Acyclic nucleoside phosphonates containing a second phosphonate group are potent inhibitors of 6-oxopurine phosphoribosyltransferases and have antimalarial activity

Article abstract of DOI:10.1021/jm301893b

Acyclic nucleoside phosphonates (ANPs) that contain a 6-oxopurine base are good inhibitors of the Plasmodium falciparum (Pf) and Plasmodium vivax (Pv) 6-oxopurine phosphoribosyltransferases (PRTs). Chemical modifications based on the crystal structure of 2-(phosphonoethoxy)ethylguanine (PEEG) in complex with human HGPRT have led to the design of new ANPs. These novel compounds contain a second phosphonate group attached to the ANP scaffold. {[(2-[(Guanine-9H-yl) methyl]propane-1,3-diyl)bis(oxy)]bis(methylene)}diphosphonic acid (compound 17) exhibited a K_i value of 30 nM for human HGPRT and 70 nM for Pf HGXPRT. The crystal structure of this compound in complex with human HGPRT shows that it fills or partially fills three critical locations in the active site: the binding sites of the purine base, the 5′-phosphate group, and pyrophosphate. This is the first HG(X)PRT inhibitor that has been able to achieve this result. Prodrugs have been synthesized resulting in IC₅₀ values as low as 3.8 μM for Pf grown in cell culture, up to 25-fold lower compared to the parent compounds.

Full text of DOI:10.1021/jm301893b

Subscriber access provided by RMIT University Library
Article
Acyclic nucleoside phosphonates containing a second
phosphonate group are potent inhibitors of 6-oxopurine
phosphoribosyltransferases and have antimalarial activity.
Dianne Therese Keough, Petr Špa#ek, Dana Hockova, Tomáš Tichý, Silvie
Vrbková, Lenka Slav#tínská, Zlatko Janeba, Lieve Naesens, Michael D. Edstein,
Marina Chavchich, Tzu Hsuan Wang, John de Jersey, and Luke W Guddat
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm301893b • Publication Date (Web): 28 Feb 2013
Downloaded from http://pubs.acs.org on March 16, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 61
Journal of Medicinal Chemistry
1
1
2
3
4
5
6
7
8
9
Acyclic nucleoside phosphonates containing a second phosphonate
group
are
potent
inhibitors
of
6-oxopurine
phosphoribosyltransferases and have antimalarial activity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
b
b*
b
b
Dianne T. Keough , Petr Špaček , Dana Hocková , Tomáš Tichý , Silvie Vrbková ,
b
b
c
d
Lenka Slavětínská , Zlatko Janeba , Lieve Naesens , Michael D. Edstein , Marina
d
a
a
a*
Chavchich , Tzu-Hsuan Wang , John de Jersey and Luke W. Guddat
a
The School of Chemistry and Molecular Biosciences, The University of Queensland,
b
Brisbane, 4072, QLD, Australia; Institute of Organic Chemistry and Biochemistry,
Academy of Sciences of the Czech Republic, v.v.i. Flemingovo nám. 2, CZ-166 10
c
Prague 6, Czech Republic; Rega Institute for Medical Research, KU Leuven,
d
Minderbroedersstraat 10, B-3000, Leuven, Belgium; Australian Army Malaria
Institute, Enoggera, Brisbane, QLD 4051, Australia
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 61
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Acyclic nucleoside phosphonates (ANPs) which contain a 6-oxopurine base are good
inhibitors of the Plasmodium falciparum (Pf) and Plasmodium vivax (Pv) 6-oxopurine
phosphoribosyltransferases (PRTs). Chemical modifications based on the crystal
structure of 2-(phosphonoethoxy)ethyl-guanine (PEE-G) in complex with human
HGPRT has led to the design of new ANPs. These novel compounds contain a second
phosphonate group attached to the ANP scaffold. [(2-((Guanine-9H-
yl)methyl)propane-1,3-diyl)bis(oxy)]bis(methylene))diphosphonic acid (compound 17)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
exhibited a K value of 30 nM for human HGPRT and 70 nM for PfHGXPRT. The
i
crystal structure of this compound in complex with human HGPRT shows that it fills
or partially fills three critical locations in the active site i.e. the binding sites of the
purine base, the 5′-phosphate group and pyrophosphate. This is the first HG(X)PRT
inhibitor that has been able to achieve this result. Prodrugs have been synthesized
resulting in IC values as low as 3.8 µM for Pf grown in cell culture, up to 25-fold
50
lower compared to the parent compounds.
Keywords: malaria, HGPRT, HGXPRT, prodrugs, crystal structure, purine salvage,
phosphoribosyltransferase
Running Title: New ANPs have antimalarial activity
ACS Paragon Plus Environment

Page 3 of 61
Journal of Medicinal Chemistry
3
1
2
3
4
5
6
7
8
9
INTRODUCTION
Malaria remains one of the most important infectious diseases in the world today,
affecting half of the world’s population, with an estimated 216 million cases and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
reports of at least 655,000 fatalities in 2010. Though there are five strains of
Plasmodium that can infect humans, the most widespread and lethal are falciparum
2
(Pf) and vivax (Pv).
Resistance to the current drugs (including artemisinin based combination therapies) for
treatment and prophylaxis of malaria is on the rise, underlining the need for the
3
discovery of new drug targets and therapeutics. One of the essential features required
for long-term use of an effective chemotherapeutic is to target proteins/enzymes that
would be difficult for the parasite to mutate without compromising its own ability to
replicate. A second important criterion is the cost of synthesis of potential drugs. The
purine salvage enzyme hypoxanthine-guanine-(xanthine) phosphoribosyltransferase
(
HG(X)PRT) and the acyclic nucleoside phosphonate (ANP) inhibitors are able to
address both these factors. HG(X)PRT has long been recognized as a malarial drug
target because its activity is essential for the synthesis of nucleoside monophosphates
required for DNA/RNA production. Figure 1 shows the reaction catalyzed by
HG(X)PRT.
It has been shown that ANPs, which inhibit PfHGXPRT, arrest parasite growth in cell
4
culture. Further, recent metabolic studies have confirmed that PfHG(X)PRT activity
is critical for the survival of Pf in cell culture. Most importantly, these studies
demonstrated that potent inhibitors of this enzyme exert their ability to arrest the
5
growth of Pf in cell culture via inhibition of the purine salvage pathway. Genomic
sequencing has suggested that, like Pf, Pv does not possess the enzymes necessary to
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 61
4
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
synthesize the purine ring de novo and it relies on the salvage pathway for the
6
production of its purine nucleoside monophosphates. Thus, it is hypothesized that
inhibitors of PvHGPRT will also be capable of arresting the growth of this parasite.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ANPs are cost-effective to produce. This is demonstrated by the fact that three
prototype ANPs with an adenine or cytosine base are in use as antiviral drugs
7
(tenofovir, adefovir and cidofovir) , with tenofovir being widely distributed to HIV-
infected individuals in third world countries. The ability of these compounds to act as
anti-viral agents is by inhibition of viral polymerases or reverse transcriptases after
they have become phosphorylated by cellular kinases. Such compounds do not inhibit
HG(X)PRT as neither pyrimidine bases nor purine bases containing a 6-amino group
are recognized by this enzyme. New ANPs were then designed where the base is a 6-
oxopurine so that they should specifically inhibit HG(X)PRT. Such compounds were
subsequently found to selectively inhibit human HGPRT, PfHG(X)PRT and
8
PvHGPRT , to arrest the in vitro growth of Pf cell lines and to have low cytotoxicity in
4b
mammalian cell lines. An important feature of the ANPs is that they contain a stable
carbon-phosphorous bond so they cannot be hydrolysed in the cell to inactive
derivatives. Schramm and colleagues have suggested that the reason why the transition
state analogs, immucillin 5′-phosphates, cannot act as antimalarials is because of
5
c
hydrolysis of this group by phosphomonoesterases. ANPs also possess a further
desirable property and this is that they can be chemically modified to produce prodrugs
9
to increase cell permeability.
Previously, 9-(2-(phosphonoethoxy)ethyl)guanine (PEEG) or –hypoxanthine (PEEHx)
(
Figure 2), were found to be good and selective inhibitors of human HGPRT and
4
b, 8a
PfHGXPRT.
The K values of PEEG for human HGPRT and PfHGXPRT are 1.0
i
and 0.1 µM, respectively and, for PEEHx, are 3.6 and 0.3 µM. Crystal structures of
ACS Paragon Plus Environment

Page 5 of 61
Journal of Medicinal Chemistry
5
1
2
3
4
5
6
7
8
9
these two compounds in complex with human HGPRT suggested possibilities for the
design of more potent inhibitors. One idea was to synthesize new compounds
i
containing a second phosphonate group designed to occupy the PP binding pocket.
10
Thus, previously synthesized bisphosphonates as well as newly designed ANPs with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a phosphonate group attached to the PEE, 2-(phosphonomethoxy)ethyl (PME) or 3-
(phosphonomethoxy)propyl scaffold could be tested as inhibitors of HG(X)PRTs.
A series of lipophilic prodrugs of the ANP inhibitors were then synthesized and their
IC50 values for the growth of Pf in cell culture determined. In parallel, the cytotoxicity
of these compounds was measured in three different cell lines, i.e. human lung
carcinoma cells, a human melanoma cell line and a derived HGPRT-deficient mutant
cell line. The crystal structure of human HGPRT in complex with [[(2-((guanine-9H-
yl)methyl)propane-1,3-diyl)bis(oxy)]bis(methylene))diphosphonic acid (compound
17), which has the highest affinity for human HGPRT and PfHGXPRT, shows how
this class of ANPs bind in the active site.
RESULTS
Chemistry. The key branched hydroxyderivative 2 for the synthesis of unsymmetrical
bisphosphonate type of ANPs (Scheme 1, compounds 7 and 8) was prepared by the
1
1
multistep sequence starting from racemic solketal. For the introduction of the acyclic
9
moiety to the N -position of 6-chloropurine or 2-amino-6-chloropurine the Mitsunobu
1
2
reaction was applied. The resulting 6-chloropurine bisphosphonate 3 was transformed
to hypoxanthine derivative 5 by nucleophilic aromatic substitution in acidic conditions
(75% aqueous trifluoroacetic acid). In the case of 2-amino-6-chloropurine
bisphosphonate 4, the Mitsunobu reaction had to be followed by heating in
13
water/tetrahydrofuran to decompose the triphenylphosphoranylidene intermediate
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 61
6
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
rising from the presence of the free amino group. The chlorine atom was next
displaced with hydroxyl in quantitative yield as described above to form guanine
9
derivative 6. This two step approach for preparation of N -substituted
hypoxanthine/guanine derivatives affords better results than the direct alkylation of 6-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
oxopurines complicated by the formation of N -regioisomers. To form free racemic
bisphosphonic acids 7 and 8, both phosphonate moieties of 5 and 6 were
simultaneously deprotected under standard conditions using Me
3
SiBr/acetonitrile
followed by hydrolysis.
For the synthesis of symmetrical bisphosphonates 16 and 17, the known precursor 11
was prepared. Although we have reported the variant synthesis of guanine derivative
1
0b
17 previously, the above described synthetic approach for the unsymmetrical ANPs
(including Mitsunobu reaction, followed by aromatic substitution and ester cleavage),
is a new synthetic method (Scheme 2) for the preparation of the guanine
bisphosphonate as well as the new hypoxanthine derivative 16.
The tetraesters of bisphosphonates 5-6 and 14-15 were used for the direct preparation
9
of the phosphoramidate prodrugs 9-10 and 18-19 by our recently published method
1
4
that is highly efficient. In the first step of the one-pot reaction, sequence the
deprotection of the phosphonate esters 5-6 and 14-15 with Me SiBr forms in situ the
3
tetra(trimethylsilyl) esters. In the second step the reaction of these intermediates with
ethyl (L)-phenylalanine in the presence of 2,2′-dithiodipyridine (Aldrithiol) and
triphenylphosphine yield the corresponding tetra-amidates 9-10 (Scheme 1) and 18-19
(
Scheme 2).
ACS Paragon Plus Environment

Page 7 of 61
Journal of Medicinal Chemistry
7
1
2
3
4
5
6
7
8
9
Inhibition of the human, Plasmodium falciparum and Plasmodium vivax 6-
oxopurine PRTases.
A series of ANPs containing a second phosphonate group attached to the acyclic linker
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
between the N atom of the purine base and the first phosphonate group were
synthesized, with the aim of increasing the potency for the 6-oxopurine PRTases
(Schemes 1 and 2). In these structures, the second phosphonate group is attached either
9
to the first or second carbon atom from N of the purine base (Figure 3). The number
9
9
of atoms between N and the first phosphorus atom is 5 or 4, and between N and the
second phosphorus atom varies between 4 and 6. In compounds 17, 16 and 20 (Figure
3
), the oxygen atom is located in a different position compared with the PEE
compounds shown in Figure 2. While in compounds 17, 16 and 20 the bisphosphonate
acyclic moiety is symmetric, compound 8 have different linkers connecting the first
i
and the second phosphonate group. The K values for these compounds are given in
Table 1.
Compound 17 (Scheme 2) is a potent inhibitor of the human and Pf enzymes. The
attachment of a second phosphonate group results in a 33-fold decrease in K value for
human HGPRT but only a 1.5-fold decrease for PfHGPRT compared with PEEG.
i
4
b, 8a
Although this compound is an excellent inhibitor of these two enzymes, it is not as
effective against PvHGPRT (Table 1). The biggest decrease in K for this compound
i
occurs for the human enzyme and, thus, the attachment of this second group is
accompanied by a loss in selectivity.
Human HGPRT and PfHGXPRT have 48% amino acid sequence identity. There are
5c, 15
only two published structures of PfHGXPRT.
Comparison of these crystal
structures with those of human HGPRT show that the residues whose side-chains that
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 61
8
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
6
enter the active site are all identical. However, the active sites contain several flexible
loops that can change their conformation depending on the inhibitor that is bound.
These changes contribute to the K
i
values. It has been demonstrated that selectivity
does exist between the enzymes. This is based on the fact that purine bases with simple
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
17
atomic substitutions have different K values. It This is one avenue that can be
m
explored to increase selectivity. Furthermore, other ANPs can also exhibit
4b, 5c, 8e
selectivity.
The attachment of a second phosphonate group decreases the K
i
8e
values as predicted from modelling but such compounds bind to human HGPRT and
PfHGPRT with similar affinities. The selectivity issue can only be addressed by further
crystal structures of PfHGXPRT in complex with ANPs containing a second
phosphonate group.
Compound 16 (Scheme 2), with hypoxanthine instead of guanine as the purine base,
i
has a higher K value for all three enzymes, compared with compound 17. This
difference is 33-fold for human HGPRT, 71-fold for PfHGXPRT and 3-fold for
PvHGPRT (Table 1). In comparison, PEEG also binds more tightly to the human and
Pf enzymes than PEEHx, but in this case the difference between the guanine and
4b, 8a
hypoxanthine analogues is only 3-fold.
Thus, for these bifunctional ANPs, the
nature of the purine base makes a marked contribution to the affinity.
Compound 8 (Scheme 2), with guanine as the base, has a decreased affinity for the
human and Pf enzymes compared with compound 17. This could be attributed to one
or both of two factors: (i) the phosphonate group which binds in the 5′-phosphate
binding pocket is too long for optimal interactions; and/or (ii) it is the positioning of
the oxygen atom in the linker in the second phosphonate tail that influences affinity.
For PvHGPRT, neither of these factors has any effect as the K is the same for
i
ACS Paragon Plus Environment

Page 9 of 61
Journal of Medicinal Chemistry
9
1
2
3
4
5
6
7
8
9
compound 17 and compound 8. In this series, compound 8 is the only one with similar
affinity for all three enzymes.
Compound 20 is a weak inhibitor of the human HGPRT and, though not a potent
inhibitor of the two parasite enzymes, it prefers PfHGXPRT and PvHGPRT (Table 1).
Compound 20 is a derivative of 2-(phosphonomethoxy)ethyl guanine (PMEG). PMEG
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
i
inhibits the human and Pf enzymes with similar K values to those found for compound
4b
2
0, i.e. 29 µM for human HGPRT and 1.6 µM for PfHGXPRT. Therefore, it is
unlikely that, in this instance, the second phosphonate group makes interactions with
any amino acids in the active site and, therefore, does not contribute to the affinity of
these ANPs for the enzymes.
Crystal
structure
of
[(2-((guanine-9H-yl)methyl)propane-1,3-
diyl)bis(oxy)]bis(methylene))diphosphonic acid (compound 17) in complex with human
HGPRT
The crystal structure of compound 17 in complex with human HGPRT has been
determined to 2.0 Å resolution, with the asymmetric unit constituting a human HGPRT
tetramer (Figure 4a; Table 2).
The electron density map in Figure 4b exemplifies that each active site is occupied by
a single molecule of compound 17, a sulfate ion (from the crystallization buffer) and
two magnesium ions. Superimposition of the four subunits shows that the purine ring,
the two phosphonate groups, the sulfate ion and the two magnesium ions align with a
high degree of precision (rmsd for all atoms < 0.2 Å). Thus, the purine ring and the
two phosphonate groups are firmly held in place. However, due to the free rotation of
the dihedral angles within the linker regions, there is variability in the location of these
atoms with differences in the atomic coordinates of up to 1.4 Å.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 61
1
0
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Factors that influence the tight binding of compound 17
There are a number of factors that contribute to holding compound 17 in the active site.
These are: (i) the purine base; (ii) the phosphonate group mimicking the 5′-phosphate
group of PRib-PP or GMP; (iii) the second phosphonate group (attached to the linker
connecting the purine base to the phosphonate group) mimicking pyrophosphate; (iv)
the chemical nature of each of the two acyclic linkers; and (v) the presence or absence
of divalent metal ions in the active site. These are discussed below.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Purine binding site
The purine base of compound 17 slots under the aromatic ring of F186 forming a Π
stacking interaction. This arrangement occurs in all the published structures of the 6-
oxopurine PRTases when a compound containing a purine base is bound. Four
hydrogen bonds contribute to the binding of compound 17. These are between (i) the
7
6-oxo group of the guanine base and the NZ atom of K165; (ii) N of the guanine base
with NZ of K165; (iii) the 6-oxo group of the guanine base and the backbone amide of
V187; and (iv) the amino group of the guanine base with the carbonyl group of D193.
The formation of bond (iv) is a likely reason why the replacement of guanine by
hypoxanthine increases the K
i
by ~30-fold as this interaction cannot be made when a
hydrogen is substituted for the exocyclic amino group.
5′-phosphate binding site
The first phosphonate group of symmetric compound 17 is designed to reach into the
5′-phosphate binding pocket in a similar way as the 5′-phosphate group of PRib-PP or
GMP. This pocket is defined by residues 137-141 which encircle this group (Figure
4c). This phosphonate group contributes to the affinity of ANPs for these enzymes and
ACS Paragon Plus Environment

Page 11 of 61
Journal of Medicinal Chemistry
1
1
1
2
3
4
5
6
7
8
9
is essential for tight binding. When compound 17 is bound, a network of hydrogen
bonds between the main-chain nitrogen atoms of D137, G139, T141 and the hydroxyl
groups of T138 and T141 and the oxygen atoms of the phosphonate group is formed
(
Figure 4c).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Binding site of the second phosphonate group
The second phosphonate group of compound 17 points down into the vicinity of the
i
PP binding pocket (Figure 4c). When binding occurs, the backbone dihedral angles
for K68 rotate such that the side chain of this amino acid residue rotates away from the
active site by 180° (Figure 4d) and occupies the same position as when the transition
2+
18
i
state analog, ImmGP and Mg .PP are bound. The NZ atom of K68 in the human
HGPRT.compound 17 complex then forms a hydrogen bond with the carbonyl group
of V96 in an adjacent subunit. In the absence of the second phosphonate group i.e.
i
when PEEG or PEEHx are bound, K68 occupies the same location as PP and is
4b
pointed upwards towards the PEE scaffold. The phosphoryl oxygen atoms of this
2
+
phosphonate group in compound 17 are anchored in position by Mg and via this ion
to the side chain of D193. This appears to be an important interaction that contributes
to the tighter binding of compound 17 compared with PEEG.
With the rotation of the side chain of K68, this area of the enzyme is then solvent
accessible and a sulfate ion, present in the crystallization solution, is able to enter the
active site. The sulfate ion is located in the same position as one of the phosphate
18
i
groups of PP when it is bound together with the transition state analog. Its
interactions with the active site residues are therefore similar to that of the phosphoryl
oxygen atoms of one of the phosphate groups in PPi. It is unlikely that the presence of
sulfate ions effects either the location or affinity of compound 17. This interpretation is
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 61
1
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
based on the observation that the addition of sulfate ions to the assay did not result in a
decrease in K value for compound 17.
i
Magnesium ions in the structure of the human HGPRT.compound 17 complex
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
It has been hypothesized that magnesium ions bind to PRib-PP prior to this substrate
entering the active site. Human HGPRT obeys a sequential mechanism with PRib-PP
i
binding first, followed by the purine base. PP then leaves followed by the nucleoside
1
9
monophosphate product, whose release is the rate-limiting step. The catalytic
mechanism of the two Plasmodium enzymes has not been determined. However, it is
assumed that they follow the same reaction pathway though this may not necessarily
be the case. There are no divalent metal ions present in the crystal structure of free
2
0
16
human HGPRT or when it is in complex with GMP or IMP. However, two
magnesium ions are located in the active site when the transition state analog, ImmGP,
i
and PP are bound. One magnesium ion is bound to the two hydroxyl groups of the
ribose ring and to two oxygen atoms of pyrophosphate, while the second is coordinated
1
8
to two oxygen atoms of pyrophosphate and to the OD1 atom of D193.
In the human HGPRT.compound 17 complex, one magnesium ion is bound to the
phosphonate group located close to the PP binding site as described above. The second
is bound to a carboxylate oxygen atom of E133 (2.1 Å) and to the OD1 atom of D134
i
2+
18
(
2.1 Å). This is a different arrangement from that found when ImmGP.Mg .PP is
i
bound. In that instance, there are no direct interactions with either of these side chains
and magnesium ion is only linked to the side chain of E133 via a water molecule. It is
unclear if, in some of the 6-oxopurine PRTases, magnesium ions bind independently
before the substrates enter. It has been proposed that, if the unliganded enzyme is
2
+
found to contain Mg , then the role of these ions may be to help to stabilize the active
ACS Paragon Plus Environment

Page 13 of 61
Journal of Medicinal Chemistry
1
3
1
2
3
4
5
6
7
8
9
2
1
site so that the catalytic reaction can proceed. Magnesium ions have been observed in
2
1
crystal structures of the two E. coli 6-oxopurine PRTases in the absence of ligands. In
E. coli XGPRT, this divalent metal ion is stabilized by bonds to surrounding water
residues and to the OD1 atom of D89 and in E. coli HPRT to the side chain atoms of
E103 and D104. The corresponding residues in human HGPRT are E133 and D134. In
the crystal structures of human HGPRT in complex with three different ANPs, i.e. 2-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
phosphonoethoxy)ethyl- guanine or -hypoxanthine (PEEG and PEEHx) and (S)-3-
4b
hydroxy-2-(phosphonomethoxy)propyl guanine [(S)-(HPMPG)], no divalent metal
ions were found in the active site. This may be because these compounds more closely
mimic that of the nucleoside monophosphate product of the reaction. Thus, it is only
when in complex with the ANP, compound 17, that two divalent metal ions have
become coordinated. The role of this second ion in human HGPRT in complex with
this inhibitor may well be to shape the active site to allow for tighter binding.
In vitro antimalarial activity and cytotoxicity studies
In order to increase cell permeability, the negative charges on the ANPs may need to
be masked. One approach is to attach hydrophobic or lipophilic groups to the parent
ANP by an ester or phosphoramidate bond. The concept for these delivery systems and
the ability to be hydrolysed enzymatically in vivo to the active compound is based on
9
the previous experiences with related ANPs that are anti-viral drugs. Initially,
22
prodrugs of PEEG (Figure 2) were synthesized. As shown in Figure 5, compound
21 contains two phosphoramidate moieties, whereas the other two prodrugs contain
one (compound 23) or two (compound 22) hexadecyloxypropyl chains. The IC50
values for the prodrugs of PEEG in the chloroquine-sensitive and the chloroquine-
resistant strains are given in Table 3. The cytostatic concentrations were determined in
three human cell lines (Table 3).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 61
1
4
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The prodrugs all have lower IC50 values compared to PEEG against Pf cell lines. The
most active prodrug (compound 23) has a single lipophilic chain attached to the
phosphonate group. However, this compound is also the most toxic in mammalian cell
lines. Addition of the second lipophilic group (compound 22) reduces the antimalarial
activity compared to the single prodrug, but this compound is still considerably more
active than PEEG. It has been demonstrated that intracellular cleavage of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
hexadecyloxypropyl prodrugs of ANPs is performed by phospholipase C. The lower
activity of the double, compared to the single ester prodrug, is consistent with a
22a
previous report.
This may be attributed to the fact that either the phosphonate
double-ester prodrug penetrates the cell membrane less rapidly than its monoester
counterpart or that the monoester is hydrolyzed more rapidly once inside the cell to the
23
active ANP or, perhaps, a combination of both these factors.
Highly polar bisphosphonates (such as compounds 17, 16, 8 and 20; Figure 3) are
unable to cross the cell membranes. A phosphoramidate type of prodrug was therefore
synthesized to try to increase their antimalarial activity. Since the alanine-based
phosphoramidate compound 21 (Table 3) was not as successful as compound 23
(
Table 3), the more hydrophobic ethyl (L)-phenylalanine was used to mask all four
2
4
hydroxy groups (Figure 6). The results of the in vitro antimalarial activity and
cytotoxicity of the tetraphosphoramidates are given in Table 4.
The prodrugs are effective in inhibiting parasite growth with a decrease in the IC50
value of at least 100-fold compared with compound 17. Compound 18 containing
hypoxanthine as the base is slightly more effective than when guanine is the base
(
compound 19). However, it is more cytotoxic in mammalian cells. Compound 10 has
similar cytotoxic properties as compound 18. Though these compounds are slightly
better inhibitors of PfHGXPRT compared with PEEG (K = 0.1 µM), the in vitro data
i
ACS Paragon Plus Environment

Page 15 of 61
Journal of Medicinal Chemistry
1
5
1
2
3
4
5
6
7
8
9
suggests that the efficient design of prodrugs, leading to good cell permeability and
efficient hydrolysis within the cell are important factors to be considered in drug
design. The corresponding prodrug of the bisphosphonate compound 20 was not
synthesized for cell based asays. This is because compound 20 has similar K values as
i
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
PMEG, the compound on which it was modeled. PMEG has an IC50 value against Pf
cell lines (14 ꢀM), but its CC50 in mammalian cells is 17 µM, resulting in a low SI
4b
value.
Discussion
i
One of the new bifunctional ANPs, compound 17, has a K of 30 nM for human
HGPRT, 70 nM for PfHGXPRT and 600 nM for PvHGPRT. The crystal structure of
compound 17 in complex with human HGPRT shows that this second phosphonate
i
group is located in the vicinity of the PP binding pocket (Figure 4c). It is held in
position by interactions with a magnesium ion (2.1 Å) and, through this, to the
carboxylate group of D193 (2.1 Å). This is the only interaction that this group forms
with active site amino acid residues and is a critical contribution to its low K
i
value. In
the human HGPRT.compound 17 complex, the side chain of K68 has moved from its
16
location found in the free structure or in the human.GMP complex. This movement
results in the PP binding site becoming vacant, making room for sulfate to enter the
i
active site. It could be argued that the reverse situation occurs and that sulfate binds
first, allowing compound 17 to bind. However, the addition of either sulfate or
phosphate has no effect on the K
i
for compound 17, making this an unlikely
proposition. The sulfate ion is located in precisely the same position as one of the
phosphate groups of pyrophosphate when PP
i
is bound in the active site together with
2
+
2+
ImmGP and Mg . One of the sulfate oxygen atoms forms interactions with Mg (2.2
Å) and the OD1 of D193 (2.9 Å) while the other is now 2.9 Å from the amide nitrogen
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 61
1
6
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of K68. The only interaction with the phosphonate group of compound 17 is through
the same magnesium ion (Figures 4c and 4d). Thus, though the presence of sulfate in
the active site is a result of its presence in the crystallization buffer and does not itself
contribute to the binding of compound 17, its location does suggest chemical
modifications to compound 17 to increase its potency.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
There is also a second magnesium ion present in the human HGPRT.compound 17
complex. This is bound to the carboxylate side chains of E133 and D134. Though the
rationale for the presence of this magnesium ion is not fully understood, it can be
speculated that its presence may help to shape the active site allowing compound 17 to
bind in its optimal position.
9
If the phosphonate group is attached one atom closer to the N atom as in compound
1
0b
i
20 (Figure 3), there is no change in the K value compared with PMEG on which
this ANP was based. This suggests that, in this case, the second phosphonate group is
9
too close to N and cannot bind in the vicinity of pyrophosphate. Thus, the interactions
between PMEG and compound 20 with active site residues are identical irrespective of
whether the second phosphonate moiety is attached or not. This accounts for the fact
i
that there is no change in the K value for either the human or Pf enzymes.
i
The K value for compound 17 for human HGPRT is 33-fold lower than for PEEG.
However, there is not such a large difference between the K values for compound 17
i
and PEEG for PfHGXPRT (1.5-fold). This suggests that the interactions between
compound 17 and human HGPRT which contribute to the high affinity may be slightly
different from those that occur when compound 17 binds to PfHGXPRT. One likely
explanation is that magnesium ions do not bind to PfHGXPRT in the presence of
compound 17. Though the free structure of human HGPRT does not contain metal
ACS Paragon Plus Environment

Page 17 of 61
Journal of Medicinal Chemistry
1
7
1
2
3
4
5
6
7
8
9
ions, they are always present in the storage buffer and appear to be necessary for this
enzyme to maintain its maximum activity. In comparison, PfHGXPRT does not require
metal ions to maintain its structure. This enzyme is stable for at least 24 months as
long as PRib-PP and hypoxanthine are present. Indeed, totally inactive PfHGXPRT,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
+
which occurs during the purification in the absence of Mg , PRib-PP and
hypoxanthine, can be restored to full activity by adding these two substrates in the
absence of divalent metal ions to the inactive enzyme. For PfHGXPRT, divalent metal
ions may only be important in catalysis. It is conceivable that compound 17 binds to
PfHGXPRT in the absence of metal ions and that the interaction between the active
site residues in human HGPRT (carboxylate group of D193), the second phosphonate
2
+,
group and Mg cannot occur. Hence, there is no dramatic reduction in the K
i
values
for PEEG and compound 17 with PfHGXPRT. There are two published structures of
2
+
5c, 15
PfHGXPRT but, in both cases, Mg .PP
i
is present.
i
As PP should bind only in the
presence of a divalent metal ion, the structure of PfHGXPRT.compound 17 is required
to determine if compound 17 can bind in their absence. An explanation for the slightly
i
higher K for compound 17 with PvHGPRT compared with the human and Pf enzymes
cannot be advanced as the structure of this enzyme is presently unknown.
The other change in the chemical structures of compound 17 and compound 16
(
Figure 3) compared with the PEE moiety is isosteric. In these new inhibitors, the
9
oxygen is moved to a different position in the linker connecting the N atom of the
purine ring to the phosphonate group binding in the 5′-phosphate pocket. This change
alters the shape of the linker so that there is one more hydrogen bond between the
phosphoryl oxygen atoms with the amino acid residues (137-141) for compound 17
than for PEEG. This is between the amide nitrogen atom of T138 and the oxygen atom
(
O1) that forms bonds with the amide atom of D137 and G139. There are also closer
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 61
1
8
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
interactions between the phosphoryl oxygen atoms of O2 and O3 and the active site
amino acid residues for compound 17 than for PEEG. The bond between O2 and the
carboxyl oxygen atom of T141 is slightly tighter for compound 17 compared with
PEEG (2.4-2.7 Å for compound 17 and 2.6-2.8 Å for PEEG). The O3 atom of
compound 17 and the amide nitrogen of T138 forms a closer interaction than in the
human.PEEG complex (2.6-2.9 Å compared with 2.9-3.14 Å). There is also a tighter
interaction between O3 and the carbonyl atom of T138 (2.4-2.8 Å compared with 2.8-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3.0 Å). Thus, these tighter interactions found for compound 17 are another factor
contributing to its affinity.
There are four amino acid side chains or backbone atoms that have been found to form
hydrogen bonds to guanine. These are the carbonyl group of V187, the OD1 atom of
D193, the NZ atom of K165 and the OD1 atom of D137. In the purine base, four atoms
1
are capable of forming hydrogen bonds to these active site residues (N , exocyclic
7
amino group at position 2, exocyclic atom at position 6 and N ). However, this
constellation of hydrogen bonds differs in the structures of human HGPRT in complex
with these two different ANPs, PEEG and compound 17. In complex with compound
1
7, there is a hydrogen bond between the 2-amino group of guanine with the carbonyl
atom of D193 (2.9-3 Å). This does not occur when PEEG binds as it is too far away
3.5-4 Å). The exocyclic atom in the 6-position of the purine base also form a weaker
(
interaction with the amide nitrogen of V187 for PEEG (cf. 2.8 -2.9 Å for compound 17
7
with 3.0-3.5 Å for PEEG). The NZ atom of K165 forms a hydrogen bond with the N
atom of the purine ring when compound 17 binds (2.5-2.7 Å). However, this hydrogen
bond is absent in the human.PEEG complex (3.2-3.5 Å). Thus, the nature of the linker
is an important contributing factor in determining the optimal orientation and location
ACS Paragon Plus Environment

Page 19 of 61
Journal of Medicinal Chemistry
1
9
1
2
3
4
5
6
7
8
9
of the purine base and the phosphonate group which binds in the 5′-phosphate binding
pocket.
There are subtle differences in the binding of guanine in the active site depending on
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
b, 16, 18
the chemical nature of the attachment to the base.
However, there are always
two common interactions with the active site amino acid residues: (i) a hydrogen bond
1
between the N atom with the carbonyl oxygen of V187; and (ii) a hydrogen bond
between the 6-oxo group with the NZ atom of K165. The differences lie in the
7
interactions of the 2-amino group and the N proton. The 2-amino group in compound
1
7 is able to form two hydrogen bonds with active site residues though, in the case of
PEEG, only one such bond occurs. For the immucillins, PEE and other ANP
derivatives, human HGPRT has higher affinity for the compounds containing guanine
instead of hypoxanthine. The exception is IMP and GMP where there is little
17
i
difference in the K value. It is clear that the loss of hydrogen bonds when the
exocyclic amino group is replaced by a proton would result in a decrease in affinity.
However, this alone is not sufficient to explain the fact that, for the PEEG and PEEHx,
this difference is only 4-fold while, for compound 17 and compound 16 (Table 1), this
difference is 33-fold. The structures of four inhibitors containing guanine as the base
(
ImmGP, GMP, PEEG and compound 17) in complex with human HGPRT show that
there are subtle differences between these structures in the binding of the base. For
7
example, the N atom in compound 17 forms a hydrogen bond with the NZ atom of
7
K165. This does not occur when GMP or ImmGP are bound. The N atom of ImmGP
rather forms a hydrogen bond with the carboxyl atom of D137 while, in the GMP
7
complex, N does not form a hydrogen bond to any of the atoms in the active site.
Another difference in binding is the hydrogen bond between the 6-oxo group in
compound 17 and the amide nitrogen of V187. Such a hydrogen bond cannot form for
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 61
2
0
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
PEEG, ImmGP or GMP. Thus, one possibility to explain the differences in affinity
between compound 17 and compound 16 (Table 1) is that, when hypoxanthine
replaces guanine, there is a subtle difference in the binding of this base weakening the
interactions with the enzyme.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The attachment of lipophilic groups through an ester or phosphoramidate bond leads to
a reduction in IC50 of the prodrugs when compared with the active ANPs (Tables 3 and
4
). The ability of the ANP prodrugs in erythrocyte cultures to act as anti-malarials can
be attributed to their first being able to enter the cells and their subsequent hydrolysis
within the cells by inherent enzymes. Whether the hydrolysis of the produg occurs
within the red blood cell and it is the ANP that enters the parasite, or, whether it is the
prodrug that is transported and then hydrolysed within the parasite itself, is at present
unknown. Also unknown is the resultant concentration of the active compound within
the parasite. These factors are currently being addressed and should lead to the design
of more effective anti-malarials.
Compounds 21 and 22 (Table 3) and the compound 17 (Tables 1 and 4) have little, if
any cytostatic effects in the A549, in the C32 and C32TG cell lines. The prodrug of
compound 17 (compound 19, Table 4) also has low though detectable cytotoxicity
values in C32 and C32TG cell lines. It is hypothesized that these cytotstatic values are
not because these prodrugs cannot enter the mammalian cells. This is based on the fact
that they must have traversed this barrier to reach their parasitic target. There are two
pieces of evidence that suggest that those compounds (compound 23, Table 3 and
compounds 18 and 10, Table 4) which are slightly cytostatic do not exert these effects
by inhibition of human HGPRT activity. The first is that compound 17 (Table 1) has
the lowest K value of all the ANPs for human HGPRT (0.03 µM) while the parent
i
i
compounds of 18 and 10 (16 and 8, Table 1) have higher K values of 1 and 0.6 µM,
ACS Paragon Plus Environment

Page 21 of 61
Journal of Medicinal Chemistry
2
1
1
2
3
4
5
6
7
8
9
respectively. Thus, if inhibition of human HGPRT was the cause of the cytotoxicity,
then compound 17 (and its prodrug 19) would be expected to have the higher CC50
value. The second piece of evidence is there is no difference in the CC50 values for
these compounds between C32 and C32TG. As C32TG cannot express HGPRT
activity, it would be expected to have lower cytotoxicity values than in C32 if human
HGPRT was the target. The explanation for the weak though measurable cytotoxicity
is complicated by the fact that these cells have a high proliferation rate with a very
active pathway of de novo purine biosynthesis. A549 cells have a higher rate of
proliferation than the C32 and C32TG cells, which may make them less sensitive. One
possible suggestion to explain the cytostatic properties is that certain ANPs can be
converted to toxic metabolites by enzymes such as kinases present within the cell.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In conclusion, a new potent inhibitor (compound 17) of the 6-oxopurine PRTases has
been discovered. The crystal structure of compound 17 in complex with human
HGPRT has suggested reasons for the affinity for human HGPRT and possible
explanations for the affinity for PfHGXPRT. This structure opens possibilities for the
design of even more potent inhibitors. Chemical modifications whereby hydrophobic
groups are linked to this compound via a phosphoramidate bonds result in an increase
in the ability to arrest parasitic growth. The prodrug of compound 17 (compound 19)
has little, if any, cytostastic properties in the mammalian cell line, A549. This data
suggests that compound 17 is a promising lead compound for further development to
increase its antimalarial efficacy.
EXPERIMENTAL SECTION
Synthesis and Analytical Chemistry. Unless otherwise stated, solvents were
2 5
evaporated at 40 °C/2 kPa, and the compounds were dried over P O at 2 kPa. NMR
1
13
spectra were recorded on Bruker Avance 500 ( H at 500 MHz, C at 125.8 MHz),
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 61
2
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
13
1
Bruker Avance 600 ( H at 600 MHz, C at 151 MHz) and Bruker Avance 400 ( H at
13
4
00 MHz, C at 100.6 MHz) spectrometers with TMS as internal standard or
referenced to the residual solvent signal. Mass spectra were measured on a ZAB-EQ
VG Analytical) spectrometer. The purity of the tested compounds was determined by
(
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the combustion analysis (C, H, N) and was higher than 95%. The chemicals were
obtained from commercial sources or prepared according to the published procedures.
THF was distilled from sodium/benzophenone under argon. Unless otherwise stated,
preparative HPLC purifications were performed on columns packed with 7 ꢀm C18
reversed phase resin (Waters Delta 600 chromatograph column), 17 × 250 mm; in ca.
200 mg batches of mixtures using gradient MeOH/H O as eluent.
2
Synthesis of the ANPs
Tetraethyl ([(3-(hydroxy)propane-1,2-
diyl)bis(oxy)]bis(ethylene))bis(phosphonate) (2)
Solketal (20.8g, 157.4 mmol) was transform to 4-((benzyloxy)methyl)-2,2-dimethyl-
11b
1
,3-dioxolane by the known procedure (yield 33.8 g). Then methanol-water (10:1,
100 ml) and DOWEX (50x8; 80 g) was added and the reaction mixture was refluxed
1
11a
for 4 h. The DOWEX was removed by filtration. The solvent was evaoprated in
vacuo and the residue was co-distilled with toluene (2x). The resulting crude 3-
(
benzyloxy)propane-1,2-diol (25.8 g, presence was confirmed by GC-MS) was used
without purification to the next step: To a well suspended mixture of 3-
(benzyloxy)propane-1,2-diol (9.1 g) and Cs CO (16.3 g. 49.8 mmol) in tert-BuOH
52.5 ml) under argon, diethyl vinylphosphonate (24 ml, 155 mmol) was added. The
2
3
1
1c
(
reaction mixture was stirred at RT for 5 days. The resulting mixture was poured to the
water (200 ml) and extracted by EtOAc (2 x 200 ml). Organic layers were collected
ACS Paragon Plus Environment

Page 23 of 61
Journal of Medicinal Chemistry
2
3
1
2
3
4
5
6
7
8
9
4
and washed by water (2 x 200 ml) and brine (1 x 200 ml) and dried with MgSO .
Solvent was evaporated in vacuo and residue (29.18 g) was used without purification
to the next step. Part of this crude tetraethyl ([(3-(benzyloxy)propane-1,2-
diyl)bis(oxy)]bis(ethylene))bis(phosphonate) (1, 5.9 g) was purified by flash
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
chromatography (CHCl :MeOH; gradient 0.5-5 %) for determining the NMR spectra
3
and UPLC-MS.
A solution of intermediate 1 (23.3 g) in MeOH (480 ml) containing a catalytic amount
2
of PdCl in HCl (aq) was hydrogenated over 10% Pd/C (2.2 g) at room temperature
overnight. The suspension was filtered over Celite and washed with methanol (50 ml)
and the filtrate was made alkaline by triethylamine. The solvent was evaporated in
vacuo and the residue was purified by flash chromatography (CHCl
3
:MeOH; gradient
1-6%) to give 2 (11 g) as colourless oil. The overall yield (from solketal) was 23.5%.
1
6
H NMR (DMSO-d ): 4.63 bt, 1 H, J(OH,1') = 5.6 (OH); 3.93-4.04 m, 8 H (Et); 3.64-
3.79 m, 2 H (H-6’); 3.54-3.60 m, 2 H (H-4’); 3.35-3.47 m, 5 H (H-1’ and H-2’ and H-
1
3
3
’); 1.98-2.08 m, 4 H (H-5’ and H-7’); 1.22 t, 12 H, J(CH , CH ) = 7.0 (Et). C NMR
2 3
(DMSO-d
6
): 79.54 (C-2’); 70.34 (C-3’); 64.87 d, J(C,P) = 1.2 (C-4’); 63.62 d, J(C,P) =
1
.0 (C-6’); 61.17 and 61.14 and 61.12 d, J(C,P) = 6.2 (Et ); 60.88 (C-1’); 26.61 d,
3
J(C,P) = 136.6 (C-7’); 26.14 d, J(C,P) = 137.0 (C-5’); 16.43 d, J(C,P) = 5.8 (Et). MS
+
(ESI): m/z = 511.3 [M+H] .
Tetraisopropyl ([(2-(hydroxymethyl)propane-1,3-
diyl)bis(oxy)]bis(methylene))bis(phosphonate) (11)
10a
Hydroxyderivative 11 was prepared according the known procedure.
Synthesis of 6-chloropurine derivates 3 and 12 via Mitsunobu reaction - General
procedure
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 61
2
4
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To a solution of triphenylphosphine (5.62 g, 21.4 mmol) in dry THF (75 ml) cooled to
30 °C under argon atmosphere diisopropylazadicarboxylate (DIAD, 3.9 ml, 19.8
–
mmol) was added. The mixture was stirred for 30 min to preformed complex. The
solution of alcohol 2 or 11 (6.42 mmol) in THF (50 ml) and the 6-chloropurine (1.6 g,
10.4 mmol) was added to the preformed complex and reaction flask was filled with
argon again. The resulting mixture was slowly warmed to room temperature and stirred
twice overnight. Solvent was evaporated and the crude mixture was purified by flash-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chromatography (silicagel, eluent CHCl
as yellowish solid.
3
:MeOH, gradient). Pure product was obtained
Diethyl (2-(3-(6-chloro-9H-purin-9-yl)-2-(2-
(diethoxyphosphoryl)ethoxy)propoxy)ethyl)phosphonate (3): starting from
1
hydroxyderivative 2, yield 64%. H NMR (DMSO-d
6
): 8.79 s, 1 H (H-2); 8.68 s, 1 H
(H-8); 4.49 dd, 1 H, J(gem) = 14.4, J(1’b, 2’) = 4.1 (H-1’b); 4.36 dd, 1 H, J(gem) =
14.4, J(1’a, 2’) = 7.0 (H-1’b); 3.85-4.05 m, 9 H (Et and H-2’); 3.67 m, 1 H (H-6’b);
3
1
1
2
.53-3.61 m, 2 H (H-4’); 3.52 m, 1 H (H-6’a); 3.42-3.46 m (3’); 2.04 dt, 2 H, J(5’,P) =
8.3, J(5’,4’) = 7.3 (H-5’); 1.89 dt, 2 H, J(7’,P) = 18.3, J(7’,6’) = 7.4 (H-7’); 1.22 t and
13
.171 t and 1.170 t, 12 H, J = 7.1 (Et). C NMR (DMSO-d
6
): 152.45 (C-4); 151.67 (C-
); 149.10 (C-6); 148.39 (C-8); 130.79 (C-5); 75.92 (C-2’); 69.54 (C-3’); 65.03 d,
J(C,P) = 1.2 (C-4’); 63.90 d, J(C,P) = 0.7 (C-6’); 61.19 d and 61.17 d and 61.15 d,
J(C,P) = 6.2 (Et); 44.76 (C-1’); 26.33 d, J(C,P) = 136.8 (C-7’); 26.00 d, J(C,P) = 136.8
(C-5’); 16.45 d and 16.40 d and 16.39 d, J(C,P) = 5.8 (Et). MS (ESI): m/z = 557.1
+
[
M+H] .
Tetraisopropyl ([(2-((6-chloro-9H-purin-9-yl)methyl)propane-1,3-
diyl)bis(oxy)]bis(methylene))bis(phosphonate) (12): starting from hydroxyderivative
1
1
1, yield 60%. H NMR (DMSO-d
6
): 8.77 s, 1 H (H-2); 8.67 s, 1 H (H-8); 4.58 dsept,
ACS Paragon Plus Environment

Page 25 of 61
Journal of Medicinal Chemistry
2
5
1
2
3
4
5
6
7
8
9
4 H, J(CH,P) = 7.7, J = 6.2 (iPr); 4.34 d, 2 H, J(1’,2’) = 7.0 (H-1’); 3.71 dd, 2 H,
J(gem) = 14.0, J(4’b,P) = 8.0 (H-4’b); 3.68 dd, 2 H, J(gem) = 14.0, J(4’a,P) = 8.0 (H-
4’a); 3.53 dd, 2 H, J(gem) = 9.8, J(3’b,2’) = 5.6 (H-3’b); 3.50 dd, 2 H, J(gem) = 9.8,
J(3’a,2’) = 5.8 (H-3’a); 2.57 tpent, 1 H, J(2’,1’) = 7.0, J(2’,3’a) = J(2’,3’b) = 5.7 (H-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
3
2’); 1.23 d and 1.22 d and 1.21 d, 24 H, J = 6.2 (iPr). C NMR (DMSO-d ): 152.40
6
(
(
(
C-4); 151.66 (C-2); 149.14 (C-6); 148.09 (C-8); 131.01 (C-5); 71.09 d, J(C,P) = 11.0
C-3’); 70.36 d and 70.34 d, J(C,P) = 6.4 (iPr); 65.28 d, J(C,P) = 164.2 (C-4’); 43.11
C-1’); 39.24 (C-2’); 24.00 d and 23.91 d, J(C,P) = 3.8 and 4.5 (iPr). MS (ESI): m/z =
+
5
99.3 [M+H] .
Synthesis of 2-amino-6-chloropurine derivates 4 and 13 via Mitsunobu reaction -
General procedure
Starting from 2-amino-6-chloropurine and hydroxyderivatives 2 and 11 the procedure
was identical as described above for 6-chloropurine. Then after the stirring of reaction
mixture for twice overnight, water (100 ml) was added and the mixture was heated at
o
70 C for 4 days. Solvent was evaporated and the residue was codistilled with toluen
(
3
3x) and purified by flash-chromatography (silica gel, eluent CHCl : MeOH,
gradient). Pure product was obtained as yellowish solid.
Diethyl (2-(3-(2-amino-6-chloro-9H-purin-9-yl)-2-(2-
(diethoxyphosphoryl)ethoxy)propoxy) ethyl)phosphonate (4): starting from
1
hydroxyderivative 2, yield 76%. H NMR (DMSO-d
6
): 8.23 s, 1 H (H-8); 7.05 bs, 2 H
(NH
2
); 4.35 dd, 1 H, J(gem) = 14.4, J(1’b, 2’) = 4.1 (H-1’b); 4.22 dd, 1 H, J(gem) =
1
4.4, J(1’b, 2’) = 7.1 (H-1’b); 4.01-4.17 m, 8 H (Et); 3.98 m, 1 H (H-2’); 3.78 m, 1 H,
(
H-6’b); 3.69-3.75 m, 2 H (H-4’); 3.66 m, 1 H (H-6’a); 3.54 m, 2 H (H-3’); 2.19 dt, 2
H, J(5’,P) = 18.2, J(5’, 4’) = 7.3 (H-5’); 2.02 m, 2 H (H-7’); 1.35 t and 1.31 t, 12 H, J =
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 61
2
6
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
3
7.1(Et). C NMR (DMSO-d6): 159.97 (C-2); 154.55 (C-4); 149.44 (C-6); 144.09 (C-
8
6
1
); 123.24 (C-5); 75.95 (C-2’); 69.62 (C-3’); 65.01 (C-4’); 63.87 (C-6’); 61.21 d and
1.19 d and 61.17 d and 61.16 d, J(C,P) = 6.2 (Et); 44.02 (C-1’); 26.35 d, J(C,P) =
36.4 (C-7’); 25.97 d, J(C,P) = 137.0 (C-5’); 16.44 d and 16.38 d, J(C,P) = 5.8 (Et).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
MS (ESI): m/z = 572.2 [M+H] .
Tetraisopropyl
([(2-((2-amino-6-chloro-9H-purin-9-yl)methyl)propane-1,3-
diyl)bis(oxy)]bis(methylene))bis(phosphonate) (13): starting from hydroxyderivative
10a
1
1, yield 73%. The spectral data were in accord with the literature.
Synthesis of 6-oxopurine derivates 5-6 and 14-15 - General procedure
The 6-chloropurine or 2-amino-6-chloro purine derivative (3, 12 or 4, 13; 4.13 mmol)
was dissolved in aqueous trifluoroacetic acid (75%, 60 ml) and stirred overnight. The
solvent was evaporated and the residue codistilled with water (3x) and toluene (2x).
The crude product was purified by flash-chromatography.
Diethyl (2-(3-(hypoxanthine-9(6H)-yl)-2-(2-
(
diethoxyphosphoryl)ethoxy)propoxy)ethyl) phosphonate (5): starting from 6-
1
chloropurine 3, yield 76%. H NMR (DMSO-d
6
): 12.30 bs, 1 H (NH); 8.05 s, 1 H (H-
8
); 8.03 s, 1 H (H-2); 4.31 dd, 1 H, J(gem) = 14.3, J(1’b, 2’) = 4.3 (H-1’b); 4.18 dd, 1
H, J(gem) = 14.3, J(1’a, 2’) = 7.0 (H-1’b); 3.87-4.04 m, 8 H (Et); 3.84 dtd, 1 H,
J(2’,1’a) = 7.0, J(2’,3’a) = J(2’,3’b) = 5.0, J(2’,1’b) = 4.3 (H-2’); 3.64 m, 1 H (H-6’b);
3
.54-3.61 m, 2 H (H-4’); 3.51 m, 1 H (H-6’a); 3.42 dd, J(gem) = 10.5, J(3’b, 2’) = 5.0
H-3’b); 3.39 dd, J(gem) = 10.5, J(3’a, 2’) = 5.1 (H-3’a); 2.06 dt, 2 H, J(5’,P) = 18.3,
J(5’,4’) = 7.3 (H-5’); 1.89 dt, 2 H, J(7’,P) = 18.3, J(7’,6’) = 7.4 (H-7’); 1.22 t and
(
13
1.184 t and 1.182 t, 12 H, J = 7.1 (Et). C NMR (DMSO-d ): 156.94 (C-6); 148.80 (C-
6
4
); 145.70 (C-2); 141.16 (C-8); 123.87 (C-5); 76.38 (C-2’); 69.70 (C-3’); 65.05 (C-4’);
ACS Paragon Plus Environment

Page 27 of 61
Journal of Medicinal Chemistry
2
7
1
2
3
4
5
6
7
8
9
63.96 (C-6’); 61.18 d and 61.17 d and 61.15 d, J(C,P) = 6.2 (Et); 44.32 (C-1’); 26.35 d,
J(C,P) = 136.2 (C-7’); 26.02 d, J(C,P) = 137.3 (C-5’); 16.45 d and 16.40 d and 16.39 d,
+
J(C,P) = 5.9 (Et). MS (ESI): m/z = 539.3 [M+H] .
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Diethyl (2-(3-(guanine-9(6H)-yl)-2-(2-(diethoxyphosphoryl)ethoxy)propoxy)ethyl)
1
phosphonate (6): starting from 2-amino-6-chloropurine 4, yield 98%. H NMR
6 2
(DMSO-d ): 10.75 s, 1 H (NH); 7.92 s, 1 H (H-8); 6.58 bs, 2 H (NH ); 4.13 dd, 1 H,
J(gem) = 14.3, J(1’b,2’) = 4.2 (H-1’b); 3.88-4.04 m, 9 H (H-1’a and Et); 3.81 dtd, 1H,
J(2’,1’a) = 7.1, J(2’,3’a) = J(2’,3’b) = 5.0, J(2’,1’b) = 4.2 (H-2’); 3.65 m, 1 H (H-6’b);
3
.56-3.61 m, 2 H (H-4’); 3.53 m, 1 H (H-6’a); 3.42 dd, 1 H, J(gem) = 10.4, J(3’b, 2’) =
.0 (H-3’b); 3.38 dd, 1 H, J(gem) = 10.4, J(3’a, 2’) = 5.0 (H-3’a); 2.07 dt, 2 H, J(5’,P)
5
= 18.3, J(5’,4’) = 7.4 (H-5’); 1.91 dt, 2 H, J(7’,P) = 18.3, J(7’,6’) = 7.4 (H-7’); 1.22 t
13
and 1.19 t, 12 H, J = 7.1 (Et). C NMR (DMSO-d
6
): 156.39 (C-6); 154.04 (C-2);
51.32 (C-4); 138.30 (C-8); 114.82 (C-5); 76.11 (C-2’); 69.66 (C-3’); 65.00 (C-4’);
3.89 (C-6’); 61.22 d and 61.20 d, J(C,P) = 6.1 (Et); 44.07 (C-1’); 26.32 d, J(C,P) =
1
6
136.3 (C-7’); 26.00 d, J(C,P) = 136.9 (C-5’); 16.44 d and 16.39 d, J(C,P) = 5.8 (Et).
+
MS (ESI): m/z = 554.1 [M+H] .
Tetraisopropyl ([(2-((hypoxanthine-9H-yl)methyl)propane-1,3-
diyl)bis(oxy)]bis(methylene))bis(phosphonate) (14): starting from 6-chloropurine 12,
1
yield 98%. H NMR (DMSO-d ): 12.30 bd, 1 H, J(NH,2) = 3.9 (NH); 8.08 s, 1 H (H-
6
8); 8.03 d, 1 H, J(2,NH) = 3.9 (H-2); 4.59 dsept, 4 H, J(CH,P) = 7.7, J(CH,CH
(iPr); 4.17 d, 2 H, J(1’,2’) = 7.0 (H-1’); 3.72 dd, 2 H, J(gem) = 14.1, J(4’b,P) = 8.0 (H-
’b); 3.69 dd, 2 H, J(gem) = 14.1, J(4’b,P) = 8.0 (H-4’a); 3.49 dd, 2 H, J(gem) = 9.8,
J(3’b,2’) = 5.7 (H-3’b); 3.47 dd, 2 H, J(gem) = 9.8, J(3’a,2’) = 5.7 (H-3’a); 2.46 m, 1
3
) = 6.2
4
13
H (H-1’); 1.24 d and 1.23 d, 24 H, J(CH ,CH) = 6.2 (iPr). C NMR (DMSO-d ):
3
6
156.78 (C-6); 148.70 (C-4); 145.76 (C-2); 140.88 (C-8); 123.95 (C-5); 70.92 d, J(C,P)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 61
2
8
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
= 11.2 (C-3’); 70.37 d and 70.35 d, J(C,P) = 6.4 (iPr); 65.29 d, J(C,P) = 164.1 (C-4’);
42.41 (C-1’); 39.44 (C-2’); 24.02 d and 23.92 d, J(C,P) = 3.7 and 4.5 (iPr). MS (ESI):
+
m/z = 581.3 [M+H] .
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Tetraisopropyl ([(2-((guanine-9H-yl)methyl)propane-1,3-
diyl)bis(oxy)]bis(methylene))bis(phosphonate) (15): starting from 2-amino-6-
chloropurine 13, yield 84%. The spectral data were in accord with the literature.
10a
Synthesis of the free bisphosphonates 7-8 and 16-17 - General procedure
A mixture of tetraesters 5, 6, 14 or 15 (1 mmol), acetonitrile (20 ml) and BrSiMe (2
3
ml) was stirred overnight at room temperature. After evaporation and codistillation
with acetonitrile (3x), the residue was stirred with MeOH:water (1:1, 100 ml) for 2 h at
room temperature. The solvent was evaporated in vacuo and the residue was purified
by preparative HPLC (water–methanol) and the product was obtained after
codistillation with methanol as white solid.
(2-(3-(Hypoxanthine -9(6H)-yl)-2-(2-
(bishydroxyphosphoryl)ethoxy)propoxy)ethyl) phosphonic acid (7): starting from
1
tetraester 5, yield 55%. H NMR (DMSO-d
6
): 12.69 bs, 1 H (NH); 8.67 s, 1 H (H-8);
8.18 s, 1 H (H-2); 4.39 dd, 1 H, J(gem) = 14.3, J(1’a,2’) = 4.2 (H-1’a); 4.26 dd, 1 H,
J(gem) = 14.3, J(1’b,2’) = 7.0 (H-1’b); 3.84 m, 1 H (H-2’); 3.69 m, 1 H (H-6’a); 3.51-
3.61 m, 3 H (H-4’ and H-6’b); 3.40-3.47 m, 2 H (H-3’); 1.69-1.90 m, 4 H (H-5’ and H-
1
3
7
1
2
6
’). C NMR (DMSO-d ): 155.40 (C-6); 148.17 (C-4); 147.15 (C-2); 141.08 (C-8);
20.37 (C-5); 75.79 (C-2’); 69.48 (C-3’); 65.96 (C-4’); 64.71 (C-6’); 45.21 (C-1’);
9.04 d, J(C,P) = 133.8 and 28.83 d, J(C,P) = 134.4 (C-5’and C-7’). HR-MS (ESI-) for
C
12
H
19
O
9
N
4
P
2
calculated: 425.0633, found: 425.0628.
ACS Paragon Plus Environment

Page 29 of 61
Journal of Medicinal Chemistry
2
9
1
2
3
4
5
6
7
8
9
(2-(3-(Guanin-9(6H)-yl)-2-(2-
(bishydroxyphosphoryl)ethoxy)propoxy)ethyl)phosphonic acid (8): starting from
1
tetraester 6, yield 62%. H NMR (DMSO-d
6
): 10.54 s, 1 H (NH); 7.65 s, 1 H (H-8);
6
.49 s, 2 H (NH ); 4.09 dd, 1 H, J(gem) = 14.3, J(1’b,2’) = 4.3 (H-1’b); 3.95 dd, 8 H,
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
J(gem) = 14.3, J(1’a,2’) = 7.2 (H-1’a); 3.75 bdtd, 1H, J(2’,1’a) = 7.2, J(2’,3’a) =
J(2’,3’b) = 5.0, J(2’,1’b) = 4.5 (H-2’); 3.66 m, 1 H (H-6’b); 3.58-3.63 m, 2 H (H-4’);
3.54 m, 1 H (H-6’a); 3.40 dd, 1 H, J(gem) = 10.6, J(3’b, 2’) = 4.8 (H-3’b); 3.35 dd, 1
H, J(gem) = 10.6, J(3’a, 2’) = 5.3 (H-3’a); 1.88 ddd, 2 H, J(5’,P) = 18.4, J(5’,4’) = 8.0
13
(
H-5’); 1.66-1.75 m, 2 H, J(7’,P) = 18.4 (H-7’). C NMR (DMSO-d ): 156.91 (C-6);
6
153.72 (C-2); 151.48 (C-4); 138.38 (C-8); 116.30 (C-5); 76.21 (C-2’); 69.77 (C-3’);
5.92 (C-4’); 64.76 (C-6’); 43.86 (C-1’); 29.22 d, J(C,P) = 133.2 (C-7’); 28.94 d,
J(C,P) = 134.1 (C-5’). HR-MS (ESI-) for C12 calculated: 440.0742, found:
40.0737.
6
20 9 5 2
H O N P
4
([(2-((hypoxanthine-9H-yl)methyl)propane-1,3-
diyl)bis(oxy)]bis(methylene))diphosphonic acid (16): starting from tetraester 14,
1
yield 73%. H NMR (DMSO-d
6
): 12.31 bs, 1 H (NH); 8.09 s, 1 H (H-8); 8.04 s, 1 H
(H-2); 4.18 d, 2 H, J(1’, 2’) = 7.2 (H-1’); 3.54 m, 4 H, J(4’, P) = 7.8 (H-4’); 3.51 dd, 2
H, J(gem) = 9.4, J(3’b, 2’) = 5.1 (H-3’b); 3.49 dd, 2 H, J(gem) = 9.4, J(3’a, 2’) = 5.8
1
3
(
H-3’a); 2.40 tpent, 1 H, J(2’,1’) = 7.2, J(2’,3’a) = J(2’,3’b) = 5.5 (H-2'). C NMR
(DMSO-d ): 156.81 (C-6); 148.70 (C-4); 145.78 (C-2); 141.07 (C-8); 123.95 (C-5);
0.16 d, J(C,P) = 8.5 (C-3’); 66.46 d, J(C,P) = 160.6 (C-4’); 42.44 (C-1’); 39.79 (C-
’). HR-MS (ESI-) for C11 calculated: 411.0476, found: 411.0473.
6
7
2
17 9 4 2
H O N P
([(2-((guanine-9H-yl)methyl)propane-1,3-
diyl)bis(oxy)]bis(methylene))diphosphonic acid (PP-P352=SV278)-(17): starting
10a
from tetraester 15, yield 68%. The spectral data were in accord with the literature.
ACS Paragon Plus Environment