Selective inhibition of nicotinamide adenine dinucleotide kinases by dinucleoside disulfide mimics of nicotinamide adenine dinucleotide analogues

Article abstract of DOI:10.1016/j.bmc.2009.06.013

Diadenosine disulfide (5) was reported to inhibit NAD kinase from Lysteria monocytogenes and the crystal structure of the enzyme-inhibitor complex has been solved. We have synthesized tiazofurin adenosine disulfide (4) and the disulfide 5, and found that these compounds were moderate inhibitors of human NAD kinase (IC₅₀ = 110 μM and IC₅₀ = 87 μM, respectively) and Mycobacterium tuberculosis NAD kinase (IC₅₀ = 80 μM and IC₅₀ = 45 μM, respectively). We also found that NAD mimics with a short disulfide (-S-S-) moiety were able to bind in the folded (compact) conformation but not in the common extended conformation, which requires the presence of a longer pyrophosphate (-O-P-O-P-O-) linkage. Since majority of NAD-dependent enzymes bind NAD in the extended conformation, selective inhibition of NAD kinases by disulfide analogues has been observed. Introduction of bromine at the C8 of the adenine ring restricted the adenosine moiety of diadenosine disulfides to the syn conformation making it even more compact. The 8-bromoadenosine adenosine disulfide (14) and its di(8-bromoadenosine) analogue (15) were found to be the most potent inhibitors of human (IC₅₀ = 6 μM) and mycobacterium NAD kinase (IC₅₀ = 14-19 μΜ reported so far. None of the disulfide analogues showed inhibition of lactate-, and inosine monophosphate-dehydrogenase (IMPDH), enzymes that bind NAD in the extended conformation.

Full text of DOI:10.1016/j.bmc.2009.06.013

Bioorganic & Medicinal Chemistry 17 (2009) 5656–5664
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Selective inhibition of nicotinamide adenine dinucleotide kinases by
dinucleoside disulfide mimics of nicotinamide adenine dinucleotide analogues
Riccardo Petrelli a, Yuk Yin Sham a, Liqiang Chen a, Krzysztof Felczak a, Eric Bennett a, Daniel Wilson a,
Courtney Aldrich a, Jose S. Yu a, Loredana Cappellacci b, Palmarisa Franchetti b, Mario Grifantini b,
Francesca Mazzola c, Michele Di Stefano c, Giulio Magni c, Krzysztof W. Pankiewicz a,
*
a Center for Drug Design, University of Minnesota, Minneapolis, MN 55455, USA
b Dipartimento di Scienze Chimiche, Universita’ di Camerino, 62032 Camerino, Italy
c Instituto di Biotecnologie Biochimiche, Universita’ Politecnica delle Marche, Via Ranieri, 60131 Ancona, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Diadenosine disulfide (5) was reported to inhibit NAD kinase from Lysteria monocytogenes and the crystal
structure of the enzyme–inhibitor complex has been solved. We have synthesized tiazofurin adenosine
disulfide (4) and the disulfide 5, and found that these compounds were moderate inhibitors of human
Received 6 May 2009
Revised 3 June 2009
Accepted 6 June 2009
Available online 13 June 2009
NAD kinase (IC50 = 110
lM and IC50 = 87
lM, respectively) and Mycobacterium tuberculosis NAD kinase
(IC50 = 80 M and IC50 = 45
l
l
M, respectively). We also found that NAD mimics with a short disulfide
(–S–S–) moiety were able to bind in the folded (compact) conformation but not in the common extended
conformation, which requires the presence of a longer pyrophosphate (–O–P–O–P–O–) linkage. Since
majority of NAD-dependent enzymes bind NAD in the extended conformation, selective inhibition of
NAD kinases by disulfide analogues has been observed. Introduction of bromine at the C8 of the adenine
ring restricted the adenosine moiety of diadenosine disulfides to the syn conformation making it even
more compact. The 8-bromoadenosine adenosine disulfide (14) and its di(8-bromoadenosine) analogue
Keywords:
NAD kinase
NADP
Enzyme inhibition
Tiazofurin
Dithioadenosine analogues, NAD
conformation
(15) were found to be the most potent inhibitors of human (IC50 = 6
lM) and mycobacterium NAD kinase
(IC50 = 14–19 V reported so far. None of the disulfide analogues showed inhibition of lactate-, and ino-
l
sine monophosphate-dehydrogenase (IMPDH), enzymes that bind NAD in the extended conformation.
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
target for the design of potential NAD kinase inhibitors. It is solvent
exposed and does not interact with the ATP closely. Chelation with
Until recently, between the two pyridine nucleotides, nicotin-
amide adenine dinucleotide (NAD) has received much more atten-
tion than its phosphorylated counterpart, nicotinamide adenine
dinucleotide 20-phosphate (NADP).1,2 NAD plays a crucial role in
cellular catabolic conversions,3,4 whereas the major role of NADPH
is to control the oxidative state of the cell and stand guard against
the ‘oxidative stress’.5,6
NAD kinase is a key enzyme that regulates supply of NADP in
the cell.7 Human NAD kinase catalyzes a magnesium-dependent
phosphorylation of the 20-hydroxyl group of the adenosine ribose
moiety of NAD using adenosine triphosphate (ATP) as phosphoryl
donor to produce NADP (Scheme 1). Thus, in contrast to the family
of ATP-dependent protein kinases that phosphorylate serine, thre-
onine, or tyrosine residues, NAD kinases bind both ATP and NAD.
Majority of protein kinase inhibitors have been designed to bind
at the ATP-binding domain,8 however, this domain is not a good
Mg2+ is needed to engage ATP at the catalytic domain. In contrast,
the NAD binding domain, as a substrate binding site, is ideal as it
plays a crucial role in molecular recognition of NAD and its
analogues.
Bacterial enzymes can use inorganic polyphosphates as phos-
phoryl donors (Scheme 1) in addition to ATP. An extensive biochem-
ical, enzymatic, and structural characterization of Mycobacterium
tuberculosis NAD kinase have been recently reported.9–11 This kinase
is of special interest as a new target for the development of potential
drugs against multi-drug resistant (MDR) tuberculosis (TB). The
mycobacterium enzyme requires a millimolar concentration of
NAD (Km = 3.3 mM), which is sixfold higher than that of human en-
zyme (0.5 mM). Thus, low micromolar inhibitors would be sufficiently
competitive to suppress NAD kinase activity, especially that of the M.
tuberculosis enzyme.
Inhibitors of the human enzyme are also of interest. They
should indirectly reduce critical supply of NADPH.1,6 Although
NAD kinase provides mainly oxidized form NADP+ (phosphoryla-
tion of NADH by the human enzyme is much less efficient), this
* Corresponding author. Tel.: +1 612 625 7968; fax: +1 612 624 825.
E-mail address: panki001@umn.edu (K.W. Pankiewicz).
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.06.013

R. Petrelli et al. / Bioorg. Med. Chem. 17 (2009) 5656–5664
5657
O
P
HO
HO OH
O
HO
O
OH
CONH2
NAD-kinase
CONH2
O
P
HO
O
P
OH
+
O
P
HO
O
P
OH
+
N
O
N
O
O
O
O
O
O
N
N
O
N
N
O
N
N
ATP or poly(P)
N
N
HO OH
NADP
HO OH
NAD
NH2
NH2
Scheme 1. Phosphorylation of NAD by NAD kinase.
nucleotide is immediately reduced into NADPH by several cellular
enzymes, such as NADP+-, isocitrate-, and glucose-6-phosphate
dehydrogenases, as well as malic enzyme.5,12 Interestingly these
enzymes are markedly increased during oxidative stress and in
cancer cells.5,13 Recently, NADPH oxidase has been suggested as a
target for anticancer therapy.14 It plays an important role in pro-
duction of reactive oxygen species (ROS) such as superoxide and
hydrogen peroxide that are found in large number of tumors.14 It
is highly expressed in human colon, prostate cancer15 as well as
in melanoma and ovarian, cancer.16–18 It was suggested, therefore,
that suppression of NADPH oxidase activity could result in anti-
angiogenesis and anticancer activities.14,18 If cancer cells require
higher supply of NADPH than normal cells, then inhibitors of hu-
man NAD kinase may show some anticancer effect.14
sible, to design any NAD-like molecules with good selectivity
against a particular enzyme. However, we and others clearly dem-
onstrated that this is not the case.20–25 It was reported that tiazofu-
rin (Fig. 1), an anticancer C-nucleoside, is converted in the cell into
an active metabolite tiazofurin adenine dinucleotide (TAD),26 an
analogue of NAD.
It inhibits IMPDH (Ki = 100 nM) without affecting other cellular
enzymes up to concentration of 100 lM. We found that TAD ana-
logues substituted at the C2 of the adenine moiety are even more
potent (low nanomolar) and selective inhibitors of IMPDH.27 Ear-
lier we reported an analogue of NAD that showed specificity
against alcohol dehydrogenase (ADH, Ki = 1 nM).20,21 Recently,
selective inhibitors of Cryptosporidium parvum IMPDH that bind
at the NAD-binding domain and do not inhibit the human enzyme
were identified.28,29
NAD kinases belong to a large family of NAD-utilizing en-
zymes.3 Although, the substrate binding domain of NAD kinases
is also conserved it is likely that NAD mimics may show some
selectivity against these enzymes without inhibition of majority
of cellular NAD-dependent enzymes.
Consequently, we report herein novel selective inhibitors of
NAD kinases that bind at the NAD-binding domain of NAD kinases,
but do not affect the majority of other NAD-dependent enzymes.
2. Results and discussion
We have examined 147 protein co-crystal structures containing
NAD or close NAD analogs, with the proteins selected for sequence
diversity.3 Most proteins bind the cofactor in an elongated (ex-
tended) conformation, such that the distance between the adenine
6-amino group and the nicotinamide amide carbon is 16 Å or more
(Figs. 2 and 3). However, there are clear exceptions to this general
trend. Several bacterial reductases bind NAD in extremely folded
conformations, where this distance is less than 6 Å (PDB entries
1RZ1, 1ZPT, and 2BKJ). Human and bacterial NAD kinases accom-
modate NAD in folded, bent (twisted) conformations forming a
cluster in the histogram around 12 Å.
2.1. Selectivity at the NAD binding domain
NAD kinases are enzymes which bind both ATP and NAD and
this distinguishes them from protein kinases that are almost exclu-
sively ATP-dependent.8 A majority of protein kinase inhibitors
have been designed to bind at the ATP-binding domain and numer-
ous small-molecule kinase inhibitors have been approved by US
Food and Drug Administration (FDA).8 The structure of the ATP cat-
alytic domain of all protein kinases is highly conserved, therefore,
selective inhibition of a single ATP-dependent enzyme, which
causes metabolic disorder or serves an unwanted biochemical
pathway(s), is an attractive therapeutic strategy. For example,
imatinib (Gleevec) was designed to bind at the ATP cleft of tyrosine
kinase encoded by aberrant gene (Bcr/Abl) expressed only in
chronic myelogenous leukemia (CML), and became the first effec-
tive and selective (rationally designed) agent for treatment of
cancer.19
As these large conformational differences illustrate (Fig. 3), the
cofactor binding pockets of these enzymes recognize a variety of
geometrically different conformations of NAD. Thus, compact
(short) NAD analogs, that do not satisfy the optimal distance
(16–21 Å) requirement, would unlikely inhibit
a majority of
NAD-dependent enzymes. A selective binding can be achieved by
distinguishing between NAD analogues fixed in the extended, com-
pact or folded conformation, as well as adopting the syn or anti
conformation [around the glycosyl bond of nicotinamide mononu-
The NAD binding domain of NAD-dependent enzymes is con-
served and it was believed that it would be difficult, if not impos-
CONH2
HO OH
HO OH
CONH2
CONH2
N
S
N
N
O
P
HO
O
P
OH
S
O
P
HO
O
P
OH
S
O
N
O
HO
O
O
O
O
O
O
O
N
X
N
O
O
N
N
N
N
N
HO OH
HO OH
HO OH
NH2
NH2
Tiazofurin
TAD; Ki = 100 nM
TAD, X = I;
TAD, X = ethynyl; Ki =10 nM
TAD, X = ethyl; Ki = 1 nM
Ki = 20 nM
Figure 1. Inhibitors of IMPDH; tiazofurin, tiazofurin adenine dinucleotide (TAD), and 2-substituted TAD analogues.

5658
R. Petrelli et al. / Bioorg. Med. Chem. 17 (2009) 5656–5664
ilar folded (twisted, bent) conformation as NAD bound at NAD
kinases, with a characteristic distance between the two 6-amino
groups of about 12 Å.
Since the crystal structure of M. tuberculosis NAD kinase is
known10,31 we docked DTA at its binding domain and found as ex-
pected that the distance between the two amino groups was in the
same range (e.g., 12.3 Å, Fig. 4) as reported for the Listeria enzyme.
2.2. Inhibitors design, syntheses, and enzymatic activities
Although the important role of NADP(H) in cellular processes is
well recognized, no potent inhibitors of human or bacterial NAD ki-
nase has ever been reported.6 It was suggested that modifications
of either nicotinamide or adenine moiety of NAD prevent binding
and phosphorylation.32 The X-ray structures of the mycobacterium
NAD kinase as well as NAD–Listeria complexes (the structure of the
human enzyme is not known) provide a good starting point in the
design process. However, due to their moderately low resolution
(2.1–2.6 Å), details are not sharply resolved. We, therefore, believe
that the additional information about the enzyme–NAD interac-
tions should emerge from studies of binding affinities of NAD ana-
logues at the NAD-binding domain. We assumed that dinucleoside
disulfide NAD mimics could serve well as molecular probes since
they can be easily synthesized, modified in the sugar and adenine
moiety, and prepared as conformationally restricted derivatives.
Figure 2. Histogram of distances between the adenine 6-amino and nicotinamide
amide carbon in 147 NAD-containing protein X-ray structures, selected for protein
sequence diversity.
Since TAD is known to inhibit IMPDH (IC50 = 0.1 lM) by binding
in the extended conformation we expected that shorter disulfide
analogue 4 would not fit to IMPDH (or other NAD-dependent en-
zymes that bind the cofactor in the extended conformation) but
may inhibit NAD kinase. Thus, tiazofurin adenine disulfide (4,
Scheme 2) was prepared in 55% yield by in situ deacetylation
(MeOH/NH3) of 50-acetylthiotiazofurin (1) and 50-acetylthioadeno-
sine (2)33 followed by iodine treatment of the mixture of the corre-
sponding di-50-thionucleosides, and chromatographic separation.
In this process we also isolated symmetric by-products such as dit-
iazofurin disulfide (3) and diadenosine disulfide (DTA, 5), and eval-
uated these compounds against mycobacterium and human NAD
kinase (Table 1).
Figure 3. Extended conformations of NAD (syn–syn and anti–anti) and folded (anti–
anti) conformation of nucleotide components of NAD.
cleotide (NMN) and adenosine monophosphate (AMP)], or ‘north’
or ‘south’ conformation of nucleotide sugars components. Such
preferred steric interactions should result in selective or even spe-
cific inhibition of a single protein.
Recently, crystal structures of NAD kinase from the human
pathogen Listeria monocytogenes (LmNADK1) in its free state, and
also bound to NAD or NADP have been solved.30 The authors found
that di-50-thio-adenosine (DTA) inhibits the Listeria enzyme
As expected, the disulfide
(IC50 = 80 M) and the human (IC50 = 110
(5) was a little more potent competitive inhibitor of both M. tuber-
culosis (IC50 = 45 M) and the human enzyme (IC50 = 87 M). The
4
inhibited M. tuberculosis
l
lM) NAD kinase. DTA
l
l
symmetric ditiazofurin analogue 3 was not active. Remarkably,
none of these compounds inhibited human IMPDH or lactate dehy-
drogenase, enzymes known to bind NAD in an extended
conformation.
As is illustrated in Figure 3, if the conformation of nucleoside
components of NAD is restricted to syn or anti, it would decrease
(Ki = 20 lM) and their X-ray analysis showed that DTA was bound
at the NAD-binding domain. Apparently, nicotinamide riboside
sub-domain of the NAD binding pocket is quite promiscuous and
tolerates well the substitution of nicotinamide riboside by adeno-
sine. Remarkably, DTA bound to the Listeria enzyme adopted a sim-
NH2
N
HO
OH
N
N
N
O
S
S
O
N
N
N
N
HO
OH
DTA
NH2
Figure 4. Structure of DTA and its folded (bent, twisted) shape when docked into M. tuberculosis NAD kinase.

R. Petrelli et al. / Bioorg. Med. Chem. 17 (2009) 5656–5664
5659
O
NH2
N
NH2
N
N
S
N
N
S
S
+
O
O
O
O
HO
OH
HO
OH
1
2
O
N
O
NH2
N
NH2
NH2
N
N
HO
OH
HO
OH
HO
OH
S
S
N
N
O
O
O
S
S
S
S
S
S
O
O
O
N
N
N
N
N
N
S
N
N
HO
OH
HO
OH
HO
OH
N
H2N
NH2
NH2
DTA (5)
3
O
4
Scheme 2. Synthesis of ditiazofurin disulfide (3), tiazofurin adenine disulfide (4), and dithioadenosine (DTA, 5).
Table 1
Inhibition of M. tuberculosis and human NAD kinase by dinucleoside disulfides
With this in mind, we decided to insert a bulky bromine atom at
the position 8 of the one or two adenine rings of DTA as in 14 or 15,
respectively (Scheme 3), which is known to force the adenosine to
adopt the syn conformation (there is not enough space for bromine
to be close to and above the ribose ring)35 (Scheme 3).
Structure
M. tuberculosis NAD Human NAD kinase
kinase IC50
(l
M)
IC50 (lM)
Ditiazofurin disulfide (3)
NAa
80
45
NA
110
87
6
Tiazofurin adenosine disulfide (4)
Diadenosine disulfide (DTA, 5)
8-Bromoadenosine
In a similar manner as described for synthesis of 15, an ana-
logue 16 with 8-phenyl group was prepared (Scheme 3). Di(8-phe-
nyladenosine) disulfide (16) also adopts the syn–syn conformation
but contains a large ‘flat’ aromatic ring. Disulfides 14–16 were pre-
pared from 20,30-di-O-isopropylidene protected adenosine deriva-
tives 6–8 in a similar manner as published for the synthesis of
5.30,33
14
adenosine disulfide (14)
Di(8-Bromoadenosine) disulfide (15)
19
6
Di(8-phenyladenosine) disulfide (16) NA
45
280
Di(20-C-methyladenosine)
disulfide (23)
500
NA
Di-(30-C-methyladenosine)
disulfide (24)
NA
90
In 1H NMR the chemical shift of the H20 signal can be used as an
indicator of the syn/anti equilibrium of nucleosides;36 a significant
downfield/upfield shift of the H20 signal is indicative of the pre-
dominance of the syn conformer or anti conformer, respectively.37
Benzamide adenine
dinucleotide (BAD)
200
a
NA—no activity.
A typical
D
value for H20 is d 5.4 ppm for compounds restricted in
syn conformation, exactly what we observed in the 1H NMR of
disulfides 14–16 (see Section 4).
or increase, respectively, the crucial distance between the adenine
6-amino group and the nicotinamide amide carbon of NAD. Confor-
mations around the glycosyl bond of the two adenosines of 5 are
not restricted allowing NAD kinases to accommodate the inhibitor
in any conformation the enzyme requires. However, fixing the con-
formation of the inhibitor to the one the enzyme prefers might be
considerably beneficial. An excellent illustration of such lowering
of the entropic barrier by restriction of conformational flexibility
is a cyclic inhibitor of penicillopepsin, which due to restricted rota-
tion around the one of the peptide bonds showed 420-fold increase
in binding affinity over its free rotational isomer (Fig. 5).34
We found that the monobromo analogue 14 and its dibromo
derivative 15 showed identical potency (IC50 = 6
human NAD kinase and similar potency against the mycobacte-
rium enzyme (IC50 = 14 and IC50 = 19 M, respectively,
Table 1). The di(8-phenyl) disulfide (16) inhibited human enzyme
(IC50 = 45 M) but was not active against M. tuberculosis NAD ki-
lM) against the
l
M
l
l
nase. These results indicate that although syn–syn binding of 8-
bromo and 8-phenyl substituted derivatives 15 and 16 is most
likely preferred by both enzymes, the substitution with large phe-
NH2
HN CH2
O
O
NH
O
P
OH
O
O
NH
O
P
OH
H
N
H
N
O
O
N
H
O
N
H
O
O
O
O
O
O
O
conformationally restricted
cyclic inhibitor, Ki = 0.1 n
acyclic inhibitor, Ki = 42 nM
M
Figure 5. Acyclic and conformationally restricted (cyclic) peptide-phosphonates.

5660
R. Petrelli et al. / Bioorg. Med. Chem. 17 (2009) 5656–5664
NH2
NH2
NH2
NH2
N
N
N
N
N
N
N
N
N
N
HO
OH
N
N
R2
R
R
R
b
a
N
O
N
c, d
N
O
N
O
O
S
HO
S
S
S
O
N
N
N
O
O
R1
O
O
N
O
O
HO
OH
HO
OH
NH2
14, R1 = Br, R2 = H
2, R = H
12, R = Br
13, R = Ph
6, R = H
9, R = H
10, R = Br
11, R = Ph
7 R = Br
,
1
2
15
16
, R = R = Br
8, R = Ph
1
2
, R = R = Ph
Scheme 3. Synthesis of 8-bromoadenosine adenosine disulfide (14), di(8-bromoadenosine) disulfide (15), and di(8-phenyladenosine disulfide (16). Reagents and conditions:
(a) DEAD, PPh3, AcSH, THF, 0 °C to rt; (b) HCO2H, H2O (c) NH3, CH3OH, H2O (d) I2.
nyl moieties is not beneficial in the case of the mycobacterium
enzyme.
tion of human NAD kinase (IC50 = 90 l
M).41 BAD like NAD can
adopt the extended conformation or be bent and twisted to be
bound in the folded conformation required by NAD kinases. DTA,
also a competitive inhibitor of NAD kinases, can fit the folded but
not the extended NAD domain.
M. tuberculosis and human NAD kinase were purified to homo-
geneity as described earlier.41 IC50 values for compounds showing
an inhibitory effect were calculated and are summarized in Table 1.
All new disulfide compounds were evaluated against two iso-
forms of human IMPDH and against lactate dehydrogenase as de-
scribed earlier.27 None of these compounds inhibited these
The encouraging activities of the conformationally restricted
compounds 14–16 prompted us to take one more step forward
and examine the activity of DTA analogues that prefer to adopt a
specific conformation around the glycosyl bond in addition to
‘south’ or ‘north’ conformation of their sugar moiety. Consequently,
we focused our attention on 20-C-methyl ‘up’- and 30-C-methyl ‘up’
adenosine.38,39 The ribose of these compounds is known to be re-
stricted to the ‘north’ and ‘south’ conformation, respectively. In
addition, the bulky methyl group in the 20-‘up’ configuration en-
forced the adenine ring to adopt the anti conformation, whereas
the 30 ‘up’ methyl group keeps the adenine moiety in the opposite
syn conformation.38,39 Thus, we prepared compounds 23 and 24
(Scheme 4) and evaluated them against the two NAD kinases
(Table 1).
enzymes up to concentration of 100 lM.
3. Conclusions
Dinucleoside disulfides such as tiazofurin adenosine disulfide
(4) and diadenosine disulfide (5) were found to be moderate inhib-
itors of M. tuberculosis and human NAD kinase (IC50’s = 45–
110 lM). A restriction of the conformation of adenine moiety to
syn by substitution with bromine atom at the C8 to give compound
14 or 15 resulted in a 3–15-fold increase of potency against myco-
bacterium (IC50 = 14–19 lM) and human enzyme (6 lM), respec-
tively. These results support our hypothesis that compact NAD
analogues that are too short to mimic NAD binding in the extended
conformation are expected to show selective inhibition of NAD ki-
Unfortunately, disulfide 23 showed only a weak inhibitory
activity against mycobacterium (IC50 = 500
lM) and human NAD
kinase (IC50 = 250 M) whereas compound 24 was inactive against
l
both enzymes. Possibly, the insertion of methyl group either at the
C20 or the C30 resulted in steric hindrance that significantly dimin-
ished or revoked binding affinity of these analogues.
We reported earlier that benzamide adenine dinucleotide
(BAD), a close 1-deaza analogue of NAD, which is a potent inhibitor
of IMPDH (IC50 = 0.8
l
M),40 showed moderate competitive inhibi-
NH2
N
NH2
N
NH2
N
N
N
N
N
N
N
N
N
N
S
S
b
c, d
HO
R2
a
O
O
O
R2
R1
R1
OH
R2
O
O
R1
HO
O
O
O
O
19 R1 = CH3, R2 = H
20 R1 = H, R2 = CH3
21 R1 = CH3, R2 = H
22 R1 = H, R2 = CH3
NH2
17 R1 = CH3, R2 = H
18 R1 = H, R2 = CH3
NH2
N
N
N
N
HO
OH
HO
OH
CH3
N
N
H3C
N
N
O
O
S
S
S
S
or
O
O
N
N
N
N
N
N
H3C
CH3
N
N
HO
OH
HO
OH
NH2
NH2
24
23
Scheme 4. Synthesis of di(20-C-methyladenosine) disulfide (23) and di(30-C-methyladenosine) disulfide (24). Reagents and conditions as in Scheme 3.

R. Petrelli et al. / Bioorg. Med. Chem. 17 (2009) 5656–5664
5661
nases. Bacterial reductases that bind NAD in extremely folded con-
formations could be especially good targets for drug design based
on compact NAD derivatives. Further studies with analogues in
which nicotinamide and adenine moieties are connected by short
linkers are in progress.
8). HRMS (ESI+) calcd for C19H24N7O7S3 558.0899 (M+H)+ found
558.0910. Anal. Calcd for C19H23N7O7S3 (557.62): C, 40.9; H, 4.1;
N, 17.5. Found: C, 40.8; H, 4.0; N, 17.6 and the slowest migrating
(Rf = 0.15) DTA (5, 11 mg, 12%), 1H NMR (600 MHz, DMSO-d6): d
3.06 (dd, 1H, J = 7.6 Hz, J = 13.8 Hz, H-50), 3.31 (dd, 1H, J = 5.6 Hz,
J = 13.8 Hz, H-500), 4.08 (m, 1H, H-40), 4.13 (m, 1H, H-30), 4.76 (pseu-
do t, 1H, J = 5.6 Hz, H-20), 5.38 (d, 1H, OH-exchangeable), 5.49
(d,1H, OH-exchangeable), 5.86 (d, 1H, J = 5.6 Hz, H-10), 7.24 (br s,
2H, NH2-exchangeable), 8.11 (s, 1H, H-2), 8.31 (s, 1H, H-8). HRMS
calcd for C20H25N10O6S2 365.1399 (M+H)+, found 365.1384. Anal.
Calcd for C20H24N10O6S2 (364.6): C, 42.5; H, 4.2; N, 24.8. Found:
C, 42.4; H, 4.3; N, 24.6.
4. Experimental
4.1. General methods
All commercial reagents (Sigma–Aldrich, Acros) were used as
provided unless otherwise indicated. An anhydrous solvent dis-
pensing system (J. C. Meyer) using two packed columns of neutral
alumina was used for drying THF, Et2O, and CH2Cl2, while two
packed columns of molecular sieves were used to dry DMF. Sol-
vents were dispensed under argon. Flash chromatography was per-
formed with Ultra Pure silica gel (Silicycle) with the indicated
solvent system. Analytical HPLC was performed on a Varian Micro-
4.1.2. (8-Bromoadenosine-50-yl) (adenosine-50-yl) disulfide (14)
20,30-O-Isopropylidene-adenosine (6) was thioacetylated to give
9 and then the isopropylidene group was removed to give 50-thio-
acetyladenosine (2) as described before.30 To an ice-cold solution
of triphenylphosphine (296 mg, 1.771 mmol) in dry THF (10 mL),
diethyl azodicarboxylate (0.41 mL, 1.771 mmol) was added over
5 min. After stirring for 30 min, 20,30-O-isopropylidene 8-bromoad-
enosine (7)42(300 mg, 0.776 mmol) was added, and stirring was
continued for 10 min. To the resulting deep yellow suspension a
sorb (C18, 5
lm, 4.6 ꢀ 250 mm) with a flow rate of 0.5 mL/min. An
isocratic or linear gradient of 0.04 M Et3N H2CO3 (TEAB) and aque-
ous MeCN (70%) was used. Nuclear magnetic resonance spectra
were recorded on a Varian 600 MHz with Me4Si, DDS, or signals
from residual solvent as the internal standard for 1H or 13C. Chem-
ical shifts are reported in ppm, and signals are described as s (sin-
glet), d (doublet), t (triplet), q (quartet), m (multiplet), br s (broad
singlet), dd (double doublet), and dt (double triplet). Values given
for coupling constants are first order. High-resolution mass spectra
were recorded on an Agilent TOF II TOF/MS instrument equipped
with either an ESI or APCI interface. Elemental analyses were
determined on an EA 1108 CHNS–O (Fisons Instruments) analyzer.
solution of thioacetic acid (121 lL, 1.771 mmol) in dry THF
(1 mL) was added dropwise and stirring was continued for another
1 h at 0 °C. At the end of the reaction the solvent was removed un-
der reduced pressure, and the resulting yellowish residue was
purified by flash chromatography on silica gel column eluting with
CHCl3/MeOH (0–10%). The fractions containing the product were
combined, the solvent was removed under reduced pressure, and
the residue was dried in vacuo to yield pure 20,30-O-isopropyli-
dene-50-acetylthio-50deoxy-8-bromoadenosine (10, 175 mg, 50%)
as a white solid: Rf 0.2 (95:5 CHCl3/MeOH). 1H NMR (600 MHz,
DMSO-d6): d 1.35 (s, 3H), 1.37 (s, 3H), 2.35 (s, 3 H, COCH3), 3.08
(dd, 1H, J = 6.7 Hz, J = 13.7 Hz, H-50), 3.19 (dd, 1H, J = 7.0 Hz,
J = 13.7 Hz, H-500), 4.12–4.16 (m, 2H, H-40), 4.93 (m, 1H, H-30),
5.48 (dd, 1H, J = 2.1 Hz, J = 6.2 Hz, H-20), 6.15 (d, 1H, J = 2.1 Hz, H-
10), 6.73 (br s, 2H, NH2), 8.12 (s, 1H, H-2). HRMS (ESI+) calcd for
C15H19BrN5O4S 445.0231 (M+H)+, found: 445.0239. Anal. Calcd
for C15H18BrN5O4S (444.30): C, 40.5; H, 4.1; N, 15.7. Found: C,
40.4; H.
A solution of 10 (150 mg, 0.337 mmol) in a mixture of formic
acid and water (10 mL, 7:3) was stirred at room temperature. After
5 h reaction time the solvent was evaporated under reduced pres-
sure, and traces of formic acid were removed by co-evaporation
(five times) with anhydrous ethanol. The obtained white powder
was purified by flash chromatography with CHCl3/MeOH (0–10%)
to give 50-acetylthio-50-deoxy-8-bromoadenosine (12, 92 mg,
66%) as a white solid: Rf 0.15 (90:10 CHCl3/MeOH). 1H NMR
(600 MHz, DMSO-d6): d 2.25 (s, 3 H, COCH3), 3.47 (m, 2H, H-50,
H-500), 4.17 (m, 1H, H-40), 4.33 (m, 1H, H-30), 5.28 (m, 1H, H-20),
5.35 (d, 1H, OH-exchangeable), 5.48 (d, 1H, OH-exchangeable),
5.77 (d, 1H, J = 5.6 Hz, H-10), 7.42 (br s, 2H, NH2), 8.09 (s, 1H, H-
2). HRMS (ESI+) calcd for C12H15BrN5O4S 405.0028 (M+H)+, found:
405.0036. Anal. Calcd for C12H14BrN5O4S (404.24): C, 35.6; H, 3.5;
N, 17.3. Found: C, 35.6; H, 3.4; N, 17.2.
4.1.1. Di(tiazofurin-50-yl) disulfide (3), (tiazofurin-50-yl)
(adenosin-50-yl) disulfide (4), and di(adenosin-50-yl) disulfide
(DTA, 5)
The 50-thioacetyltiazofurin (1) was prepared according to the
published procedure for synthesis of adenosine derivative 2.33
Thus, 1 (51 mg, 0.157 mmol) was dissolved in 7 N methanolic
ammonia (4 mL) and stirred at 0 °C for 30 min. In another flask
the 50-thioadenosine (2, 50 mg, 0.157 mmol) was dissolved in 7 N
methanolic ammonia and stirred at 0 °C for 30 min. The resulting
reaction mixtures were mixed together and a catalytic amount of
iodine (5 mg, 0.012 mmol) in ethanol (0.5 mL) was added drop-
wise. The resulting mixture was allowed to stir for 12 h at rt and
the solvent was evaporated under reduced pressure. The residue
was purified by flash chromatography with EtOAc/acetone/EtOH/
H2O (6:1:1:0.5) as the eluent affording the fast (Rf = 0.35) migrat-
ing disulfide 3 (28 mg, 32%) as a white solid 1H NMR (600 MHz,
DMSO-d6): d 2.97 (dd, 1H, J = 7.3, 13.8 Hz, H-50), 3.18 (dd, 1H,
J = 4.4, 13.8 Hz, H-500), 3.85 (m, 1H, H-40), 4.06–4.14 (m, 2H, H-20,
H-30), 4.95 (d, 1H, J = 4.4 Hz, H-10), 5.22 (d, 1H, OH-exchangeable),
5.45 (d, 1H, OH- exchangeable), 7.53 (s, 1H, CONH2, exchangeable),
7.65 (s, 1H, CONH2, exchangeable), 8.11 (s, 1H, H-5). HRMS (ESI+)
calcd for C18H23N4O8S4 51.0398 (M+H)+ found 551.0409. Anal.
Calcd for C22H26N4O10S2 (570.59): C, 39.2; H, 4.0; N, 10.1. Found:
C, 39.3; H, 4.1; N, 10.0, followed by tiazofurin adenosine disulfide
4 (Rf = 0.25) as a white solid (48 mg, 55%) 1H NMR (600 MHz,
DMSO-d6): d 2.86 [dd, 1H, J = 7.9 Hz, J = 14.1 Hz, H-50 (T)], 3.05
[m, 2H, H-500 (T), H-50 (A)], 3.16 [dd 1H, J = 13.8 Hz, J = 5.4 Hz, H-
50 (A)], 3.82 [m, 1H, H-40 (T)], 4.10–4.19 [m, 4H, H-40 (A), H-30 (T),
H-20 (T), H-30 (A)], 4.78 [m, 1H, H-20(A)], 4.94 [d, 1H, J = 4.1 Hz,
H-10 (T)], 5.22 [br s, 1H, OH-exchangeable], 5.37 [br s, 1H, OH-
exchangeable], 5.45 [d, 1H, OH-exchangeable], 5.51 [d, 1H, OH-
exchangeable], 5.88 [d, 1H, J = 5.8 Hz, H-10 (A)], 7.25 [br s, 2H,
NH2-exchangeable], 7.52 and 7.65 [two brs, 1H each, CONH2(T)-
exchangeable], 8.12 (s, 1H, H-5), 8.16 (s, 1H, H-2), 8.32 (s, 1H, H-
The thionucleoside 12 (60 mg, 0.148 mmol) was dissolved in
methanolic ammonia 7 N (6 mL) and stirred at 0 °C for 30 min. In
another flask nucleoside 2 (50 mg, 0.148 mmol) was dissolved in
methanolic ammonia 7 N and stirred at 0 °C for 30 min. The result-
ing reaction mixtures are mixed together and iodine (5 mg,
0.018 mmol) in ethanol (0.5 mL) was added dropwise. The result-
ing mixture was allowed to stir at rt for 14 h and the solvent was
evaporated under reduced pressure. The residue was purified by
flash chromatography with EtOAc/acetone/EtOH/H2O (6:1:1:0.5)
to give (8-bromoadenosine-50-yl) (adenosine-50-yl) disulfide 14
as a white solid (52 mg, 55%): Rf 0.25 (6:1:1:0.5 EtOAc/acetone/

5662
R. Petrelli et al. / Bioorg. Med. Chem. 17 (2009) 5656–5664
EtOH/H2O), 1H NMR (600 MHz, DMSO-d6): d 3.05–3.12 (m, 2H, H-
50, H-500), 3.25–3.36 (m, 2H, H-50, H-500), 4.09 (m, 1H, H-40), 4.13
(m, 1H, H-40), 4.24 (m, 1H, H-30), 4.35 (m, 1H, H-30), 4.76 (m, 1H,
H-20), 5.31 (m, 1H, H-20), 5.34 (d, 1H, OH-exchangeable), 5.41 (d,
1H, OH-exchangeable), 5.49 (d, 1H, OH-exchangeable), 5.51 (d,
1H, OH-exchangeable), 5.82 (d, 1H, J = 5.8 Hz, H-10), 5.87 (d, 1H,
J = 5.5 Hz, H-10), 7.25 (br s, 2H, NH2), 7.58 (br s, 2H, NH2), 8.12 (s,
1H, H-2), 8.14 (s, 1H, H-2), 8.31 (s, 1H, H-2). HRMS (ESI+) for
C20H24BrN10O6S2 644.3145 (M+H)+ found 644.3162. Anal. Calcd
for C20H23BrN10O6S2 (643.39): C, 42.5; H, 4.3; N, 21.7. Found: C,
42.6; H, 4.2; N, 21.8; followed by di(8-bromoadenosine-50-yl)
disulfide (15) as a white solid (10 mg, 21%): Rf 0.4 (6:1:1:0.5
EtOAc/acetone/EtOH/H2O) and DTA (9 mg, 23%): Rf 0.15
(6:1:1:0.5 EtOAc/acetone/EtOH/H2O).
Calcd for C32H32N10O6S2 (716.79): C, 53.6; H, 5.5; N, 19.5. Found:
C, 53.5; H, 4.4; N, 19.6.
4.1.5. Di(20-C-methyladenosine-50-yl) disulfide (23)
To an ice-cold solution of triphenylphosphine (140 mg,
0.819 mmol) in abs THF (10 mL), diethyl azodicarboxylate
(0.19 mL, 0.819 mmol) was added over 5 min. After stirring for
30 min, 20-C-methyl-20,30-O-isopropylidene-adenosine 1738 (120 mg,
0.373 mmol) was added, and stirring was continued for 10 min.
To the resulting yellow suspension, a solution of thioacetic acid
(6 lL, 0.819 mmol) in abs THF (0.5 mL) was added dropwise and
stirring was continued for another 4 h at 0 °C. During this time
the yellow suspension cleared, and an orange solution was ob-
tained. At the end of the reaction the solvent was removed under
reduced pressure, and the resulting yellowish residue was purified
by flash chromatography on silica gel CHCl3/MeOH (0–5%). The
fractions containing the product were combined, the solvent was
removed under reduced pressure, and the residue was dried in va-
cuo to yield pure 50-thionucleoside 19 (105 mg, 75%) as a white so-
lid: Rf 0.25 (95:5 CHCl3/MeOH). 1H NMR (600 MHz, DMSO-d6): d
1.18 (s, 3H, CH3), 1.39 (s, 3H, CH3), 1.53 (s, 3H, CH3), 2.41 (s, 3H,
COCH3), 3.63 (m, 2H, H-50, H-500), 4.39 (m, 1H, H-40), 4.58 (d, 1H,
J = 2.64 Hz, H-30), 6.17 (s, 1H, H-10), 7.53 (br s, 2H, NH2), 8.14 (s,
1H, H-2), 8.35 (s, 1H, H-8). HRMS (ESI+) calcd for C16H22N5O4S
380.1392 (M+H)+, found 380.1405. A solution of 50-thionucleoside
19 (70 mg, 0.184 mmol) in a mixture of formic acid and water
(10 mL, 2:1) was stirred at room temperature. After 12 h reaction
time the solvent was evaporated under reduced pressure, and
traces of formic acid were removed by co-evaporating five times
with abs ethanol. The obtained white powder was purified by flash
chromatography on silica gel with CHCl3/MeOH (5–20%). The frac-
tions containing the product were combined, the solvent was re-
moved under reduced pressure, and the product was dried in
vacuo to yield 21 (58 mg, 92%) as a white powder. Rf 0.15 (90:10
CHCl3/MeOH). 1H NMR (600 MHz, DMSO-d6): d 1.08 (s, 3H, CH3),
2.32 (s, 3H, COCH3), 3.41 (m, 2H, H-50, H-500), 4.21 (m, 1H, H-40),
4.58 (d, 1H, J = 2.64 Hz, H-30), 5.12 (d, 1H, OH-exchangeable),
5.28 (d, 1H, OH-exchangeable), 6.11 (s, 1H, H-10), 7.23 (br s, 2H,
NH2), 8.11 (s, 1H, H-2), 8.27 (s, 1H, H-8). HRMS (ESI+) calcd for
C13H18N5O4S 340.1079 (M+H)+, found 340.1087.
The thionucleoside 21 (50 mg, 0.147 mmol) was dissolved in
methanolic ammonia 7 N (5 mL) and stirred at 0 °C for 30 min.
After stirring for 30 min, iodine (5 mg, 0.012 mmol) in ethanol
(0.5 mL) was added dropwise. The resulting mixture was allowed
to stir at rt for 12 h and the solvent was evaporated under reduced
pressure. Purification by flash chromatography with EtOAc/ace-
tone/EtOH/H2O (6:1:1:0.1) afforded 23 as a white solid (54 mg,
62%): Rf 0.15 (6:1:1:0.1 EtOAc/acetone/EtOH/H2O). 1H NMR
(600 MHz, DMSO-d6): d 1.20 (s, 3H, CH3), 3.16–3.18 (m, 2H, H-50,
H500), 4.09–4.18 (m, 2H, H-30, H-40), 5.22 (br s, 1H, OH-exchange-
able), 5.35 (d, 1H, OH-2-exchangeable), 5.95 (s, 1H, H-10), 7.25
(br s, 2H, NH2), 8.11 (s, 1H, H-2), 8.21 (s, 1H, H-8). HRMS (ESI+)
calcd for C22H29N10O6S2 593.1712 (M+H)+, found 593.1698. Anal.
Calcd for C22H28N10O6S2 (592.6): C, 44.5; H, 4.7; N, 23.6. Found:
C, 44.4; H, 4.8; N, 23.5.
4.1.3. Di(8-bromoadenosine-50-yl) disulfide (15)
Deacetylation of 12 (40 mg, 0.098 mmol) with methanolic
ammonia 7 N followed by reaction with iodine overnight (4 mg,
0.011 mmol) and purification by flash chromatography with
EtOAc/acetone/EtOH/H2O (6:1:1:0.2) afforded the desired di(8-
bromoadenosine-50-yl) disulfide 15 (33 mg, 51%) as a white solid:
Rf 0.25 (6:1:1:0.5 EtOAc/acetone/EtOH/H2O). 1H NMR (600 MHz,
DMSO-d6): d 3.20-3.38 (m, 2H, H-50, H-500), 4.13 (m, 1H, H-40),
4.35 (m, 1H, H-30), 5.31 (m, 1H, H-20), 5.32 (d, 1H, OH-exchange-
able), 5.47 (d, 1H, OH-exchangeable), 5.97 (d, 1H, J = 5.5 Hz, H-10),
7.45 (br s, 2H, NH2), 8.12 (s, 1H, H-2). HRMS (ESI+) calcd for
C20H23Br2N10O6S2 723.2643 (M+H)+, found: 723.2635. Anal. Calcd
for C20H22Br2N10O6S2 (722.39): C, 33.2; H, 3.1; N, 19.4. Found C,
33.3; H.
4.1.4. Di(8-Phenyladenosine-50-yl) disulfide (16)
In a similar manner as described above for 8-bromo analogue 7,
compound 8 (180 mg, 0.469 mmol) was converted into 50-acetyl-
thio-20,30-O- isopropylidene 8-phenyladenosine 11. Purification
by flash chromatography with CH2Cl2/MeOH (0–3%) afforded 11
(190 mg, 92%) as a white solid: Rf 0.2 (98:2 CH2Cl2/MeOH). 1H
NMR (600 MHz, DMSO-d6): d 1.35 (s, 3H, CH3), 1.56 (s, 3H, CH3),
2.44 (s, 3H, COCH3), 3.42 (m, 1H, H-50), 3.53 (m, 1H, H-500), 3.77
(m, 1H, H-40), 4.11 (m, 1H, H-30), 5.19 (m, 1H, H-20), 5.79 (d, 1H,
J = 7.1 Hz, H-10), 7.44 (br s, 2H, NH2), 7.61 (m, 3H, Ph), 7.82 (m,
2H, Ph), 8.13 (s, 1H, H-2). HRMS (ESI+) for C21H24N5O4S calcd for
441.1545 (M+H)+, found: 441.1552.
A solution of 11 (70 mg, 0.158 mmol) in a mixture of formic acid
and water (12 mL, 2:1) was stirred at room temperature. After 8 h
reaction time the solvent was evaporated under reduced pressure,
and traces of formic acid were removed by co-evaporation (five
times) with anhydrous ethanol. Purification by flash chromatogra-
phy with CHCl3/MeOH (0–5%) afforded de-isopropylidenated com-
pound 13 (45 mg, 72%) as a white solid: Rf 0.2 (95:5 CHCl3/MeOH).
1H NMR (600 MHz, DMSO-d6): d 2.47 (s, 3H, COCH3), 3.28 (m, 1H,
H-50), 3.39 (m, 1H, H-500), 4.01 (m, 1H, H-40), 4.09 (m, 1H, H-30),
5.27 (m, 1H, H-20), 5.77 (d, 1H, J = 6.6 Hz, H-10), 7.39 (br s, 2H,
NH2), 7.58 (m, 3H, Ph), 7.73 (m, 2H, Ph), 8.19 (s, 1H, H-2). HRMS
(ESI+) for C18H20N5O4S calcd for 401.1278 (M+H)+, found:
401.1284.
Deacetylation of 13 (32 mg, 0.079 mmol) with methanolic
ammonia 7 N followed by reaction with iodine (4 mg, 0.009 mmol)
overnight and purification by flash chromatography with EtOAc/
acetone/EtOH/H2O (6:1:1:0.2) afforded the title compound 16
(34 mg, 62%) as a white solid: Rf 0.15 (6:1:1:0.1 EtOAc/acetone/
EtOH/H2O).1H NMR (600 MHz, DMSO-d6): d 3.18 (m, 1H, H-50),
3.34 (m, 1H, H-500), 4.09–4.19 (m, 2H, H-40, H-30), 5.31 (br s, 1H,
OH-exchangeable), 5.41 (br s, 1H, OH-exchangeable), 5.45 (m,
1H, H-20), 5.71 (d, 1H, J = 6.2 Hz, H-10), 7.34 (br s, 2H, NH2), 7.52
(m, 3H, Ph), 7.67 (m, 2H, Ph), 8.15 (s, 1H, H-2). HRMS (ESI+) for
C32H33N10O6S2 calcd for 717.1923 (M+H)+, found: 717.1935. Anal.
4.1.6. Di(30-C-methyladenosine-50-yl) disulfide (24)
As described above compound 1838,39 (100 mg, 0.311 mmol)
was converted into thionucleoside 20 (65 mg, 55%) as a white so-
lid: Rf 0.25 (95:5 CHCl3/MeOH). 1HNMR (600 MHz, DMSO-d6): d
1.34 (s, 3H, CH3), 1.48 (s, 3H, CH3), 1.58 (s, 3H, CH3), 2.34 (s, 3H,
COCH3), 3.01 (dd, 1H, J = 9.4 Hz, J = 13.8 Hz, H-50), 3.20 (dd, 1H,
J = 4.4 Hz, J = 13.8 Hz, H-500), 4.00 (dd, 1H, J = 4.4 Hz, J = 9.4 Hz, H-
40), 4.94 (d, 1H, J = 2.1 Hz, H-20), 6.03 (d, 1H, J = 2.1 Hz, H-10), 7.31
(br s, 2H, NH2), 8.13 (s, 1H, H-2), 8.29 (s, 1H, H-8). HRMS (ESI+)
calcd for C16H22N5O4S 380.1397 (M + H)+, found 380.1407.

R. Petrelli et al. / Bioorg. Med. Chem. 17 (2009) 5656–5664
5663
A solution of 20 (60 mg, 0.158 mmol) in a mixture of formic acid
and water (10 mL, 2:1) was stirred at room temperature. After 6 h
reaction time the solvent was evaporated under reduced pressure,
and traces of formic acid were removed by co-evaporation (five
times) with anhydrous ethanol. The obtained white powder was
purified by flash chromatography with CHCl3/MeOH (0–10%) to
give 50-acetylthio-30-C-methyl-50-deoxyadenosine 22 (35 mg,
66%) as a white solid: Rf 0.15 (90:10 CHCl3/MeOH). 1HNMR
(600 MHz, DMSO-d6): d 1.27 (s, 3H, CH3), 2.27 (s, 3H, COCH3),
3.20 (dd, 1H, J = 4.1 Hz, J = 13.8 Hz, H-50), 3.22 (dd, 1H, J = 9.9 Hz,
H-500), 3.82 (dd, 1H, J = 4.1 Hz, 9.9 Hz, H-40), 4.62 (d, 1H, J = 7.6 Hz,
H-20), 5.05 (s, 1H, OH-exchangeable), 5.51 (d, 1H, OH-exchange-
able) 5.77 (d, 1H, J = 7.6 Hz, H-10), 7.18 (brs, 2H, NH2), 8.09 (s, 1H,
H-2), 8.33 (s, 1H, H-8). HRMS (ESI+) calcd for C13H18N5O4S
340.1084 (M+H)+, found 340.1086.
with buffer A for 4.5 min prior to the next run. The flow rate was
1.3 mL/min, and temperature 25 °C. The eluate absorbance was
monitored at 260 nm. If necessary, other elution conditions, or
other columns, were employed in order to separate the synthe-
sized compounds from the nucleotides of the reaction. Control
reaction mixtures have been prepared to assess the stability of
the various compounds during the incubation at 37 °C. The possi-
bility that some of the synthesized compounds could be used as
alternative substrates in the NAD kinase catalyzed reaction was
considered and appropriate reaction mixtures were analyzed.
Acknowledgments
These studies were funded by the Center for Drug Design, Uni-
versity of Minnesota.
Deacetylation of 22 (40 mg, 0.114 mmol) with methanolic
ammonia 7 N followed by reaction with iodine 12 h and purifica-
tion by flash chromatography with EtOAc/acetone/EtOH/H2O
(6:1:1:0.1) afforded 24 (38 mg, 56%) as a white solid: Rf 0.15
(6:1:1:0.1 EtOAc/acetone/EtOH/H2O). 1HNMR (600 MHz, DMSO-
d6): d 1.17 (s, 3H, CH3), 2.91 (dd, 1H, J = 2.6 Hz, J = 13.6 Hz, H-50),
3.14 (dd, 1H, J = 10.6 Hz, J = 13.6 Hz, H-500), 3.97 (dd, 1H, J = 2.6,
J = 10.6 Hz, H-40), 4.61 (d, 1H, J = 7.9 Hz, H-20), 4.99 (s, 1H, OH-
exchangeable), 5.46 (d, 1H, OH-exchangeable) 5.81 (d, 1H,
J = 7.9 Hz, H-10), 7.22 (br s, 2H, NH2), 8.11 (s, 1H, H-2), 8.33 (s,
1H, H-8). C22H29N10O6S2 593.1718 (M+H)+, found 593.1728. Anal.
Calcd for C22H28N10O6S2 (592.6): C, 44.5; H, 4.7; N, 23.6. Found:
C, 44.6; H, 4.4; N, 23.7.
References and notes
1. Pollak, N.; Niere, M.; Ziegler, M. J. Biol. Chem. 2007, 282, 33562.
2. Pollak, N.; Dolle, C.; Ziegler, M. Biochem. J. 2007, 402, 205.
3. Chen, L.; Petrelli, R.; Gao, G.; Bonnac, L.; Yu, J. S.; Bennett, E. M.; Pankiewicz, K.
W. Curr. Med. Chem. 2008, 15, 650.
4. Magni, G.; Orsomando, G.; Raffelli, N.; Ruggieri, S. Front Biosci. 2008, 13, 6135.
5. Singh, R.; Mailloux, R. J.; Puiseux-Dao, S.; Appanna, V. D. J. Bacteriol. 2007, 189,
6665.
6. Grose, J. H.; Joss, L.; Velick, S. F.; Roth, J. R. Proc. Natl. Acad. Sci. U.S.A. 2006, 103,
7601.
7. Magni, G.; Orsomando, G.; Raffaelli, N. Mini-Rev. Med. Chem. 2006, 6, 739.
8. Liao, J. J. J. Med. Chem. 2007, 50, 409.
9. Raffaelli, N.; Finaurini, L.; Mazzola, F.; Pucci, L.; Sorci, L.; Amici, A.; Magni, G.
Biochemistry 2004, 43, 7610.
10. Garavaglia, S. R.; Finaurini, L.; Magni, G.; Rizzi, M. J. Biol. Chem. 2004, 279,
40980.
11. Mori, S.; Kawai, S.; Shi, F.; Mikami, B.; Murata, K. J. Biol. Chem. 2005, 280, 24104.
12. Kawai, S.; Murata, K. Biosci., Biotechnol., Biochem. 2008, 72, 919.
13. Sun, J. S.; Tsuang, Y. H.; Huang, W. C.; Chen, L. T.; Hang, Y. S.; Lu, F. J. Cell Mol.
Life Sci. 1997, 53, 967.
4.2. Molecular modeling
Pair wise sequence alignment was carried out using Clustal W
v1.843 to identify conserved sequence regions. Molecular model-
ling was carried out using the Schrodinger modelling software
package (Schrodinger, LLC, Portland, OR 2007). The structure of hu-
man NAD kinase is homology modelled using Prime based on the
Clustal W sequence alignment and the recently solved X-ray struc-
ture of Mycobacterium tuberculosis NAD kinase–NAD complex
(PDB ID: 1Y3I). DTA were docked into the binding site of both hu-
man and M. tuberculosis NAD kinase by superposition using the
solved X-ray structure of L. monocytogenes NAD kinase-DTA com-
plex (PDB ID:2I2C). All residues within a 9 Å shell radius of each of
the top selected bound NAD and DTA conformation were subjected
to a series of restraint energy minimizations using the generalized
born solvent accessible (GB/SA) continuum solvent model to
achieve final optimized bound structure.
14. Ushio-Fukai, M.; Nakamura, Y. Cancer Lett. 2008.
15. Lim, S. D.; Sun, C.; Lambeth, J. D.; Marshall, F.; Amin, M.; Chung, L.; Petros, J. A.;
Arnold, R. S. Prostate 2005, 62, 200.
16. Govindarajan, B.; Sligh, J. E.; Vincent, B. J.; Li, M.; Canter, J. A.; Nickoloff, B. J.;
Rodenburg, R. J.; Smeitink, J. A.; Oberley, L.; Zhang, Y.; Slingerland, J.; Arnold, R.
S.; Lambeth, J. D.; Cohen, C.; Hilenski, L.; Griendling, K.; Martinez-Diez, M.;
Cuezva, J. M.; Arbiser, J. L. J. Clin. Invest. 2007, 117, 719.
17. Brar, S. S.; Corbin, Z.; Kennedy, T. P.; Hemendinger, R.; Thornton, L.;
Bommarius, B.; Arnold, R. S.; Whorton, A. R.; Sturrock, A. B.; Huecksteadt, T.
P.; Quinn, M. T.; Krenitsky, K.; Ardie, K. G.; Lambeth, J. D.; Hoidal, J. R. Am. J.
Physiol. 2003, 285, C353.
18. Xia, C.; Meng, Q.; Liu, L. Z.; Rojanasakul, Y.; Wang, X. R.; Jiang, B. H. Cancer Res.
2007, 67, 10823.
19. Druker, B. J.; Tamura, S.; Buchdunger, E.; Ohno, S.; Segal, G. M.; Fanning, S.;
Zimmermann, J.; Lydon, N. B. Nat. Med. 1996, 2, 561.
20. Goldstein, B. M.; Li, H.; Jones, J. P.; Bell, J. E.; Zeidler, J.; Pankiewicz, K. W.;
Watanabe, K. A. J. Med. Chem. 1994, 37, 392.
21. Li, H.; Hallows, W. H.; Punzi, J. S.; Pankiewicz, K. W.; Watanabe, K. A.;
Goldstein, B. M. Biochemistry 1994, 33, 11734.
22. Pankiewicz, K. W. Pharmacol. Ther. 1997, 76, 89.
4.3. M. tuberculosis and human NAD kinase inhibitory activity
23. Pankiewicz, K. W.; Zatorski, A.; Watanabe, K. A. Acta Biochim. Pol. 1996, 43, 183.
24. Pankiewicz, K. W.; Patterson, S. E.; Black, P. L.; Jayaram, H. N.; Risal, D.;
Goldstein, B. M.; Stuyver, L. J.; Schinazi, R. F. Curr. Med. Chem. 2004, 11, 887.
25. Shu, Q.; Nair, V. Med. Res. Rev. 2008, 28, 219.
26. Gebeyehu, G.; Marquez, V. E.; Van Cott, A.; Cooney, D. A.; Kelley, J. A.; Jayaram,
H. N.; Ahluwalia, G. S.; Dion, R. L.; Wilson, Y. A.; Johns, D. G. J. Med. Chem. 1985,
28, 99.
27. Chen, L.; Gao, G.; Felczak, K.; Bonnac, L.; Patterson, S. E.; Wilson, D.; Bennett, E.
M.; Jayaram, H. N.; Hedstrom, L.; Pankiewicz, K. W. J. Med. Chem. 2007, 50,
5743.
28. Umejiego, N. N.; Gollapalli, D.; Sharling, L.; Volftsun, A.; Lu, J.; Benjamin, N. N.;
Stroupe, A. H.; Riera, T. V.; Striepen, B.; Hedstrom, L. Chem. Biol. 2008, 15, 70.
29. Riera, T. V.; Wang, W.; Josephine, H. R.; Hedstrom, L. Biochemistry 2008, 47,
8689.
30. Poncet-Montange, G.; Assairi, L.; Arold, S.; Pochet, S.; Labesse, G. J. Biol. Chem.
2007, 47, 33925.
31. Mori, S.; Yamasaki, M.; Maruyama, Y.; Momma, K.; Kawai, S.; Hashimoto, W.;
Mikami, B.; Murata, K. Biochem. Biophys. Res. Commun. 2005, 327, 500.
32. Lerner, F.; Niere, M.; Ludwig, A.; Ziegler, M. Biochem. Biophys. Res. Commun.
2001, 288, 69.
33. Pignot, M.; Pljevaljcic, G.; Weinhold, E. Eur. J. Org. Chem. 2000, 3, 549.
34. Khan, A. R.; Parrish, J. C.; Fraser, M. E.; Smith, W. W.; Bartlett, P. A.; James, M. N.
G. Biochemistry 1998, 37, 16839.
The enzymatic activity was determined by a slightly modified
HPLC based assay.44 The assay mixture, consisting of 50 mM Tris/
HCl, pH 8.0, 20 mM MgCl2, ATP and NAD substrates, and an appro-
priate amount of enzymatic preparation, was incubated both in the
absence and in the presence of various concentrations of the com-
pounds to be tested. After suitable time incubation at 37 °C, the
reaction was stopped by HClO4 treatment, following neutralization
with K2CO3. HPLC analyses was performed by using the chromato-
graphic separation conditions previously set up by us in order to
separate a large number of different nucleotides, nucleosides and
bases. Briefly, samples were loaded onto a Supelcosil LC-18 re-
versed-phase column (250 ꢀ 4.6 mm, 5
lm), and elution condi-
tions was as follows: 9 min at 100% buffer A (0.1 M potassium
phosphate, pH 6.0), 6 min at up to 12% buffer B (buffer A, contain-
ing 20% methanol), 2.5 min at up to 45% buffer B, 2.5 min at up to
100% buffer B, and hold at 100% buffer B for 5.5 min; finally the
gradient returns to 100% buffer A in 5 min. The column was flushed
35. Tavale, S. S.; Sobell, H. M. J. Mol. Biol. 1970, 48, 109.

5664
R. Petrelli et al. / Bioorg. Med. Chem. 17 (2009) 5656–5664
36. Stolarski, R.; Dudycz, L.; Shugar, D. Eur. J. Biochem. 1980, 108, 111.
37. Rosemeyer, H.; Toth, G.; Golankiewicz, B.; Kazimierczuk, Z.; Bourgeois, W.;
Kretchmer, U.; Muth, H. P.; Seela, F. J. Org. Chem. 1990, 55, 5784.
38. Franchetti, P.; Cappellacci, L.; Marchetti, S.; Trincavelli, L.; Martini, C.; Mazzoni,
M. R.; Lucacchini, A.; Grifantini, M. J. Med. Chem. 1998, 41, 1708.
39. Cappellacci, L.; Franchetti, P.; Pasqualini, M.; Petrelli, R.; Vita, P.; Lavecchia, A.;
Novellino, E.; Costa, B.; Martini, C.; Klotz, K. N.; Grifantini, M. J. Med. Chem.
2005, 48, 1550.
41. Bonnac, L.; Chen, L.; Pathak, R.; Gao, G.; Ming, Q.; Bennett, E.; Felczak, K.;
Kullberg, M.; Patterson, S. E.; Mazzola, F.; Magni, G.; Pankiewicz, K. W. Bioorg.
Med. Chem. Lett. 2007, 17, 1512.
42. Butora, G.; Schmitt, C.; Levorse, D. A.; Streckfuss, E.; Doss, G. A.; MacCoss, M.
Tetrahedron 2007, 63, 3782.
43. Thompson, J. D.; Higgins, D. G.; Gibson, T. J. Nucleic Acid Res. 1994, 22,
4673.
44. Balducci, E.; Emanuelli, M.; Raffaelli, N.; Ruggieri, S.; Amici, A.; Magni,
G.; Orsomando, G.; Polzonetti, V.; Natalini, P. Anal. Biochem. 1995, 228,
64.
40. Zatorski, A.; Watanabe, K. A.; Carr, S. F.; Goldstein, B. M.; Pankiewicz, K. W. J.
Med. Chem. 1996, 39, 2422.