Design, synthesis, and biological evaluation of novel human 5′-deoxy-5′-methylthioadenosine phosphorylase (MTAP) substrates

Article abstract of DOI:10.1016/j.bmcl.2005.03.107

The structure-based design, chemical synthesis, and biological evaluation of novel MTAP substrates are described. These compounds incorporate various C5′-moieties and are shown to have different k_cat/K_m values compared with the natural MTAP substrate (MTA).

Full text of DOI:10.1016/j.bmcl.2005.03.107

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2829–2833
Design, synthesis, and biological evaluation of novel human
50-deoxy-50-methylthioadenosine phosphorylase (MTAP) substrates
Pei-Pei Kung,* Luke R. Zehnder, Jerry J. Meng, Stanley W. Kupchinsky,
Donald J. Skalitzky, M. Catherine Johnson, Karen A. Maegley, Anne Ekker,
Leslie A. Kuhn, Peter W. Rose and Laura A. Bloom
Pfizer Global Research and Development-La Jolla, 10770 Science Center Drive, San Diego, CA 92121, USA
Received 6 January 2005; revised 18 March 2005; accepted 25 March 2005
Available online 4 May 2005
Abstract—The structure-based design, chemical synthesis, and biological evaluation of novel MTAP substrates are described. These
compounds incorporate various C50-moieties and are shown to have different kcat/Km values compared with the natural MTAP sub-
strate (MTA).
ꢀ 2005 Elsevier Ltd. All rights reserved.
Cytotoxic chemotherapeutic agents are a proven class of
anti-cancer treatments.1 A subset of these agents func-
tion by blocking the activity of key enzymes required
for de novo purine biosynthesis including glycinamide
ribonucleotide transformylase (GARFT, EC 2.1.2.2)
and aminoimidazolecarboxamide ribonucleotide trans-
formylase (AICARFT, EC 2.1.2.3) (Fig. 1).2 This block-
age deprives cancer cells of critical building blocks
required for DNA synthesis (AMP and GMP), thereby
limiting the potential for tumor growth. Unfortunately,
these chemotherapeutic agents often exert cytotoxic ef-
fects against non-cancerous (e.g., healthy) cells resulting
in a number of toxicities, limiting their utility. We there-
fore sought to identify a biological mechanism to reduce
the toxicity of anti-purine chemotherapeutic agents to-
ward non-cancer cells while preserving their cytotoxic
effects against tumors.
for generation of essentially all the free adenine in
human cells,3 and accounts for the majority of AMP
generation when the de novo synthesis pathway is
blocked.
In addition, although MTAP is present in all healthy
cells, certain cancers are known to have an incidence
of MTAP-deficiency.4 We therefore speculated that co-
administration of a MTAP substrate with a de novo
purine biosynthesis inhibitor (a chemotherapeutic agent)
would reduce the inhibitorÕs toxicity in MTAP-compe-
tent non-tumor cells while retaining its cytotoxic proper-
ties against MTAP-deficient cancers. This hypothesis is
supported by extensive in vitro studies, which combine
MTA with L-alanosine, an inhibitor of AMP synthesis,
as well as combined with the de novo purine synthesis
inhibitors Lometrexol and AG2037.5,6a,b It was our in-
tent to develop MTAP substrates as anti-toxicity agents,
which have different biological and physical properties
than the enzyme natural substrate in an effort to increase
the maximally tolerated dose of AG2037, a cytotoxic
inhibitor of GARFT currently in development.6c Ide-
ally, these novel MTAP substrates would be converted
to adenine by MTAP, which can further enter the de
novo AMP synthesis pathway and thereby increase
AMP production in MTAP(+) containing (non-tumor)
cells, but not in MTAP(ꢀ) cancer cells (Fig. 1).1 The
synthesis of such novel MTAP substrates is described
below along with an assessment of their kcat and Km
values.7
Human 50-deoxy-50-methylthioadenosine phosphorylase
(MTAP) functions in the purine salvage pathway
where it catalyzes the reversible phosphorolysis of 50-
deoxy-50-methylthioadenosine (MTA) and thereby
facilitates its conversion to adenine and 5-(methylthio)-
ribose-1-phosphate (MTRP) (Fig. 1). Adenine is then
converted to AMP via purine salvage pathways, and
MTRP is converted via an independent mechanism
to methionine.1 MTAP activity is thus responsible
*
Corresponding author. e-mail: peipei.kung@pfizer.com
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.107

2830
P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2829–2833
Figure 1.
A previously reported co-crystal structure8 of MTA
(compound 1) and MTAP was utilized to design the
desired substrates although it does not reflect the
transition state of the MTA cleavage process. Initial de-
signs explored compounds containing 20 and 30 moieties
other than hydroxyl groups. Several molecules were syn-
thesized, which replaced either the 20-OH or 30-OH sub-
stituents with potential hydrogen bonding acceptors and
donors such as F, N3, and NH2. Unfortunately, these
compounds were poorer MTAP substrates as compared
with the natural substrate, MTA (compounds 3-8, Table
1).
In the MTA/MTAP co-crystal structure, there is visible
space in the vicinity of the MTA C50 position between
the ligand and the protein. Accordingly, we decided to
examine whether modification of the MTA C50 position
would afford novel MTAP substrates. The kcat and Km
values of several such MTA analogues are listed in
Table 2. Compounds 2, 9 (ETA), 10, and 11 are litera-
ture-reported MTAP substrates,9 and their kcat/Km val-
ues determined in our assays were poorer than that
exhibited by MTA (compound 1). Compounds contain-
0
ing other C5 alkyl chain moieties (compounds 14 and
15) gave similar catalytic activity compared with ETA
Table 1.
kcat (sꢀ1
)
Km (lM)
kcat/Km (lMꢀ1 sꢀ1
)
0
0
0
R2
Compound #
R5
R3
1 (MTA)
CH3S
H
OH
OH
OH
H
OH
OH
H
4.6 0.2
4.8 0.4
1.8 0.3
6
3.2
0.8
2
3
4
5
6
7
8
2
CH3S
CH3S
CH3S
CH3S
CH3S
CH3S
0.28 0.005
0.004 0.0001
0.0004
3.1 0.4
3.9 0.5
ND
0.09
0.001
ND
1.1 · 10ꢀ5
0.0001
0.022
OH
F
OH
OH
OH
OH
N3
0.04 0.01
0.0065 0.0002
0.67 0.01
3600 1500
OHa
NH2
106
30
9
3
a Stereochemistry is R.

P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2829–2833
2831
Table 2.
Table 3.
Compound # R
kcat
(sꢀ1
Km
kcat/Km
(lMꢀ1 sꢀ1
)
Compound # R
kcat
(sꢀ1
Km
(lM)
kcat/Km
(lMꢀ1 s1)
)
(lM)
)
1 (MTA)
9 (ETA)
CH3
CH3CH2
4.6 0.2
2.9 0.1
1.8 0.3 3.2
1.8 0.4 1.6
19
27
1.5 0.1
3
7
0.7 0.5
0.11
0.8 0.07
2
10
11
12
2.3 0.2
5.9 0.3
1.46 0.3
3.1
1
0.74
4
0.8 1.48
28
29
0.3 0.02 1.7 0.6 0.17
1.4 0.07 1.5 0.4 0.93
25
17
2
2
0.058
0.05
13
0.85 0.2
14
15
5.5 0.4
4.2 0.2
2.8 0.8 1.96
0.84
30
31
2.4 0.1
2 0.4 1.2
5
1
0.18 0.007 5.6 0.7 0.03
16
17
0.23 0.004 0.88 0.07 0.26
3.44 0.08
5.2 0.7 0.66
(20, 21), quinoline (18), benzyl (22, 23, 24), and satu-
0
rated carbocyclic rings (25, 26) at the C5 position were
also evaluated (Table 2). Several novel MTAP sub-
strates were discovered, which exhibited kcat/Km values
between 0.3 and 0.8. Compounds containing saturated
carbocyclic moieties were no better or worse substrates
than their aromatic counterparts (compare 25 with 19,
and 26 with 20 and 21).
18
19
3.16 0.2
1.5 0.1
6.9
3
1
0.46
0.7 0.5
20
21
22
23
24
3.8 0.2
2.49 0.09
2.5 0.1
9
2
0.42
Due to the ready synthetic availability of derivatives and
analogues, we decided to focus additional SAR studies
on improving compound 19. In Table 3, compounds
27–31, which have different substituents on the phenyl
3.4 0.6 0.73
4.6 1.1 0.53
Table 4.
2.14 0.09
5
1
1
0.43
0.09
1.01 0.02 10.2
25
26
1.57 0.03
1.5 0.1
3.5 0.2 0.45
3.2 0.5 0.47
0
Compound # R5
kcat
(sꢀ1
Km
kcat/Km
(lMꢀ1 sꢀ1
)
)
(lM)
19
1.5 0.1
3
0.7 0.5
(compound 9). Compounds containing alkyl chains with
a polarized group (compounds 12) gave reduced
catalytic activity. The MTA/MTAP co-crystal structure
32
1.11 0.01 12
1
0.09
0
also showed that the MTA C5 position was surrounded
33
34
2.02 0.03 8.5 0.6 0.24
4.6 0.3 1.3 0.3 3.5
by hydrophobic residues, which created a space larger
than that expected to be occupied by a simple alkyl car-
bon chain. Therefore, some MTA analogues containing
phenyl (19), pyridyl (16, 17), five-membered heterocycles

2832
P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2829–2833
moiety were synthesized and compared. Compounds 29
and 30 displayed similar kcat/Km values that were
approximately 2-fold greater than that exhibited by 19
(note however that all these values are within the exper-
imental error of the assay). Compounds 27 and 28 con-
tain polarized substituents at the para position of the
phenyl moiety and displayed reduced catalytic activities
compared with compound 19. The 10-fold decrease in
kcat displayed by compound 31 may be due to the bulky
methyl group present at the ortho position of the phenyl
moiety.
Scheme 2. Reagents and conditions (R see Tables 2 and 3): (a) K+ꢀOt-
Bu (1 M in THF), RSH, DMF, 40 ꢁC, 12 h, 40–80%.
by Montgomery et al.15 Intermediate C was then reacted
with a Wittig reagent to give intermediate D. Intermedi-
ate D was reduced by hydrogenation using Pd/C and
then fully deprotected to give compound 34 (Scheme 4).
We then chose compound 19 to explore the possibility of
replacing the C50-sulfur atom with oxygen, nitrogen,
and carbon atoms. As shown in Table 4, the carbon ana-
logue of compound 19 (compound 34) gave the greatest
kcat/Km value (3.5 0.3), which closely resembles that
exhibited by MTA (kcat/Km = 3.2).
0
In conclusion, the C5 position of adenosine could toler-
0
ate a variety of substituents, however, the C2 and C3
0
positions are relatively conserved with respect to sub-
strate activity. During the process of probing the C5
space for MTAP enzyme, we discovered several novel
MTAP substrates, especially the carbon analogue (com-
pound 34), which displays kcat/Km values similar to the
natural substrate (MTA). Those novel MTAP substrates
might provide alternative therapeutic strategies for the
use of cytotoxic anti-cancer agents.
Compounds 3, 4, 5, and 7 were synthesized by following
the reported literature methods.10 The synthesis of
compounds 6 and 8 are described in Scheme 1. Interme-
diate A was prepared via the method described by
Gavagnin and Sodano.11 The C2 -hydroxy group was in-
0
verted by a SN2 reaction with sodium acetate then was
inverted again with sodium azide. Following deprotec-
tion with tert-butyl ammonium fluoride (TBAF) to give
compound 6. Compound 6 was then reduced with tri-
phenylphosphine to give compound 8.
Methylthioadenosine phosphorylase (MTAP) enzymatic
assay: Human MTAP containing an N-terminal six-his-
tidine tag was expressed in E. coli BL21 DE3 cells. The
protein was purified to homogeneity by Ni2+ affinity
chromatography. Enzymatic activity was measured
using a coupled spectrophotometric assay designed to
monitor the reaction product adenine.16 Various con-
centrations of the indicated 50-deoxy-50-methylthioaden-
osine (MTA) substrate analog were incubated in assay
buffer (40 mM potassium phosphate buffer, 1 mM, and
DTT 0.8 units/ml xanthine oxidase coupling enzyme)
for 5 min at 37 ꢁC. The reaction was initiated by the
addition of MTAP. The exact concentration of enzyme
used varied for each substrate tested and ranged from
2 to 500 nM. Activity as a function of enzyme concen-
tration was determined for each substrate tested to en-
sure that the appropriate enzyme concentration was
used. Activity was detected by continuous monitoring
of absorbance at 305 nm for 10 min (DE = 15,500 Mꢀ1).
Initial velocities were calculated by linear regression. Km
The synthetic approach to the described series of 50-thio-
ether analogues involves (1) preparation of an alkyl
chloride, that is, 50-deoxy-50-chloroadenosine; interme-
diate B, which was prepared via the method reported
by Robins.12 (2) reaction of this chloro derivative with
appropriate thioalcohols or thiophenols (Scheme 2).
The preparation of compound 32 began with the Mits-
unobu reaction of the commercially available 20,30-O-
isopropylideneadenosine and phenol. Subsequent acetic
acid deprotection of the isopropylidene group afforded
compound 32 in 40% yield (Scheme 3).13
Compound 33 was synthesized from intermediate B
using a modified literature condition.14 Compound 34
was synthesized through the aldehyde intermediate C,
which was prepared by following the procedure reported
Scheme 1. Reagents and conditions: (a) triflic anhydride, DMAP, pyridine, DCM, 0 ꢁC, 1 h, 40%; (b) NaOAc, HMPA, 40 ꢁC, 1 h, 22%; (c) NH3/
MeOH, 23 ꢁC, 20 min; (d) triflic anhydride, DMAP, pyridine, DCM, 0 ꢁC, 3 h; (e) NaN3, DMF, 23 ꢁC, 16 h; (f) TBAF, THF, 23 ꢁC, 30 min, 18%
over three steps; (g) PPh3, pyridine, 23 ꢁC, 18 h, 40%.

P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2829–2833
2833
Scheme 3. Reagents and conditions: (a) PPh3, PhOH, diisopropyl azodicarboxylate (DIAD), THF, 23 ꢁC, 18 h, 40%; (b) 80% acetic acid, 100 ꢁC, 6 h,
40%.
Scheme 4. Reagents and conditions: (a) benzyltriphenyl phosphonium chloride, sodium bis(trimethylsilyl)amide (1 M in THF), THF, 55 ꢁC, 3 h,
46%; (b) ethyl acetate, Pd/C, H2, 4 h; (c) ethanol, CH3NH–NH2, 55 ꢁC, 12 h, 88%; (d) 80% acetic acid, 100 ꢁC, 6 h, 42.6%.
6. (a) Hori, H.; Tran, P.; Carrera, C. J.; Hori, Y.; Rosen-
bach, M. D.; Carson, D. A.; Nobori, T. Cancer Res. 1996,
56, 5653; (b) Bloom, L. A.; Piraino, J. S.; Lorenzana, E.
G.; Boritzki, T. J. Proc. Am. Assoc. Cancer Res. 2004,
Abstract no.: 4000; (c) Robert, F.; Garrett, C.; Dinwoodie,
W. R.; Sullivan, D. M.; Bishop, M.; Amantea, M.; Zhang,
M.; Reich, S. D. Proc. Am. Assoc. Clin. Oncol. 2004,
Abstract no.: 3075.
7. Ghoda, L. Y.; Savarese, T. M.; Northup, C. H.; Parks, R.
E.; Garofalo, J.; Katz, L.; Ellenbogen, B. B.; Bacchi, C. J.
Mol. Biochem. Parasitol. 1988, 27, 109.
8. Appleby, T. C.; Erion, M. D.; Ealick, S. E. Structure 1999,
7, 629.
and kcat values were determined by fitting initial velocity
data to the Henri–Michaelis–Menton equation.
Acknowledgements
We thank Jay Srirangam, Tom Goetzen, and Khanh
Tran for the large-scale preparation of several synthetic
intermediates. We also thank Simon Bailey and Ted
Boritzki for proof-reading the manuscript.
9. White, M. W.; Vandenbark, A. A.; Barney, C. L.; Ferro,
A. Biochem. Pharmacol. 1982, 31, 503.
References and notes
10. (a) Compounds 3 and 4: Guillerm, D.; Guillerm, G.;
Witkowski-Vandenplas, C. Nucleosides Nucleotides 2001,
20, 689; (b) compound 5: Kawasaki, A. M.; Casper, M.
D.; Freier, S. M.; Lesnik, E. A.; Zounes, M. C.; Cummins,
L. L.; Cook, P. D. J. Med. Chem. 1993, 36, 831; (c)
compound 7: Montgomery, J. A.; Shortnacy, A. T.;
Thomas, H. J. J. Med. Chem. 1974, 17, 1197.
11. Gavagnin, M.; Sodano, G. Nucleosides Nucleotides 1989,
8, 1319.
12. Robins, M. J.; Hansske, F.; Wnuk, S. F.; Kanai, T. Can. J.
Chem. 1991, 69, 1468.
13. Li, H.; Miller, M. J. J. Org. Chem. 1999, 64, 9289.
14. Kolb, M.; Danzin, C.; Barth, J.; Claverie, N. J. Med.
Chem. 1982, 25, 550.
1. (a) Rustum, Y. M.; Takita, H.; Gomez, G. Antibiot.
Chemother. 1980, 28, 86; (b) Natsumeda, Y.; Prajda, N.;
Donohue, J. P.; Glover, J. L.; Weber, G. Cancer Res.
1984, 44, 2475.
2. Benkovic, S. J. Trends Biochem. Sci. 1984, 9, 320.
3. Kamatani, N.; Carson, D. A. Biochim. Biophys. Acta 1981,
675, 344.
4. (a) Fitchen, J. H.; Riscoe, M. K.; Dana, B. W.; Lawrence,
H. J. Cancer Res. 1986, 46, 5409; (b) Nobori, T.; Szinai, I.;
Amox, D.; Parker, B.; Olopade, O. I.; Buchhagen, D. L.;
Carson, D. A. Cancer Res. 1993, 53, 1098; (c) Nobori, T.;
Takabayashi, K.; Tran, P.; Orvis, L.; Batova, A.; Yu, A.
L. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 6203.
5. Batova, A.; Diccianni, M. B.; Omura-Minamisawa, M.;
Yu, J.; Carrera, C. J.; Bridgeman, L. J.; Kung, F. H.;
Pullen, J.; Amylon, M. D.; Yu, A. L. Cancer Res. 1999, 59,
1492.
15. Montgomery, J. A.; Laseter, A. G.; Hewson, K. J.
Heterocycl. Chem. 1974, 11, 211.
16. Savarese, T. M.; Crabtree, G. W.; Parks, R. E., Jr.
Biochem. Pharmacol. 1980, 30, 189.