives
cells of essential levels of thymidine. Intriguingly, the activity of the ProTides was cell line-dependent, and
activation of p16 was manifest only in Cf-Pac-1 pancreatic ductal adenocarcinoma cells.
Introduction
have been sought by the use of drugs that alter the transcriptional
status of the affected tumor suppressor genes by inhibiting DNA
methylation.3,4 Therefore, silenced tumor suppressor genes
present themselves as obvious targets for reactivation by DNA
methylation inhibitors such as 5-azacytosine nucleosides (2a,b)
and more recently zebularine (1a).5,6
The initiation and progression of cancer are driven by both
genetic and epigenetic changes. Epigenetic changes, which are
independent of the primary DNA sequence, have received a
great deal of attention for their key role in cancer initiation and
tumor progression.1,2 In the epigenomics landscape, one of the
most studied processes is DNA methylation, an event that
controls the transcriptional silencing of a number of tumor
suppressor genes. In cancer, such transcriptional repression is
associated with the abnormal methylation of cytosines in CpG
islands near the promoter regions of the genes and maintained
through cellular division. Because these methylated CpG islands
are incapable of initiating transcription unless the methylation
signal is removed, therapeutic strategies to revert this process
Zebularine [1-(ꢀ-D-ribofuranosyl)-1,2-dihydropyrimidin-2-
one] is a cytidine analogue that was initially developed as a
cytidine deaminase (CDA)a inhibitor7,8 and recently discovered
to also inhibit DNA methylation.5 The simple removal of the
4-amino group from cytidine increases the electrophilicity of
the resulting 2-(1H)-pyrimidinone aglycon, which explains the
ease of nucleophilic attacks at C4 or C6 of the 2-oxopyrimidine
ring that are responsible for zebularine’s activity as a potent
inhibitor of both CDA and DNA methyltransferases, respec-
tively.9 Once incorporated into DNA, the 2-(1H)-pyrimidinone
aglycon forms a covalent complex with DNA methyltransferases
via nucleophilic attack at C6 from a key conserved cysteine
residue (Cys81)10 that results in the depletion of DNA meth-
yltransferase 1 (DNMT1),11,12 the reactivation of hypermethy-
lated genes in yeast and solid tumor cells,5 and antitumor
activities in mouse xenografts5 and radiation-induced T-cell
lymphomas in mice.13
* To whom correspondence should be addressed. P.A.J.: tel, 323-865-
0816; fax, 323-865-0102; e-mail, jones_p@ccnt.hsc.usc.edu. C.M.: tel, +44
(0)29 20874537; fax, +44 (0)29 20874537; e-mail, mcguigan@cf.ac.uk.
V.E.M.: tel, 301-846-5954; fax, 301-846-6033; e-mail, marquezv@
mail.nih.gov.
†
University of Southern California.
Cardiff University.
National Cancer InstitutesFrederick.
‡
§
a Abbreviations: CDA, cytidine deaminase; dCK, deoxycytidine kinase;
dCMP, deoxycytidine monophosphate deaminase; DNMT1, DNA methyl-
transferase 1; dUMP, deoxyuridine monophosphate; dZMP, 2′-deoxyzebu-
larine 5′-monophosphate; dZTP, 2′-deoxyzebularine 5′-triphosphate; FCS,
fetal calf serum; GAPDH, glyceraldehyde 3-phosphate dehydrogenase;
HCV, hepatitis C virus; HIV, human immunodeficiency virus; IMDM,
Iscove’s modified Dulbecco’s medium; MAGEA1, melanoma antigen family
A1; MOE SLog P, molecular operating environment SMART log P; Ms-
SNuPE, methylation-sensitive single-nucleotide primer extension; NMI,
N-methylimidazole; NMR, nuclear magnetic resonance; P, partition coef-
ficient; PBS, phosphate-buffered saline; PCR, polymerase chain reaction;
PD, population doubling; RNR, ribonucleotide-diphosphate reductase; RT-
PCR, reverse transcriptase PCR, SAR, structure-activity relationship;
SMART, Smiles arbitrary target specification; SMILES, simplified molecular
input line entry specification; THF, tetrahydrofuran; TMP, thymidine
monophosphate; TS, thymidylate synthase; Z, zebularine; ZMP, zebularine
5′-monophosphate; ZDP, zebularine 5′-diphosphate; ZTP, zebularine 5′-
triphosphate; ZDP-Chol, zebularine-5′-diphosphocholine.
In Vitro inhibition of bacterial M.HhaI DNA methyltransferase
activity by double-stranded oligodeoxynucleotides containing
either 2′-deoxy-5-azacytidine (2b) or 2′-deoxyzebularine (1b)
10.1021/jm8005965 CCC: $40.75
2008 American Chemical Society
Published on Web 11/12/2008

7594 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Yoo et al.
in the hemimethylated recognition sequence [5′-GXGC-3′/3′-
CGMG-5′ (M ) 5-methylcytosine, X ) 5-azacytosine or
2-(1H)-pyrimidinone)] confirmed that the two heterocyclic
moieties have comparable reactivity.14 In contrast, much higher
doses of zebularine were required to achieve comparable levels
of p16 reactivation in T24 cells relative to the 5-azacytidine
nucleosides, which suggested that the differences in potency
were due to other factors.5 Indeed, zebularine (1a) required doses
10-100-fold higher than 5-azacytidine (2a) and 2′-deoxy-5-
azacytidine (2b), respectively, to induce comparable levels of
p16 expression.5 On the other hand, the stability and reduced
toxicity of zebularine allowed it to be administered continuously
to cells resulting in marked p16 expression.5
The incorporation of zebularine into DNA necessitates critical
levels of 2′-deoxyzebularine 5′-triphosphate (dZTP) which is
formed by a complex metabolic route that may explain its
weaker potency.15 A quantitative assessment of the phospho-
rylation and DNA incorporation of zebularine in T24 cells using
2-[14C]-zebularine revealed that the drug is readily phosphory-
lated to the corresponding 5′-mono- (ZMP), 5′-di- (ZDP), and
5′-triphosphate (ZTP) in a dose- and time-dependent manner.15
Two additional zebularine-containing metabolites were also
observed and identified as diphosphocholine (ZDP-Chol) and
diphosphoethanolamine adducts. Intracellular concentrations of
ZTP and ZDP-Chol were comparable and greatly exceeded those
of the other metabolites.15
When DNA and RNA levels of incorporation were compared,
RNA incorporation surpassed DNA incorporation by at least
7-fold.15 Thus, formation of zebularine riboside metabolites
appears to be quite robust, but conversion of zebularine 5′-
diphosphate (ZDP) to 2′-deoxyzebularine 5′-diphosphate (dZDP),
which is catalyzed by ribonucleotide-diphosphate reductase
(RNR), seems to be a rate-limiting step that could explain
zebularine’s weaker potency.15 Unfortunately, the use of 2′-
deoxyzebularine (1b) was unable to overcome this deficiency
because it was completely ineffective, perhaps owing to lack
of recognition by the activating enzyme, deoxycytidine kinase
(dCK).5,15 In a biochemical sense zebularine behaves like a
cytidine analogue,16 but its metabolic activation to form the
dZTP required for DNA incorporation seems to be complex
and inefficient.
In considering the limitations of zebularine described above
and the inactivity of 2′-deoxyzebularine, we decided to inves-
tigate a prodrug strategy for dZMP that could simultaneously
bypass both the dCK step (needed for 2′-deoxyzebularine) and
the RNR step (involved in the case of zebularine at its
5′-diphosphate anabolite stage) as a means of increasing DNA
incorporation and enhancing potency. Thus, the intracellular
delivery of dZMP in a membrane-permeable “ProTide” form
was explored by the well-known aryloxyphosphoramidate
approach of McGuigan et al.17,18 Through this technology the
lipophilicity of the nucleotide is augmented, making the ProTide
able to penetrate the interior of the cell by passive diffusion.
Once inside the cell, the nucleotide monophosphate (dZMP) is
expected to be released at effective intracellular levels. The
McGuigan ProTide approach already has produced significant
enhancements in the antiviral activity of various nucleosides
against HIV and HCV viruses compared to the parent nucleoside
analogues in Vitro.19,20 As far as we know, this work is the first
attempt to utilize the ProTide technology for an anticancer agent.
parameters with increases in lipophilicity calculated according
to the atom-based program MOE SLog P21 as shown in Table
1. The objective was to quickly perform a limited structure-
activity analysis to identify the most potent candidates for proof
of concept. The biological response used to gauge our struc-
ture-activity analysis was the reexpression of the tumor
suppressor gene p16 in Cf-Pac-1 pancreatic tumor cells using
real time PCR analysis.11,22,23 For the purposes of our initial
screen, the compounds in Table 1 were given a qualitative “yes”
or “no” response to p16 reactivation.
In the first and simplest ProTide (3) we combined a phenoxy
group with the L-alanine methyl ester. The same two groups
were also attached to a modified 2′-ꢀ-fluoro-2′-deoxyzebularine
analogue (4) which was expected to function as a 2′-deoxy
surrogate with increased lipophilicity. However, none of these
compounds showed activity. A further increase in log P achieved
with the p-chlorophenoxy group and the L-alanine ethyl ester
in the next compound (5a) seems to suggest that a log P
approaching 2 is a good threshold for activity as compound 5a
was indeed able to induce p16 reactivation. Interestingly, the
3′,5′-bis[4-chlorophenoxy(ethoxy-L-alanyl)] phosphoramidate
(5b, log P ) 4.15) obtained as a byproduct also showed activity.
This was unexpected on several grounds. First, the presence of
the bulky group at the 3′ position might be anticipated to
possibly impede enzyme processing of the ProTide. Second, if
only the 5′-ProTide moiety were processed, as intended, the
dZMP-3′-ProTide would not be expected to be active per se.
Third, if both ProTide motifs were processed, the resulting 3′,5′-
bisphosphate would not be expected to be the ideal pharma-
cophore. Thus, the data seem to imply that 3′,5′-bis-ProTides
may have a surprising efficacy by a mechanism to be deter-
mined. The next two entries in Table 1 (compounds 6 and 7)
showed that contributions from the aryloxy and L-alanine ester
sides are different and potency does not necessarily correlate
with log P. In compound 6, a rise in log P to 2.60, achieved by
increasing the lipophilicity of the L-alanine moiety from an ethyl
to a benzyl ester, while leaving the phenoxy group intact,
resulted in an inactive compound. On the other hand, augment-
ing the log P from the aryloxy side with the naphthyloxy group
and adding an extra methyl group on the ester side with the
dimethylglycine ethyl ester to reach a similar log P of 2.70
produced a very potent compound (7a) which together with the
two diastereoisomers of the 3′,5′-bis[naphthyloxy(ethyloxydim-
ethylglycinyl)] phosphoramidate (7b and 7b′, log P ) 5.93)
induced strong p16 activation (Vide infra). The next compound
(8), which differs from 6 by the simple addition of fluorine,
was also inactive despite having an adequate log P of 2.97.
Notwithstanding the fact that compound 6 failed to reactivate
p16, the simple change of the L-alanine benzyl ester to the
dimethylglycine benzyl ester was accompanied by a modest rise
in log P to produce 9a, which was active. Similarly to compound
5b, the 3′,5′-bis[phenoxy(benzyloxydimethylglycinyl)] phos-
phoramidate 9b (log P ) 6.51) was active. The final compound
of the initial series was compound 10, which evolved from 9a
by replacing the phenyloxy moiety with the naphthyloxy group.
This compound is the most lipophilic member of the 5′-ProTide
series (log P ) 4.15).
Chemistry
The synthesis of the phosphoramidates (3-10) was performed
using either the THF/N-methylimidazole (NMI) or the tert-
butylmagnesium chloride protocol (see Scheme 1 and Experi-
mental Section). Because of the stereochemistry of the phos-
phorus center, each of the 5′-monophosphoramidates 3, 4, 5a,
Selection of ProTide Targets and Preliminary Activity
A rapid survey with different aryloxyphosphoramidates was
explored with the idea of correlating some key structural

ActiVation of p16 Gene by DeriVatiVes of 2′-Deoxyzebularine
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7595
Table 1. Target Aryloxyphosphoramidates Synthesized for Initial SAR Analysis
a Compounds 5b, 7b, 7b′, and 9b are 3′,5′-bisphosphoramidates. b Calculated log P value. c p16 activation in the presence of 100 mM thymidine.
Scheme 1. Synthesis of 2′-Deoxyzebularine Phosphoramidates
chromatographic separation of some of the 3′,5′-bisphosphora-
midate isomers, and each fraction, 7b and 7b′, gave only two
31P NMR signals, this being consistent with each being a single
diastereoisomer. As discussed later these compounds possessed
different potencies in the activation of p16.
The four most potent compounds 7a, 7b, 7b′, and 10 from
Table 1 that were identified in the preliminary screen as strong
promoters of p16 reexpression were fully characterized by
chemical and spectroscopic means (see Experimental Section)
and used for additional biological studies.
Biological Studies
All ProTides were initially tested in T24 cells by treating
continuously with the compounds for 8 days. No induction of
the p16 gene was observed at 1, 10, and 100 µM concentrations,
indicating that the compounds were unable to inhibit DNA
methylation in T24 cells (data not shown). Similarly, all
compounds failed to induce the p16 gene in Cf-Pac-1 pancreatic
cancer cells (data not shown).
6, 7a, 8, 9a, and 10 was isolated as pairs of diastereoisomers
and tested as such. Such isomerism was most evident from the
presence of two closely spaced signals in the 31P NMR in a
ratio of roughly 1:1. Other signals were also duplicated in the
proton and 13C NMR spectra in each case. For the 3′,5′-
bisphosphoramidates, four stereoisomers (eight 31P NMR peaks)
are possible, and multiple peaks were observed, consistent with
mixtures being generated. In the case of 7 we achieved
When the compounds were tested with Cf-Pac-1 cells
supplemented with 100 µM thymidine, activities began to

7596 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Yoo et al.
Figure 1. Effect of 2′-deoxyzebularine ProTides 7a, 7b, 7b′, and 10 on p16 expression in Cf-Pac-1 pancreatic cancer cells. Cf-Pac-1 cells were
treated with ProTides continuously for 8 days. RNA was collected, and the levels of p16 expression were measured by real time RT-PCR. GADPH
was used as a reference gene. All compounds induce a strong p16 expression in the presence of 100 µM thymidine.
Table 2. Percent Increase in Doubling Time of Cf-Pac-1 Pancreatic
emerge (Table 1). The rationale for including thymidine in the
Carcinoma Cells Treated with dZMP Prodrugs Continuously for 8 Days
in the Presence of 100 µM Thymidine
treatment was to maintain a pool of thymine monophosphate
(TMP) to allow DNA synthesis to progress without cell cycle
arrest. 2′-Deoxyzebularine monophosphate (dZMP) is also
known to inhibit two critical enzymes: (1) thymidylate synthase
(TS),24 the enzyme responsible for converting dUMP into TMP
via the salvage pathway, and (2) dCMP deaminase,25 which
prevents formation of dUMP, the substrate for TS to make TMP.
By inhibiting TS and dCMP deaminase, dZMP could slow DNA
synthesis, which is essential for inhibition of DNA methylation.
We therefore supplemented the cells with 100 µM thymidine
to allow the cell cycle to progress without interference from
the inhibition of TS and dCMP deaminase. Strong induction of
p16 expression was observed in Cf-Pac-1 cells in the presence
of 100 µM thymidine for compounds 7a and 10 at 100 µM
(Figure 1). Importantly, these compounds induced measurable
p16 expression still at 10 µM and displayed a significantly
stronger response than zebularine at an equivalent 100 µM
concentration. However, every active compound in Cf-Pac cells
failed to induce p16 expression in T24 and HCT15 cancer cells
even in the presence of thymidine (data not shown). Zebularine
alone, in contrast, was effective in T24 cells and HCT15 cells.
Of the two 3′,5′-bis[naphthyloxy(ethyloxydimethylglycine)]
phosphoramidate diastereoisomers 7b was the most potent
compound of the entire series showing stronger induction of
p16 at 10 µM compared to zebularine at 100 µM (Figure 1).
Also, this diastereoisomer was ca. 3-fold more potent than 7b′,
indicating that the enzymatic release of dZMP was different
and perhaps dependent on the stereochemistry of the phosphorus
center.
Since thymidine is also known to block DNA synthesis at
millimolar concentrations (1-2 mM),26 we then tested to see
if lower concentrations of thymidine were effective in facilitating
the demethylating activity of the ProTides. However, the
induction of the p16 gene was observed only at 100 µM
thymidine and not at lower concentrations, indicating that these
cells must be supplemented with a 1:1 ratio of thymidine to
ProTide. Replacement of 100 µM thymidine with 100 µM
uridine did not aid the ProTides in promoting the induction of
p16.
time in cell culture (Table 2). However, 7b and 7b′ caused
severe toxicity at 100 µM while 7a and 10 were less toxic at
the same concentrations (data not shown).
To confirm that these phosphoramidates were indeed hypom-
ethylating the DNA in these cells, we analyzed the DNA
methylation status of Cf-Pac-1 cells after treatment with 7a,
10, 7b, and 7b′ at three different loci by Ms-SNuPE.27 The
regions analyzed were the CpG-rich promoter region of p16,
the CpG-poor promoter region of MAGEA1, and a repeat
sequence known as D4Z4. All three loci are hypermethylated
The activity of compounds 7a and 10, as well as diastereoi-
somers 7b and 7b′, correlated well with increases in doubling

ActiVation of p16 Gene by DeriVatiVes of 2′-Deoxyzebularine
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7597
Figure 2. DNA methylation analysis of D4DZ in Cf-Pac-1 pancreatic cancer cells treated with 2′-deoxyzebularine ProTides 7a, 7b, 7b′, and 10.
Cf-Pac-1 cells were treated with ProTides continuously for 8 days, and DNA was collected. DNA methylation status of the D4DZ repeats expressed
in percent methylation was analyzed by quantitative Ms-SNuPE. Treatment with these compounds caused demethylation at the D4DZ locus.
and have previously been shown to lose DNA methylation after
zebularine treatment.11 The results of DNA methylation analyses
at all three regions were similar: the dZMP ProTides caused a
decrease in the methylation level in the presence of thymidine.
The pattern is maintained whether the region is a CpG-rich,
CpG-poor, or repetitive element containing regions. Figure 2
summarizes the DNA methylation status at D4Z4; all four
compounds tested showed demethylation. As before, these drugs
worked best in Cf-Pac-1 pancreatic cancer cells and marginally
in HCT15 colon cancer cells (data not shown).
they are very cell line-specific and require the presence of
thymidine, these drugs provide proof of concept that two
enzymes, dCK and RNR, can be efficiently circumvented by
these ProTides, which represent a promising start for the
development of future prodrug strategies for 2′-deoxyzebularine.
Experimental Section
Materials and Methods. Cell Lines and Drug Treatment. T24
bladder carcinoma cells and HCT116 colon carcinoma cells were
obtained from the American Type Culture Collection (Rockville,
MD) and cultured in McCoy’s 5A medium supplemented with 10%
heat-inactivated fetal calf serum (FCS), 100 units/mL penicillin,
and 100 µg/mL streptomycin (Invitrogen, Carlsbad, CA). Cf-Pac-1
pancreatic carcinoma cells also from American Type Culture
Collection were cultured in IMDM medium (Gibco/Life Technolo-
gies Inc., Palo Alto, CA) supplemented with 10% heat-inactivated
FCS, 100 units/mL penicillin, and 100 µg/mL streptomycin
(Invitrogen, Carlsbad, CA). All cells were cultured in a humidified
incubator at 37 °C in 5% CO2. Zebularine and 5-aza-CdR (Sigma-
Aldrich, St. Louis, MO) were dissolved in PBS. Zebularine prodrugs
were dissolved in PBS or ethanol, depending on solubility, and
stored at -80 °C. Cells were seeded (3 × 105 cells/10 cm dish)
and treated with various compounds after 24 h. The medium was
changed every 2-3 days after the initial treatment and supplemented
with a fresh dose of drug.
Determination of Doubling Time. The cell number/dish was
counted with a Z1 Coulter particle counter (Beckman Coulter Corp.,
Hialeh, FL) every 2-3 days. Untreated cells were analyzed under
similar conditions as a control. The average cell number from two
plates was determined, and the mean cell numbers were plotted to
define the cell population doubling times, where population doubling
PD ) log (number of cells harvested/number of cells seeded)/log
2. Initial drug treatment was started 24 h after seeding.
Discussion
Zebularine is an effective inhibitor of DNA methylation
capable of inducing the expression of methylation-silenced genes
upon treatment. It has many properties that are advantageous
as a chemotherapeutic agent, including stability in aqueous
solution as well as low toxicity in animals and cultured
cells.5,11,28 However, zebularine must undergo several metabolic
steps before DNA incorporation,15 which subsequently weakens
the demethylating effect of the drug. A number of aryloxy-
phosphoramidate nucleotide prodrugs (ProTides) of 2′-deoxy-
zebularine (dZMP) were tested in search of a demethylating
agent more potent and efficient than zebularine. The ProTides
that were effective displayed good activity in Cf-Pac-1 pancre-
atic cancer cells only in the presence of 100 µM thymidine;
they were marginally active in HCT15 colon cancer cells and
inactive in T24 bladder cancer cells, even in the presence of
thymidine. Such cell line specificity of the phosphoramidates
may be due to the availability and differential activities of
enzymes involved in the release of the dZMP moiety.
The unexpected finding of this investigation was that the
phosphoramidates were not able to induce p16 expression in
the absence of thymidine. Thymidine was used to prevent the
inhibition of TS24 and dCMP deaminase25 by dZMP, which
would deplete TMP levels. Since zebularine depends on the
synthesis of new DNA to inhibit methylation and the inhibition
of TS and dCMP would block DNA synthesis due to the
shortage of TMP, it was imperative to supplement the cells with
an outside source of thymidine.
Nucleic Acid Isolation. RNA was collected from T24 and
HCT116 cells with the RNeasy mini kit (Qiagen, Valencia, CA)
according to the manufacturer’s protocol. DNA was collected using
the DNeasy tissue kit (Qiagen, Valencia, CA) according to the
manufacturer’s protocol.
Quantitative RT-PCR. Total RNA (5 µg) was reverse tran-
scribed with Moloney murine leukemia virus reverse transcriptase
(Invitrogen, Carlsbad, CA) and random primers (Invitrogen, Carls-
bad, CA). The reverse transcription was carried out in a total volume
of 50 µL as previously described.23 The quantitation of mRNA
levels was carried out by a real time fluorescence detection method
as described previously.11,22 All samples were normalized to the
reference gene, GAPDH. The primer and probe sequences are as
follows: for p16, sense 5′-CTG CCC AAC GCA CCG A-3′, probe
5′ 6-FAM-TGG ATC GGC CTC CGA CCG TAA CT BHQ-1 3′,
and antisense 5′-CGC TGC CCA TCA TCA TGA C-3′; for
SPANXA1, sense 5′-TGT GAT TCC AAC GAG GCC A-3′, probe
5′ 6-FAM-CGA GAT GAT GCC GGA GAC CCC A BHQ-1 3′,
Although these compounds demonstrated conceptually that
the treatment with dZMP ProTides inhibits DNA methylation
efficiently, the required administration of thymidine might
particularly limit their use in a potential clinical setting.
Nevertheless, it might be feasible to use this approach in the
future.
In summary, some dZMP ProTides were capable of inducing
greater expression of the p16 gene than zebularine. Although

7598 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Yoo et al.
and antisense 5′-GCG GGT CTG AGT CCC CA-3′; for MAGEA1,
sense 5′-GAA CCT GAC CCA GGC TCT GTG-3′, probe 5′
6-FAM-CAA GGT TTT CAG GGG ACA GGC CAA C BHQ-1
3′, and antisense 5′-CCA CAG GCA GAT CTT CTC CTT C′-3′;
for MAGEB2, sense 5′-CGG CAG TCA AGC CAT CAT G-3′,
probe 5′ 6-FAM-TCG TGG TCA GAA GAG TAA GCT CCG
TGC BHQ-1 3′, and antisense 5′-GCG GGT CTG AGT CCC CA-
3′; for GAGE, sense 5′-GCT GAT AGC CAG GAA CAG GG-3′,
probe 5′ 6-FAM-CAC CCA CAG ACT GGG TGT GAG TGT GA
BHQ-1 3′, and antisense 5′-CCT GCC CAT CAG GAC CAT C-3′;
for XAGEA1, sense 5′-TCC CCA GAC GGG ACC AG-3′, probe
5′ 5-FAM-AGA GGG ACG GCA TGA GCG ACA CAC BHQ-1
3′, and antisense 5′-CTG GCT GTG TGG TTC TGT GTT T-3′;
for GAPDH, sense 5′-TGA AGG TCG GAG TCA ACG G-3′, probe
5′ 6-FAM -TTT GGT CGT ATT GGG CGC CTG G BHQ-1 3′,
and antisense 5′-AGA GTT AAA AGC AGC CCT GGT G-3′. The
conditions for real time RT-PCR are 94 °C for 9 min followed by
45 cycles at 94 °C for 15 s and 60 °C for 1 min.
Quantitation of DNA Methylation. Genomic DNA (4 µg) was
treated with sodium bisulfite as previously described.11 Methylation
analysis was performed using the methylation-sensitive single-
nucleotide primer extension (Ms-SNuPE) assay for the p16 5′ region
as previously described.27 The PCR primers used are as follows:
for p16, sense 5′-TTT GAG GGA TAG GGT-3′ and antisense 5′-
TCT AAT AAC CAA CCA ACC CCT CC-3′; for MAGEA1, sense
5′-GTT TAT TTT TAT TTT TAT TTA GGT AGG A-3 and
antisense 5′-TTA CCT CCT CAC AAA ACC TAA A-3′; for
MAGEB2, sense 5′-TTG AGG GAG GTG GGG GTA TTG T-3′
and antisense 5′-CTT CAA TTT ACA CTC AAA ATC CTC ACC
T-3′; for D4Z4, sense 5′-GGG TTG AGG GTT GGG TTT AT-3′
and antisense 5′-AAC TTA CAC CCT TCC CTA CA-3′. An initial
denaturation at 94 °C for 3 min was followed by 94 °C for 45 s,
annealing for 45 s, and 72 °C for 45 s for 40 cycles. The annealing
temperatures for each locus are as follows: for p16, 65 °C, for
MAGEA1, 53 °C, for MAGEB2, 62 °C, and for D4Z4, 58 °C.
Primers used for Ms-SNuPE analysis are as follows: for p16, 5′-
TTT TTT TGT TTG GAA AGA TAT-3′, 5′-TTT TAG GGG TGT
TAT ATT-3′, and 5′-GTA GAG TTT AGT T-3′; for MAGEA1,
5′-AGG TTT TTA TTT TGA GGG A-3′, 5′-TGG GGT AGA GAG
AAG-3′ and 5′-TTT TAT TTT TAT TTA GGT AGG ATT-3′; for
MAGEB2, 5′-ATT GTT TGG AGG TTG G-3′, 5′-GAG GAT TTT
TAG TGA AGA-3′, and 5′-GAT GTG GTT TAT TTT GAT TTT-
3′; for D4Z4, 5′-TGA GGG TTG GGT TTA TAG T-3′, 5′-TAT
ATT TTT AGG TTT AGT TTT GTA A-3′, 5′-GTG GTT TAG
GGA GTG GG-3′, and 5′-GAA AGG TTG GTT ATG T-3′.
Conditions for primer extension were 94 °C for 1 min, 50 °C for
30 s, and 72 °C for 20 s.
LCT time-of-flight mass spectrometer coupled to a Waters M600
HPLC pump. Samples were dissolved in methanol and injected into
the solvent stream via a Rheodyne injector. The mobile phase used
was methanol at a flow rate of 200 µL/min. The electrospray source
was operated at a temperature of 130 °C with a desolvation
temperature of 300 °C, a capillary voltage of 3 kV, and cone voltage
of 30 V. Data were collected in the continuum mode over the mass
range 100-2000 amu and processed using Masslynx 4.1 software.
Accurate mass measurements were facilitated by the introduction
of a single lockmass compound of known elemental composition
into the source concurrently with sample. The purity of crude
compounds 7a, 7b, 7b′, and 10 was established by HPLC chroma-
tography which ranged from 85% to 93%. Single peaks of pure
compounds were collected dried and weighed accurately.
Standard Procedure A: Synthesis of Phosphorodichloridate.
Phosphorus oxychloride (1.0 mmol) was added to a stirred solution
of the appropriate phenol or naphthol (1.0 mmol) in dry ether. Then
the solution was stirred at -78 °C, and anhydrous triethylamine
(1.0 mmol) was added dropwise. After 1 h the reaction was left to
rise to room temperature and stirred for about 3 h, monitoring the
formation of the desired compound by 31P NMR. The triethylamine
hydrochloride salt was filtered off and the filtrate reduced to dryness
to give a crude oil as product that was used without further
purification for the next step.
Standard Procedure B: Synthesis of Phosphorochloridate.
Anhydrous triethylamine (2.0 mol equiv) was added dropwise at
-78 °C to a stirred solution of the appropriate phosphorodichlo-
ridate (1.0 mol equiv) and the appropriate amino acid hydrochloride
salt (1.0 mol equiv) in anhydrous dichloromethane (15-30 mL).
After 1 h the reaction was allowed to slowly warm to room
temperature, and the formation of the desired compound was
monitored by 31P NMR. The solvent was removed under reduced
pressure, and the crude residue was resuspended in anhydrous ether
and filtered under nitrogen. The corresponding filtrate was reduced
to dryness and purified by flash chromatography (ethyl acetate/
hexane 7/3) to give the product as an oil.
Standard Procedure C: Synthesis of Phosphoramidate
(tBuMgCl Method). tert-Butylmagnesium chloride (1 M solution
in THF, 1.2 mmol) was added to a stirring suspension of the
appropriate nucleoside (1.0 mmol) in dry THF (10 mL) under argon
atmosphere. The appropriate phosphorochloridate (1.2 mmol)
dissolved in dry THF was added dropwise, and the reaction was
left stirring overnight. Volatiles were evaporated, and the residue
was purified by flash chromatography (CH2Cl2/CH3OH) to give the
desired product.
Standard Procedure D: Synthesis of Phosphoramidate
(NMI Method). N-Methylimidazole (NMI, 5.0 mmol) was added
to a stirring suspension of the appropriate nucleoside (1.0 mmol)
in dry THF (10 mL) under argon atmosphere. The appropriate
phosphorochloridate (3.0 mmol) dissolved in THF was added
dropwise and the reaction left stirring overnight. Volatiles were
evaporated, and the residue was dissolved in CH2Cl2 and washed
with 0.5 M HCl. The organic layer was dried over MgSO4, filtered,
reduced to dryness, and purified by flash chromatography (CH2Cl2/
CH3OH).
Standard Procedure E: Synthesis of Phosphoramidate
(Variant of the NMI Method). A stirred solution of the appropriate
nucleoside (1.0 mmol) in anhydrous pyridine (5 mL) was treated
with the appropriate phosphorochloridate (3.4 mmol) in dry THF
(4 mL) under argon atmosphere. N-Methylimidazole (NMI, 6.3
mmol) was then added, and the reaction was stirred overnight.
Additional amounts of phosphorochloridate (3.4 mmol) and NMI
(6.3 mmol) were added, and stirring was continued for an additional
4 h. Volatiles were evaporated, and the residue was dissolved in
CH2Cl2 and chromatographed twice by flash column chromatog-
raphy (CH2Cl2/CH3OH).
Synthesis. General Procedures. All experiments involving
water-sensitive compounds were conducted under scrupulously dry
conditions. Anhydrous tetrahydrofuran (THF) and dichloromethane
were purchased from Aldrich and used directly. Column chroma-
tography refers to flash column chromatography carried out using
Merck silica gel 60 (40-60 µm) as stationary phase. Proton, carbon,
and phosphorus nuclear magnetic resonance (1H, 13C, 31P NMR)
spectra were recorded on Bruker Avance spectrometers operating
either at 500, 125, and 202 MHz or at 300, 75, and 121 MHz. For
compounds 3 and 4 the spectrometer was a Varian Unity Inova
instrument operating at 400, 100, and 161.9 MHz. The solvents
used are indicated for each compound. All 13C and 31P spectra
1
were recorded proton decoupled. Chemical shifts for H and 13C
spectra are in parts per million downfield from tetramethylsilane.
Coupling constants are referred to as J values. Signal splitting
patterns are described as singlet (s), doublet (d), triplet (t), quartet
(q), broad signal (br), doublet of doublet (dd), doublet of triplet
(dt), or multiplet (m). Chemical shifts for 31P spectra are in parts
per million relative to an external phosphoric acid standard. Many
proton and carbon NMR signals were split because of the presence
of (phosphate) diastereoisomers in the samples. The mode of
ionization for mass spectrometry for compounds 3 and 4 was fast
atom bombardment (FAB) using MNOBA (m-nitrobenzyl alcohol)
as matrix. Electrospray mass spectra were obtained using a Waters
2′-Deoxyzebularine 5′-[Phenyl(methoxy-L-alaninyl)]phosphate
(3). Prepared according to standard procedure E, using 2′-deoxy-
zebularine (0.09 g, 0.42 mmol), methylchlorophenylphosphoryl-L-
alaninate (2 × 0.40 g, 2.88 mmol), and NMI (2 × 0.24 mL, 6
mmol). The crude was purified twice by column chromatography

ActiVation of p16 Gene by DeriVatiVes of 2′-Deoxyzebularine
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7599
(CH2Cl2/CH3OH 93/7) and coevaporated with ether (2 × 5 mL) to
165.6 (C-6), 173.4, 173.5, 173.7 (CO2Et). HPLC (H2O/CH3CN from
70/30 to 0/100 in 10 min): retention time 7.21 min.
give the pure product as a white foamy solid (57 mg, 29.7% yield).
1
31P NMR (CDCl3, 161.9 MHz): δ 3.23, 3.36. H NMR (CDCl3,
2′-Deoxyzebularine 5′-[Phenyl(benzoxy-L-alaninyl)]phosphate
(6). Prepared according to standard procedure D, from 2′-deoxy-
zebularine (0.20 g, 0.94 mmol), phenyl(benzoxy-L-alaninyl) phos-
phorochloridate (1.00 g, 2.83 mmol), NMI (0.39 g, 4.70 mmol,
0.37 mL), and dry THF (10 mL). The crude was purified by column
chromatography(CH2Cl2/CH3OH96/4)andpreparativeTLC(CH2Cl2/
CH3OH 96/4) to give the pure product as a white foamy solid (70
mg, 14.1% yield). 31P NMR (CDCl3, 121 MHz): δ 4.09, 4.28. 1H
NMR (CDCl3, 300 MHz): δ 1.42-1.45 (3H, m, CHCH3), 1.99-2.12
(1H, m, one of H-2′), 2.69-2.78 (1H, m, one of H-2′), 4.03-4.48
(7H, m, H-3′, H-4′, H-5′, OH-3′, CHNH, CHNH), 5.14-5.22 (2H,
m, PhCH2), 6.18-6.32 (2H, m, H-5, H-1′), 7.20-7.38 (10H, m,
PhO, PhCH2), 8.21-8.25 (1H, m, H-4), 8.46-8.54 (1H, m, H-6).
13C NMR (CDCl3, 75 MHz): δ 19.6, 19.7, 19.8 (CH3), 40.2 (C-
2′), 40.3, 40.4, 40.5 (CHCH3), 64.6, 64.7, 64.8 (C-5′), 66.2, 66.3
(CH2Ph), 68.8, 68.9 (C-3′), 84.4, 84.5, 84.6, 84.7 (C-4′), 86.6 (C-
1′), 103.3 (C-5), 119.0, 119.1, 119.3, 124.2, 127.2, 127.5, 127.6,
127.7, 128.7, 128.8 (Ph), 134.2 (“ispo” CH2Ph), 142.5, 142.6 (C-
4), 149.4, 149.5 (“ipso” OPh), 164.6, 164.8 (C-6), 172.3, 172.4
(COOCH2Ph). HPLC (H2O/CH3CN from 100/0 to 0/100 in 20 min):
retention time 12.37 min.
400 MHz): δ 1.43 (3H, m, CH3), 2.00-2.10 (1H, m, H-2′),
2.60-2.70 (1H, m, H-2′), 3.50-3.70 (4H, m, H-4′, CO2CH3),
3.90-4.05 (1H, m, CH), 4.10 (1H, m, H-3′), 4.25-4.40 (2H, m,
H-5′), 6.10 (1H, m, H-5), 6.22 (1H, m, H-1′), 7.10-7.30 (5H, m,
PhO), 8.14 (1H, m, H-6), 8.50 (1H, br s, H-4). 13C NMR (CDCl3,
100 MHz): δ 19.8 (CH3CH), 40.2 (C-2′), 49.3, 51.6 (CHCH3), 64.7
(CH3O), 68.7 (C-5′), 68.9 (C-3′), 86.6 (C-4′), 103.3, 109.1, 118.9,
119.1, 124.2, 128.8, 142.6, 149.4, 154.5, 164.6, 172.9, 173.1
(CO2Me). FAB-MS m/e: 454.1 (MH+, 15%); Anal. Calcd for
C19H24N3O8P: C, 50.33; H, 5.34; N, 9.27. Found: C, 50.04; H, 5.59;
N, 9.03.
2′-ꢀ-Fluoro-2′-deoxyzebularine 5′-[Phenyl(methoxy-L-alani-
nyl)]phosphate (4). Prepared according to standard procedure E,
using 2′-fluoro-2′-deoxyzebularine (0.230 g, 1 mmol), methylchlo-
rophenylphosphoryl-L-alaninate (2 × 0.95 g, 6.8 mmol), and NMI
(2 × 0.5 mL, 12.54 mmol). The crude was purified twice by column
chromatography (CH2Cl2/CH3OH 93/7) and coevaporated with ether
(2 × 5 mL) to give the pure product as a white foamy solid (160
mg, 34% yield). 31P NMR (CDCl3, 161.9 MHz): δ 3.46, 3.82. 1H
NMR (CDCl3, 400 MHz): δ 1.30-1.36 (3H, 2 d, J ) 5.4 Hz, CH3),
3.63 and 3.65 (3H, singlets, CO2CH3), 3.70-3.85 (1H, dt, J ) 37.6,
10.0 Hz, H-2′), 3.91-4.04 (1H, m, CH), 4.22-4.40 (3H, m, H-5′,
H-4′), 5.28 (1H, dm, J ) 50.8 Hz, H-3′), 6.26 (1H, dt, J ) 19.9,
2.9 Hz, H-1′), 6.33 (1H, br t, H-5), 7.05-7.30 (5H, m, PhO),
8.08-8.14 (1H, ddd, J ) 12.7, 6.6, 1.9 Hz, H-6), 8.54 (1H, br s,
H-4). 13C NMR (CHCl3, 100 MHz): δ 19.6, 19.7, (CH3CH), 49.2
(CHCH3), 51.52, 64.6, 64.8 (C-5′), 73.4, 73.6 (C4′), 83.0, 83.1,
83.2, 83.3 (C3′), 86.12 (t), 93.0 (d, J ) 195 Hz, C-2′), 103.3 (C-
1′), 119.0, 119.1, 119.2, 119.3, 124.1, 128.7, 128.8, 144.3, 149.4,
149.5, 149.6, 154.3, 165.3, 172.9 (CdO). FAB-MS m/e: 472.4
(MH+, 70%). Anal. Calcd for C19H23FN3O8P·0.5H2O: C, 47.50;
H, 5.04; N, 8.75. Found: C, 47.54; H, 5.06; N, 8.57.
2′-Deoxyzebularine 5′-[Naphthyl(ethoxydimethylglycinyl)]-
phosphate (7a). Prepared adopting the standard procedure D using
2′-deoxyzebularine (0.13 g, 0.60 mmol) in dry THF (10 mL), NMI
(0.24 mL, 3.0 mmol), naphthyl(ethoxydimethylglycinyl) phospho-
rochloridate (0.64 g, 1.8 mmol). The use of procedure D minimized
the amount of 3′,5′-bisphosphoramidates, whereas procedure C
failed to provide 7a. The crude was purified by column chroma-
tography (CH2Cl2/CH3OH 96/4) to give the pure product as a white
solid (35.0 mg, 11.0%). 31P NMR (MeOD, 202 MHz): δ 2.93, 2.94.
1H NMR (MeOD, 500 MHz): δ 1.12-1.15 (3H, m, CH2CH3)
1.42-1.45 (6H, m, CH3), 1.58-1.67 (1H, m, H-2′), 2.30-2.36 (1H,
m, H-2′), 4.03-4.37 (6H, m, H-3′, H-4′, H-5′, CH2CH3), 5.39-5.97
(1H, m, H-1′), 6.08-6.17 (1H, m, H-5), 7.26-8.07 (7H, m, Naph),
8.09-8.12 (1H, m, H-4), 8.31-8.37 (1H, m, H-6). 13C NMR
(MeOD, 125 MHz): δ 14.45 (CH3CH2), 27.56, 27.59, 27.99, 28.05,
28.11 (CH3), 42.21, 42.29 (C-2′), 62.66 (CH3CH2), 67.54, 67.58,
67.63 (C-5′), 71.73, 71.87 (C-3′), 87.57, 87.63, 87.69 (C-4′), 89.39,
89.46 (C-1′), 106.06 (C-5), 116.29, 116.31, 116.76, 116.78, 122.83,
122.95, 126.00, 126.10, 126.53, 126.58, 127.48, 127.53, 127.89,
128.95 (Naph), 136.31 (ipso Naph), 145.56 (C-4), 157.04 (C-2),
167.03, 167.12 (C-6), 176.81 (CdO). MS (ES) m/e: 554.4 (MNa+,
100%). Accurate mass: C25H30N3O8NaP required 554.1688, found
554.1664. HPLC (H2O/CH3CN from 100/0 to 0/100 in 20 min):
retention time 11.92 min.
2′-Deoxyzebularine 5′-[4-chlorophenyl(ethoxy-L-alaninyl)]phos-
phate (5a). Prepared according to standard procedure D, from 2′-
deoxyzebularine (0.19 g, 0.88 mmol), 4-chlorophenyl(ethoxy-L-
alaninyl) phosphorochloridate (0.86 g, 2.64 mmol), NMI (0.36 g,
4.40 mmol, 0.35 mL), and dry THF (10 mL). The crude was purified
by column chromatography (CH2Cl2/CH3OH 95/5) to give the pure
product as a white foamy solid (88 mg, 20.0% yield). 31P NMR
1
(CDCl3, 121 MHz): δ 3.23, 3.32. H NMR (CDCl3, 300 MHz): δ
1.11-1.20 (3H, m, OCH2CH3), 1.39-1.42 (3H, m, CHCH3),
1.99-2.09 (1H, m, one of H-2′), 2.61-2.73 (1H, m, one of H-2′),
3.83-4.40 (9H, m, H-3′, H-4′, H-5′, OH-3′, CHNH, CHNH,
OCH2CH3), 6.10-6.30 (2H, m, H-5, H-1′), 7.13-7.22 (4H, m, pCl-
Ph), 8.09-8.19 (1H, m, H-4), 8.45-8.53 (1H, m, H-6). 13C NMR
(CDCl3, 75 MHz): δ 14.1 (CH3CH2O), 20.8 (CH3CH), 41.2 (C-
2′), 50.3, 50.4 (CHCH3), 61.8 (CH3CH2O), 65.9 (C-5′), 69.8, 70.0
(C-3′), 85.4, 85.5 (C-4′), 87.7 (C-1′), 121.4, 121.5, 121.6, 129.8
(C-5, pCl-Ph), 130.5 (C-1, pCl-Ph), 143.6 (C-4), 149.0 (C-4, pCl-
Ph), 155.5 (C-2), 165.7 (C-6), 173.6 (CO2Et). HPLC (H2O/CH3CN
from 70/30 to 0/100 in 10 min): retention time 4.61 min.
2′-Deoxyzebularine 3′,5′-Bis[naphthyl(ethoxydimethylgly-
cinyl)]phosphate (7b and 7b′). Prepared adopting the standard
procedure C, using 2′-deoxyzebularine (0.15 g, 0.70 mmol) in dry
t
THF (10 mL), BuMgCl (1 M solution in THF, 0.84 mL, 0.84
mmol), and naphthyl(ethoxydimethylglycinyl) phosphorochloridate
(0.30 g, 0.84 mmol). The crude was purified by column chroma-
tography (CH2Cl2/CH3OH 96/4) to give the pure product as a white
solid (fast eluting fraction ) 24.4 mg, 4.1%; slow eluting fraction
) 19.4 mg, 3.2%).
2′-Deoxyzebularine 3′,5′-Bis[4-chlorophenyl(ethoxy-L-alani-
nyl)]phosphate (5b). Isolated through column chromatography
purification of 5a crude reaction mixture and purified by preparative
reverse-phase HPLC (40 mg, 9.1% yield) and obtained as a white,
foamy solid. 31P NMR (CDCl3, 121 MHz): δ 1.84, 1.89, 2.42, 2.47,
3.04, 3.09, 3.11. 1H NMR (CDCl3, 300 MHz): δ 1.11-1.37 (12H,
m, 2 CHCH3 2 OCH2CH3), 1.98-2.14 (1H, m, one of H-2′),
2.78-2.97 (1H, m, one of H-2′), 3.82-4.42 (9H, m, H-4′, H-5′, 2
CHNH, 2 CHNH, 2 OCH2CH3), 4.99-5.21 (1H, m, H-3′),
6.03-6.30 (2H, m, H-5, H-1′), 7.01-7.28 (8H, m, 2 pCl-Ph),
8.03-8.18 (1H, m, H-4), 8.42-8.51 (1H, m, H-6). 13C NMR
(CDCl3, 75 MHz): δ 14.1 (2 CH3CH2O), 20.6, 20.8, 20.9 (2
CH3CH), 39.8, 39.9, 40.0 (C-2′), 50.2, 50.3, 50.4, 50.5 (2 CHCH3),
61.8, 61.9 (2 CH3CH2O), 65.6, 65.7, 66.8, 66.9 (C-5′), 76.7, 77.0,
77.1 (C-3′), 84.2, 84.3 (C-4′), 87.3, 87.4 (C-1′), 121.3, 121.5, 121.6,
129.7, 129.8 (C-5, 2 pCl-Ph), 130.7 (C-1, pCl-Ph), 143.8, 144.0,
144.1 (C-4), 149.1, 149.2 (2 C-4, pCl-Ph), 155.3, 155.4 (C-2), 165.5,
Fast Eluting Fraction (7b). 31P NMR (MeOD, 202 MHz): δ
2.60, 2.93. 1H NMR (MeOD, 500 MHz): δ 1.17-1.26 (6H, m,
CH2CH3), 1.42-1.57 (12H, m, CH3), 1.75-1.78 (1H, m, H-2′),
2.75-2.80 (1H, m, H-2′), 4.09-4.17 (4H, m, CH2CH3), 4.32-4.41
(3H, m, H-4′, H-5′), 5.22-5.24 (1H, m, H-3′), 5.95-5.98 (1H, m,
H-1′), 6.06-6.08 (1H, m, H-5), 7.34-8.24 (15H, m, Naph, H-4),
8.39-8.40 (1H, m, H-6). 13C NMR (MeOD, 125 MHz): δ 14.41,
14.47 (CH3CH2), 27.46, 27.55, 27.58, 27.96, 28.02, 28.13, 28.18
(CH3), 40.70, 40.74 (C-2′), 62.58, 62.66 (CH3CH2), 66.81, 66.85
(C-5′), 77.81, 77.85 (C-3′), 86.00, 86.07, 86.10 (C-4′), 89.14 (C-
1′), 106.08 (C-5), 116.95, 123.00, 123.07, 126.13, 126.20, 126.53,
126.61, 127.51, 127.58, 127.84, 127.93, 128.89, 128.94 (Naph),
136.25, 136.37 (ipso Ph), 145.32 (C-4), 156.78 (C-2), 167.12 (C-
6), 176.70, 176.88 (CdO). MS (ES) m/e: 873.4 (MNa+, 100%).

7600 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Yoo et al.
Accurate mass: C41H48N4O12NaP2 required 873.2642, found 873.2661.
HPLC (H2O/CH3CN from 100/0 to 0/100 in 20 min): retention time
18.04 min.
202 MHz): δ 1.50, 1.59, 1.62, 1.88, 2.12, 2.26, 2.31, 2.37. 1H NMR
(MeOD, 500 MHz): δ 1.35-1.40 (12H, m, CH3), 1.76-1.86 (1H,
m, H-2′), 2.57-2.73 (1H, m, H-2′), 4.15-4.27 (3H, m, H-4′, H-5′),
5.00-5.08 (5H, m, H-3′, CH2Ph), 5.92-6.02 (1H, m, H-1′),
6.26-6.37 (1H, m, H-5), 7.00-7.27 (20H, m, Ph), 8.16-8.21 (1H,
m, H-4), 8.40-8.45 (1H, m, H-6). 13C NMR (MeOD, 125 MHz):
δ 27.43, 27.49, 27.52, 27.64, 27.81, 27.87, 27.90, 27.94, 27.96,
28.00, 28.08, 28.16 (CH3), 40.77, 40.81, 40.89, 40.92 (C-2′), 66.79,
66.83, 67.15, 67.19, 67.28 (C-5′), 68.22, 68.24, 68.27, 68.30
(CH2Ph), 78.36, 78.40, 78.56, 78.59, 78.67, 78.72 (C-3′), 85.95,
86.05, 86.12, 86.17, 86.33, 86.38 (C-4′), 89.21, 89.32, 89.43, 89.48
(C-1′), 106.27, 106.32 (C-5), 121.51, 121.55, 121.60, 121.62,
121.63, 121.66, 121.70, 121.75, 121.76, 121.80, 126.31, 126.35,
129.16, 129.19, 129.24, 129.29, 129.32, 129.34, 129.37, 129.64,
130.88 (Ph), 137.32, 137.38 (ipso Ph), 145.73, 145.84 (C-4), 152.07,
152.10, 152.13 (ipso OPh), 157.01, 157.03, 157.06 (C-2), 167.48,
167.52 (C-6), 176.48 (CdO). MS (ES) m/e: 1771.4 (2M + Na+,
100%).
Slow Eluting Fraction (7b′). 31P NMR (MeOD, 202 MHz): δ
2.14, 2.69. 1H NMR (MeOD, 500 MHz): δ 1.04-1.14 (6H, m,
CH2CH3), 1.36-1.45 (12H, m, CH3), 1.78 1.84 (1H, m, H-2′),
2.69-2.73 (1H, m, H-2′), 3.95-4.05 (4H, m, CH2CH3), 4.22-4.30
(3H, m, H-4′, H-5′), 5.09-5.11 (1H, m, H-3′, 5.92-5.96 (1H, m,
H-1′), 6.19-6.22 (1H, m, H-5), 7.21-8.13 (15H, m, Naph, H-4),
8.39-8.40 (1H, m, H-6). 13C NMR (MeOD, 125 MHz): δ 14.41,
14.46 (CH3CH2), 27.55, 27.58, 27.61, 27.94, 28.00, 28.07, 28.13
(CH3), 40.83, 40.87 (C-2′), 62.61, 62.64 (CH3CH2), 67.31, 67.36
(C-5′), 78.81, 78.85 (C-3′), 86.37, 86.43, 86.47 (C-4′), 89.35, 89.45
(C-1′), 106.21 (C-5), 116.31, 116.33, 116.85, 116.88, 122.81,
123.03, 125.98, 126.14, 126.52, 127.51, 127.81, 127.84, 127.96,
128.01, 128.87, 128.91 (Naph), 136.27, 136.34 (ipso Ph), 145.55
(C-4), 156.99 (C-2), 167.40 (C-6), 176.72, 176.81 (CdO). MS (ES)
m/e: 873.2 (MNa+, 100%). Accurate mass: C41H48N4O12NaP2
required 873.2642, found 873.2669. HPLC (H2O/CH3CN from
100/0 to 0/100 in 20 min): retention time 17.75 min.
2′-Deoxyzebularine 5′-[Naphthyl(benzoxydimethylglycinyl)]-
phosphate (10). Prepared adopting the standard procedure D, using
2′-deoxyzebularine (0.13 g, 0.60 mmol), in dry THF (10 mL), NMI
(0.24 mL, 3.0 mmol), and naphthyl(ethoxydimethylglycinyl) phos-
phorochloridate (0.76 g, 1.8 mmol). The crude was purified by
column chromatography (CH2Cl2/CH3OH 96/4) to give the pure
product as a white solid (7.0 mg, 1.7%). 31P NMR (MeOD, 202
2′-Deoxy-2′-ꢀ-fluorozebularine 5′-[Phenyl(benzoxy-L-alani-
nyl)]phosphate (8). Prepared adopting the standard procedure D,
using 2′-deoxy-2′-ꢀ-fluorozebularine (200 mg, 0.87 mmol), in dry
THF (10 mL), NMI (345 µL, 4.35 mmol), phenyl(benzoxydimeth-
ylglycininyl) phosphorochloridate, 1 M in dry THF (2.61 mL, 2.61
mmol). The crude mixture was purified by flash chromatography
(CH2Cl2/CH3OH gradient elution from 98/2 to 95/5) and semi-
preparative TLC to give the pure product as a white foamy solid
1
MHz): δ 2.92, 2.89. H NMR (MeOD, 500 MHz): δ 1.56-1.59
(6H, m, CH3), 1.62-1.72 (1H, m, H-2′), 2.38-2.44 (1H, m, H-2′),
4.13-4.37 (4H, m, H-3′, H-4′, H-5′), 5.13-5.20 (2H, m, CH2Ph),
6.02-6.06 (1H, m, H-1′), 6.17-6.25 (1H, m, H-5), 7.29-8.19 (8H,
m, H-4, Naph), 8.42-8.47 (1H, m, H-6). 13C NMR (MeOD, 125
MHz): δ 27.51, 27.54, 27.99, 28.06, 28.13 (CH3), 42.20, 42.28 (C-
2′), 67.51, 67.56, 67.60 (C-5′), 68.32 (CH2Ph), 71.69, 71.83 (C-
3′), 87.52, 87.59, 87.65 (C-4′), 84.37, 89.43 (C-1′), 106.06 (C-5),
116.33, 116.80, 116.83, 122.82, 122.94, 125.99, 126.09, 126.50,
126.56, 127.46, 127.51, 127.87, 128.93, 129.33, 129.59 (Naph, Ph),
136.29 (ipso Ph), 137.31 (ipso Naph), 145.52 (C-4), 156.98 (C-2),
166.99, 167.08 (C-6), 176.56 (CdO). MS (ES) m/e: 616.3 (MNa+,
100%). Accurate mass: C30H32N3O8NaP required 616.1825, found
616.1818.
1
(110 mg, 23%). 31P NMR (MeOD, 121 MHz): δ 5.00, 5.21. H
NMR (MeOD, 300 MHz): δ 1.30-1.43 (3H, m, CH3CH), 3.96-4.40
(5H, m, H-3′, H-4′, H-5′, CHCH3), 5.06-5.26 (3H, m, H-2′,
CH2Ph), 6.15-6.30 (1H, m, H-1′), 6.45-6.55 (1H, m, H-5),
7.13-7.43 (10H, m, Ph), 8.26-8.36 (1H, m, H-4), 8.56-8.65 (1H,
m, H-6). 13C NMR (MeOD, 75 MHz): δ 20.5, 20.6, 20.7 (CH3CH),
52.0 (CHCH3), 66.9, 67.0 (C-5′), 68.2 (CH2Ph), 75.3, 75.4, 75.6,
75.7 (C-4′), 85.4, 85.5, 85.6 (C-3′), 88.6, 88.8 (C-5), 95.7 (d, J )
195 Hz, C-2′), 106.2 (C-1′), 121.6, 121.7, 121.8, 126.5, 129.5, 129.6,
129.8, 131.0, 131.1, 131.2, 137.5 (Ph), 147.0 (C-4), 152.3, 152.4
(“ipso” OPh), 157.1 (C-2), 168.2, 168.4 (C-6), 174.8, 174.9, 175.0,
175.1 (CdO). HPLC (H2O/CH3CN from 100/0 to 0/100 in 10 min):
retention time 7.05 min.
Acknowledgment. This research was supported in part by
the Intramural Research Program of the NIH, Center for Cancer
Research, NCIsFrederick.
2′-Deoxyzebularine 5′-[Phenyl(benzoxydimethylglycininyl)]-
phosphate (9a). Prepared adopting the standard procedure D, using
2′-deoxyzebularine (135 mg, 0.63 mmol), in dry THF (10 mL),
NMI (250 µL, 3.15 mmol), and phenyl(benzoxydimethylglycinyl)
phosphorochloridate, 1 M in dry THF (1.89 mL, 1.89 mmol). As
before, the use of procedure D minimized the amount of 3′,5′-
bisphosphoramidates, whereas procedure C failed to provide 9a.
The crude mixture was purified with flash chromatography (CH2Cl2/
CH3OH gradient elution from 98/2 to 95/5), semipreparative TLC,
and eventually HPLC (isocratic elution water/acetonitrile, 60/40)
to give the pure product as a white foamy solid (25 mg, 7%). 31P
NMR (MeOD, 121 MHz): δ 2.52, 2.60. 1H NMR (MeOD, 300
MHz): δ 1.49-1.55 (6H, m, (CH3)2C), 1.80-2.03 (1H, m, H-2′),
2.45-2.63 (1H, m, H-2′), 4.13-4.40 (4H, m, H-3′, H-4′, H-5′),
5.10-5.20 (2H, m, CH2Ph), 6.10-6.20 (1H, m, H-1′), 6.40-6.53
(1H, m, H-5), 7.13-7.43 (10H, m, Ph), 8.30-8.40 (1H, m, H-4),
8.50-8.60 (1H, m, H-6). 13C NMR (MeOD, 75 MHz): δ 27.5,
27.8, 27.9, 28.0 ((CH3)2C), 42.3 (C-2′), 58.2 ((CH3)2C), 66.9 (C-
5′), 67.3, 67.4 (OCH2Ph), 71.7 (C-3′), 87.5, 87.6, 87.7 (C-4′), 89.5,
89.6 (C-1′), 106.2 (C-5), 121.5, 121.6, 126.3, 128.0, 128.3, 129.3,
129.4, 129.5, 129.6, 129.9, 130.8, 132.4, 133.6, 137.4 (Ph), 145.9
(C-4), 152.1, 152.2, 152.3 (“ipso” OPh), 157.2 (C-2), 167.2, (C-6),
176.5 (CdO). HPLC (H2O/CH3CN 60/40): retention time 3.21 min.
2′-Deoxyzebularine 3′,5′-Bis[phenyl(benzoxydimethylgly-
cinyl)]phosphate (9b). Prepared adopting the standard procedure
C, using 2′-deoxyzebularine (0.15 g, 0.70 mmol), in dry THF (10
mL), NMI (0.28 mL, 3.50 mmol), and phenyl(benzoxydimethylg-
lycinyl) phosphorochloridate (0.77 g, 2.10 mmol). The crude was
purified by column chromatography (CH2Cl2/CH3OH 96/4) to give
the pure product as a white solid (60 mg, 9.8%). 31P NMR (MeOD,
References
(1) Baylin, S. B.; Ohm, J. E. Epigenetic gene silencing in cancersA
mechanism for early oncogenic pathway addiction. Nat. ReV. Cancer
2006, 6, 107–116.
(2) Jones, P. A.; Baylin, S. B. The epigenomics of cancer. Cell 2007,
128, 683–692.
(3) Baylin, S. B. DNA methylation and gene silencing. Nat. Clin. Pract.
Oncol. 2005, 2, S4-S11.
(4) Yoo, C. B.; Jones, P. A. Epigenetic therapy of cancer: past, present
and future. Nat. ReV. Drug DiscoVery 2006, 5, 37–50.
(5) Cheng, J. C.; Matsen, C. B.; Gonzalez, F. A.; Ye, W.; Greer, S.;
Marquez, V. E.; Jones, P. A.; Selker, E. U. Inhibition of DNA
methylation and reactivation of silenced genes by zebularine. J. Natl.
Cancer Inst. 2003, 95, 399–409.
(6) Yoo, C. B.; Cheng, J. C.; Jones, P. A. Zebularine: a new drug for
epigenetic therapy. Biochem. Soc. Trans. 2004, 32, 910–912.
(7) McCormack, J. J.; Marquez, V. E.; Liu, P. S.; Vistica, D. T.; Driscoll,
J. S. Inhibition of cytidine deaminase by 2-oxopyrimidine riboside
and related compounds. Biochem. Pharmacol. 1989, 29, 830–832.
(8) Frick, L.; Yang, C.; Marquez, V. E.; Wolfenden, R. Binding of
pyrimidin-2-one ribonucleoside by cytidine deaminase as a transition-
state analogue 3,4-dihydrouridine and contribution of the 4-hydroxyl
group to its binding activity. Biochemistry 1989, 28, 9423–9430.
(9) Marquez, V. E.; Kelley, J. A.; Agbaria, R.; Ben-Kasus, T.; Cheng,
J. C.; Yoo, C. B.; Jones, P. A. Zebularine: A unique molecule for an
epigenetically based strategy in cancer chemotherapy. Ann. N.Y. Acad.
Sci. 2005, 1058, 246–254.
(10) Zhou, L.; Cheng, X.; Connolly, B. A.; Dickman, M. J.; Hurd, P. J.;
Hornby, D. P. Zebularine: a novel DNA methylation inhibitor that
forms a covalent complex with DNA methyltransferases. J. Mol. Biol.
2002, 321, 591–599.

ActiVation of p16 Gene by DeriVatiVes of 2′-Deoxyzebularine
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7601
(11) Cheng, J. C.; Weisenberger, D. J.; Gonzalez, F. A.; Liang, G.; Xu,
G.-L.; Hu, Y,-G.; Marquez, V. E.; Jones, P. A. Continuous zebularine
treatment effectively sustains demethylation in human bladder cells.
Mol. Cell. Biol. 2004, 24, 1270–1278.
(12) Cheng, J. C.; Yoo, C. B.; Weisenberger, D. J.; Chuang, J.; Wozniak,
C.; Liang, G.; Marquez, V. E.; Greer, S.; Orntoft, T. F.; Thykjaer, T.;
Jones, P. A. Preferential response of cancer cells to zebularine. Cancer
Cell 2004, 6, 151–158.
(13) Herranz, M.; Caballero, J.-M.; Fraga, M. F.; Ruiz-Cabello, J.; Flores,
J. M.; Marquez, V. E.; Esteller, M. Zebularine is a non-toxic DNA
demethylating drug effective against the development of lymphoma.
Blood 2006, 107, 1174–1177.
(14) Marquez, V. E.; Rao, K. V. R.; Barchi, J. J., Jr.; Kelley, J. A.; Agbaria,
R.; Ben-Kasus, T.; Cheng, J. C.; Yoo, C. B.; Jones, P. A. Zebularine:
A unique molecule for an epigenetically-based strategy in cancer
chemotherapy. The magic of its chemistry and biology. Nucleosides,
Nucleotides Nucleic Acids 2005, 24, 305–318.
(20) Perrone, P.; Daverio, F.; Valente, R.; Rajyaguru, S.; Martin, J. A.;
Le´veˆque, V.; Le Pogam, S.; Najera, I.; Klumpp, K.; Smith, D. B.;
McGuigan, C. First example of phosphoramidate approach applied to
a 4′-substituted purine nucleoside (4′-azidoadenosine): Conversion of
an inactive nucleoside to a submicromolar compound versus hepatitis
C virus. J. Med. Chem. 2007, 50, 5463–5470.
(21) Wildman, S. A.; Crippen, G. M. Prediction of physicochemical
parameters by atomic contributions. J. Chem. Inf. Comput. Sci. 1999,
39, 868–873. (http://iss.nci.nih.gov/cgi-bin/moe-front/moe.tcl).
(22) Eads, C. A.; Danenberg, K. D.; Kawakami, K.; Saltz, L. B.; Danenberg,
P. V.; Laird, P. W. CpG island hypermethylation in human colorectal
tumors is not associated with DNA methyltransferase overexpression.
Cancer Res. 1999, 59, 2302–2306.
(23) Gonzalez-Zulueta, M.; Bender, C. M.; Yang, A. S.; Nguyen, T.; Beart,
R. W.; Van Tornout, J.; Jones, P. Methylation of the 5′-CpG island
of the p16/CDKN2 tumor suppressor gene in normal and transformed
human tissues correlates with gene silencing. Cancer Res. 1995, 55,
4531–4535.
(24) Votruba, I.; Holy, A.; Wightman, R. H. The mechanism of inhibition
of DNA synthesis in Escherichia coli by pyrimidine-2-one ꢀ-D-
ribofuranoside. Biochim. Biophys. Acta 1973, 324, 14–23.
(25) Barchi, J. J., Jr.; Cooney, D. A.; Hao, Z.; Weinberg, M.; Taft, C.;
Marquez, V. E.; Ford, H., Jr. Improved synthesis of zebularine [1-(ꢀ-
D-ribofuranosyl)-dihydroxypyrimidin-2-one] nucleosides as inhibitors
of human deoxycytidylate deaminase. J. Enzyme Inhib. 1995, 9, 147–
162.
(26) Bjursell, G.; Reichard, P. Effects of thymidine on deoxyribonucleoside
triphosphate pools and deoxyribonucleic acid synthesis in Chinese
hamster ovary cells. J. Biol. Chem. 1973, 248, 3904–3909.
(27) Gonzalgo, M. L.; Jones, P. A. Quantitative methylation analysis using
methylation-sensitive single-nucleotide primer extension (Ms-SNuPE).
Methods 2002, 27, 128–133.
(15) Ben-Kasus, T.; Ben-Zvi, Z.; Marquez, V. E.; Kelley, J. A.; Agbaria,
R. Metabolic activation of zebularine, a novel DNA methylation
inhibitor in human bladder carcinoma cells. Biochem. Pharmacol.
2005, 70, 121–133.
(16) Dowd, C.; Sutch, B. T.; Haworth, I. S.; Eritja, R.; Marquez, V. E.;
Yang, A. S. Incorporation of zebularine from its 2′-deoxynucleoside
triphosphate derivative and activity as a template-coding nucleobase.
Nucleosides, Nucleotides Nucleic Acids 2008, 27, 131–145.
(17) Balzarini, J.; Kruining, J.; Wedgwood, O.; Pannecouque, C.; Aquaro,
S.; Perno, C. F.; Naesens, L.; Witvrouw, M.; Heijtink, R.; De Clercq,
E.; McGuigan, C. Conversion of 2′,3′-dideoxyadenosine (ddA) and
2′,3′-didehydro-2′,3′-dideoxyadenosine (d4A) to their corresponding
aryloxyphosphoramidate derivatives markedly potentiates their activity
against human immunodeficiency virus and hepatitis B virus. FEBS
Lett. 1997, 410, 324–328.
(18) McGuigan, C.; Cahard, D.; Sheeka, H. M.; De Clercq, E.; Balzarini,
J. Aryl phosphoramidate derivatives of d4T have improved anti-
HIV efficacy in tissue culture and may act by the generation of a
novel intracellular metabolite. J. Med. Chem. 1996, 39, 1748–1753.
(19) McGuigan, C.; Wedgwood, O. M.; De Clercq, E.; Balzarini, J.
Phosphoramidate derivatives of 2′,3′-didehydro-2,3′-dideoxyadenosine
(d4A) have markedly improved anti-HIV potency and selectivity.
Bioorg. Med. Chem. Lett. 1996, 6, 2359–2362.
(28) Yoo, C.; Chuang, J. C.; Byun, H.-M.; Egger, G.; Yang, A. S.; Dubeau,
L.; Long, T.; Laird, P. W.; Marquez., V. E.; Jones, P. A. Long-term
epigenetic therapy with oral zebularine has minimal side effects and
prevents intestinal tumors in mice. Cancer PreV. Res. 2008, 1,
233–240.
JM8005965