Novel mycophenolic adenine bis(phosphonate) analogues as potential differentiation agents against human leukemia

Article abstract of DOI:10.1021/jm0104116

Novel mycophenolic adenine dinucleotide (MAD) analogues have been prepared as potential inhibitors of inosine monophosphate dehydrogenase (IMPDH). MAD analogues resemble nicotinamide adenine dinucleotide binding at the cofactor binding domain of IMPDH; ho

Full text of DOI:10.1021/jm0104116

J . Med. Chem. 2002, 45, 703-712
703
Novel Mycop h en olic Ad en in e Bis(p h osp h on a te) An a logu es As P oten tia l
Differ en tia tion Agen ts a ga in st Hu m a n Leu k em ia 1
Krzysztof W. Pankiewicz,*,† Krystyna B. Lesiak-Watanabe,†,‡ Kyoichi A. Watanabe,† Steven E. Patterson,†
Hiremagalur N. J ayaram,§ J oel A. Yalowitz,§ Michael D. Miller,§ Michael Seidman,| Alokes Majumdar,|
Gerd Prehna, and Barry M. Goldstein
Pharmasset Inc., 1860 Montreal Road, Tucker, Georgia 30084, Department of Biochemistry and Molecular Biology,
Indiana University School of Medicine, 635 Barnhil Drive, Indianapolis, Indiana 46202, Laboratory of Molecular Genetics,
National Institute of Aging, NIH, 5600 Nathan Shock Drive, Baltimore, Maryland 21224, and Department of Biochemistry and
Biophysics, University of Rochester Medical Center, 601 Elmwood Avenue, Rochester, New York 14642
Received August 30, 2001
Novel mycophenolic adenine dinucleotide (MAD) analogues have been prepared as potential
inhibitors of inosine monophosphate dehydrogenase (IMPDH). MAD analogues resemble
nicotinamide adenine dinucleotide binding at the cofactor binding domain of IMPDH; however,
they cannot participate in hydride transfer and therefore inhibit the enzyme. The methylenebis-
(phosphonate) analogues C2-MAD and C4-MAD were obtained by coupling 2′,3′-O-isopropy-
lideneadenosine 5′-methylenebis(phosphonate) (22) with mycophenolic alcohols 20 and 21 in
the presence of diisopropylcarbodiimide followed by deprotection. C2-MAD was also prepared
by coupling of mycophenolic methylenebis(phosphonate) derivative 30 with 2′,3′-O-isopropy-
lideneadenosine. Compound 30 was conveniently synthesized by the treatment of benzyl-
protected mycophenolic alcohol 27 with a commercially available methylenebis(phosphonic
dichloride) under Yoshikawa’s reaction conditions. C2-MAD and C4-MAD were found to inhibit
the growth of K562 cells (IC50 ) 0.7 µM and IC50 ) 0.1 µM, respectively) as potently as
mycophenolic acid (IC50 ) 0.3 µM). In addition, C2-MAD and C4-MAD triggered vigorous
differentiation of K562 cells an order of magnitude more potently than tiazofurin, and MAD
analogues were resistant to glucuronidation in vitro. These results show that C2-MAD and
C4-MAD may be of therapeutic interest in the treatment of human leukemias.
In tr od u ction
type p53 induces differentiation in a variety of cells,
resulting in G1 arrest and slow doubling time.5 Tumor
suppressor proteins inhibit cell proliferation, not by
killing cells but by activation of tumor cell differentia-
tion.6 Induction of p53 results in reduced IMPDH
expression. Inhibitors of IMPDH such as tiazofurin (1,
TR; Scheme 2) and mycophenolic acid (MPA, 10; Scheme
3) cause cell cycle arrest at the G1/S boundary and
induce differentiation in several cell lines, including HL-
60 and K562 cells.7 Consequently, it appears that the
inhibition of IMPDH mimics the effects of p53 expres-
sion4 and would be of therapeutic importance in cancer
treatment.8
IMPDH (EC 1.1.1.205) catalyzes the rate-limiting step
in the de novo biosynthesis of guanine nucleotides, the
nicotinamide adenine dinucleotide (NAD)-dependent
conversion of inosine 5′-monophosphate (IMP) to xan-
thosine 5′-monophosphate (XMP; Scheme 1). After
substrate addition, nucleophilic attack of an active site
cysteine (Cys-331) on the C2 of IMP base forms a
covalent intermediate. The binding of cofactor NAD
results in hydride transfer to the B-side of the nicoti-
namide ring, and subsequent hydrolysis leads to the
formation of XMP. XMP is then aminated by guanosine
5′-monophosphate (GMP)-syntase in the next step to
form GMP.
Many of the strategies for the treatment of neoplasia
have focused on the prevention of cell proliferation.
Unfortunately, cytotoxic drugs, as well as radiation
therapy, often cause debilitating side effects, and fre-
quently, these therapies fail. An attractive alternative
to the antiproliferative approach is to induce cell dif-
ferentiation in order to convert malignant cells into
mature resting cells. Not all cells are phenotypically able
to differentiate; however, it has been established that
leukemia-derived cell lines, such as the human chronic
myelogenous leukemia K562 and the human promy-
elomonocytic leukemia HL-60, easily undergo drug-
induced differentiation.2 Thus, stimulation of cell dif-
ferentiation is particularly promising for the treatment
of human leukemias.
Cellular differentiation, in both prokaryotes and
eukaryotes, is controlled by the level of guanine nucle-
otides.3 The concentrations of guanine nucleotides in the
cell are in turn regulated by the activities of p53 tumor
suppressor protein4 and inosine monophosphate dehy-
drogenase (IMPDH), the key enzyme in the de novo
synthesis of purine nucleotides. Overexpression of wild-
* To whom correspondence should be addressed. Tel: 678-395-0027.
Fax: 678-395-0030. E-mail: kpankiewicz@pharmasset.com.
† Pharmasset Inc.
Two isoforms of IMPDH are found in mammalian
cells, each encoded by distinct genes.9 Type I is ex-
pressed in normal cells while the levels of type II are
markedly increased in tumor cells and activated lym-
‡ Deceased.
§ Indiana University School of Medicine.
| National Institute of Aging.
University of Rochester Medical Center.
10.1021/jm0104116 CCC: $22.00 © 2002 American Chemical Society
Published on Web 01/08/2002

704 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Pankiewicz et al.
Sch em e 1
Sch em e 2
Sch em e 3
phocytes. When tumor cells are induced to differentiate,
transcripts of type II decline to below those of type I.10
Therefore, it is expected that specific inhibition of
IMPDH type II may induce differentiation while provid-
ing a significant therapeutic advantage by eliminating
toxicity caused by the inhibition of the type I isoform.
The differentiation-inducing activity of potential drugs
is commonly studied in vitro in K562 cells, derived from
a human chronic myeloid leukemia (CML) cell line,
which is a widely used system for differentiation that
can be induced by various agents to produce hemoglo-
bin.11,12 The antitumor agent TR (1) has proved to be
one of the most effective inducers of differentiation in
K562 cells among numerous compounds studied.13 The
biological activity of TR is due to its unique metabolism.
TR is phosphorylated in cells to TR monophosphate (4;
Scheme 2), which is coupled with adenosine 5′-triphos-
phate by nicotinamide mononucleotide (NMN) adenyl-
transferase to form the active metabolite, thiazole-4-
carboxamide adenine dinucleotide (7, TAD).17
TR induced complete hematologic remissions in pa-
tient’s acute myeloid leukemia (AML) and in blast crisis
of CML (CML-BC), but the remissions were of short
duration. Tricot et al.14,15 reported that TR was effective
in patients with CML-BC in clinical trials. The observa-
tion that the therapeutic effect of TR was due to the
induction of cell differentiation in humans was con-
firmed by Wright et al.16 in their Phase II clinical

Mycophenolic Adenine Bis(phosphonate) Analogues
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 705
F igu r e 1.
studies. Unfortunately, TR was found to be too toxic for
a broad clinical application.
TAD (7) is an analogue of NAD (9) in which TR
replaces the nicotinamide riboside moiety of NAD. TAD
mimics the natural cofactor, but the thiazole ring cannot
participate in hydride transfer, resulting in potent
inhibition of both isoforms of IMPDH (Ki ) 0.1 µM).
Unfortunately, TAD, which is a pyrophosphate, is
metabolically unstable. Thus, in vivo, TAD is degraded
to the corresponding mononucleotides and then to the
nucleosides by the action of cellular phosphodiesterases
and phosphatases.18 Resistance to TR is associated both
with impaired accumulation of TAD in resistant cells19
and with excessive degradation of TAD by phospho-
diesterases, including a specific TADase.20 A new in-
hibitor of IMPDH, benzamide riboside (2, BR),21 is
metabolized to benzamide adenine dinucleotide (8,
BAD)22 in a manner similar to the conversion of TR to
TAD. This compound is currently being examined as an
alternative to TR.23
In 1986, a methylenebis(phosphonate) analogue of
TAD (11) was synthesized in which the unstable pyro-
phosphate moiety (P-O-P) was replaced by a P-CH2-P
linkage that is stable to enzymatic and chemical hy-
drolysis.24 This compound inhibited IMPDH as well as
TAD and, additionally, showed good activity in TR-
resistant cell lines. These findings prompted us to
synthesize and examine such bis(phosphonate) inhibi-
tors as potential drugs. We developed25,26 an efficient
method of synthesis of P1, P2-disubstituted bis(phos-
phonate)s and synthesized a number of methylenebis-
(phosphonate) and difluoromethylenebis(phosphonate)
analogues of TAD and BAD27 (11-14; Figure 1). We
found that these compounds are potent inhibitors of
IMPDH, can penetrate cells, and are chemically and
metabolically stable in serum and cell extracts.18,28 We
also found that neither TAD nor BAD (pyrophosphates),
nor their bis(phosphonate) analogues 11-14, show any
specificity against isoforms of IMPDH.29,30
F igu r e 2.
leaves the adenosine 5′-monophosphate (AMP) subsite
of the inhibitor-enzyme complex empty.31 Thus, we
assumed that attachment of AMP to MPA would result
in a mixed anhydride (Figure 2) that extends into the
AMP binding site. This may offer the possibility of
enhancing binding affinity. Furthermore, addition of the
bulky AMP to the side chain of MPA as well as
introduction of the negatively charged phosphorus group
was expected to reduce or prevent the undesired glu-
curonidation of MPA (below).
Despite these potential advantages, a mixed anhy-
dride would not be stable in protic solvents. Therefore,
we replaced the unstable P(O)-O-C(O) anhydro group
of the mixed anhydride with a stable methylenebis-
(phosphonate) [P(O)-CH2-P(O)] linkage. Thus, reduc-
tion of the carboxyl group of MPA afforded the known
mycophenolic alcohol (19, MPAlc; Scheme 2),32 which
was then coupled with 2′,3′-O-isopropylideneadenosine
5′-methylenebis(phosphonate) to give, after deprotec-
tion, the desired bis(phosphonate) analogue of myco-
phenolic adenine dinucleotide (15, C6-MAD; Figure 2).33
This analogue of MAD contains six carbon atoms and
the oxygen in the linker between the aromatic ring of
mycophenolic alcohol and the â-phosphorus atom of the
adenosine bis(phosphonate) moiety. C6-MAD is a potent
inhibitor of IMPDH type II (Ki ) 0.3 µM); however, it
is 30-fold less potent than MPA (Ki ) 0.01 µM).33 Its
linker is two atoms longer than that in the ideal mixed
anhydride.
Compound 10 (MPA) is one of the few inhibitors to
show modest (4-fold) specificity for the type II isoform9
(Ki ∼ 8 and 33 nM for type II and type I, respectively).
The crystal structure of the complex of Chinese hamster
IMPDH with MPA revealed that MPA binds at the
NMN subsite of the cofactor pocket of IMPDH.31 This
In an attempt to optimize the efficacy of this hybrid
compound, we synthesized two MAD analogues with
shorter linkers (16, C4-MAD and 17, C2-MAD; Figure

706 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Pankiewicz et al.
Sch em e 4
2). We report here the syntheses of these analogues and
their ability to inhibit both isoforms of IMPDH. In
addition, we examine both the in vitro cell differentia-
tion-inducing and the growth inhibitory activities of the
MAD analogues in K562 cells. These results are com-
pared with activities of TR (1), BR (2), and the bis-
(phosphonate) analogues of TAD (11 and 12) and BAD
(13) prepared by us earlier.27,30,33
ylidene)propionaldehyde followed by reduction with
NaBH4 afforded the desired alcohol 21.32 The aldehyde
precursor of 21 (see Experimental Section) and conse-
quently the alcohol 21 were obtained exclusively as E
isomers. Irradiation of the aldehyde proton (at 9.36 ppm)
resulted in a significant nuclear Overhauser effect
increasing the intensity of the olefinic proton (triplet
at 6.51 ppm) of the precursor. This indicates that the
olefinic proton and the aldehyde group are in cis
configuration, e.g., E stereochemistry of the double bond.
Resu lts a n d Discu ssion
1. Ch em ica l Syn th esis. In our synthetic strategy,
we used MPA (10) as the starting material for the
preparation of mycophenolic alcohols 20 and 21 with
side chains containing two and four carbon atoms,
respectively (Scheme 3). We modified the known pro-
cedure32 of synthesis of the aldehyde 18. Instead of
ozonolysis of 10, which worked poorly in our hands, we
used a one pot OsO4 oxidation of the double bond of 10
followed by oxidative cleavage (NaIO4). The desired
product precipitated from the reaction mixture in 96%
yield. The product was found to be an approximately
2.5:1 mixture of the aldehyde 18 and its hemiacetal 18a
formed by intramolecular attack of the neighboring
7-OH group. The 1H nuclear magnetic resonance (NMR)
of the product showed all resonances characteristic for
both 18 and 18a . The ratio of 18:18a was determined
by a relative integration of the resonance signals of the
corresponding 4-methyl groups of both compounds.
Reduction of the mixture of 18 and 18a with NaBH4
gave the alcohol derivative 20 in high yield, whereas
treatment of the mixture with 2-(triphenylphosphoran-
The coupling of 20 with the 2′,3′-O-isopropylidene-
adenosine 5′-methylenebis(phosphonate) (22; Scheme 4)
was accomplished by treatment of 22 with 3.5 equiv of
diisopropylcarbodiimide (DIC) in dry pyridine followed
by the addition of the mycophenolic alcohol 20 according
to our method for preparation of P1, P2-disubstituted
methylenebis(phosphonate)s.25,26 Thus, dehydration of
22 afforded the active bicylic trisanhydride intermediate
24, which under reaction with 20 gave a product 25.
Hydrolysis of 25 with water afforded the desired deriva-
tive 26 in 65% yield, after high-performance liquid
chromatography (HPLC) separation. A small amount of
unreacted tetraphosphonate derivative 23 (5% yield)
was also obtained along with some other unidentified
products. Compound 23 can be recycled, e.g., used as a
starting material for the reaction with DIC and 20.
Deisopropylidenation of 26 with trifluoroacetic acid
afforded the desired C2-MAD (17).
The assignment of the structure of 17 was confirmed
1
1
by H and 31P NMR. The H NMR resonance signal of

Mycophenolic Adenine Bis(phosphonate) Analogues
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 707
Sch em e 5
Ta ble 1. Inhibition of IMPDH and K562 Cell Growth by MPA
(10), TR (1), BR (2), and Methylenebis(phosphonate) Analogues
of TAD (11, 12), BAD (13), and MAD (15-17)
the -CH2- of the side chain at 3.73 ppm was recorded
as a quartet (J H-P ) 7.0 Hz, J H-H ) 7.0 Hz) showing
the coupling with the â-phosphorus atom of the bis-
(phosphonate) moiety. Heteronuclear shift correlation
experiments also confirmed the P-CH2 coupling with
phosphorus.
IMPDH
type I
Ki (µM)
IMPDH
type II
Ki (µM)
K562 cell
growth
IC50 (µM)
inhibitors
MPA (10)
TR (1)
BR (2)
TAD (11)
TAD (12)
BAD (13)
C6-MAD (15)
C4-MAD (16)
C2-MAD (17)
bis(phosphonate) 31
bis(phosphonate) 33
0.04
0.01
0.3
3.0
1.5
18.0
11.0
60.0
1.5
0.1
0.7
43.2
2089.7
Because mycophenolic alcohol 20 contains a free
phenolic group at C7, the formation of 7-O-linked
regioisomer and/or double-substituted derivative (at the
O7 and the C2′) could be expected. Therefore, the
protection of the phenol function may be necessary in
order to improve efficiency of the coupling procedure.
We used a benzyl group for this protection. The reaction
of mycophenolic alcohol 20 with benzyl bromide in a 1
M solution of tetrabutylammonium fluoride in tetrahy-
drofuran (THF) afforded the desired 7-O-benzyl deriva-
tive 27 (Scheme 5) in crystalline form and high yield.
Contrary to expectations, we found that the coupling of
adenosine bis(phosphonate) 22 with the protected my-
cophenolic alcohol 27 required a longer reaction time
and the yield was similar to that observed for the
coupling with the unprotected compound 20.
We also examined an alternative synthetic strategy
by using 7-O-benzyl derivative 27 for the preparation
of the methylenebis(phosphonate) analogue 30 and the
subsequent coupling of 30 with 2′,3′-O-isopropylidene-
adenosine (Scheme 5). Mesylation of alcohol 27 afforded
a good yield of mesylate 28. However, further reaction
with the tris(tetra-n-butylammonium) salt of methyl-
enebis(phosphonic) acid according to Poulter’s proce-
dure34 gave exclusively the elimination product 29.
Nevertheless, the desired bis(phosphonate) was ob-
tained in very good yield by the reaction of 27 with
commercially available methylenebis(phosphonic dichlo-
ride) in conditions of Yashikawa’s phosphorylation in
(EtO)3P(O). Although a small amount of bis(mycophe-
nolic)bis(phosphonate) derivative 32 (e.g., the P1, P2-
disubstituted analogue) was also formed in the reaction,
compounds 30 and 32 were easily separated by pre-
parative HPLC. The benzyl protecting group was then
removed quantitatively from 30 and 32 by hydrogenoly-
sis on Pd/C to give bis(phosphonate) derivatives 31 and
33, respectively. These compounds were also examined
as potential inhibitors of IMPDH.
not active
not active
0.11
0.30
0.66
0.33
0.52
0.33
0.89
not active
not active
0.11
0.30
0.95
0.29
0.38
0.26
0.40
2.3
0.47
Finally, DIC coupling of the 7-O-benzyl-protected bis-
(phosphonate) derivative 30 with the 2′,3′-O-isopropy-
lideneadenosine (iPrAd) afforded, after hydrogenolytic
debenzylation (Pd/C) and deisopropylidenation (CF3-
COOH/H2O), the desired analogue 17. We found, how-
ever, that the overall yield of this new procedure was
not superior to that depicted in Scheme 4. Both syn-
thetic approaches were almost equally efficient.
The similar coupling of the 7-unprotected mycophe-
nolic alcohol derivative 21 with bis(phosphonate) 22
gave, after hydrolysis and deisopropylidenation with
CF3COOH/H2O, the desired C4-MAD (16; Figure 2) in
moderate yield. The 1H NMR resonance signal of the
-CH2- group of the side chain at 4.20 ppm was
recorded as a doublet (J H-P ) 7.0 Hz) showing the
coupling with the â-phosphorus atom of the bis(phos-
phonate) moiety. The similar P-H coupling confirming
the attachment of the bis(phosphonate) moiety to the
-CH2- group of the side chain of the mycophenolic
alcohol in compounds 30-34 has been also recorded.
2. Biologica l Activity. The MAD analogues 16 and
17 as well as the bis(phosphonate)s 31 and 33 were
assayed for inhibitory activity against human IMPDH
type I and type II. The pattern of inhibition in each case
is uncompetitive. Ki values are given in Table 1. It is
worthwhile to note that Km values for NAD are much
higher and similar for the type I and the type II
isoforms, -46 and 32 µM, respectively.9 Antiprolifera-

708 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Pankiewicz et al.
tive activity of MAD analogues against K562 cells was
also evaluated, and the corresponding IC50 values are
given in Table 1. For comparison, previously published
Ki and IC50 values of MPA (10), TR (1), BR (2), and the
bis(phosphonate) analogues of TAD (11, 12) and BAD
(13) are also presented in Table 1.27,30
Comparison of these data indicates that MAD ana-
logues are potent inhibitors of IMPDH but show little
specificity against the two isoforms of the human
enzyme. Regardless of the linker size, MAD analogues
are an order of magnitude less potent than MPA.
Contrary to expectations, the attachment of the adenos-
ine moiety to MPA did not enhance binding affinity.
Indeed, a bis(phosphonate) derivative 31, which does
not contain the adenosine moiety, binds IMPDH with
comparable affinity to the dinucleotide analogues. In-
terestingly, this compound still showed a 2-fold better
inhibitory activity against IMPDH type II than type I.
Surprisingly, bis(mycophenolic)bis(phosphonate) ana-
logue 33, which contains a mycophenolic moiety at-
tached to the P1 and the P2, showed a similar selectivity
(5-fold) in the inhibition of the type II isoform to that of
MPA (4-fold).
In retrospect, the fact that the MAD analogues do not
show improved binding over MPA is perhaps not
surprising. These compounds have more degrees of
freedom than MPA. Thus, any gain in binding enthalpy
is likely offset by the entropic penalty of constraining
the longer dinucleotide analogue. The penalty for the
constraint of even a single rotational degree of freedom
in a ligand can be severe.35 Furthermore, Digits and
Hedstrom36 have shown recently that binding between
the nicotinamide and the adenosine 5′-diphosphate
components of the NAD site in human IMPDH type II
is not synergistic.
F igu r e 3.
sulted in inactive compounds.39 In the form of a prodrug,
mycophenolic mofetil (MMF), MPA is used in clinics as
an immunosuppressant. Although there is no glucu-
ronidation in activated lymphocytes, high doses (2-3 g
per day) of MMF are needed in order to maintain the
therapeutic level of MPA. Because the level of glucuron-
yltransferase activity is rather high in many cancer cells
(in fact, human colorectal carcinoma cells were used in
vitro as a means to assess the metabolism of analogues
of MPA),39 it is not surprising that MPA lacks any
significant anticancer activity.
We found that in contrast to MPA, MAD analogues
were resistant to glucuronidation in vitro.33 The incuba-
tion of MAD analogues or MPA with uridine 5′-diphos-
phoglucuronyltransferase and uridine 5′-diphosphoglu-
curonic acid resulted in extensive (70%) glucuronidation
of MPA, whereas no formation of the corresponding
MAD glucuronides was observed under similar condi-
tions.33 These results indicate that if MAD analogues
are not glucuronidated in vivo, they would have a good
therapeutic potential against cancer.
Interestingly, MAD analogues, specifically C2-MAD
and C4-MAD, were also found to be potent inducers of
K562 cell differentiation (Table 2), an order of magni-
tude more potent than TR. Taken together, their
resistance to glucuronidation, potent inhibition of cell
growth, and differentiation-inducing abilities make the
MAD analogues potentially valuable antineoplastic
agents. Synthesis of gram amounts of the above meth-
ylenebis(phosphonate) analogues of MAD for pharma-
cokinetic and toxicological studies is now in progress.
3. Con clu sion s. Known inhibitors of IMPDH such
as MMF, mizoribine, and ribavirin were introduced to
the drug market some time ago, and they continue to
serve as clinically useful immunosuppressants or anti-
viral drugs. However, no anticancer drug, based on the
inhibition of IMPDH, has been approved for clinical use.
The data in Table 1 also show that MAD analogues
are approximately as potent inhibitors of IMPDH as
methylenebis(phosphonate) analogues of TAD and BAD
(11-13). However, MAD analogues were found to be
superior to the bis(phoshonate)s of TAD and BAD 11-
13 in the inhibition of the growth of K562 leukemic cells
by 2 orders of magnitude. Compounds 11-13 were less
effective than their parent nucleosides (1 or 2) in the
inhibition of K562 cells growth (IC50 values). In contrast,
the most potent MAD analogue 16 shows more potent
growth inhibition than MPA itself, although it is less
active than MPA as an inhibitor of the isolated enzyme.
This is probably due to the efficient (active) transport
of MAD analogues into the cells. Indeed, all MAD
analogues inhibit cell growth much more potently than
analogues of TAD and BAD, or even TR.
Although MPA is the most potent inhibitor of IMPDH,
it is not active against cancer.37 This drug’s therapeutic
potential is limited by its undesirable metabolism. In
humans, MPA is rapidly metabolized (via its C7-
phenolic function) into the inactive glucuronide (MPAG;
Figure 3), and as much as 90% of the drug circulates in
this inactive form, which is then excreted into the bile.38
In addition, MPAG is deglucuronidated by colonic
bacteria, and MPA is adsorbed in the lower gastrointes-
tinal track causing local damage to the intestinal
epithelium. Protection of the C7 function or replacement
of the phenol group with a fluorine atom, amino, or
nitrile group in order to prevent glucuronidation re-

Mycophenolic Adenine Bis(phosphonate) Analogues
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 709
Ta ble 2. Effect of TR (1), TAD (7), Methylenebis(phosphonate)
Analogues of TAD and BAD (11-13), and MAD Analogues
15-17 on Differentiation of K562 Cells
and a solution of OsO4 (250 mg) in acetone (10 mL) was added.
The reaction mixture was stirred for 3.5 h, and the disappear-
ance of MPA was monitored by TLC with CH2Cl2-MeOH
(20%) as the eluent. The mixture was diluted with water (300
mL) and cooled in an ice bath, and a solution of NaIO4 (21.4
g) in water (200 mL) was added in portions. The formed
precipitate was filtered, washed with cold water (3 × 50 mL),
and dried to give a 2.5:1 mixture of 18 and 18a (6.8 g, 96.0%)
% of benzidine
concn % of benzidine
positive cells
inhibitors
TR (1)
TAD (7)
TAD (11)
TAD (12)
BAD (13)
C6-MAD (15)
C4-MAD (16)
C2-MAD (17)
bis(phosphonate) 31
(µM)
positive cells
at 1.0 µM
15.0
14.0
14.3
14.0
15.0
10
1.5
1.6
20.0
80.0
35.0
35.0
48.0
9.5
90.0
95.0
65.4
67.0
4.4
5.3
2.5
2.4
3.4
0.6
1
as a white solid. H NMR (CDCl3): δ 2.04 (18a , s, 3H, CH3),
2.16 (18, s, 3H, CH3), 3.29 (18a , dd, 1H, H1, J 1-2 ) 2.0 Hz,
J 1-1′ ) 16.2 Hz), 3.52 (18a , dd, 1H, H1′, J 1′-2 ) 6.5 Hz), 3.72
(18, s, 3H, OCH3), 3.76 (18, d, 2H, CH2, J 1-2 ) 6.0 Hz), 3.98
(18a , s 3H, OCH3), 5.13 (18a , s, 2H, CH2-lactone), 5.23 (18,
s, 2H, CH2-lactone), 6.31 (18a , dd, 1H, H2), 7.71 (1, s, 1H,
7-OH, exchange), 9.75 (18, t, 1H, aldehyde). Anal. (C12H12O5)
C, H.
9.0
63.3
40.9
3.3
bis(phosphonate) 33 100.0
0.4
7-H yd r oxy-6-(2-h yd r oxyet h yl)-5-m et h oxy-4-m et h yl-
p h th a la n -1-on e (20). The solution of a mixture of 18 and 18a
(10 mmol, 2.36 g) in EtOH (100 mL) was treated with NaBH4
(1.0 g) in portions over 0.5 h. The mixture was kept at room
temperature until TLC (a small sample for TLC analysis was
neutralized with 3 N HCl, extracted with EtOAc, and chro-
matographed with CH2Cl2 (2%) as the eluent) showed complete
reduction (approximately 2 h). The mixture was concentrated,
and the residue was treated with 3 N HCl (50 mL) and
extracted with EtOAc to give after concentration the alcohol
20 (2.36 g). 1H NMR (CDCl3): δ 2.15 (s, 3H, CH3), 2.98 (t, 2H,
CH2, J ) 6.4 Hz), 3.79 (s, 3H, OCH3), 3.85 (t, 2H, CH2, J ) 6.4
Hz), 5.20 (s, 2H, CH2-actone), 7.71 (s, 1H, 7-OH, exchange).
Anal. (C12H14O5) C, H.
7-Hydr oxy-6-(4-h ydr oxy-3-m eth ylbu t-2-en yl)-5-m eth oxy-
4-m eth ylp h th a la n -1-on e (21). The mixture of 18 and 18a
(4.0 g) and Ph3PdC(Me)CHO (6.0 g) in benzene (300 mL) was
heated under reflux for 24 h. The reaction mixture was
concentrated, and the residue was chromatographed on a silica
gel column with n-hexanes-EtOAc (20%), followed by n-hex-
anes-EtOAc (30%) affording a white solid, which was crystal-
lized from EtOH to give an aldehyde (3.6 g, 77%). 1H NMR
(CDCl3): δ 1.90 (s, 3H, CH3-side chain), 2.15 (s, 3H, CH3),
3.71 (d, 2H, CH2, J ) 7.0 Hz), 3.77 (s, 3H, OCH3), 5.21 (s, 2H,
CH2-lactone), 6.51 (t, 1H, H2), 7.75 (s, 1H, 7-OH, exchange),
9.36 (s, 1H, aldehyde).
A solution of the aldehyde (3.6 g) in EtOH (220 mL) was
treated with NaBH4 (900 mg) added in portions over 45 min.
The reaction mixture was kept overnight, concentrated, and
treated with a mixture of 3 N HCl and EtOAc. The organic
layer was washed with water and concentrated, and the
residue was chromatographed on a silica gel column with CH2-
Cl2-acetone (2.5%) as the eluent to give a white solid.
Crystallization from benzene-petrol ether afforded 21 (3.4 g).
1H NMR (CDCl3): δ 1.83 (s, 3H, CH3-side chain), 2.14 (s, 3H,
CH3), 3.43 (d, 2H, CH2 (1), J ) 7.0 Hz), 3.78 (s, 3H, OCH3),
3.98 (s, 2H, CH2 (4)), 5.19 (s, 2H, CH2-lactone), 5.48 (t, 1H,
H2, J ) 7.0 Hz), 7.68 (s, 1H, 7-OH, exchange). Anal. (C15H18O5)
C, H.
P 1-(2′,3′-O-Isopr opyliden eaden osin -5′-yl)-P 2-[7-h ydr oxy-
6-(2-h yd r oxyeth yl)-5-m eth oxy-4-m eth ylp h th a la n -1-on e-
2-yl]m eth ylen ebis(p h osp h on a te) (26). To a solution of 22
(226 mg, 0.34 mmol) in pyridine (2 mL) was added DIC (156
mL, 2 mmol), and the mixture was left overnight until the
intermediate 24 was formed (multisignal resonances in 31P
NMR).25 Compound 20 (95 mg, 0.4 mmol) was then added, and
the reaction mixture was heated at 55-60 °C for 14 h and at
70 °C for 6 h. At that time, the 31P NMR of the reaction mixture
showed two broad signals at 8 and 25 ppm, characteristic for
the presence of tetrasubstituted derivative 25. Then, a mixture
of water (400 µL) and Et3N (200 µL) was added, and the
reaction mixture was kept at 75 °C for 30 h. HPLC purification
afforded a faster migrating tetraphosphonate 23 (19 mg, 5%),
1H NMR identical to that of the original sample,25 and a slower
migrating compound 26 (200 mg, 65%) as the triethylammo-
nium salt. 1N NMR (D2O): δ 1.24 (t, 18H, Et3N), 1.35 and 1.60
(two s, 3H each, isopropylidene) 1.72 (s, 3H, CH3), 2.17 (t, 2H,
P-CH2-P, J ) 19.8 Hz), 2.65 (t, 2H, CH2 (1), J ) 7.2 Hz),
TR is still in clinical trials; however, it was found to be
too toxic for a broad clinical application. MAD analogues
show interesting potential as anticancer drug candi-
dates. Their resistance to glucuronidation and potent
differentiation-inducing activity indicate that MAD
analogues may be of therapeutic interest in the treat-
ment of human leukemias.
MPA, MAD analogues, and other bis(phosphonate)s
discussed herein do not show significant selectivity
against an isoform of the human enzyme. However, an
extension of the inhibitor into the AMP site would offer
the advantage to design the isoform-specific inhibitor.
The crystal structure of the type II isoform complexed
with the NAD analogue selenazole-4-carboxamide ad-
enine dinucleotide revealed the location of the adenosine
end of the cofactor binding site.40 In this region, the
adenine moiety interacts with several residues that are
not conserved between isoforms. Thus, modification of
the adenine moiety of the MAD analogue would allow
for exploitation of these isoform-dependent interactions
to enhance specificity for the type II isoform. The
synthesis of such analogues is now under way in our
laboratory.
Exp er im en ta l Section
Gen er a l Meth od s. Preparative HPLC was performed on
a Dynamax-300A C18-83-243-C column with a flow rate 20 or
25 mL/min of 0.1 M triethylammonium bicarbonate (TEAB)
by a linear gradient of 0.1 M TEAB-aqueous MeCN (70%).
The homogeneity of dinucleotides was determined by analyti-
cal HPLC (reverse phase column) in the linear gradient and/
or in the isocratic mixture of 0.1 M TEAB-aqueous MeCN
(70%) in the ratio of 4:1. Analytical thin-layer chromatography
(TLC) was performed on Analtech Uniplates with short
wavelength UV light for visualization. Column chromatogra-
phy was performed on silica gel G60 (70-230 mesh, ASTM,
Merck). NMR spectra were recorded on a J EOL Eclipse EX-
270 or Varian 400 MHz instrument. Chemical shifts are
reported in ppm (δ), and signals are described as s (singlet), d
(doublet), dd (double doublet), t (triplet), q (quartet), m
(multiplet), or bs (broad singlet). Values given for coupling
constants are first-order. Chemical shifts for 31P NMR are
referred to H3PO4. All exchangeable protons were detected by
disappearance on the addition of D2O. Elementary analyses
were performed by Atlantic Microlab, Inc., Norcross, GA.
Negative-ion fast atom bombardment mass spectroscopy
was performed by Emory University Mass Spectrometry
Center.
7-H yd r oxy-6-(for m ylm e t h yl)-5-m e t h oxy-4-m e t h yl-
p h th a la n -1-on e (18) a n d Its Hem ia ceta l 18a . MPA (30
mmol, 9.60 g) and 4-methylmorfoline-N-oxide (60 mmol, 7.03
g) was dissolved in a 3:1 (v/v) mixture of THF/H2O (120 mL),

710 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Pankiewicz et al.
3.08 (q, 12H, Et3N), 3.52 (s, 3H, OCH3), 3.75 (q, 2H, CH2 (2),
J H,H ) 7.0 Hz, J H,P ) 7.0 Hz), 4.07 (m, 1H, H5′, J 5′,5” ) 12.4
Hz), 4.10 (m, 1H, H5′′), 4.19 (m, 1H, H4′), 4.34 (t, 1H, H3′,
J 2′,3′ ) 4.8 Hz, J 3′,4′ ) 5.8 Hz), 4.49 (t, 1H, H2′, J 1′,2′ ) 4.4 Hz),
4.79 (s, 2H, CH2-lactone), 5.77 (d, 1H, H1′), 7.90, 8.21 (two
1H singlets, H2, H8-adenine). 31P NMR (D2O): δ 16.20-17.50
(AB system, J ) 13.8 Hz). MS (FAB): m/z 684 (M - H)- 688
(100%) and the sodium adduct (644 + 22).
ylp h th a la n -1-on e (28). To a mixture of compound 27 (1.95
g, 6.0 mmol), Et3N (1.67 mL, 12.0 mmol), and 4-(dimethylami-
no)pyridine (DMAP, 732 mg, 6.0 mmol) in CH2Cl2 (15 mL), a
solution of mesyl chloride (MsCl, 930 µL, 12.0 mmol) in CH2-
Cl2 (5 mL) was added dropwise. The reaction mixture was
stirred for 10 min, diluted with MeOH (20 mL), and concen-
trated. The residue was chromatographed on a silica gel
column with hexane-EtOAc (30%), followed by hexanes-
EtOAc (40%), and then hexanes-EtOAc (50%) as the eluent
P 1-(Ad en osin -5′-yl)-P 2-[7-h yd r oxy-6-(2-h yd r oxyeth yl)-
5-m eth oxy-4-m eth yl-p h th a la n -1-on e-2-yl]m eth ylen ebis-
(p h osp h on a te) (17, C2-MAD). (A) From 26. Compound 26
(90 mg, 0.1 mmol, as the triethylammonium salt) was dissolved
in a mixture of MeOH (4 mL) and water (4 mL) containing
CF3COOH (2 mL). The mixture was kept at room temperature
for 1 h and was heated at 50 °C for 30 min. Methanol was
removed in vacuo, and the mixture was diluted with water (8
mL), neutralized with concentrated ammonia, and concen-
trated. The residue was chromatographed on HPLC to give
17 as the trietylammonium salt. This compound was converted
into a sodium salt (57 mg, 83%) by passing through a column
of Dowex 50WX8/Na+ form. 1H NMR (D2O): δ 1.69 (s, 3H,
CH3), 2.07 (t, 2H, P-CH2-P, J ) 20.0 Hz), 2.61 (t, 2H, CH2
(1), J ) 6.8 Hz), 3.52 (s, 3H, OCH3), 3.73 (q, 2H, CH2 (2), J H.H
) 6.4 Hz, J H.P ) 6.4 Hz), 4.02 (m, 1H, H5′, J 5′,5” ) 12.4 Hz),
1
to give 28 (2.2 g, 91%) as a white solid. H NMR (CDCl3): δ
2.20 (s, 3H, CH3), 2.82 (s, 3H, Ms), 3.04 (t, 2H, CH2 (1), J )
7.2 Hz), 3.80 (s, 3H, OCH3), 4.25 (t, 2H, CH2 (2), J ) 7.2 Hz),
5.18 (s, 2H, CH2-lactone), 5.32 (s, 2H, CH2Ph), 7.35-7.49 (m,
5H, Ph). Anal. (C20H22O7S), C, H.
7-Ben zyloxy-6-(2-eth en e)-5-m eth oxy-4-m eth ylph th alan -
1-on e (29). A mesyl derivative 28 (404 mg, 1.0 mmol) was
dissolved in CH3CN (1 mL) and treated with the tris(tetrabu-
tylammonium) salt of methylenebis(phosphonic) acid according
to Poulter’s procedure.34 The reaction mixture was partitioned
between EtOAc-water, and the organic layer was separated,
dried, and concentrated to give 29 as a white solid (200 mg,
91%). 1H NMR (CDCl3): δ 2.19 (s, 3H, CH3), 3.74 (s, 3H,
OCH3), 5.16 and 5.19 (two 2H s, CH2-lactone, CH2Ph), 5.54
(dd, 1H, H1, J 1-2 ) 2.4 Hz, J 1-2′ ) 12.0 Hz), 6.11 (dd, 1H, H2,
J 2-2′ ) 18.0 Hz), 6.79 (dd, 1H, H2′), 7.35-7.55 (m, 5H, Ph).
Anal. (C19H18O4) C, H.
4.10 (m, 1H, H5′′), 4.17 (m, 1H, H4′), 4.34 (t, 1H, H3′, J 2′,3′
)
4.8 Hz, J 3′,4′ ) 5.8 Hz), 4.47 (t, 1H, H2′, J 1′,2′ ) 4.4 Hz), 4.74 (s,
2H, CH2-lactone), 5.77 (d, 1H, H1′), 7.85, 8.21 (two 1H
singlets, H2, H8-adenine). 31P NMR (D2O): δ 17.20-17.50
two overlapping doublets. MS (FAB): m/z 644 (M - H)-
(55%).
[7-Ben zyloxy-6-(2-h yd r oxyet h yl)-5-m et h oxy-4-m et h -
ylp h t h a la n -1-on e-2-yl]m et h ylen eb is(p h osp h on ic) Acid
(30) a n d P 1-[7-Ben zyloxy-6-(2-h yd r oxyeth yl)-5-m eth oxy-
4-m e t h y lp h t h a la n -1-o n e -2-y l]-P 2 -[7-b e n zy lo x y -6-(2-
h yd r oxyeth yl)-5-m eth oxy-4-m eth ylp h th a la n -1-on e-2-yl]-
m eth ylen ebis(p h osp h on a te) (32). A cold solution of 27 (587
mg, 1.65 mmol) in triethyl phosphate (2 mL) was added to a
cold solution of methylenebis(phosphonic) dichloride (450 mg,
1.8 mmol) in triethyl phosphate (2 mL), and the reaction
mixture was kept in the refrigerator overnight and then at
room temperature for 4 h, and the reaction mixture was added
dropwise into the solution of 1 M TEAB (16 mL). The mixture
was diluted with water and extracted with EtOAc, and the
water layer was concentrated. The residue was chromato-
graphed on an HPLC column to give a faster migrating bis-
(phosphonic) acid 30 (985 mg, 87%) as a triethylammonium
salt. A small sample of this compound was converted into a
sodium salt by passing through a column of Dowex 50/Na+
(B) From 34. Compound 34 (94 mg, 0.1 mmol) was dissolved
in MeOH (20 mL), palladium on activated carbon (10%) was
added, and the mixture was shaken under a hydrogen atmo-
sphere (20 psi) in a Parr hydrogenation apparatus for 1 h. The
mixture was filtered and concentrated in vacuo to give a white
solid (84 mg). This compound was treated with a mixture of
MeOH-H2O-CF3COOH as described for compound 26 to give
17 (62 mg, 90%), identical to the compound obtained from
26.
P 1-(Ad en osin -5′-yl)-P 2-[7-h yd r oxy-6-(4-h yd r oxy-3-m e-
th ylbu t-2-en yl)-5-m eth oxy-4-m eth ylp h th a la n -1-on e-4-yl]-
m et h ylen eb is(p h osp h on a t e) (16, C4-MAD). In a similar
manner as described for the synthesis of 17, compound 21 (124
mg, 0.5 mmol) was coupled with bis(phosphonate) 22 (300 mg,
0.45 mmol, triethylammonium salt). The hydrolysis and
formation of the isopropylidene-protected derivative (major
compound) were monitored by analytical HPLC. The reaction
mixture was concentrated and treated with a mixture of
MeOH-H2O-CF3COOH as described for compound 26. HPLC
purification followed by conversion into the sodium salt
afforded 16 (148 mg, 45%). 1H NMR (D2O): δ 1.78 (s, 3H, CH3-
side chain), 2.08 (s, 3H, CH3), 2.14 (t, 2H, P-CH2-P, J ) 19.5
Hz), 3.27 (d, 2H, CH2 (1), J ) 7.0 Hz), 3.68 (s, 3H, OCH3),
4.12 (m, 2H, H5′,5′′), 4.20 (d, 2H, CH2 (4), J H-P ) 6.8 Hz), 4.27
(M, 1H, H4′), 4.42 (m, 1H, H3′), 4.65 (m, 1H, H2′), 5.19 (s, 2H,
CH2-lactone), 5.28 (t, 1H, (2), J ) 7.0 Hz), 5.81 (d, 1H, H1′,
J 1′-2′ ) 5.0 Hz), 8.00 and 8.35 (2 1H singlets, H2, H8). 31P NMR
(D2O): δ 17.80 and 18.10 (AB system, J ) 11.5 Hz). MS
(FAB): m/z 684 (M - H)-.
1
form. H NMR (D2O): δ 1.85 (t, 2H, P-CH2P, J ) 19.6 Hz),
2.07 (s, 3H, CH3), 2.76 (t, 2H, CH2 (1), J 1-2 ) 6.8 Hz), 3.69 (s,
3H, OCH3), 3.79 (q, 2H, CH2 (2), J H,H ) J H,P ) 6.8 Hz), 5.06
and 5.20 (two 2H s, CH2-lactone, CH2Ph), 7.29-7.37 (m, 5H,
Ph). 31P NMR (D2O): δ 15.84 (d, 1P, J ) 7.7 Hz), 17.71 (d,
1P). MS (FAB): m/z 485 (M - H)-.
The slower migrating 32 (73 mg, 9%) was then eluted and
converted into a sodium salt. 1H NMR (D2O): δ 1.96 (s, 3H,
CH3), 2.04 (t, 2H, P-CH2P, J ) 20.0 Hz), 2.82 (t, 2H, CH2 (1),
J ) 6.8.0 Hz), 3.69 (s, 3H, OCH3), 3.88 (q, 2H, CH2 (2), J H,H
)
J H.P ) 6.8 Hz), 4.87 and 4.92 (two 2H s, CH2-lactone, CH2-
Ph), 7.11-7.26 (m, 5H, Ph). 31P NMR (D2O): δ 17.21 (s).
[7-H yd r oxy-6-(2-h yd r oxyet h yl)-5-m et h oxy-4-m et h yl-
ph th alan -1-on e-2-yl]m eth ylen ebis(ph osph on ic) Acid (31).
Compound 30 (69 mg, 0.1 mmol) was dissolved in MeOH (20
mL), palladium on activated carbon (10%) was added, and the
mixture was shaken under a hydrogen atmosphere (20 psi) in
a Parr hydrogenation apparatus for 1 h. The mixture was
filtered and concentrated in vacuo, and the residue was
dissolved in water and passed through a column of Dowex
7-Ben zyloxy-6-(2-h yd r oxyeth yl)-5-m eth oxy-4-m eth yl-
p h th a la n -1-on e (27). A solution of compound 20 (3.0 g, 12.7
mmol,) in 1 M Bu4NF in THF (26 mL) was treated with benzyl
bromide (1.65 mL). The mixture was stirred for 1 h and
analyzed by TLC with CH2Cl2-EtOH (3%). An additional
amount of benzyl bromide (0.3 mL) was added to complete the
reaction. After saturated NaHCO3 was added, the mixture was
extracted with EtOAc, washed with water, dried (Na2SO4), and
concentrated. The residue was chromatographed on a silica
gel column with CH2Cl2-EtOH (1%) to give 27 (4. 0 g, 97%)
1
50WX8/Na+ form to give 31 (45 mg, 99%). H NMR (D2O): δ
2.07 (t, 2H, P-CH2-P, J ) 19.6 Hz), 2.07 (s, 3H, CH3), 2.76
(t, 2H, CH2 (1), J 1-2 ) 7.0 Hz), 3.69 (s, 3H, OCH3), 3.79 (q,
2H, CH2 (2′), J H,H ) J H,P ) 7.0 Hz), 5.20 (s, 2H, CH2-lactone,
CH2Ph). 31P NMR (D2O): δ 15.84 (d, 1P, J ) 7.7 Hz), 17.71 (d,
1P). MS (FAB): m/z 395 (M - H)- (10%); the sodium adduct:
m/z 417 (5%).
1
as a white solid. H NMR (CDCl3): δ 2.20 (s, 3H, CH3), 2.90
(t, 2H, CH2 (1), J ) 6.5 Hz), 3.70 (t, 2H, CH2 (2), J ) 6.5 Hz),
3.79 (s, 3H, OCH3), 5.17 (s, 2H, CH2-lactone), 5.32 (s, 2H, CH2-
Ph), 7.39-7.49 (m, 5H, Ph). Anal. (C19H20O5) C, H.
P 1-[7-Hyd r oxy-6-(2-h yd r oxyeth yl)-5-m eth oxy-4-m eth -
ylp h th a la n -1-on e-2-yl]-P 2-[7-h yd r oxy-6-(2-h yd r oxyeth yl)-
5-m et h oxy-4-m et h ylp h t h a la n -1-on e-2-yl]m et h ylen eb is-
7-Ben zyloxy-6-(2-m esyloxyet h yl)-5-m et h oxy-4-m et h -

Mycophenolic Adenine Bis(phosphonate) Analogues
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 711
(p h osp h on a te) (33). Compound 32 (101 mg, 0.1 mmol) was
dissolved in MeOH (20 mL), palladium on activated carbon
(10%) was added, and the mixture was shaken under a
hydrogen atmosphere (20 psi) in a Parr hydrogenation ap-
paratus for 1 h. The mixture was filtered and concentrated in
vacuo, and the residue was dissolved in water and passed
through a column of Dowex 50WX8/Na+ form to give 33 (65
mg, 95%). 1H NMR (D2O): δ 2.10 (s, 3H, CH3), 2.14 (t, 2H,
P-CH2-P, J ) 16.6 Hz), 2.93 (t, 2H, CH2 (1), J ) 7.2 Hz),
3.85 (s, 3H, OCH3), 4.03 (q, 2H, CH2 (2), J H,H ) J H,P ) 7.2 Hz),
5.11 (s, 2H, CH2-lactone). 31P NMR (D2O): δ 17.72 (s). MS
(FAB): m/z 615 (M - H)- (55%) and the sodium adduct 637
(100%).
from NIH and NSF (Emory University Mass Spectrom-
etry Center).
Refer en ces
(1) NAD analogues 20. For part 19, see: Pankiewicz, K. W.;
Patterson, S. E.; J ayaram, H. N.; Goldstein, B. M. Cofactor
analogues as inhibitors of IMP dehydrogenase: Design and new
synthetic approaches. In Inosine Monophosphate Dehydrogenase:
Perspectives on a Major Therapeutic Target; Goldstein, B. M.,
Pankiewicz, K. W., Eds.; ACS Symposium Series, in press.
(2) Beere, H. M.; Hickman, J . A. Differentiation: a suitable strategy
for cancer chemotherapy? Anti-Cancer Drug Des. 1993, 8, 299-
322.
P 1-(2′,3′-O-Isop r op ylid en ea d en osin -5′-yl)-P 2-[7-b en z-
yloxy-6-(2-h ydr oxyeth yl)-5-m eth oxy-4-m eth ylph th alan -1-
on e-2-yl]m eth ylen ebis(p h osp h on a te) (34). (A) From 30. To
a solution of compound 30 (345 mg, 0.5 mmol) in pyridine (2
mL) was added DIC (156 mL, 2 mmol), and the mixture was
left overnight until the multisignal resonances were observed
in 31P NMR.23 The 2′,3′-O-isopropylideneadenosine (153 mg,
0.5 mmol) was then added, and the reaction mixture was
heated at 55-60 °C for 48 h. At that time, the 31P NMR of the
reaction mixture showed two broad signals at 8 and 25 ppm,
characteristic for the presence of the tetrasubstituted deriva-
tive. Then, a mixture of water (400 µL) and Et3N (200 µL) was
added, and the reaction mixture was kept at 75 °C for 30 h.
HPLC purification afforded compound 34 (297 mg, 60%) as
the triethylammonium salt. 1N NMR (D2O): δ 1.18 (t, 18H,
Et3N), 1.45 and 1.60 (two s, 3H each, isopropylidene), 2.05 (s,
3H, CH3), 2.51 (dt, 2H, P-CH2-P, J ) 21.2 Hz, J ) 8.8 Hz),
3.10 (q, 12H, Et3N), 3.40-3.48 (m, 2H, H5′, 5′′), 3.60 (t, 2H,
CH2 (1), J ) 6.4 Hz), 3.62 (s, 3H, OCH3), 4.10-4.12 (m, 3H,
CH2 (2), H4′), 4.54 (m, 1H, H3′), 4.67 (m, 1H, H2′), 4.95 (s,
2H, CH2-lactone), 5.14, (s, 2H, CH2-benzyl), 6.01 (d, 1H, H1′,
J 1′,2′ ) 2.4 Hz), 7.94 (t, 2H, phenyl, J ) 7.6 Hz), 7.96, 8.27
(two 1H singlets, H2, H8-adenine), 8.42 (t, 1H, phenyl, J )
7.6 Hz), 8.69-8.71 (m. 2H, phenyl). 31P NMR (D2O): δ 16.06-
17.79 (AB system, J ) 16.6 Hz).
(3) Weber, G.; Prajda, N.; Abonyi, M.; Look, K. Y.; Tricot, G.
Tiazofurin: Molecular and clinical action. Anticancer Res. 1996,
16, 3313-3322.
(4) Sherley, J . L. Guanine nucleotide biosynthesis is regulated by
the cellular p53 concentration. J . Biol. Chem. 1991, 266, 24815-
24828.
(5) (a) Feinstein, E.; Gale, R. P.; Canaani, E. Expression of the
normal p53 gene induces differentiation of K562 cells. Oncogene
1992, 1, 1853-1857. (b) Chylicki, K.; Ehinger, M.; Svedberg, K.;
Gullberg, U. Characterization of the molecular mechanism for
p53-mediated differentiation. Cell Growth Differ. 2000, 11, 561-
571. (c) Cui, X. S.; Donehover, L. A. Differential gene expression
in mouse mammary adenocarcinomas in the presence and
absence of wild-type p53. Oncogene 2000, 19, 5988-5996.
(6) Li, Z.-R.; Hormchak, R.; Mudipalli, A.; Bloch, A. Tumor sup-
pressor proteins as regulators of cell differentiation. Cancer Res.
1998, 58, 4282-4287.
(7) Franklin, T. J .; Edwards, G.; Hedge, P. Inosine 5′-monophos-
phate dehydrogenase as a chemotherapeutic target. In Purine
and Pyrimidine Metabolism in Man VII; Sahota, A., Taylor, M.,
Eds.; Plenum Press: New York, 1995; pp 155-160.
(8) Liu, Y.; Shirley, B. A.; Sherley, J . L. Inosine-5′-monophosphate
dehydrogenase is a rate-determining factor for p53-dependent
growth regulation. Mol. Biol. Cell. 1998, 9, 15-28.
(9) (a) Carr, S. F.; Papp, E.; Wu, J . C.; Natsumeda, Y. Characteriza-
tion of human type I and type II dehydrogenases. J . Biol. Chem.
1993, 268, 27286. (b) Glesne, D.; Collart, F.; Drabkin, H.;
Huberman, E. Chromosomal localization and structure of the
human type II IMP dehydrogenase gene (IMPDH2). Genomics
1993, 16, 274-277. (c) Gu, J . J .; Kaiser-Rogers, K.; Rao, K.;
Mitchell, B. S. Assignment of the human type I IMP dehydro-
genase gene (IMPDH1) to chromosome 7q31.3-q32. Genomics
1994, 24, 179-181.
(10) Nagai, M.; Natsumeda, Y.; Konno, Y.; Hoffman, R.; Irino, S.;
Weber, G. Selective up-regulation of type II inosine 5′-mono-
phosphate dehydrogenase messenger RNA expression in human
leukemias. Cancer Res. 1991, 51, 3886-3890.
(B) The coupling of compounds 22 and 27. The reaction of
22 (332 mg, 0.5 mmol) and 27 (197 mg, 0.6 mmol) was
performed as described for the coupling of 22 with 20, except
the reaction mixture was heated at 55-60 °C for 2 days. After
hydrolysis and HPLC purification, compound 34 (314 mg, 67%)
was obtained as a triethylammonium salt. Proton and phos-
phorus NMR were identical as reported above.
(11) Andersson, L. C.; J okinen, M.; Gahamberg, C. G. Induction of
erythroid differentiation in the human leukemia cell line K562.
Nature 1979, 278, 364-365.
(12) Rowley, P. T.; Ohlsson-Wilhelm, B. M.; Farley, B. A.; LaBella,
S. Inducers of erythroid differentiation in K562 human leukemia
cells. Exp. Hematol. 1981, 9, 32-37.
En zym e Assa ys. The new compounds were tested as
inhibitors of IMPDH type I and type II as described by us
previously.29 Briefly, substrate, cofactor, and inhibitor are
mixed, and the reaction started by the addition of the enzyme.
The course of the reaction is followed by measuring the
changing absorbance at 340 nm of the absorbance peak of
NADH.
(13) Olah, E.; Csokay, B.; Prajda, N.; Kote-J ari, Z.; Yeh, Y. A.; Weber,
G. Molecular mechanisms in the antiproliferative action of taxol
and tiazofurin. Anticancer Res. 1996, 16, 2469-2478.
(14) Tricot, G.; Weber, G. Biochemically targeted therapy of refractory
leukemia and myeloid blast crisis of chronic granulocytic leu-
kemia with tiazofurin, a selective blocker of inosine 5′phosphate
dehydrogenase activity. Anticancer Res. 1996, 16, 3341-3348.
(15) (a) Tricot, G. J .; J ayaram, H. N.; Nichols, C. R.; Pennington, K.;
Lapis, E.; Weber, G.; Hoffman, R. Hematologic and biochemical
action of tiazofurin (NSC 286193) in a case of refractory acute
myeloid leukemia. Cancer Res. 1987, 47, 4988-4991. (b) Tricot,
G. J .; J ayaram, H. N.; Lapis, E.; Natsumeda, Y.; Nichols, C. R.;
Kneebone, P.; Heerema, N.; Weber, G.; Hoffman, R. Biochemi-
cally directed therapy of leukemia with tiazofurin, a selective
blocker of inosine 5′-monophosphate dehydrogenase activity.
Cancer Res. 1989, 49, 3696-3701.
(16) Wright, D. G.; Boosalis, M. S.; Waraska, K.; Oshry, L. J .;
Weintraub, L. R.; Vosburgh. Tiazofurin effects on IMP-dehydro-
genase activity and expression in the leukemia cells of patients
with CML blas crisis. Anticancer Res. 1996, 16, 3349-3354.
(17) Cooney, D. A.; J ayaram, H. N.; Glazer, R. A.; Kelly, J . A.;
Marquez, V. A.; Gebeyehu, G.; van Cott, A. C.; Zwelling, L. A.;
J ohns, D. G. Studies on the mechanism of action of tiazofurin
metabolism to an analogue of NAD with potent IMP dehydro-
genase-inhibitory activity. Adv. Enzyme Regul. 1983, 21, 271-
303.
Mea su r em en t of Cellu la r Differ en tia tion . Human leu-
kemia K562 cells were treated with TR, C2-MAD, C4-MAD,
and bis(phosphonate) analogues 31 and 33 and then examined
for their differentiating activity. K562 cells were grown and
assayed as described by us earlier.30
Resista n ce to Glu cu r on id a tion . In the same manner as
described by us previously,33 the C2-MAD and C4-MAD
were examined as potential substrates for glucuronidation in
vitro. Under conditions in which MPA was completely glu-
coronidated, MAD analogues did not produce any glucuron-
ides.
Ack n ow led gm en t. The authors thank Dr. Takashi
Tsuji of Ajinomoto Co., Inc., J apan for a gift of MPA.
This investigation was supported in part by funds from
the National Cancer Institute, NIH (SBIR Grant CA-
84828), and National Institute of Allergy and Infectious
Diseases, NIH (SBIR Grant AI-46093). We acknowledge
the use of Shared Instrumentation provided by grants
(18) Pankiewicz, K. W.; Lesiak-Watanabe, K. Novel mycophenolic
adenine bis(phosphonate)s as potent anticancer agents and
inducers of cells differentiation. Nucleosides Nucleotides 1999,
18, 927-932.

712 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Pankiewicz et al.
(19) J ayaram, H. N.; Zhen, W.; Gharehbaghi, K. Biochemical conse-
quences of resistance to tiazofurinin human myelogenous leu-
kemic K562 cells. Cancer Res. 1993, 15, 2344-2348.
(20) Ahluwalia, G. S.; Cooney, D. A.; Marquez, V. E.; J ayaram, H,
N.; J ohns. Biochem. Pharmacol. 1986, 35, 3783-3790.
(21) Krohn, K.; Heins, H.; Wielckens, K. Synthesis and cytotoxic
activity of C-glycosidic nicotinamide riboside analogues. J . Med.
Chem. 1992, 35, 511-517.
(22) Gharehbaghi, K.; Paull, K. D.; Kelley, J . A.; Barchim, J . J ., J r.;
Marquez, V. E.; Cooney, D. A.; Monks, A.; Scudiero, D.; Krohn,
K.; J ayaram, H. N. Cytotoxicity and characterization of an active
metabolite of benzamide riboside, a novel inhibitor of IMP
dehydogenase. Int. J . Cancer 1994, 56, 892-899.
(23) J ayaram, H. N.; Grusch, M.; Cooney, D, A.; Krupitza, G.
Consequences of IMP dehydrogenase inhibition and its relation-
ship to cancer and apoptosis. Curr. Med. Chem. 1999, 6, 561-
575.
(24) Marquez, V. E.; Tseng, C. K. H.; Gebeyehu, G.; Cooney, D. A.;
Ahluwalia, G. S.; Kelly, J . A.; Dalal, M.; Fuller, R. W.; Wilson,
Y. A,; J ohns, D. G. Thiazole-4-carboxamide adenine dinucleotide
(TAD). Analogues stable to phosphodiesterase hydrolysis. J .
Med. Chem. 1986, 29, 1726-1731.
(25) Pankiewicz, K. W.; Lesiak, K.; Watanabe, K. A. Efficient
synthesis of methylenebis(phosphonate) analogues of P1, P2-
disubstituted pyrophosphates of biological interest. A novel
plausible mechanism. J . Am. Chem. Soc. 1997, 119, 3691.
(26) Lesiak, K.; Watanabe, K. A.; George, J .; Pankiewicz, K. W. 2-(4-
(30) Pankiewicz, K. W.; Lesiak, K.; Zatorski, A.; Goldstein, B. M.;
Carr, S. F.; Sochacki, M.; Majumdar, A.; Seidman, M.; Watanabe,
K. A. The practical synthesis of methylenebis(phosphonate)
analogue of benzamide adenine dinucleotide. Inhibition of hu-
man inosine monophosphate dehydrogenase (type I and type II).
J . Med. Chem. 1997, 40, 1287-1291.
(31) Sintchak, M. D.; Fleming, M. A.; Futer, O.; Raybuck, S. A.;
Chambers, S. P.; Caron, P. R.; Murcko, M. A.; Wilson, K. P.
Structure and mechanism of inosine monophosphate dehydro-
genase in complex with the immunosuppressant mycophenolic
acid. Cell 1996, 85, 921.
(32) J ones, D. F.; Mils, S. D. Preparation and antitumor properties
of analogues and derivatives of mycophenolic acid. J . Med. Chem.
1971, 14, 307-311.
(33) Lesiak, K.; Watanabe, K. A.; Majumdar, A.; Powell, J .; Seidman,
M.; Vanderveen, K.; Goldstein, B. M.; Pankiewicz, K. W.
Synthesis of a methylenebis(phosphonate) analogue of myco-
phenolic adenine dinucleotide: A glucuronidation-resistant MAD
analogue of NAD. J . Med. Chem. 1998, 41, 618.
(34) Davisson, V. J .; Davis, D. R.; Dixit, V. M.; Poulter, C. D.
Synthesis of nucleotide 5′-diphosphates from 5′-O-tosyl nucleo-
sides. J . Org. Chem. 1987, 52, 1794-1801.
(35) Khan, A. R.; Parrish, J . C.; Fraser, M. E.; Smith, W. W.; Bartlett,
P. A.; J ames, M. N. G. Lowering the entropic barrier for binding
conformationally flexible inhibitors to enzymes. Biochemistry
1998, 37, 16839-16845.
(36) Digits, J . A.; Hedstrom, L. Drug selectivity is determined by
coupling across the NAD+ sine of IMP dehydrogenase. Biochem-
istry 2000, 39, 1771-1777.
(37) Wu, J . C. Mycophenolic mofetil: Molecular mechanism of action.
Perspect. Drug Discovery Des. 1994, 2, 185.
(38) Papageorgiu, C. Enterohepatic Recirculation: A powerful Incen-
tive for Drug Discovery in the Inosine Monophosphate Dehy-
drogenase Field. Mini-Review in Medicinal Chemistry 2001, 1,
0000 or www.bentham.org/mrmc/.
(39) (a) Franklin, T. J .; J acobs, V.; J ones, G.; Ple, P.; Bruneau, P.
Glucuronidation associated with intrinsic resistance to myco-
phenolic acid in human colorectal carcinoma cells. Cancer Res.
1996, 56, 984. (b) Franklin, T. J .; J acobs, V. N.; J ones, G.; Ple,
P. Human colorectal carcinoma cells in vitro as a means to assess
the matabolism of analogues of mycophenolic acid. Drug Metab.
Dispos. 1997, 25, 367.
(40) Colby, T. D.; Vanderveen, K.; Stricler, M. D.; Markham, G. D.;
Goldstein, B. M. Crystal structure of human type II inosine
monophosphate dehydrogenase: implications for ligand binding
and drug design. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
3531-3536.
Nitrophenyl)ethyl methylenebis (phosphonate):
A versatile
reagent for the synthesis of nucleoside 5′-methylenebis-(phos-
phonate)s. J . Org. Chem. 1998, 63, 1906.
(27) Lesiak, K.; Watanabe, K. A.; Majumdar, A.; Seidman, M.;
Venderveen, K.; Goldstein, B. M.; Pankiewicz, K. W. Synthesis
of nonhydrolyzable analogues of thiazole-4-carboxamide and
benzamide adenine dinucleotide containing fluorine atom at the
C2′ of adenine nucleoside: Induction of K562 differentiation and
inosine monophoshate dehydrogeanse inhibitory activity. J . Med.
Chem. 1997, 40, 2533-2538.
(28) Our HPLC studies of the stability of TAD vs the methylenebis-
(phosphonate) analogue of TAD in serum and cell extracts
showed that within 1-2 h TAD was converted into tiazofurin
mononucleotide and AMP, which was further converted into
tiazofurin and adenosine. The last nucleoside was deaminated
to give inosine. In contrast, the methylenebis(phosphonate)
analogue of TAD was found unchanged after 24 h of incubation.
Similarly, the methylenebis(phosphonate) analogue was found
to be stable against the action of snake venom phosphos-
diesterase whereas TAD was a substrate.
(29) Zatorski, A.; Watanabe, K. A.; Carr, S. F.; Goldstein, B. M.;
Pankiewicz, K. W. The chemical synthesis of benzamide adenine
dinucleotide: Inhibition of inosine monophosphate dehydroge-
nase (type I and type II). J . Med. Chem. 1996, 39, 2422-2426.
J M0104116