AMP deaminase inhibitors. 2. Initial discovery of a non-nucleotide transition-state inhibitor series

Article abstract of DOI:10.1021/jm990447m

A series of N3-substituted coformycin aglycon analogues are described that inhibit adenosine 5'-monophosphate deaminase (AMPDA) or adenosine deaminase (ADA). The key steps involved in the preparation of these compounds are (1) treating the sodium salt of 6,7-dihydroimidazo[4,5-d][1,3]diazepin- 8(3H)-one (4) with an alkyl bromide or an alkyl mesylate to generate the N3- alkylated compound 5 and (2) reducing 5 with NaBH₄. Selective inhibition of AMPDA was realized when the N3-substituent contained a carboxylic acid moiety. For example, compound 7b which has a hexanoic acid side chain inhibited AMPDA with a K(i) = 4.2 μM and ADA with a K(i) = 280 μM. Substitution of large lipophilic groups α to the carboxylate provided a moderate potency increase with maintained selectivity as exemplified by the α-benzyl analogue 7j (AMPDA K(i) = 0.41 μM and ADA K(i) > 1000 μM). These compounds, as well as others described in this series of papers, are the first compounds suitable for testing whether selective inhibition of AMPDA can protect tissue from ischemic damage by increasing local adenosine concentrations at the site of injury and/or by minimizing adenylate loss.

Full text of DOI:10.1021/jm990447m

J . Med. Chem. 2000, 43, 1495-1507
1495
AMP Dea m in a se In h ibitor s. 2. In itia l Discover y of a Non -Nu cleotid e
Tr a n sition -Sta te In h ibitor Ser ies1
Brett C. Bookser,* Srinivas Rao Kasibhatla, J ames R. Appleman, and Mark D. Erion
Metabasis Therapeutics Inc., 9390 Towne Centre Drive, San Diego, California 92121
Received August 31, 1999
A series of N3-substituted coformycin aglycon analogues are described that inhibit adenosine
5′-monophosphate deaminase (AMPDA) or adenosine deaminase (ADA). The key steps involved
in the preparation of these compounds are (1) treating the sodium salt of 6,7-dihydroimidazo-
[4,5-d][1,3]diazepin-8(3H)-one (4) with an alkyl bromide or an alkyl mesylate to generate the
N3-alkylated compound 5 and (2) reducing 5 with NaBH4. Selective inhibition of AMPDA was
realized when the N3-substituent contained a carboxylic acid moiety. For example, compound
7b which has a hexanoic acid side chain inhibited AMPDA with a Ki ) 4.2 µM and ADA with
a Ki ) 280 µM. Substitution of large lipophilic groups R to the carboxylate provided a moderate
potency increase with maintained selectivity as exemplified by the R-benzyl analogue 7j
(AMPDA Ki ) 0.41 µM and ADA Ki > 1000 µM). These compounds, as well as others described
in this series of papers, are the first compounds suitable for testing whether selective inhibition
of AMPDA can protect tissue from ischemic damage by increasing local adenosine concentrations
at the site of injury and/or by minimizing adenylate loss.
Sch em e 1. AMP Metabolic Pathway
In tr od u ction
AMP deaminase (AMPDA) is a potential new target
for the treatment of ischemia-related diseases based on
initial studies showing that AMPDA is important in the
regulation of the production of adenosine and sparing
of adenylates specifically during periods of net ATP
breakdown.1 AMPDA catalyzes the hydrolytic deami-
nation of AMP to IMP (Km ∼ 1 mM) (Scheme 1). With
this pathway blocked, AMP may be dephosphorylated
to adenosine via an endonucleotidase or reverted back
to ATP via adenylate kinase. Moreover, flux through the
enzyme is reported to be significant only during periods
of net ATP breakdown when the concentration of AMP
is increased manyfold.2 Hence, inhibitors of this enzyme
represent potential site- and event-specific cardio- and
neuroprotective agents since they would only affect
purine metabolism during periods of ischemia and
would be pharmacologically silent otherwise.3 A recent
published study further supports this contention.4 Pa-
tients diagnosed with congestive heart failure (CHF)
were found to have a significantly longer time period of
survival without heart transplant (7.6 vs 3.2 years; P
< 0.001) if they had a genetic deficiency in skeletal
muscle AMPDA. This suggests a potential use of AMP-
DA inhibitors for the treatment of CHF.5
AMPDA, however, represents a particularly difficult
target for medicinal chemists since no compounds are
known that can bind with high affinity and specificity
to AMP binding sites and also diffuse into cells. In
general, a phosphate or phosphonate functionality is the
crucial component necessary for a compound to have
affinity to a nucleotide binding site, but this feature also
prevents cell penetration. Also, compounds containing
a phosphate are subject to rapid dephosphorylation in
vivo by ectonucleotidases. Nevertheless, when a nucle-
otide is the active drug, a common approach is admin-
istration of nucleotide precursors (e.g. nucleosides and
bases) and to rely on intracellular metabolic conversion
to the active nucleotide form.6 Alternatively, phosphate
mimetics and prodrug strategies that mask charge or
combinations thereof are used.7 However, these strate-
gies posed significant challenges when applied to AMP-
DA, and the historically poor results from such strate-
gies forced us to look for alternatives.
Coformycin (1)8 and coformycin 5′-monophosphate (2)
are potent AMPDA inhibitors with Kis of 3 and 0.0001
µM,9 respectively. Coformycin (1) has been used to
evaluate the role of AMPDA in purine metabolism,10 but
since it is a much more potent ADA inhibitor (Ki )
0.00001 µM)11 it has limited utility for defining the
importance of AMPDA in nucleotide metabolism. We
also sought to avoid dual inhibitors because of the
association of ADA deficiency with severe combined
immunodeficiency.12 Coformycin 5′-monophosphate (2)
is completely selective, but its phosphate substituent
makes it an unsuitable tool compound (vida supra).
The potency of coformycin (1) at AMPDA is notable
since it does not require the highly charged phosphate
group normally necessary for affinity to nucleotide
binding sites. Coformycin (1) can be defined as a
transition-state inhibitor based on the similarity of its
* To whom correspondence should be addressed. Phone: 858-622-
5512. Fax: 858-622-5573. E-mail: bookser@mbasis.com.
10.1021/jm990447m CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/25/2000

1496 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Bookser et al.
F igu r e 1. AMPDA inhibitor design strategy.
Sch em e 2. General Method for Coformycin Aglycon
structure to the structure of the tetrahedral intermedi-
ate of the deamination reaction (Chart 1). Coformycin
(1) compensates for the binding not accessed in the
phosphate pocket by having good transition-state (TS)
mimicry from the aglycon portion of the molecule.13 In
the active site of these enzymes the aglycon moiety of
coformycin (1) can be positioned to have close interac-
tions with the zinc and other catalytic residues which
stabilize the TS of the deamination.1,14 The aglycon
binding regions of AMPDA and ADA are likely to be
very similar because of the high active site residue hom-
ology they share.1 TS mimicry is also enzyme specific
so coformycin (1) or analogues containing the aglycon
moiety are unlikely to interact with other adenosine-
or AMP-binding proteins. Thus, we reasoned that ana-
logues of coformycin (1) could potentially be selective
cell-penetrable inhibitors. Still, the challenge was to find
a replacement for the ribose group which would allow
discriminatory binding to AMPDA relative to ADA.
In Figure 1 is shown our inhibitor design strategy
which entailed replacing the ribose monophosphate of
2 with a lipophilic group capped with an anionic
substituent Y.1 We envisioned that the anionic group
would interact with the cationic, phosphate binding site
of AMPDA and thereby provide potent and selective
inhibition of AMPDA vs ADA. The anionic Y group was
envisioned to have physicochemical properties which
would allow the molecule to cross cell membranes. Thus
a carboxylate was chosen as the Y group.
Analogue Preparationa
a
(a) NaH, DMF, 60 °C, 30 min; R-X, NaI, 90 °C; (b) NaBH4,
MeOH, CH2Cl2.
6,7-dihydroimidazo[4,5-d][1,3]diazepin-8(3H)-one (4; pre-
pared in eight steps from 4-methylimidazole)16a was
alkylated employing sodium hydride as a base and an
alkyl bromide or mesylate as the electrophile in the
presence of sodium iodide in hot DMF. This produced
mainly a chromatographically separable mixture of N3-
and N1-alkyl products 5 and 6 in yields typically on the
order of 30% and 15%, respectively. Products of dialky-
lation were occasionally obtained, especially if the base
and electrophile were in excess of the heterocycle. The
ketone 5 was not stable to aqueous acid or aqueous
base.16,17 It was typically reduced as the next step with
NaBH4 to provide a 60-90% yield of the N3-alkylcofor-
mycin aglycon product 7 as a 1:1 mixture of C8 epimers.
In general, for reactivity considerations, the alcohol 7
had good stability to aqueous base but had a limited
half-life in mild aqueous acid.18,19 Tables 1 and 2 list
analogues 7.
Herein we report that replacement of the ribose 5′-
monophosphate of 2 with an alkylcarboxy group re-
sulted in a series of potent and selective AMPDA
inhibitors. These compounds and additional efforts to
optimize the series15 for the first time enable investiga-
tion of inhibition of AMPDA as a potential route to drugs
targeting ischemia-related diseases.
The regiochemistry of the alkylation was proven in
part by chemical means as shown in Scheme 3. The
analogue 5f ′ could be prepared either by alkylation of
4 with benzyl bromide or by cyclization of compound 8
(a precursor in the synthesis of compound 4) with
Ch em istr y
Scheme 2 illustrates the general method of coformycin
aglycon analogue synthesis.16 The known heterocycle
Ch a r t 1. AMPDA- and ADA-Catalyzed Deamination Reaction

AMP Deaminase Inhibitors. 2
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1497
Sch em e 3. Regiochemistry of Major Alkylation Product
Sch em e 4. Representative Substituted Hexanoate
Electrophile Preparation and Representative Final
Product Ester Hydrolysisa
Defineda
a
(a) NaH, DMF; NaI, BnBr; (b) (EtO)3CH, DMSO, 70 °C; (c)
NaBH4.
a
(a) NaOH, H2O, dioxane; (b) EtI, K2CO3, DMF; (c) TBSCl,
DMF, imidazole; (d) LDA, MeI, THF, DMPU; (e) LDA, (3-
BrPh)CH2Br, THF, DMPU; (f) Bu4NF, THF; (g) MsCl, NEt3,
CH2Cl2; (h) 4, NaH, DMF; 11d , NaI; (i) NaBH4, MeOH, CH2Cl2;
(j) 0.5 M NaOH, dioxane; DOWEX 1×8-400 acetate. DMPU )
N,N′-dimethylpropyleneurea.
to this triad, especially C3a and C7. Thus structure 6
in Scheme 2 is drawn with a partial double bond
between the three atoms N4, C5, and N6. Finally, since
the UV/vis λmax value for 7f (R ) (CH2)5CO2Et), λmax
(methanol) 283 nm (ꢀ 6545), is nearly identical to that
reported for coformycin (1),8a
max(H2O) 282 nm (ꢀ 8250),
-
λ
it must be the N3-alkyl analogue. In comparison the
value for the N1-alkyl analogue 6f′ (R ) (CH2)5CO2Et)
is significantly shifted, λmax(methanol) 278 nm (ꢀ 4680).
F igu r e 2. 13C NMR (DMSO-d6) assignment of compounds 5f
and 6f.
Some of the electrophiles, R-X, used in the coupling
reaction of Scheme 2 were commercially available alkyl
bromides or were mesylates of available alcohols.21
Scheme 4 shows the general method for construction of
some of the R-substituted hexanoate electrophiles as
exemplified by the preparation of 11d . ꢀ-Caprolactone
(9) was hydrolyzed and protected to provide the acyclic
ester 10c (92% for three steps). Two enolate alkylations
were followed by conversion of the resulting silyl ether
11b to the mesylate 11d (47% for four steps). Ketone 4
was alkylated with 11d to provide a 31% yield of 5u
(not shown) which was reduced to analogue 7u (74%).
Scheme 4 also shows the final step used to prepare
many of the carboxylic acids in Table 1. Sodium hy-
droxide hydrolysis of ester 7u followed by purification
of the product with DOWEX 1×8-400 acetate ion-
exchange resin provided carboxylate 7v (69%). Some
sodium carboxylates in the series (compounds 7a , 7b,
and 7i) and free acids (7c and 7d ) were obtained by
treatment of the resulting hydrolysis solution with
Amberlite CG-50 resin to absorb excess sodium hydrox-
ide. However, some of these sodium salts were prohibi-
tively deliquescent, and this necessitated the use of
DOWEX 1×8-400 acetate ion-exchange resin to isolate
the more stable free acids. Alternatively, some carboxy-
lates (i.e. 7y, 7z, and 7a ′) were prepared in good yields
by hydrogenolysis of their respective benzyl esters.
triethyl orthoformate. Since the structure of 8 was
previously defined16a it follows that the alkylation of 4
provides 5 as the primary product.
The structure assignments were further confirmed by
13C NMR (Figure 2), 1H NMR, and UV/vis spectral
1
1
analysis. H-1H and H-13C 2-dimensional NMR cor-
relation experiments in DMSO-d6 enabled carbon and
proton shift assignments for compounds 5f and 6f. The
bridgehead carbons C8a and C3a were particularly
diagnostic for each isomer. The bridgehead carbon
resonance shifts upfield when the neighboring nitrogen
is substituted. N1-Substitution results in C8a being
shielded by 11 ppm in compound 6f compared to 5f. N3-
Substitution results in a 7.1 ppm upfield shift in the
C3a resonance in 5f compared to 6f. This parallels 13C
NMR shift variances seen with N9- vs N7-methyl-
1
substituted adenines.20 The H NMR in DMSO-d6 also
exhibits distinctly different spectra for the isomers. Most
notable are those protons connected to the diazepinone
section of the molecule. For the N3-isomer 5f, protons
on C5 and C7 are sharp doublets with coupling con-
stants of J ) 4 Hz for each and the N6-proton is well-
defined with a broad peak at δ 8.4-8.5. By contrast the
C5- and C7-protons of 6f are defined by broad singlets.
This line broadening can be explained by considering
6f as an equilibrating 1:1 tautomeric mixture of N4d
C5-N6 and N4-C5dN6 forms in DMSO-d6. Compari-
son of the N6-H (δ 8.2) integration for the former to the
N4-H (δ 10.6) integration for the latter supports this
ratio assignment. Further evidence for this equilibration
is the broad 13C NMR resonances for those carbons close
Electrophile 11d is a racemic mixture. To evaluate
whether one enantiomeric form of this side chain has a
higher affinity for AMPDA, the undefined single enan-
tiomer 11e was prepared by crystallization of the (-)-

1498 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Bookser et al.
Sch em e 5. Preparation of a Hindered Ester
by reduction and finally SEM group removal with fluor-
ide ion in DMF provided the analogue 7b′. The phos-
phonates 20b, 21b, and 21c were prepared by reacting
the individual phosphites 20a and 21a with NaH and
then with the specific 1,ω-dibromoalkane. These elec-
trophiles were used to prepare 7c′, 7d ′, and 7e′.
Electrophilea
a
(a) LICA, THF, -78 °C; (CH3)2CHCH2I, DMSO, 23 °C; (b) 3
equiv LDA, THF, DMPU, 0 °C; 4 equiv I(CH3)4Cl, 55 °C; TFA,
CH2Cl2, reflux; BnBr, K2CO3, DMF, 23 °C.
brucine salt from acetonitrile (34% yield). Compound
1
11e was judged to be >95% ee on the basis of the H
NMR (CDCl3) of the complex of with (S)-(-)-R-methyl-
benzylamine. This in turn was used to prepare 7w ,
which is a 1:1 8R/S-diastereomeric mixture of two of the
four compounds that make up compound 7v.
Resu lts
Tables 1 and 2 list the coformycin aglycon analogues
7 and their Ki values for inhibition of either porcine
heart or human L-type AMPDA and calf intestinal ADA.
A cross-comparison of AMPDA Ki values determined
from both enzyme sources indicated the values have
e10% variation. Ki values of the 8R-isomers are likely
one-half the indicated value since the compounds are
1:1 8R/S-isomer mixtures and only the 8R-isomer is
likely to be inhibitory based on earlier work with (8S)-
2′-deoxycoformycin monophosphate27 and (8S)-2′-deoxy-
coformycin.28 The inhibitor initial screening concentra-
tions in general were chosen so that the upper limits of
the Ki values determined for AMPDA and ADA inhibi-
tion were 125 and 7.5 µM, respectively. In select
examples, to evaluate the full extent of the selectivity
of the enzyme inhibition, ADA inhibition was evaluated
at higher concentrations. Table 1 shows that potency
and selectivity can be achieved when the carbon tether
connecting the aglycon to the CO2H was at least five
atoms in length (e.g. for compound 7b, AMPDA Ki )
4.2 µM and ADA Ki ) 280 µM). Longer chain lengths
provided similar potency and selectivity (i.e. compounds
7c and 7d ), but the CO2H group was crucial to selectiv-
ity and potency as analysis of the various esters
demonstrates. For example, the ethyl ester 7f was
approximately a 10-fold weaker AMPDA inhibitor and
a 100-fold better ADA inhibitor compared to its deriva-
tive carboxylate 7b. Examination of other anionic
groups revealed the tetrazole as in analogue 7b′ to be a
satisfactory replacement for the carboxylate, but phos-
phonates 7c′, 7d ′, and 7e′ were at least 10-fold weaker
than the corresponding carboxylate analogues.
The hexanoate electrophiles 13 and 15 were prepared
by reacting precursors 12 and 1422 with LDA and 2
equiv of 1,4-dibromobutane in THF/DMPU.
The hindered ester 17 was prepared using successive
alkylations of tert-butyl acetate (16) as shown in Scheme
5.23,24 R-Alkylation under conditions of excess LDA in
THF-DMPU (2-1) at 0 °C for deprotonation and then
addition of 4-chloro-1-iodobutane at elevated tempera-
tures25 provided the desired hindered ester which after
ester exchange provided the electrophile 18.
Other desired electrophiles were also prepared by
monoalkylating 1,ω-dibromoalkanes. Tetrazole was
treated with NaH and then N-protected with SEM-Cl
(SEM ) (CH3)3SiCH2CH2OCH2). This produced a sepa-
rable 1:1 mixture of N1- and N2-protected tetrazoles.
The N1-protected compound decomposed upon treat-
ment with t-BuLi, whereas the N2-protected compound
19a upon successive treatment with t-BuLi and 1,5-
dibromopentane produced the desired electrophile 19b.26
Alkylation of the base 4 with this electrophile followed
Substitution of large hydrophobic groups R to the
carboxylate group resulted in increased potency and
maintained selectivity (see 7j: AMPDA Ki 0.41 µM and
ADA Ki > 1000 µM). This substitution also introduced
a second chiral center into the molecule in addition to
the C8 center. Thus it was of interest to examine a
resolved R-substituted analogue for evidence of a pre-
ferred configuration. In this regard, compound 7w
showed little difference from the corresponding diaster-
eomeric mixture 7v. Since little preference was evident,
R,R-disubstituted analogues 7y, 7z, and 7a ′ were ex-

AMP Deaminase Inhibitors. 2
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1499
Ta ble 1. Coformycin Aglycon Analogues with Anionic and Ester N3-Substituents

1500 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Ta ble 1 (Continued)
Bookser et al.
a
b
Analyses for C, H, N were correct within (0.4% unless otherwise stated. H, N; C: calcd, 53.97; found, 54.38. c Not determined, very
deliquescent compound. C, H; N: calcd, 11.37; found, 10.57. e C; H: calcd, 6.03; found, 5.56; N: calcd, 11.40; found, 10.74. f C; H: calcd,
10.32; found, 9.79; N: calcd, 11.75; found, 12.35. C, H; N: calcd, 14.50; found, 14.02. C, N; H: calcd, 6.78; found, 6.33. C, H; N: calcd,
24.83; found, 23.80.
d
g
h
i
amined and found to be modest and selective inhibitors
of AMPDA with Kis of ca. 1 µM each.
site. For example, compound 7j, the most potent of
this series, has an AMPDA Ki ) 0.41 µM and an ADA
Ki > 1000 µM. As a point of reference, the aglycon itself
(22) is not an AMPDA inhibitor (Ki > 1000 µM) and a
very weak ADA inhibitor (Ki ) 50 µM).1 Thus the N3-
substituent is very important for proper positioning of
the aglycon moiety in the active site as a TS inhibitor,
and importantly, it does not have to be a ribose
analogue. Prior to this work, no inhibitors of AMPDA
were described with properties appropriate to study
their effects on cellular function.29 In addition to having
poor affinity to AMPDA, those compounds all contain a
phosphate or phosphonate functionality and thus are
relatively impermeable to cells. Coformycin (1) was the
best non-nucleotide with an AMPDA Ki ) 3 µM;
however, it is also a very potent ADA inhibitor (Ki )
0.00001 µM) and thus is preempted as a tool to study
the biological effects of selective AMPDA inhibition.
Inhibitory potency of analogues with neutral or non-
anionic N3-substituents are described in Table 2. These
compounds with alkyl, alkylhydroxy, alkylcarboxamide,
and alkylcyano N3-substituents were in general much
weaker inhibitors of AMPDA than the carboxylates and
nonselective vs ADA. A few n-alkyl-substituted ana-
logues did inhibit AMPDA modestly, but these com-
pounds also proved to be good ADA inhibitors (e.g. 7k ′
with an AMPDA Ki of 10 µM and an ADA Ki of 0.35
µM).
The carboxylate 7m was evaluated for its ability to
penetrate cells in vitro. This compound was able to
equilibrate rapidly across cell membranes (t1/2 e 5 min)
for a variety of cell types which included rat hepatocytes,
rabbit erythrocytes, and cultured bovine endothelial
cells.
Discu ssion
The first potent, selective, and cell-penetrating inhibi-
tors of AMPDA have been prepared by (1) utilizing the
affinity and selectivity that the coformycin aglycon
moiety exhibits presumably as a result of it TS mimicry
and (2) rationalizing that an N3-anionic group, namely
an alkyl-CO2H group, would provide preferred affinity
to AMPDA by interacting with its phosphate binding
We propose that the CO2H group of these AMPDA
inhibitors does not lie inside the phosphate-binding

AMP Deaminase Inhibitors. 2
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1501
Ta ble 2. Coformycin Aglycon Analogues with Nonanionic N3-Substituents
a
b
Analyses for C, H, N were correct within (0.4% unless otherwise stated. Elemental analysis not determined. c >95% HPLC pure,
detector wavelength set at 280 nM. Calcd: C, 48.97; H, 6.17; N, 28.55. Found: C, 47.09; H, 5.96; N, 25.70. e Calcd: C, 51.41; H, 6.70;
d
N, 26.65. Found: C, 47.25; H, 6.47; N, 21.74. f Calcd: C, 55.41; H, 7.61; N, 23.51. Found: C, 52.58; H, 7.19; N, 21.09. Calcd: C, 55.17;
g
H, 8.04; N, 21.45. Found: C, 56.00; H, 8.13; N, 20.98.
pocket of the active site but rather that it provides
binding to the pocket with a through-space electrostatic
interaction. A modeling study comparing the energy-
minimized structure of a more advanced AMPDA in-
hibitor to the crystal structure of AMP in the anti
conformation supports this assertion.1 This study re-
vealed the critical CO2H to overlay in close proximity
to the 5′-phosphate of AMP. The inhibitor could not
assume a low-energy conformation with direct CO2H/
PO4H2 overlay.30 Neutral N3-substituents containing
groups such as alkyl esters, alkyls, alkylphenyls, alkyl-
hydroxys, alkyl-C(O)NH2, and alkyl-CN are generally
nonselective inhibitors of AMPDA which underlines the
importance of the CO2H to potency and selectivity.
Other anionic groups such as tetrazole and phospho-
nate, while not improving potency relative to the
CO2H analogues, also show selective inhibition of
AMPDA relative to ADA. A tether length of at least five
atoms was optimal (compare 7b, 7c, and 7d to 7a )
possibly because shorter tethers cannot position the
CO2H close enough to the phosphate binding site. This
minimum five-atom tether length further reinforces the
similarity of these inhibitors to AMP since it equals the
shortest atom-chain path from N9 of AMP to its 5′-
phosphate.
It is useful to note when considering the further
optimization of this series that the active site allows
large hydrophobic substituents close to the CÃ2Η group

1502 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Bookser et al.
did not settle rapidly were decanted off and discarded. This
was repeated until no significant suspension of fine particles
was evident (3-4 times). The resin was transferred as an
aqueous slurry to a coarse glass fritted funnel and gravity
eluted with 10 volumes of 3 M aqueous NaOH, 10 volumes of
3 M aqueous acetic acid and 20 volumes of water (final pH
6-7). The resin was stored at 5 °C in water.
Eth yl 2-(3-Br om op h en ylm eth yl)-6-m eth a n esu lfon oxy-
2-m eth ylh exa n oa te (11d ). Step 1. To ꢀ-caprolactone (50.0
g, 0.44 mol) in 1.7 L of dioxane was added NaOH (0.88 L of a
1 M aqueous solution) and the mixture stirred at 25-35 °C
for 16 h. The mixture was diluted with 1 L of water and
extracted with EtOAc (3 × 0.7 L). Under ice bath cooling, the
pH of the aqueous layer was lowered to 1 with the addition of
96 mL of concentrated HCl then extracted with EtOAc (5 ×
0.7 L). The combined extracts were dried (MgSO4) and evap-
orated to provide 53.4 g (95%) of 6-hydroxyhexanoic acid (10a )
as an oil: 1H NMR (D2O/NaOD with 3-(trimethylsilyl)propri-
onic acid, sodium salt as internal standard) 1.25-1.45 (m, 2),
1.5-1.7 (m, 4), 2.20 (t, 2, J ) 7 Hz), 3.60 (t, 2, J ) 7 Hz).
Step 2. A mixture of 6-hydroxyhexanoic acid (27.9 g, 0.21
mol), iodoethane (20.8 mL, 0.26 mol) and K2CO3 (43.7 g, 0.32
mol) in 0.4 L of DMF was stirred at 80 °C for 4 h, cooled, di-
luted with 1 L of water and extracted with EtOAc (3 × 0.5 L).
The combined extracts were dried (MgSO4) and evaporated to
provide 40.0 g of ethyl 6-hydroxyhexanoate (10b) as an oil con-
taminated by residual DMF. It was used directly in the next
reaction without further purification: 1H NMR (DMSO-d6) 1.17
(t, 3, J ) 7 Hz), 1.2-1.6 (m, 6), 2.26 (t, 2, J ) 7 Hz), 3.37 (q,
2, J ) 7 Hz), 4.04 (q, 2, J ) 7 Hz), 4.36 (t, 1, J ) 7 Hz).
Step 3. A mixture of ethyl 6-hydroxyhexanoate (6.12 g, 38
mmol), tert-butyldimethylsilyl chloride (5.76 g, 0.38 mmol) and
imidazole (2.72 g, 39.9 mmol) in 100 mL of DMF was stirred
at 23 °C for 16 h, then diluted with 300 mL of Et2O and washed
with 500 mL of water. The aqueous layer was separated and
extracted with 100 mL of Et2O. The combined organic extracts
were washed with 5% aqueous HCl, water, and brine, dried
(MgSO4) and evaporated to provide 10.09 g (97%) of ethyl
6-tert-butyldimethylsilyloxyhexanoate (10c) as an oil: 1H NMR
(DMSO-d6) 0.01 (s, 6), 0.85 (s, 9), 1.13 (t, 3, J ) 7 Hz), 1.2-1.6
(m, 6), 2.26 (t, 2, J ) 7 Hz), 3.56 (t, 2, J ) 7 Hz), 4.03 (q, 2, J
) 7 Hz).
Step 4. To a solution of diisopropylamine (10.0 mL, 71.2
mmol) in 77 mL of THF at -78 °C was added n-BuLi (7.12
mL of a 10 M solution in hexanes, 71.2 mmol) and the mixture
stirred at -78 °C for 15 min. To this solution was added a
solution of ethyl 6-tert-butyldimethylsilyloxyhexanoate (17 g,
61.9 mmol) in 30 mL of THF over a period of 40 min by syringe
pump injection. After stirring for an additional 45 min at -78
°C, 9 mL of DMPU was added followed by the addition of
iodomethane (4.8 mL, 77 mmol) and then the temperature
raised to -15 °C and stirring continued for 1 h when it was
quenched with 4 M NH4Cl. The solvent was concentrated to
ca. 40 mL by rotary evaporation and then diluted with 200
mL of Et2O and extracted with water. The aqueous layer was
separated and extracted with 100 mL of Et2O. The combined
ether extracts were washed with 5% aqueous HCl, water, and
brine, dried (MgSO4) and evaporated to provide 17.2 g
(96%) of ethyl 6-tert-butyldimethylsilyloxy-2-methylhexa-
noate (11a ) as an oil: 1H NMR (DMSO-d6) 0.01 (s, 6), 0.85 (s,
9), 1.05 (d, 3, J ) 7 Hz), 1.17 (t, 3, J ) 7 Hz), 1.2-1.7 (m, 6),
2.38 (q, 2, J ) 7 Hz), 3.55 (t, 2, J ) 7 Hz), 4.04 (q, 2, J )
7 Hz).
Step 5. In a manner identical to step 4, ethyl 6-tert-
butyldimethylsilyloxy-2-methylhexanoate (4.47 g, 15.5 mmol)
and 3-bromobenzyl bromide (4.65 g, 18.6 mmol) were com-
bined to produce 7.36 g of crude ethyl 2-(3-bromobenzyl)-6-
tert-butyldimethylsilyloxy-2-methylhexanoate (11b ) as an
oil which was used directly in the next reaction: 1H NMR
(DMSO-d6) 0.01 (s, 6), 0.85 (s, 9), 0.98 (s, 3), 1.17 (t, 3, J
) 7 Hz), 1.2-1.7 (m, 6), 2.66 (d, 1, J ) 13 Hz), 2.93 (d, 1, J )
13 Hz), 3.56 (t, 2, J ) 6 Hz), 4.04 (q, 2, J ) 7 Hz), 7.09 (d, 1,
J ) 8 Hz), 7.19 (t, 1, J ) 8 Hz), 7.27 (s, 1), 7.36 (d, 1, J )
8 Hz).
which incidentally might not be expected if the CO2H
was inside the very hydrophilic phosphate binding site.
Compound 7j, with an R-benzyl substituent, is 10-fold
more potent than the initial lead 7b and still AMPDA
selective. Even more hydrophobic substitutions near the
CO2H are tolerated by AMPDA. The R,R-diisobutyl and
R,R-diphenyl analogues 7y and 7a ′ were moderate
inhibitors (AMPDA Ki ) 0.86 and 0.92, respectively).
The moderate increases in potency realized from R-sub-
stituents suggest that some hydrophobic binding is oc-
curring and much more optimization might be possible
by a different approach. Further work on compound
optimization can be found in the following papers.15
Su m m a r y
The coformycin aglycon is an excellent template from
which to construct potent, selective, and cell-penetrating
AMPDA inhibitors. The best compounds result from
substitution of an alkylcarboxy group at N3 of the
aglycon, wherein the alkyl chain is at least five atoms
long. The CO2H group is critical to potency and selectiv-
ity vs ADA. We propose that the CO2H group has
through-space electrostatic interactions with the phos-
phate binding site of AMPDA. Affinity gained through
TS mimicry from the inhibitors probably compensates
for the affinity not accessed from direct binding into the
phosphate pocket. TS mimicry also makes the com-
pounds selective vs other AMP- and adenosine-binding
proteins.1 These compounds provide an entry point from
which to examine the effects of AMPDA inhibition on
cellular function. Inhibitors of AMPDA in this series
have been shown to increase both intra- and extracel-
lular adenosine and preserve the adenylate pool in
hepatocytes subjected to conditions that induce ATP
breakdown.1 The inhibitors did not alter adenosine and
adenylate concentrations under unstressed or normal
physiological conditions. This indicates that these com-
pounds may be potential site- and event-specific anti-
ischemic drugs.
Exp er im en ta l Section
Gen er a l Meth od s. Glassware for moisture-sensitive reac-
tions was flame-dried and cooled to room temperature in a
desiccator and all reactions were carried out under an atmo-
sphere of nitrogen. Anhydrous solvents were purchased from
Aldrich and stored over 4A molecular seives. THF was freshly
distilled from Na/benzophenone ketyl under nitrogen. Flash
chromatography was performed on 230-400 mesh EM Science
silica gel 60. Melting points were recorded on a Thomas-Hoover
capillary melting point apparatus and are uncorrected. 1H and
13C NMR were obtained on a Varian Gemini-200 operating at
1
200 and 50 MHz, respectively. H and 13C NMR spectra were
recorded in units δ with tetramethylsilane (δ 0.00) and CDCl3
(δ 77.0) or DMSO-d6 (δ 39.9) as reference line internal
standards, respectively. UV spectra (200-350 nM) of metha-
nolic solutions were recorded on a Kontron Uvikon 860.
Analytical HPLC was performed on a 4.6- × 250-mm YMC
ODS-AQ 5-µm column eluting with a 0.1% aqueous AcOH/
MeOH gradient and the detector set at 280 nM. C, H, N
microanalyses were performed by NuMega Resonance Labs,
Inc., San Diego, CA, or by Robertson Microlit Laboratories,
Inc., Madison, NJ . Low-resolution mass spectral (LRMS)
analyses were performed at Mass Consortium, San Diego, CA.
En zym e Assa ys. The AMPDA and ADA Ki determinations
were performed as previously described.1
DOWEX 1×8-400 Aceta te Ion -Exch a n ge Resin . DOWEX
1×8-400 ion-exchange resin (chloride form) was first slurried
with about 3-4 volumes of water and the fine particles which

AMP Deaminase Inhibitors. 2
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1503
Step 6. A mixture of crude ethyl 2-(3-bromobenzyl)-6-tert-
butyldimethylsilyloxy-2-methylhexanoate (7.36 g, ca. 15.5
mmol) and tetrabutylammonium fluoride (24 mL of a 1 M
solution in THF) in 55 mL THF was stirred at 23 °C for 4 h
and then diluted with Et2O, washed with water and brine,
dried (MgSO4) and evaporated. Chromatography of the resi-
due and elution with a hexane/EtOAc gradient of 5:1, 4:1,
3:1 and 2:1 provided 2.63 g (49%) of ethyl 2-(3-bromoben-
zyl)-6-hydroxy-2-methylhexanoate (11c) as an oil: 1H NMR
(DMSO-d6) 1.02 (s, 3), 1.21 (t, 3, J ) 7 Hz), 1.1-1.8 (m, 6),
2.69 (d, 1, J ) 13 Hz), 2.98 (d, 1, J ) 13 Hz), 3.42 (q, 2,
J ) 7 Hz), 4.09 (q, 2, J ) 7 Hz), 4.41 (t, 1, J ) 7 Hz), 7.14 (d,
1, J ) 8 Hz), 7.28 (t, 1, J ) 8 Hz), 7.32 (s, 1), 7.45 (d, 1, J )
8 Hz).
Step 7. To an ice bath-cooled solution of ethyl 2-(3-
bromobenzyl)-6-hydroxy-2-methylhexanoate (2.5 g, 7.3 mmol)
and triethylamine (1.4 mL, 9.8 mmol) in 10 mL of CH2Cl2
was added methanesulfonyl chloride (0.71 mL, 9.1 mmol).
The mixture was then stirred at 23 °C for 1 h, diluted with
CH2Cl2, washed with water and brine, dried (MgSO4) and
evaporated to provide 3.06 g (99%) of ethyl 2-(3-bromophenyl-
methyl)-6-methanesulfonoxy-2-methylhexanoate (11d ) as an
oil: 1H NMR (DMSO-d6) 1.03 (s, 3), 1.20 (t, 3, J ) 7 Hz), 1.2-
1.8 (m, 6), 2.70 (d, 1, J ) 13 Hz), 3.00 (d, 1, J ) 13 Hz), 4.04
(q, 2, J ) 7 Hz), 4.16 (t, 2, J ) 7 Hz), 7.15 (d, 1, J ) 8 Hz),
7.27 (t, 1, J ) 8 Hz), 7.30 (s, 1H), 7.47 (d, 1, J ) 8 Hz).
5-Ca r bob en zyloxy-1-ch lor o-7-m et h yl-5-(2-m et h ylp r o-
p yl)octa n e (18). Step 1. A mixture of isopropylcyclohexyl-
amine (5.1 mL, 31.1 mmol) and n-BuLi (3.1 mL of a 10 M
solution in hexanes, 31 mmol) in 30 mL of THF was stirred at
-78 °C for 15 min (preparation of LICA) and then tert-butyl
acetate 16 (4 mL, 29.6 mmol) was added over a period of 30
min by syringe pump injection. This mixture was stirred for
an additional 30 min at -78 °C while the solution went from
cloudy to pale yellow and homogeneous. This solution was
cannulated into a solution of 1-iodo-2-methylpropane (3.6 mL,
31.1 mmol) in 30 mL of DMSO at room temperature which
produced a white suspension which was stirred for 50 min and
then diluted with 20 mL of 4 M NH4Cl. This mixture was
extracted with ether (2 × 200 mL) and the ether extract
washed with 5% aqueous HCl, water, and brine, dried (MgSO4)
and evaporated to provide 3.13 g (61%) of tert-butyl 4-meth-
ylpentanoate as an oil: 1H NMR (CDCl3) 0.86 (d, 6, J ) 7 Hz),
1.4-1.6 (m, 3), 1.41 (s, 9), 2.1-2.25 (m, 2).
Step 2. This reaction was conducted as for step 1 except
that tert-butyl 4-methylpentanoate (2.87 g, 16.6 mmol) was
added as a solution in 6 mL of THF slowly to the LICA (17.5
mmol) in 16 mL of THF at -78 °C and after stirring this for
30 min it was added to the 1-iodo-2-methylpropane (2.0 mL,
17.5 mmol) in 16 mL of DMSO. Workup as for step 1 provided
3.25 g (86%) of tert-butyl 2-(2-methylpropyl)-4-methylpen-
tanoate (17) as an oil sufficiently pure for the next step: 1H
NMR (CDCl3) 0.87 (t, 12, J ) 7 Hz), 1.4-1.6 (m, 6), 1.41 (s, 9),
2.38 (m, 1).
This general method was also used to prepare the electro-
philes used as coupling partners for the preparation of
compounds 7j-7p and 7u -7x.
Step 3. To a mixture of diisopropylamine (1.7 mL, 12 mmol)
in 12 mL of THF and 6 mL of DMPU at 5 °C was added n-BuLi
(1.16 mL of a 10 M solution in hexanes, 11.64 mmol). To this
solution was added slowly a solution of tert-butyl 2-(2-meth-
ylpropyl)-4-methylpentanoate (886 mg, 3.88 mmol) in 3 mL of
THF and this mixture was allowed to stir at 5 °C for 2 h to
provide an orange solution. Then the solution was warmed to
55 °C and 4-chloro-1-iodobutane (1.9 mL, 15.52 mmol) was
added which initiated a reflux and dispersed the color to
yellow. The temperature was maintained at 55 °C for 4 h and
then the mixture was cooled to room temperature, diluted with
aqueous 4 M NH4Cl and extracted with ether (2 × 100 mL);
the ether extract was washed with 5% aqueous HCl, water,
and brine, dried (MgSO4) and evaporated to provide 1.17 g of
residue which by 1H NMR appeared to be a 3:1 mixture of
product to starting ester. This residue was refluxed in a
mixture of 19 mL of CH2Cl2 and 1.5 mL of TFA for 3 h and
the mixture cooled, evaporated, diluted with EtOAc and
extracted with aqueous 10% K2CO3; the aqueous extract was
washed with EtOAc. The pH of the basic aqueous layer was
lowered to pH 2 with aqueous 5% HCl and then extracted with
EtOAc (3 × 100 mL). These combined EtOAc extracts were
washed with brine, dried (MgSO4) and evaporated to provide
940 mg of a residue predominantly 5-carboxy-1-chloro-7-
methyl-5-(2-methylpropyl)octane: 1H NMR (CDCl3) 0.85 (d, 6,
J ) 6 Hz), 0.88 (d, 6, J ) 6 Hz), 1.1-1.9 (m, 12), 3.55 (t, 2, J
) 7 Hz). This residue (0.94 g) was mixed with benzyl bromide
(0.64 mL, 5.4 mmol) and K2CO3 (1 g, 7.2 mmol) in 20 mL of
DMF for 18 h at room temperature. Then the mixture was
diluted with water and extracted with ether (2 × 100 mL)
and the ether extract washed with water and brine, dried
(MgSO4) and evaporated to provide 1.03 g of residue which
was subjected to chromatography eluting with hexane/
EtOAc mixtures of 100:0 (300 mL), 100:1 (450 mL) and 75:1
(500 mL) which provided 360 mg (26% from 17) of the title
ester 18 as an oil: 1H NMR (CDCl3) 0.78 (d, 6, J ) 6 Hz), 0.86
(d, 6, J ) 6 Hz), 1.4-1.8 (m, 12), 3.52 (t, 2, J ) 7 Hz), 5.05 (s,
2), 7.36 (m, 5); MS found [MH+] 353/355. Anal. (C21H33ClO2)
C, H.
In a manner identical to step 4 of the preparation of com-
pound 11d , compounds 12a , 12b, 14a and 14b were combined
with 2 equiv of 1,4-dibromobutane to provide respectively com-
pounds 13a , 13b, 15a and 15b.
Resolu tion of 2-(3-Br om op h en ylm eth yl)-6-h yd r oxy-2-
m eth ylh exa n oic Acid . (2R or S)-2-(3-Br om op h en ylm eth -
yl)-6-h yd r oxy-2-m eth ylh exa n oic Acid (11e). A solution of
methyl 2-(3-bromophenylmethyl)-6-hydroxy-2-methylhexanoate
(12.6 g, 33 mmol, prepared as in the preparation of 11d ) in
130 mL of dioxane and 66 mL of 2 M aqueous NaOH was
stirred for 16 h at 100 °C, then cooled and diluted with water
and extracted with EtOAc. The pH of the aqueous layer was
lowered to 2 with aqueous HCl and extracted with EtOAc (4
× 0.3 L) and the organic layer dried (MgSO4) and evaporated
to provide 6.60 g (63%) of 2-(3-bromophenylmethyl)-6-hydroxy-
2-methylhexanoic acid: 1H NMR (CDCl3) 1.10 (s, 3, R-CH3),
1.1-1.8 (m, 6, C3, C4 and C5 Hs), 2.69 and 3.02 (d, 1 each, J
) 14 Hz, benzylic Hs), 3.66 (t, 2, J ) 7 Hz, C6 Hs), 5.37 (br s,
1, O-H), 7.0-7.4 (m, 4, Ar Hs).
A mixture of 2-(3-bromophenylmethyl)-6-hydroxy-2-meth-
ylhexanoic acid (2.07 g, 6.6 mmol) and (-)-brucine (2.46 g, 6.24
mmol) in 35 mL of CH3CN was heated at 80 °C until solution
was obtained and then cooled to 30 °C and the resulting
crystals were collected by filtration to provide 1.6 g of the
partially resolved carboxylate (-)-brucine salt. This was
redissolved in 30 mL of CH3CN at 80 °C and allowed to
gradually cool to 30 °C to provide a white powder (mp 167-
169 °C) which was collected and dissolved in water and the
pH lowered to 2. Extraction of the aqueous mixture with
EtOAc, drying (MgSO4) the organic extract and evaporation
provided 701 mg (34%) of the resolved carboxylate 11e as a
1
gum. H NMR (CDCl3) of a 1:1 complex of this acid with (S)-
(-)-R-methylbenzylamine, characteristic shifts: 0.96 (s, 3,
R-CH3), 2.57 and 2.94 (d, 1 each, J ) 13 Hz, benzylic Hs). The
first CH3CN filtrate was processed in a manner similar to
provide a sample enriched in the opposite enantiomer. 1H
NMR (CDCl3) of a 1:1 complex of this acid mixture with (S)-
(-)-R-methylbenzylamine showed characteristic shifts of the
enantiomer which does not crystallize first with (-)-brucine:
0.98 (s, 3, R-CH3), 2.54 and 2.90 (d, 1 each, J ) 13 Hz, ben-
zylic Hs).
This compound was converted to the resolved analogue of
11d by standard procedures (step 2 with iodomethane and step
7 in the preparation of compound 11d ) and used to synthesize
compound 7w .
5-(5-Br om op en tyl)-2-(2-(tr im eth ylsilyl)eth oxym eth yl)-
tetr a zole (19b). Step 1. A 60% suspension of NaH in min-
eral oil (1.26 g, 31.4 mmol) was slurried with 4 mL of dry
hexane, the hexane decanted by syringe under N2 (2×) and
the solid dried under a high vacuum. To the solid was added
100 mL of DMF and while the suspension stirred at 5 °C a
solution of tetrazole (2.00 g, 28.5 mmol) in 40 mL of DMF was

1504 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Bookser et al.
added. After this mixture stirred for 30 min, H2 evolution had
ceased. The reagent 2-(trimethylsilyl)ethoxymethyl chloride
(5.30 mL, 29.9 mmol) was added and the mixture allowed to
stir at room temperature for 16 h. Then the mixture was
diluted with 250 mL of ether, washed with water (2 × 200
mL) and brine, dried (MgSO4) and evaporated to provide an
oil which was subjected to chromatography eluting with a
hexane/EtOAc gradient of 20:1, 15:1, and 10:1 to provide 1.3
g (23%) of the less polar regioisomer assigned as 2-(2-
(trimethylsilyl)ethoxymethyl)tetrazole, 19a : 1H NMR (CDCl3)
0.04 (s, 9), 0.96 (t, 2, J ) 7 Hz), 3.72 (t, 2, J ) 7 Hz), 5.95 (s,
2), 8.60 (s, 1).
Step 2. To a solution of 19a (500 mg, 2.5 mmol) in 10 mL
of THF and 10 mL of DMPU at -78 °C was added t-BuLi (3.67
mL of a 1.7 M solution in hexanes, 6.24 mmol) which produced
an orange suspension. This was allowed to stir at -78 °C for
5 min and then 1,5-dibromopentane (1.71 mL, 12.5 mmol) was
added; this mixture stirred at -23 °C for 7 h to produce a
yellow solution. Aqueous NH4Cl was added and the mixture
extracted with ether (2 × 100 mL). The combined ether
extracts were washed with water (2 × 200 mL) and brine, dried
(MgSO4) and evaporated to provide an oil which was sub-
jected to chromatography eluting with a hexane/EtOAc gradi-
ent of 20:0, 20:1, 10:1 and 5:1 which provided 468 mg of 75%
pure 19b with the 25% impurity being 19a . This material
was used to alkylate compound 4 without further purifi-
cation. Compound 19b: 1H NMR (CDCl3) 0.02 (s, 9), 0.97 (t,
2, J ) 7 Hz), 1.57 (quintet, 2, J ) 7 Hz), 1.8-2.1 (m, 4), 2.98
(t, 2, J ) 7 Hz), 3.45 (t, 2, J ) 7 Hz), 3.71 (t, 2, J ) 7 Hz), 5.86
(s, 2).
Diben zyl 5-Br om op en tylp h osp h on a te (21b). A 60%
suspension of NaH in mineral oil (2.00 g, 50 mmol) was
slurried with 5 mL of dry hexane, the hexane decanted by
syringe under N2 (2×) and the solid dried under a high
vacuum. To the solid was added 250 mL of DMF and while
this suspension stirred at 5 °C, dibenzyl phosphite (21a ) (10.0
mL, 45.5 mmol) was added slowly. After this mixture stirred
for 1 h, 1,5-dibromopentane (7.5 mL, 55 mmol) was added at
5 °C and the resulting mixture allowed to stir at room
temperature for 16 h, then it was diluted with ether, washed
with water (2×) and brine, dried (MgSO4) and evaporated to
provide an oil which was subjected to chromatography eluting
first with CH2Cl2 and then with 18:1 CH2Cl2/MeOH to provide
9.03 g (49%) of the title compound as an oil: 1H NMR (CDCl3)
1.3-2.0 (m, 8), 3.46 (t, 2, J ) 7 Hz), 4.9-5.1 (m, 4), 7.25-7.45
(m, 10).
This method was used to prepare all compounds 7 or the
ester precursors of compounds 7 and compound 6f′ (the
reduced product of 6f in Figure 2).
1
Compound 7f: mp 115-118 °C; H NMR (DMSO-d6) 1.17
(t, 3, J ) 7 Hz), 1.2-1.8 (m, 6), 2.26 (t, 2, J ) 7 Hz), 3.15 (br
s, 2), 3.84 (t, 2, J ) 7 Hz), 4.03 (q, 2, J ) 7 Hz), 4.80 (br s, 1),
4.91 (d, 1, J ) 6 Hz), 6.95 (d, 1, J ) 4 Hz), 7.28 (s, 1), 7.45 (d,
1, J ) 4 Hz); UV (methanol) λmax 283 nm (ꢀ 6545). Anal.
(C14H22N4O3) C, H, N.
Compound 6f′: 1H NMR (DMSO-d6) 1.17 (t, 3, J ) 7 Hz),
1.2-1.8 (m, 6), 2.27 (t, 2, J ) 7 Hz), 3.36 (br s, 2), 3.79 (t, 2, J
) 7 Hz), 4.04 (q, 2, J ) 7 Hz), 4.62 (br s, 1), 5.30 (br s, 1), 7.12
(s, 1), 8.00 (s, 1); UV (methanol) λmax 278 nm (ꢀ 4680); APCI
MS (positive mode) M + 1 calcd 295, found m/z 295.
Isola tion of Rep r esen ta tive N1- a n d N3-Alk yla tion
P r od u cts. 3-(5-Ca r beth oxyp en tyl)-6,7-d ih yd r oim id a zo-
[4,5-d ][1,3]d ia zep in -8(3H)-on e (5f) a n d 1-(5-Ca r beth oxy-
p en t yl)-6,7-d ih yd r oim id a zo[4,5-d ][1,3]d ia zep in -8(1H )-
on e (6f). Reaction of the base 4 (37.8 mmol) with ethyl
6-bromohexanoate (37.8 mmol) as previously described1 pro-
vided an N3/N1-alkylation mixture which was subjected to
chromatography on SiO2 eluting with a CH2Cl2/methanol
gradient of 20:1, 17.5:1 and 15:1 to provide a fraction which
was reduced to a minimal volume and diluted with ether. The
resulting solid was collected to provide 1.30 g of the N3-isomer
5f as a tan solid. The 1H and 13C NMR assignments were made
1
1
based in part on H-1H coupling (COSY) and H-13C coupling
(HETCOR) two-dimensional NMR experiments. Compound
5f: mp 128-131 °C; 1H NMR (DMSO-d6) 1.17 (t, 3, J ) 7 Hz,
ester CH3), 1.25 (quin, 2, J ) 7 Hz, C3′ Hs), 1.55 (quin, 2, J )
7 Hz, C4′ Hs), 1.71 (quin, 2, J ) 7 Hz, C2′ Hs), 2.28 (t, 2, J )
7 Hz, C5′ Hs), 3.78 (d, 2, J ) 4 Hz, C7 Hs), 3.96 (t, 2, J ) 7
Hz, C1′ Hs), 4.04 (q, 2, J ) 7 Hz, ester CH2), 7.45 (d, 1, J ) 4
Hz, H5), 7.67 (s, 1, H2), 8.4-8.5 (m, 1, H6); 13C NMR (DMSO-
d6) 14.1 (ester CH3), 24.0 (C4′), 25.4 (C3′), 29.4 (C2′), 33.3 (C5′),
42.4 (C1′), 51.5 (C7), 59.7 (ester CH2), 128.7 (C8a), 136.3 (C2),
146.4 (C3a), 151.1 (C5), 172.8 (C6′), 181.1 (C8); UV (methanol)
λmax 231 nm (ꢀ 14300), 235 (14300), 302 (4,700). Anal.
(C14H20N4O3) C, H, N.
Further elution with 15:1 CH2Cl2/methanol provided 2.04
g of a mixture of N3- and N1-isomers (5f and 6f). This was
subjected to MPLC eluting with a CH2Cl2/methanol gradient
of 17.5:1 and 15:1 which provided 875 mg of additional 5f. The
total yield of 5f was 2.17 g (20%). Further elution with 12.5:1
provided a fraction which was reduced to a minimal volume
and diluted with ether. The resulting solid was collected to
provide 676 mg (6%) of the N1-isomer 6f as orange crystals.
The 1H and 13C NMR assignments were made based in part
on 1H-1H coupling (COSY) and 1H-13C coupling (HETCOR)
two-dimensional NMR experiments. Compound 6f: mp 113-
This method was also used to prepare the phosphonate
electrophiles 20b and 21c.
3-(6-(3-Br om op h en yl)-5-ca r beth oxy-5-m eth ylh exyl)co-
for m ycin Aglycon (7u ). Step 1, Alk yla tion . According to
the method previously described1 ketone 4 (1.09 g, 7.3 mmol)
was alkylated with 11d (3.07 g, 7.3 mmol). Chromatography
on SiO2 with a CH2Cl2/methanol gradient of 20:1, 18:1 and 16:1
provided 1.08 g (31%) of 3-(6-(3-bromophenyl)-5-carbethoxy-
5-methylhexyl)-6,7-dihydroimidazo[4,5-d][1,3]diazepin-8(3H)-
one (5u ) as a light yellow gum: 1H NMR (DMSO-d6) 0.96 (s,
3), 1.0-1.6 (m, 4), 1.13 (t, 3H, J ) 7 Hz), 1.70 (quin, 2, J ) 7
Hz), 2.65 (d, 1, J ) 13 Hz), 2.92 (d, 1, J ) 13 Hz), 3.73 (d, 2,
J ) 4 Hz), 3.94 (t, 2, J ) 7 Hz), 4.02 (q, 2, J ) 7 Hz), 7.08 (d,
1, J ) 8 Hz), 7.23 (t, 1, J ) 8 Hz), 7.27 (s, 1), 7.42 (d, 1, J ) 8
Hz), 7.43 (d, 1, J ) 4 Hz), 7.62 (s, 1), 8.34 (br m, 1).
1
116 °C; H NMR (DMSO-d6) 1.17 (t, 3, J ) 7 Hz, ester CH3),
1.25 (quin, 2, J ) 7 Hz, C3′ Hs), 1.53 (quin, 2, J ) 7 Hz, C4′
Hs), 1.67 (quin, 2, J ) 7 Hz, C2′ Hs), 2.27 (t, 2, J ) 7 Hz, C5′
Hs), 3.8 (br s, 2, C7 Hs), 4.04 (q, 2, J ) 7 Hz, ester CH2), 4.18
(t, 2, J ) 7 Hz, C1′ Hs), 7.37 (br s, 1, H5), 7.84 (s, 1, H2), 8.2
(br s, 1/2, H6), 10.6 (br s, 1/2, H4); 13C NMR (DMSO-d6) 14.1
(ester CH3), 23.9 (C4′), 25.1 (C3′), 30.1 (C2′), 33.3 (C5′), 46.3
(C1′), 55.5 (br s, C7), 59.6 (ester CH2), 117.7 (C8a), 141.5 (C2),
153.5 (br s, C3a), 148.5 (C5), 172.7 (C6′), 181.2 (C8); UV
(methanol) λmax 231 nm (ꢀ 17100), 235 (16900), 301 (5020).
Anal. (C14H20N4O3) C, H, N.
St ep 2, R ed u ct ion . According to the method previously
described1 5u (1.08 g, 2.27 mmol)was reduced with NaBH4 (106
mg, 2.72 mmol). Chromatography on SiO2 with a CH2Cl2/
methanol/triethylamine gradient of 25:1:0.25 and 20:1:0.2
provided a fraction which was reduced to a minimal volume
and diluted with ether. The resulting solid was collected to
provide 802 mg (74%) of 7u as a white solid: mp 133-136 °C;
1H NMR (DMSO-d6) 0.97 (s, 3), 1.15 (t, 3H, J ) 7 Hz), 1.2-
1.8 (m, 6), 2.65 (d, 1, J ) 13 Hz), 2.93 (d, 1, J ) 13 Hz), 3.15
(br s, 2), 3.84 (t, 2, J ) 7 Hz), 4.03 (q, 2, J ) 7 Hz), 4.80 (br s,
1), 4.90 (br s, 1), 6.96 (d, 1, J ) 4 Hz), 7.08 (d, 1, J ) 8 Hz),
7.24 (t, 1, J ) 8 Hz), 7.28 (s, 1), 7.42 (d, 1, J ) 8 Hz), 7.45 (br
s, 1). Anal. (C22H29BrN4O3) C, H, N.
Alter n a te Meth od for th e P r ep a r a tion of 3-Alk yl-6,7-
d ih yd r oim id a zo[4,5-d ][1,3]d ia zep in -8(3H)-on es. 3-Ben -
zyl-6,7-dih ydr oim idazo[4,5-d][1,3]diazepin -8(3H)-on e (5f ′).
To a suspension of 2-amino-1-(1-benzyl-5-amino-1H-imidazol-
4-yl)ethanone dihydrochloride (8)16a (1.00 g, 3.3 mmol) in
16.5 mL of DMSO at 70 °C was added triethyl orthoformate
(2.74 mL, 16.5 mmol). After 15 min the resulting solution was
cooled to room temperature and stirred with Darco G-60
charcoal and filtered. The filtrate was concentrated under high
vacuum distillation to a 5 mL volume and then poured onto
ether. The solvent was decanted off of the resulting red oil
and it was mixed with methanol. The resulting white precipi-
tate was filtered and the filtrate concentrated and then diluted

AMP Deaminase Inhibitors. 2
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1505
with CH2Cl2 and applied to the head of a chromatography
column previously eluted with 10:1 CH2Cl2/methanol. Elu-
tion with 10:1 CH2Cl2/methanol provided 150 mg (19%) of 5f ′
as a tan solid: 1H NMR (DMSO-d6) 3.76 (d, 2, J ) 4 Hz), 5.20
(s, 2), 7.1-7.4 (m, 5), 7.44 (d, 1, J ) 4 Hz), 7.74 (s, 1), 8.41 (br
s, 1).
provided a fraction which was lyophilized to provide 20 mg
(22%) of the desired monoacid 7c′ as a solid: mp 67-68 °C;
1H NMR (DMSO-d6) 1.2-1.8 (m, 10), 3.2-3.4 (m, 2), 3.53 (d,
3, J ) 10 Hz), 4.19 (t, 2, J ) 7 Hz), 4.9-5.0 (m, 1), 7.36 (d, 1,
J ) 4 Hz), 9.03 (s, 1). Anal. (C13H23N4O4P‚2H2O) C, H, N.
3-(6-Hyd r oxyh exyl)cofor m ycin Aglycon (7q′). To 3-(6-
acetoxyhexyl)coformycin aglycon (90 mg, 0.32 mmol) was
added 1.6 mL of a freshly prepared 0.5 M MeONa solution in
MeOH (prepared by dissolving 230 mg of sodium in 10 mL of
MeOH) and the mixture stirred for 30 min at room tempera-
ture, then neutralized with Amberlite CG-50 resin, and
filtered; the filtrate was evaporated. The residual water was
removed by lyophilization and the residue precipitated from
a concentrate in methanol by the addition of ether. The solid
was collected by filtration to provide 12 mg (15%) of the alcohol
7q′ as a white solid: mp 138-139 °C; 1H NMR (DMSO-d6)
1.2-1.7 (m, 8), 3.17 (m, 2), 3.85 (t, 2, J ) 7 Hz), 4.35 (br s, 2),
4.82 (br s), 4.91 (d, 1, J ) 6 Hz), 6.97 (d, 1, J ) 4 Hz), 7.29 (s,
1), 7.4-7.5 (m, 1).
3-(5-Ca r boxy-5-m eth yl-6-(3-br om op h en yl)h exyl)cofor -
m ycin Aglycon (7v). The ester 7u (614 mg, 1.29 mmol) was
hydrolyzed and isolated with DOWEX 1×8-400 acetate ion-
exchange resin as previously described1 to provide 440 mg
(69%) of the carboxylate 7v as a pink solid: Rf 0.51 (3:1 CH3-
CN/0.2 N NH4Cl); mp 83-88 °C; 1H NMR (DMSO-d6) 0.92 (s,
3), 1.0-1.7 (m, 6), 2.61 (d, 1, J ) 12 Hz), 2.92 (d, 1, J ) 12
Hz), 3.16 (br s, 2), 3.84 (t, 2, J ) 7 Hz,), 4.81 (br s, 1), 6.97 (d,
1, J ) 4 Hz), 7.13 (d, 1, J ) 8 Hz), 7.21 (t, 1, J ) 8 Hz), 7.33
(s, 1,), 7.29 (s, 1), 7.40 (d, 1, J ) 8 Hz). Anal. (C20H25
BrN4O3‚0.8CH3CO2H‚1H2O) C, H, N.
-
This method was also used to prepare carboxylates 7j, 7k ,
7m -7p , 7r , 7t and 7v-7x. Note: in general, hydrolysis of
methyl esters was more rapid.
This method was also used to prepare compounds 7m ′-7p ′.
3-(5-(Ca r ba m oyl)p en tyl)cofor m ycin Aglycon (7r ′). A
solution of 3-(5-carbethoxypentyl)coformycin aglycon (7f) (60
mg, 0.20 mmol) in 2 mL of 15% NH3/MeOH was heated in a
closed bomb apparatus at 100 °C for 72 h, then cooled and
evaporated. The product was precipitated from a concentrate
in methanol by the addition of ether. The resulting solid was
collected to provide 48 mg (87%) of the primary amide 7r ′ as
Alternatively, in a limited number of examples the aqueous
layer was neutralized with Amberlite CG-50 resin and filtered
and the filtrate lyophilized to provide the sodium salt of the
acid as a deliquescent solid. This solid was handled as a
suspension under ether or acetone; then the solvent was
decanted off and the solid dried under a high vacuum. This
method was used to prepare 7a , 7b and 7i. This method
produced compounds 7c and 7d as methanol-insoluble free
acids.
1
a tan solid: mp 159-163 °C; H NMR (DMSO-d6) 1.21 (quin,
2, J ) 7 Hz), 1.48 (quin, 2, J ) 7 Hz), 1.64 (quin, 2, J ) 7 Hz),
2.02 (t, 2, J ) 7 Hz), 3.1-3.2 (m, 2), 3.83 (t, 2, J ) 7 Hz), 4.80
(t, 1, J ) 3 Hz), 4.9 (br s, 1), 6.7 (br s, 1), 6.96 (d, 1, J ) 4 Hz),
7.23 (br s, 1), 7.28 (s, 1), 7.45 (br d, 1, J ) 4 Hz). Anal.
(C12H19N5O2‚0.4H2O) C, H, N.
Rep r esen ta tive Ben zyl Ester Hyd r ogen olysis P r oce-
d u r e. P r ep a r a tion of 3-(5-Ca r boxy-5,5-d ip h en ylp en tyl)-
cofor m ycin Aglycon (7a ′). 3-(5-Carbobenzyloxy-5,5-diphen-
ylpentyl)coformycin aglycon (181 mg, 0.36 mmol) was hydro-
genated (30 psi H2) in 5 mL of methanol in the presence of
10% Pd(OH)2 on carbon (25 mg) for 4 h. The mixture was
filtered over Celite and solvent concentrated and then diluted
with ether. The resulting solid was collected to provide 124
Ack n ow led gm en t. We thank Ms. Maureen Cottrell,
Mr. Patrick McCurley, Mr. Colin Ingraham, and Dr.
Ken Takabayashi for their scientific contributions to this
work, and we give special thanks to Dr. Harry Gruber
for his encouragement and suggestions.
1
mg (74%) of 7a ′ as a white solid: mp 200-201 °C; H NMR
(DMSO-d6) 0.9-1.1 (m, 2), 1.5-1.7 (m, 2), 2.2-2.4 (m, 2), 3.13
(br s, 2), 3.76 (t, 2, J ) 7 Hz), 4.77 (br s, 1), 6.93 (d, 1, J ) 4
Hz), 7.1-7.5 (M, 11), 7.50 (m, 1). Anal. (C24H26N4O3‚2.75H2O)
C, H, N.
Su p p or tin g In for m a tion Ava ila ble: Lists of key inter-
1
mediates and electrophiles and H NMR data of key interme-
diates and electrophiles. This material is available free of
charge via the Internet at http://pubs.acs.org.
This method was also used to prepare carboxylates 7y and
7z and the phosphonates 7d ′ and 7e′.
Refer en ces
3-(5-(Tetr a zol-5-yl)p en tyl)cofor m ycin Aglycon Dicy-
cloh exyla m m on iu m Sa lt (7b′). A mixture of 3-(5-(2-(2-
(trimethylsilyl)ethoxymethyl)tetrazol-5-yl)pentyl)coformycin
aglycon (14 mg, 0.033 mmol) (prepared according to the
method described for the preparation of 7u ), n-Bu4NF (0.05
mL of a 1 M solution in THF) and CsF (10 mg, 0.066 mmol) in
0.5 mL of DMF was stirred at 50 °C for 2 h and then the sol-
vent removed by rotary evaporation. The residue was dissolved
in water and slurried with DOWEX 1×8-400 acetate for 1 h
and filtered. The resin was washed with water (3×) and then
slurried with 0.05 Μ ΑcOH for 15 min and filtered (3×). The
combined filtrates were lyophilized; the residue was dissolved
in methanol and dicyclohexylamine (13 µL, 0.065 mmol) added.
The mixture was evaporated and then suspended in acetone,
the acetone decanted and suspended in ether, the ether
decanted and the resulting solid dried under vacuum to
provide 6 mg (39%) of 7b′ as a pale orange solid: mp 220 °C;
1H NMR (DMSO-d6) 1.0-2.0 (m, 26), 2.64 (t, 2, J ) 7 Hz), 2.96
(m, 2), 3.15 (br s, 2), 3.82 (t, 2, J ) 7 Hz), 4.79 (br s, 1), 6.96
(d, 1, J ) 4 Hz), 7.27 (s, 1), 7.43 (m, 1). Anal. (C12H18N8O‚
1C12H23N‚2H2O) C, H; N: calcd, 24.83; found, 23.80.
3-(6-(Mon om e t h ylp h osp h on oxy)h e xyl)cofor m ycin
Aglycon (7c′). A mixture of 3-(6-(dimethylphosphonoxy)-
hexyl)coformycin aglycon (98 mg, 0.28 mmol) and lithium
methoxide (162 mg, 4.3 mmol) was stirred for 72 h at 50 °C
and then the pH reduced to 1-2 with the addition of metha-
nolic HCl. Reverse-phase (C-18) silica gel (450 mg) was added
and the solvent evaporated. The resulting powder was loaded
onto a column packed with 5 g of reverse phase silica gel in
water. Elution with water and then with 4% methanolic water
(1) For AMP Deaminase Inhibitors. 1, see: Erion, M. D.; Kasibhatla,
S. R.; Bookser, B. C.; van Poelje, P. D.; Reddy, M. R.; Gruber,
H. E.; Appleman, J . R. Discovery of AMP Mimetics that Exhibit
High Inhibitory Potency and Specificity for AMP Deaminase. J .
Am. Chem. Soc. 1999, 121, 308-319.
(2) Skladanowski, A. C. In Myocardial Energy Metabolism; de J ong,
J . W., Ed.; Martinus Nijhoff: Dordrecht, 1988; pp 53-65.
(3) For lead references on adenosine-related therapeutics research,
see: (a) Adenosine and Adenosine Receptors; Williams, M., Ed.;
Humana Press: Clifton, NJ , 1990. (b) Purinergic Approaches in
Experimental Therapeutics; J acobson, K. A., J arvis, M. F., Eds.;
Wiley-Liss: New York, 1997.
(4) Loh, E.; Rebbeck, T. R.; Mahoney, P. D.; DeNofrio, D.; Swain, J .
L.; Holmes, E. W. Common Variant in AMPD1 Gene Predicts
Improved Clinical Outcome in Patients With Heart Failure.
Circulation 1999, 99, 1422-1425.
(5) In addition to its potential antiischemic uses, recent reports
indicate AMPDA inhibitors may be worthwhile to investigate
as herbicidal and anti-Alzheimer’s agents: (a) Lindell, S. D.;
Moloney, B. A.; Hewitt, B. D.; Earnshaw, C. G.; Dudfield, P. J .;
Dancer, J . E. The Design and Synthesis of Inhibitors of Adeno-
sine-5′-Phosphate Deaminase. Bioorg. Med. Chem. Lett. 1999,
9, 1985-1990. (b) Sims, B.; Powers, R. E.; Sabina, R. L.;
Theibert, A. B. Elevated Adenosine Monophosphate Deaminase
Activity in Alzheimer’s Disease Brain. Neurobiol. Aging 1998,
19, 385-391.
(6) (a) Hapke, D. M.; Stegmann, A. P. A.; Mitchell, B. S. Retroviral
Transfer of Deoxycytidine Kinase into Tumor Cell Lines En-
hances Nucleoside Toxicity. Cancer Res. 1996, 56, 2343-2347.
(b) Maehara, Y.; Kusumoto, T.; Yasunori, E.; Anai, H.; Sakagu-
chi, Y.; Kohnoe, S.; Sugimachi, K. 5-Fluorouracil Is Converted
to F-Nucleotides More Extensively and is More Cytotoxic in
Poorly Differentiated than in Well Differentiated Human Gastric
Carcinoma. Anticancer Res. 1990, 10, 1091-1094.

1506 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Bookser et al.
(7) (a) Starrett, J . E., J r.; Tortolani, D. R.; Hitchcock, M. J . M.;
Martin, J . C.; Mansuri, M. M. Synthesis and in vitro evaluation
of a phosphonate prodrug: bis(pivaloyloxymethyl) 9-(2-phospho-
nylmethoxyethyl)adenine. Antiviral Res. 1992, 19, 267-273. (b)
Srivastva, D. N.; Farquhar, D. Bioreversible Phosphate Protec-
tive Groups: Synthesis and Stability of Model Acyloxymethyl
Phosphates. Bioorg. Chem. 1984, 12, 118-129. (c) Krise, J . P.;
Stella, V. J . Prodrugs of phosphates, phosphonates and phos-
phinates. Adv. Drug Delivery Rev. 1996, 19, 287-310. (d) J ones,
R. J .; Bischofberger, N. Minireview: nucleotide prodrugs. An-
tiviral Res. 1995, 27, 1-17.
(8) Isolation and identification: (a) Ohno, M.; Yagisawa, N.; Shiba-
hara, S.; Kondo, S.; Maeda, K.; Umezawa, H. Structure of
Coformycin, an Unusual Nucleoside of Microbial Origin. J . Am.
Chem. Soc. 1974, 96, 4327-4328. The 8R/S-isomer was prepared
for this study according to literature methods: (b) Hawkins, L.
D.; Hanvey, J . C.; Boyd, F. L. J r.; Baker, D. C. Inhibitors of
Adenosine Deaminase. Synthesis of Coformycin and 3′-Deoxy-
coformycin. Nucleosides Nucleotides 1983, 2, 479-494. (c) Show-
alter, H. D. H.; Putt, S. R.; Baker, D. C. In Nucleic Acid
Chemistry; Townsend, L. B., Tipson, R. S., Eds.; Wiley: New
York, 1991; pp 308-313.
Biosynthesis and Studies Related to Compounds of Interest in
Cancer Chemotherapy. Ph.D. Dissertation, University of Ala-
bama, 1985. (d) Thomas, H. J .; Riordan, J . M.; Montgomery, J .
A. The Synthesis of Coformycin from 5-Amino-1-â-D-ribofura-
nosylimidazole-4-carboxamide. Nucleosides Nucleotides 1986, 5,
431-439. (e) Saville-Stones, E. A.; Turner, R. M.; Lindell, S. D.;
J ennings, N. S.; Head, J . C.; Carver, D. S. Total Synthesis of
Carbocyclic Analogues of Coformycin. Tetrahedron 1994, 50,
6695-6704. (f) Truong, T. V.; Rapoport, H. Chirospecific Syn-
thesis of the Tetrahydroimidazodiazepinol Aglycon of Pentostatin
and Its Analogues. J . Org. Chem. 1993, 58, 6090-6096. (g)
Truong, T. V. Chirospecific Synthesis of Pentostatin and Its
Analogues. Ph.D. Dissertation, University of California, Berke-
ley, 1989.
(17) The t1/2 values for the decomposition of example ketone i were
determined at low pHs at T ) 30 °C by following the change in
UV absorbance: pH (t1/2 min) 1 (0.3), 2 (3).
(9) Frieden, C.; Kurz, L. C.; Gilbert, H. R. Adenosine Deaminase
and Adenylate Deaminase: Comparative Kinetic Studies with
Transition State and Ground-State Analogue Inhibitors. Bio-
chemistry 1980, 19, 5303-5309.
(10) (a) Hohl, C. M.; Wimsatt, D. K.; Brierley, G. P.; Altschuld IMP
Production by ATP-Depleted Adult Rat Heart Cells Effects of
Glycolysis and R1-Adrenergic Stimulation. Circ. Res. 1989, 65,
754-760. (b) Bennett, L. L., J r.; Brockman, R. W.; Allan, P. W.;
Rose, L. M.; Shaddix, S. C. Alterations in Nucleotide Pools
Induced by 3-Deazaadenosine and Related Compounds Role of
Adenylate Deaminase. Biochem. Pharmacol. 1988, 37, 1233-
1244. (c) Zoref-Shani, E.; Shainberg, A.; Sperling, O. Pathways
of adenine nucleotide catabolism in primary rat muscle cultures.
Biochim. Biophys. Acta 1987, 926, 287-295. (d) Ashby, B.;
Wernick, E.; Holmsen, H. Coformycin Inhibition of Platelet AMP
Deaminase Has No Effect on Thrombin-induced Platelet Secre-
tion nor on Glycolysis or Glycogenolysis. J . Biol. Chem. 1983,
258, 321-325. (e) Fishbein, W. N.; Davis, J . I.; Winkert, J . W.;
Strong, D. M. Levels of Adenosine Deaminase, AMP Deaminase,
and Adenylate Kinase in Cultured Human Lymphoblast Lines:
Exquisite Sensitivity of AMP Deaminase to Adenosine Deami-
nase Inhibitors. Biochem. Med. 1981, 26, 377-386. (f) Van den
Berghe, G.; Bontemps, F.; Hers, H.-G. Purine Catabolism in
Isolated Rat Hepatocytes Influence of Coformycin. Biochem. J .
1980, 188, 913-920.
(18) Al-Razzak, L. A.; Benedetti, A. E.; Waugh, W. N.; Stella, V. J .
Chemical Stability of Pentostatin (NSC-218321), a Cytotoxic and
Immunosuppressant Agent. Pharm. Res. 1990, 7, 452-460.
(19) The t1/2 values for the decomposition of example alcohol 7k ′ were
determined at low pHs at T ) 35 °C by following the change in
UV absorbance: pH (t1/2 min) 1 (23), 2 (47). The pKas for
bisprotonated 7k ′ were determined to be 6.5 and 2.1.
(20) (a) Iyer, R. S.; Voehler, M. W.; Harris, T. M. Adenine Adduct of
Aflatoxin B1 Epoxide. J . Am. Chem. Soc. 1994, 116, 8863-8869.
(b) Pugmire, R. J .; Grant, D. M.; Townsend, L. B.; Robins, R. K.
Carbon-13 Magnetic Resonance. XXII. The N-Methylpurines. J .
Am. Chem. Soc. 1973, 95, 2791-2796.
(21) The electrophile used to prepare 7i was synthesized according
to a literature method: Shchaveleva, I. L.; Kochetkova, S. V.;
Tsilevich, T. L.; Vladyko, G. V.; Korobchenko, L. V.; Boreko, E.
I.; Smirnov, I. P.; Khorlin, A. A.; Gottikh, B. P.; Florent’ev, V.
L. Preparation of Chloromethylalkylcarboxy Esters. Sov. J .
Bioorg. Chem. 1990, 15, 523-527.
(22) Compound 14b was prepared according to the method described
by: (a) Chen, B.-C.; Lue, P. A Convenient Preparation of 5,5-
Dialkyl Meldrum’s Acids. Org. Prep. Proced. Int. 1992, 24, 185-
188 and (b) Oikawa, Y.; Hirasawa, H.; Yonemitsu, O. Meldrum’s
Acid in Organic Synthesis. 1. A Convenient One-Pot Synthesis
of Ethyl Indolepropionates. Tetrahedron Lett. 1978, 1759-1762.
(23) Rathke, M. W.; Lindert, A. The Reaction of Lithium N-Isopro-
pylcyclohexylamide with Esters. A Method for the Formation
and Alkylation of Ester Enolates. J . Am. Chem. Soc. 1971, 93,
2318-2320.
(11) Cha, S.; Agarwal, R. P.; Parks, R. E., J r. Tight Binding
Inhibitors-II Non-Steady-State Nature of Inhibition of Milk
Xanthine Oxidase by Allopurinol and Alloxanthine and of
Human Erythrocytic Adenosine Deaminase by Coformycin.
Biochem. Pharmacol. 1975, 24, 2187-2197.
(12) Kredich, N. M.; Hershfield, M. S. In The Metabolic Basis of
Inherited Disease; Scrivner et al., Eds.; McGraw-Hill: New York,
1989; pp 1045-1075.
(13) Wolfenden, R. Analogue Approaches to the Structure of the
Transition State in Enzyme Reactions. Acc. Chem. Res. 1972,
5, 10-18.
(14) (a) Marrone, T. J .; Straatsma, T. P.; Briggs, J . M.; Wilson, D.
K.; Quiocho, F. A.; McCammon, J . A. Theoretical Study of
Inhibition of Adenosine Deaminase by (8R)-Coformycin and (8R)-
Deoxycoformycin. J . Med. Chem. 1996, 39, 277-284. (b) Kline,
P. C.; Schramm, V. L. Electrostatic Potential Surfaces of the
Transition State for AMP Deaminase and for (R)-Coformycin, a
Transition State Inhibitor. J . Biol. Chem. 1994, 269, 22385-
(24) The application of the method used to prepare compounds 13
and 15 to the very hindered ester iii resulted in apparent
deprotonation of the benzylic methylene, and the resulting anion
Claisen-condensed with another molecule of ester to produce the
unusual product iv in high yielda:
22390. (c) Kati, W. M.; Acheson, S. A.; Wolfenden, R.
A
Transition State in Pieces: Major Contributions of Entropic
Effects to Ligand Binding by Adenosine Deaminase. Biochem-
istry 1992, 31, 7356-7366.
(15) See the following papers in this issue: (a) Kasibhatla, S. R.;
Bookser, B. C.; Probst, G.; Appleman, J . R.; Erion, M. D. AMP
Deaminase Inhibitors. 3. SAR of 3-(Carboxyarylalkyl)coformycin
Aglycon Analogues. J . Med. Chem. 2000, 43, 1508-1518. (b)
Bookser, B. C.; Kasibhatla, S. R.; Erion, M. D. AMP Deaminase
Inhibitors. 4. Further N3-Substituted Coformycin Aglycon Ana-
logues. N3-Alkylmalonates as Ribose 5’-Monophosphate Mimet-
ics. J . Med. Chem. 2000, 43, 1519-1524.
(16) For literature relating to the synthesis of coformycin analogues,
see ref 8 and: (a) Chan, E.; Putt, S. R.; Showalter, H. D. H.;
Baker, D. C. Total Synthesis of (8R)-3-(2-Deoxy-â-D-erythro-
pentofuranosyl)-3,6,7,8-tetrahydroimidazo[4,5-d][1,3]diazepin-8-
ol (Pentostatin), the Potent Inhibitor of Adenosine Deaminase.
J . Org. Chem. 1982, 47, 3457-3464. (b) Showalter, H. D. H.;
Putt, S. R.; Borondy, P. E.; Shillis, J . L. Adenosine Deaminase
Inhibitors. Synthesis and Biological Evaluation of (()-3,6,7,8-
Tetrahydro-3-[(2-hydroxyethoxy)methyl]imidazo[4,5-d][1,3]-
diazepin-8-ol and Some Selected C-5 Homologues of Pentostatin.
J . Med. Chem. 1983, 26, 1478-1482. (c) Smal, E. Synthesis,

AMP Deaminase Inhibitors. 2
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1507
Compound iv: 1H NMR (CDCl3) 0.8-1.0 (m, 24, methyls), 1.1-
1.4 (m, 4, i-Bu methines), 1.5-1.8 (m, 8, methylenes), 2.55-2.8
(m, 2, C-R methines); MS found [MH+] 417, [MNa+] 439, [MK+]
455. For a similar observation see: Upton, C. L.; Beak, P. Dipole
Stabilized Carbanions. Reactions of Benzoate Esters with Lithium
2,2,6,6-Tetramethylpiperidide. J . Org. Chem. 1975, 40, 1094-
1098.
(29) (a) Hampton, A.; Sasaki, T.; Perini, F.; Slotin, L. A.; Kappler, F.
Design of Substrate-Site-Directed Irreversible Inhibitors of
Adenosine 5′-Phosphate Aminohydrolase. Effect of Substrate
Substituents on Affinity for the Substrate Site. J . Med. Chem.
1976, 19, 1029-1033. (b) Merkler, D. J .; Schramm, V. L.
Catalytic Mechanism of Yeast Adenosine 5′-Monophosphate
Deaminase. Zinc Content, Substrate Specificity, pH Studies and
Solvent Isotope Effects. Biochemistry 1993, 32, 5792-5799.
(30) For X-ray crystallographic examples of this type of interaction,
see: (a) Erion, M. D.; Niwas, S.; Rose, J . D.; Ananthan, S.; Allen,
M.; Secrist, III, J . A.; Babu, Y. S.; Bugg, C. E.; Guida, W. C.;
Ealick, S. E.; Montgomery, J . A. Structure-Based Design of
Inhibitors of Purine Nucleoside Phosphorylase. 3. 9-Arylmethyl
Derivatives of 9-Deazaguanine Substituted on the Methylene
Group. J . Med. Chem. 1993, 36, 3771-3783. (b) Guida, W. C.;
Elliot, R. D.; Thomas, H. J .; Secrist, III, J . A.; Babu, Y. S.; Bugg,
C. S.; Erion, M. D.; Ealick, S. E.; Montgomery, J . A. Structure-
Based Design of Inhibitors of Purine Nucleoside Phosphorylase.
4. A Study of Phosphate Mimics. J . Med. Chem. 1994, 37, 1109-
1114.
(25) MacPhee, J . A.; Dubois, J .-E. Steric Limits to Ester Alkylation;
Synthesis of Highly Hindered Esters via Hexamethylphosphora-
mide-favoured Enolization. J . Chem. Soc., Perkin Trans. 1 1977,
694-696.
(26) For prior 5-lithiotetrazole chemistry see: (a) Yi, K. Y.; Yoo, S.-
e. Synthesis of 5-Aryl and Vinyl Tetrazoles by the Palladium-
Catalyzed Cross-Coupling Reaction. Tetrahedron Lett. 1995, 36,
1679-1682. (b) Satoh, Y.; Marcopulos, N. Application of
5-Lithiotetrazoles in Organic Synthesis. Tetrahedron Lett. 1995,
36, 1759-1762.
(27) Merkler, D. J .; Brenowitz, M.; Schramm, V. L. The Rate
Constant Describing Slow-Onset Inhibition of Yeast AMP Deam-
inase by Coformycin Analogues Is Independent of Inhibitor
Structure. Biochemistry 1990, 29, 8358-8364.
(28) Schramm, V. L.; Baker, D. C. Spontaneous Epimerization of (S)-
Deoxycoformycin and Interaction of (R)-Deoxycoformycin, (S)-
Deoxycoformycin, and 8-Ketodeoxycoformycin with Adenosine
Deaminase. Biochemistry 1985, 24, 641-646.
J M990447M