Acyclic Nucleoside Phosphonates Containing 9-Deazahypoxanthine and a Five-Membered Heterocycle as Selective Inhibitors of Plasmodial 6-Oxopurine Phosphoribosyltransferases

Article abstract of DOI:10.1002/cmdc.201700293

Acyclic nucleoside phosphonates (ANPs) are an important class of therapeutic drugs that act as antiviral agents by inhibiting viral DNA polymerases and reverse transcriptases. ANPs containing a 6-oxopurine unit instead of a 6-aminopurine or pyrimidine base are inhibitors of the purine salvage enzyme, hypoxanthine-guanine-[xanthine] phosphoribosyltransferase (HG[X]PRT). Such compounds, and their prodrugs, are able to arrest the growth of Plasmodium falciparum (Pf) in cell culture. A new series of ANPs were synthesized and tested as inhibitors of human HGPRT, PfHGXPRT, and Plasmodium vivax (Pv) HGPRT. The novelty of these compounds is that they contain a five-membered heterocycle (imidazoline, imidazole, or triazole) inserted between the acyclic linker(s) and the nucleobase, namely, 9-deazahypoxanthine. Five of the compounds were found to be micromolar inhibitors of PfHGXPRT and PvHGPRT, but no inhibition of human HGPRT was observed under the same assay conditions. This demonstrates selectivity of these types of compounds for the two parasitic enzymes compared to the human counterpart and confirms the importance of the chemical nature of the acyclic moiety in conferring affinity/selectivity for these three enzymes.

Full text of DOI:10.1002/cmdc.201700293

Accepted Article
Title: Acyclic nucleoside phosphonates containing 9-
deazahypoxanthine and a 5-membered heterocycle as selective
inhibitors of plasmodial 6-oxopurine phosphoribosyltransferases
Authors: Martin Maxmilian Kaiser, Ondřej Baszczyňski, Dana Hocková,
Lenka Poštová-Slavětínská, Martin Dračínský, Dianne T.
Keough, Luke W. Guddat, and Zlatko Janeba
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.201700293
Link to VoR: http://dx.doi.org/10.1002/cmdc.201700293
A Journal of
www.chemmedchem.org

10.1002/cmdc.201700293
ChemMedChem
Acyclic nucleoside phosphonates containing 9-deazahypoxanthine and a 5-membered heterocycle
as selective inhibitors of plasmodial 6-oxopurine phosphoribosyltransferases
Dr. Martin Maxmilian Kaiser,[a] Dr. Ondřej Baszczyňski,[a] Dr. Dana Hocková,[a] Dr. Lenka Poštová-
Slavětínská,[a] Dr. Martin Dračínský,[a] Dr. Dianne T. Keough,[b] Dr. Luke W. Guddat,*[b] and Dr. Zlatko
Janeba*[a]
aThe Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences, Flemingovo
nám. 2, CZ-16610 Prague 6, Czech Republic
bSchool of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, Queensland,
4068, Australia
Keywords: malaria; acyclic nucleoside phosphonates; 9-deazahypoxanthine; 6-oxopurine
phosphoribosyltransferase; purine salvage
Abstract
Acyclic nucleoside phosphonates (ANPs) are an important class of therapeutic drugs that act as antiviral
agents by inhibiting viral DNA polymerases and reverse transcriptases. ANPs containing a 6-oxopurine
instead of 6-aminopurine or pyrimidine bases are inhibitors of the purine salvage enzyme, hypoxanthine-
guanine-[xanthine]-phosphoribosyltransferase (HG[X]PRT). Such compounds, and their prodrugs, are
able to arrest the growth of Plasmodium falciparum (Pf) in cell culture. A new series of ANPs has been
synthesized and tested as inhibitors of human HGPRT, PfHGXPRT and Plasmodium vivax (Pv) HGPRT.
The novelty of these compounds is that they contain a 5-membered heterocycle (imidazoline, imidazole
or triazole) inserted between the acyclic linker(s) and the nucleobase, namely 9-deazahypoxanthine. Five
of the compounds were found to be micromolar inhibitors of PfHGXPRT and PvHGPRT but no
inhibition of human HGPRT was observed under the same assay conditions. This demonstrates
selectivity of this type of compounds for the two parasitic enzymes compared to the human counterpart
and confirms the importance of the chemical nature of the acyclic moiety in conferring affinity/selectivity
for these three enzymes.
1
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Introduction
Malaria remains one of the most important of all human protozoan borne infections. This disease is
caused by one of five plasmodial species of which Plasmodium falciparum (Pf) and Plasmodium vivax
(Pv) are the most deadly and widespread.1,2 Limitations of current vaccines3 and increasing development
of resistance to antimalarial drugs underline the need to develop new antimalarial therapeutics against
previously unused molecular targets.4,5
Parasites of the genus Plasmodium lack the enzymes for de novo synthesis of purine bases and rely on
the transport of these nucleobases from the host. Hypoxanthine-guanine-[xanthine]
phosphoribosyltransferase (HG[X]PRT) is the key enzyme for the production of the essential metabolites
(GMP, IMP and AMP) which, in malarial parasites, can only be produced from IMP.6 Thus, these
parasites depend on the activity of HG[X]PRT for their survival and reproduction.
Acyclic nucleoside phosphonates (ANPs)7 constitute a distinguished class of antivirals.8 A number of
structurally modified ANPs were also shown to be potent and/or selective inhibitors of PfHGXPRT and
PvHGPRT.9 These include acyclic nucleoside bisphosphonates (ANbPs)10 and aza-acyclic nucleoside
phosphonates (aza-ANPs).11 Crystal structures in complex with human HGPRT confirmed that
bisphosphonate analogues are able to fit into all three key binding pockets of the enzyme, namely the
binding site of the purine base, the 5’-phosphate group, and pyrophosphate.10,11b Furthermore, several
bisphosphonates also exhibited potent inhibitory properties against bacterial (Escherichia coli and
Mycobacterium tuberculosis) 6-oxopurine PRTs.12
Transition state analogues known as immucillins (e.g. immucillinHP, 1, Fig. 1)13 and chemically more
stable acyclic immucillin phosphonates (e.g. compounds 2 and 3, Fig. 1),14, are potent inhibitors of 6-
oxopurine PRTs. Compounds 2 and 3 have Ki values of 10.6 nM and 0.65 nM for PfHGXPRT and Ki
values for human HGPRT of 4.9 M and 385 nM, a decrease of 466- and 592-fold, respectively.14a
However, the structure of compound 2 in complex with PfHGXPRT does not provide an explanation for
the wide disparity in Ki values between the Pf and human enzymes. When the immucillin phosphates,
immucillin phosphonates or their acyclic analogues bind in the active site, the side chain of D148 (Pf) or
D137 (human) moves close to the purine ring and forms a hydrogen bond with N7. This aspartic acid
residue is located at the top of the 5-phosphate binding pocket and points away from the nucleobase
when GMP15 or ANPs with either guanine of hypoxanthine bind.9a
2
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Figure 1. Chemical structure of a) immucillinHP (ImmHP, 1),13 b) acyclic immucillin phosphonates 2
and 3,14 and c) pyrrolidine based nucleotide analogue 4.16
For the study reported herein, 9-deazahypoxanthine was chosen as the base since the acyclic immucilin
phosphonates that contain this nucleobase are selective and potent inhibitors of the malarial enzymes.14
It has been previously reported16 that ANPs with a locked conformation via a pyrrolidine ring (e.g.
compound 4, Fig. 1) are excellent inhibitors of the 6-oxopurine PRTs. We have speculated that the
increased flexibility of the acyclic moiety of ANPs allows them to bind optimally into the active sites of
these enzymes potentially leading to the decrease of selectivity. On the other hand, an introduction of
some rigidity to the acyclic chain of ANPs could help to fine-tune the selectivity among the parasitic and
human enzymes. Thus, novel series of ANPs (A) and ANbPs (B) were designed (Fig. 2), where different
5-membered heterocyclic moieties, either partially saturated (imidazoline) or aromatic (imidazole or
triazole), were inserted between 9-deazahypoxanthine and an aliphatic linker(s) bearing the phosphonate
moiety to examine the effect of such inclusion on selectivity. In total, 18 compounds were synthesized
and evaluated as potential inhibitors of Pf HGXPRT, PvHGPRT and human HGPRT.
Figure 2. General structures of target acyclic nucleoside phosphonates (ANPs, A) and acyclic nucleoside
bisphosphonates (ANbPs, B) with the 5-membered heterocycle (in red) attached to the C-9 position of
9-deazahypoxanthine.
3
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Results and discussion
Chemistry
Compounds 6 and 8 (Scheme 1) are the ideal synthetic precursors for the synthesis of N-
(phosphonoalkyl)imidazoline derivatives. Compound 6 was prepared using aza-Michael addition of
diethyl vinylphosphonate (DEVP, 5) to ethylendiamine (EDA) according to a recently published
methodology using only water without additional catalyst or organic solvent.17 Reaction of diisopropyl
(3-bromopropyl)phosphonate (7) with EDA, serving both as a reactant and solvent, was used for the
synthesis of compound 8.
Scheme 1. Synthesis of N-(phosphonoalkyl)ethylendiamines 6 and 8. Reagents and conditions: a)
EDA, H2O, RT, overnight, 95%; b) EDA, RT, 2 h, 96%.
Compounds 10 and 11 (Scheme 2) were prepared from 6-chloro-9-deazapurine 9 on a large scale
according to the previously reported procedure.18 Compound 11 was selectively protected at position N-
7 using SEMCl under basic conditions (NaH in THF) providing alcohol 12 which was subsequently
oxidized with Dess-Martin reagent in acetic acid to form aldehyde 13 in a high yield. Compound 13 can
be subsequently transformed to 2-imidazoline derivatives. A number of different synthetic procedures
for the formation of imidazoline ring from an aldehyde have been reported.19 However, in this case, these
approaches failed because of the low reactivity of aldehyde 13 and, apparently, harsher reaction
conditions were required. We decided to modify the previously described cyclisation,19b and MW-
assisted heating (90 or 120 °C) of 13 for 1 h with either EDA, compound 6, or compound 8 gave
imidazoline derivatives 14, 16, and 17 (Scheme 2), respectively, in good yields. Aromatization of
compounds 16 and 17 with KMnO4 in dioxane20 afforded corresponding imidazole analogues 22 and 23.
All protected 9-deazapurine derivatives 14, 16, 17, 22, and 23 were converted into 9-deazahypoxanthine
analogues 15, 18, 19, 24, and 25, respectively, using 75% aqueous trifluoroacetic acid. The unreacted N-
7-hydroxymethyl side-products can be alternatively eliminated by treatment of the reaction mixture with
4
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
7 M ammonia in methanol followed by evaporation to dryness. Removal of ester groups from
phosphonates 18, 19, 24, and 25 under standard conditions (Me3SiBr/MeCN),21 followed by hydrolysis
gave free phosphonic acids 20, 21, 26, and 27 (Scheme 2). 2,6-Lutidine was added to neutralize HBr
present in commercial Me3SiBr and/or formed during the reaction progress.
Scheme 2. Synthesis of the imidazoline and imidazole derivatives. Reagents and conditions: a) BnOH,
NaH, 90 °C, overnight, 69%; b) CH2=O, K2CO3, H2O/dioxane (1/1), 90 °C, 18 h, 85%; c) 1. BSA,
Me3SiBr, MeCN, RT, 5 min.; 2. NaH, SEMCl, THF, 0 °C – RT, 2 h, 81%; d) Dess-Martin reagent,
AcOH, RT, 1 h, 89%; e) EDA, I2, K2CO3, t-BuOH, MW, 120 °C, 1 h, 61%; f) 75% CF3COOH, RT,
overnight, 71–91%; g) 6, I2, K2CO3, t-BuOH, MW, 90 °C, 1 h, 76%; h) Me3SiBr, 2,6-lutidine, MeCN,
RT, overnight, 38–63%; i) 8, I2, K2CO3, t-BuOH, MW, 90 °C, 1 h, 72%; j) KMnO4, dioxane, 80 °C,
overnight 61–63%.
Next, phosphonoalkylazides 29 and 31 (Scheme 3), precursors for “click-chemistry”, were prepared by
treatment of corresponding bromoalkyphosphonates 28 and 30 with NaN3 in toluene at 90 °C and in the
presence of tetrabutylammonium bromide (Bu4NBr) as a phase-transfer catalyst.
5
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Scheme 3. Synthesis of azidoalkylphosphonates 29 and 31. Reagents and conditions: a) NaN3,
Bu4NBr, toluene, 90 °C, overnight, 78% (29) and 92% (31).
For the synthesis of 1,4- and 1,5-disubstituted 1,2,3-triazole analogues, compound 9 (Scheme 4) was first
converted into 9-iodo-9-deazapurine derivative 32 with NIS in THF and then into SEM-derivative 33
with SEMCl and NaH in DMF. Treatment of compound 33 with TMS-acetylene under standard
Sonogashira reaction conditions22 afforded TMS-derivative 34 in excellent yield. The SEM group of
compound 34 was removed with tetrabutylammonium fluoride (TBAF) in THF to give the unprotected
acetylene derivative 35.
Two methods were used for the formation of the 1,2,3-triazole ring: a) Cu(I)-catalyzed22 azide alkyne
Huisgen24 “click” cycloaddition (CuAAC) for the regioselective assembly of 1,4-disubstituted 1,2,3-
triazoles, and b) pentamethylcyclopentadienyl ruthenium chloride [Cp*RuCl] complexes25 catalyzed
azide alkyne cycloaddition (RuAAC) leading to the formation of 1,5-disubstituted 1,2,3-triazoles with
high regioselectivity.
Thus, acetylene derivative 35 (or TMS-acetylene derivative 34 after in-situ TMS removal) was reacted
with azides 29 or 31 in DMF under presence of CuI at room temperature to obtain 1,4-disubstituted
triazoles 36 (28%) and 42 (67% of a crude product), respectively (Scheme 4).
For the second method, a variety of [Cp*RuCl] complexes can be employed. In our case, Cp*Ru(COD)Cl
exhibiting excellent activities at room temperature,26 was chosen as a selective catalyst for preparation
of 1,5-disubstituted 1,2,3-triazoles. Reaction of intermediate 35 (or 34 after in-situ TMS removal) with
azides 29 and 31 was carried out in dry toluene, as a convenient aprotic solvent for [Cp*RuCl]-catalyzed
cycloadditions,26 and gave the desired products 37 (67%) and 43 (88%), respectively, in excellent yields.
Formed triazoles 36, 37, 42 and 43 were treated with 75% aqueous trifluoroacetic acid at 60 °C to form
compounds 38, 39 and 44, 45 (Scheme 4), respectively. Finally, the desired free phosphonic acids 40,
41, 46, and 47 were obtained in good yields after standard removal of the phosphonate ester moieties.
6
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Scheme 4. Synthesis of monosubstituted triazole derivatives. Reagents and conditions: a) NIS, THF,
r.t., 1 h, 84%; b) NaH, SEMCl, DMF, RT, overnight, 92%; c)TMS-acetylene, Pd(PPh3)2Cl2, CuI, Et3N,
THF, 0 °C – RT, 3 h, 94%; d) TBAF, THF, RT, 1 h, 59%; e) 1. TBAF, THF, RT, 30 min; 2. 29, CuI,
Et3N, DMF, RT, 4 h, 28%; f) 75% CF3COOH, 60 °C, 5-16 h, 64–93%; g) Me3SiBr, 2,6-lutidine,
MeCN, RT, overnight, 49–93%; h) 1. TBAF, THF, RT, 30 min; 2. 29, Cp*Ru(COD)Cl, toluene, RT,
16 h, 67%; i) 31, CuI, Et3N, DMF, RT, 16 h, 67%; j) 31, Cp*Ru(COD)Cl, toluene, RT, 16 h, 88%.
The CuAAC methodology was also used for the synthesis of carboxylic acids 49 and 51 (Scheme 5),
bearing a carboxylic acid moiety in the side chain. Cu(I)-catalyzed reaction of acetylene derivative 35
with
(S)-2-azido-4-(tert-butoxy)-4-oxobutanoic
acid
or
4-azido-2-((tert-
butoxycarbonyl)amino)butanoic acid in DMF gave in both cases complex reaction mixtures which, after
laborious isolation and purification, afforded intermediates 48 (13%) and 50 (12%). Their subsequent
treatment with aqueous 75% trifluoroacetic acid at 60 °C gave the desired products 49 and 51 in good
yields (Scheme 5).
7
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Scheme 5. Synthesis of monosubstituted triazole derivatives 49 and 51. Reagents and conditions: a)
(S)-2-azido-4-(tert-butoxy)-4-oxobutanoic acid, CuI, Et3N, DMF, RT, 16 h, 13%; b) 75% CF3COOH,
60 °C, 5 h, 61% (49) and 87% (51); c) 4-azido-2-((tert-butoxycarbonyl)amino)butanoic acid, CuI,
Et3N, DMF, RT, 16 h, 12%.
SEM-protected 6-(benzyloxy)-9-iodo-9-deazapurine 52 (Scheme 6) was selected as another suitable
starting material for cross-coupling reactions. Compound 52 was obtained by the treatment of chloro
derivative 33 with BnOH and NaH (Scheme 6). The benzyloxy group was chosen for its stability under
alkaline conditions during cross-coupling reactions. The reaction of compound 52 with 3-butyn-1-ol
under the standard Sonogashira conditions (PdCl2(PPh3)2, CuI, Et3N, THF, RT) proceeded only with
approximately 40% conversion (UPLC-MS) and 9-deiodo derivative was the major product in the
reaction mixture. Thus, we decided to modify the procedure using Pd(PPh3)4, CuI, and Cs2CO3 in DMF
and the desired alcohol 53 was obtained in a 91% yield.
Scheme 6. Synthesis of triazole derivatives bearing two aliphatic linkers. Reagents and conditions: a)
BnOH, NaH, 90 °C, overnight, 91%; b) 3-butyn-1-ol, Pd(PPh3)4, CuI, Cs2CO3, DMF, RT, 3 h, 91%; c)
8
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
31, Cp*Ru(PPh3)2Cl, toluene, 100 °C, 2 days, 22% (54) and 41% (55); d) 75% CF3COOH, RT,
overnight, 44–49%; e) Me3SiBr, 2,6-lutidine, MeCN, RT, overnight, 75–77%.
As already mentioned above, cycloaddition of alkyl azides with terminal alkynes catalyzed by copper(I)
ion leads exclusively to 4-substituted 1,2,3-triazols in high yields. However, this type of reaction cannot
be applied for the cycloadditions with internal alkynes. This fact was confirmed by mechanistic studies
showing that such reactions involve terminal copper acetylide and proceed through a stepwise non-
concerted process.27 In contrast to CuAAC, the Ru-complexes were reported to catalyze the cycloaddition
reactions of internal alkynes. We decided to utilize the [Cp*RuCl] catalytic system for the cycloaddition
of internal alkynes with organic azides to form 1,4,5-trisubstituted 1,2,3-triazoles.26 Initial experiments
were conducted with alkyne derivative 53, azide 31, and Cp*Ru(COD)Cl as a catalyst in toluene at room
temperature. But the reaction afforded only traces of desired product 55, even when the reaction time
was prolonged to 3 days or when the reaction mixture was heated to 100 °C. The problem with the
elevated temperature may lie in the fact that the cyclooctadiene (COD) ligand in Cp*Ru(COD)Cl is labile
at higher temperatures and heating of the reaction results in rapid deactivation of the catalyst.26 Compared
to Ru-catalyst with COD ligand, Cp*Ru(PPh3)2Cl with more stable phosphine ligands in the
bis(triphenylphosphine) complex enables to perform cycloadditions at temperatures up to 110 °C.26 But
the cycloaddition of internal alkyne 53 with azide 31 and Cp*Ru(PPh3)2Cl as catalyst in toluene at 100
°C for 2 days afforded only a mixture of products 54 and 55 (Scheme 6), however, with higher amounts
of 1,4,5-trisubstituted 1,2,3-triazole 55. Compounds 54 and 55 were successfully separated using silica
gel column chromatography and subsequently treated with aqueous 75% trifluoroacetic acid to afford 6-
oxo-9-deazapurine products 56 and 57. To prepare the final free phosphonic acids 58 and 59 (Scheme
6), the phosphonate ester groups of phosphonates 56 and 57 were cleaved under the standard conditions
using Me3SiBr/MeCN followed by hydrolysis.
For the synthesis of other triazole bisphosphonate derivatives, key internal alkyne derivative 62 (Scheme
7) was prepared. Alkylation of 3-butyn-1-ol with triflate 60in THF using BuLi as a base afforded terminal
alkyne 61 in high yield. The treatment of compound 61 with intermediate 52 under the above described
Sonogashira conditions gave the protected phosphonate 62 in a high yield (Scheme 7). Compound 62
was also prepared by the treatment of alcohol 53 with triflate 60 and NaH in THF at room temperature,
but in a lower yield (52%) and this approach was less suitable for the scale-up. The cycloaddition of
compound 62 with azide 31 using catalyst Cp*Ru(PPh3)2Cl was not selective again and only an
9
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
inseparable mixture of 3,4,5-trisubtituted triazol 63 and expected 1,4,5-trisubstituted triazol 64. After
treatment of the mixture of 63 and 64 with aqueous 75% trifluoroacetic acid, bisphosphonates 65 and 66
(Scheme 7) were successfully separated from each other using C18 silica gel flash chromatography. The
standard use Me3SiBr/MeCN for removal of phosphonate ester moieties followed by hydrolysis led to
the formation of bisphosphonic acids 67 and 68 (Scheme 7).
Scheme 7. Synthesis of bisphosphonates 67 and 68. Reagents and conditions: a) 3-butyn-1-ol, BuLi,
THF, -78 °C – RT, overnight, 87%; b) 52, Pd(PPh3)4, CuI, Cs2CO3, DMF, RT, 3 h, 79%; c) 31,
Cp*Ru(PPh3)2Cl, toluene, 100 °C, 2 days, 45% (63 + 64); d) 75% CF3COOH, RT, overnight, 36% (65)
and 29% (66); e) Me3SiBr, 2,6-lutidine, MeCN, RT, overnight, 80–81%.
A different approach was employed for the preparation of prolonged bisphosphonates, where the PME
linker is replaced with PEE linker. The first reaction step involved modified oxa-Michael addition on
diethyl vinylphosphonate (DEVP, 5) to introduce the ethyl phosphonate moiety.28 Treatment of above
prepared alcohol 53 with 5 and a catalytic amount of KOH in dioxane afforded desired phosphonate 69
together with vinylacetylen derivative 70 as a by-product (Scheme 8). Compound 69 was further treated
with azide 31 and Cp*Ru(PPh3)2Cl in toluene at 100 °C to give an inseparable mixture of protected
phosphonates 71 and 72. Treatment of the mixture with aqueous 75% trifluoroacetic acid led to the
formation of phosphonates 73 and 74 which were subsequently separated on the C18 silica gel column.
The phosphonate ester groups of 73 and 74 were cleaved under the standard conditions using
Me3SiBr/MeCN followed by hydrolysis to form final bisphosphonic acids 75 and 76 (Scheme 8).
10
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Scheme 8. Synthesis of bisphosphonates 75 and 76. Reagents and conditions: a) DEVP (5), KOH,
dioxane, RT, overnight, 59% (69) and 12% (70); b) 31, Cp*Ru(PPh3)2Cl, toluene, 100 °C, 2 days, 56%
(71 + 72); c) 75% CF3COOH, RT, overnight, 35% (73) and 25% (74); d) Me3SiBr, 2,6-lutidine, MeCN,
RT, overnight, 82–85%.
Finally, we intended to compare the above rigid analogues, containing triazole ring attached directly to
the C-9 position of 9-deazahypoxanthine, with more flexible compounds where the triazole moiety is
attached to 9-deazahypoxanthine via a two-carbon linker. For that reason, model phosphonic acid 82
(Scheme 9) has been designed.
11
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Scheme 9. Synthesis of compound 82. Reagents and conditions: a) PdCl2(dppf), CuI, DMA, 80 °C,
overnight, 41%; b) TBAF, THF, RT, 3 h, 93%; c) N3CH2P(O)(OEt)2, CuI, Et3N, DMF, RT, overnight,
75%; d) 75% CF3COOH, RT, overnight, 47%; e) Me3SiBr, 2,6-lutidine, MeCN, RT, overnight, 78%.
Preliminary attempts to introduce (trimethylsilyl)butynyl moiety to compound 52 (Scheme 9) by its
treatment with (4-(trimethylsilyl)but-3-yn-1-yl)lithium or (4-(trimethylsilyl)but-3-yn-1-yl)copper(I)
failed. Subsequently, Negishi coupling under the reaction conditions reported by Storch et al.29 was
successfully applied. Zinc reagent 77,29 prepared using highly reactive Rieke zinc,30 was heated with
compound 52 in the presence of PdCl2(dppf) and CuI in dry DMA to give desired compound 78 in a 41%
yield. The benzyl group was cleaved during the reaction but this partial deprotection enabled us to
separate product 78 from the deiodo by-product, 6-O-benzyl-7-((2-trimethylsilyl)ethoxy)methyl-9-
dezapurine, which were formed in about 1:1 ratio. Removal of the trimethylsilyl group with TBAF gave
terminal alkyne 79 in high yield. Subsequent cycloaddition of compound 79 with diethyl
(azidomethyl)phosphonate31 was performed using CuI in DMF to afford 1,4-disubstituted triazole
derivative 80 in a 75% yield. An attempt to prepare the corresponding 1,5-disubstituted triazole,
regioisomer of compound 80, failed as the treatment of alkyne 79 with Cp*Ru(PPh3)2Cl in toluene
provided only a complex reaction mixture with very low conversion (~5% of the desired product
according to UPLC-MS analysis). Treatment of triazole 80 with aqueous 75% trifluoroacetic acid led to
the formation of phosphonate 81 (47%) which was subsequently converted to final free phosphonic acid
82 (78%) with Me3SiBr/MeCN followed by hydrolysis (Scheme 9).
Inhibition studies of novel 9-deazahypoxanthine acyclic nucleoside phosphonates
Five of the six triazole analogues bearing a single acyclic linker, namely compounds 40, 41 and 46
(Scheme 4), 49 and 51 (Scheme 5), do not inhibit any of the three HG[X]PRTs at concentrations between
60 - 100 M. Similarly, triazole bisphosphates 67 and 68 (Scheme 7) and 75 and 76 (Scheme 8), as well
as analogues where the 5-membered ring in the linker is imidazole, as in compounds 26 and 27 (Scheme
2), do not inhibit any of the three enzymes. In these cases, the rigid heterocycles do not appear to allow
either the purine base or the acyclic linker with polar groups to position themselves properly in the active
sites.
Table 1 shows the five new derivatives with build in heterocycle that proved to be selective micromolar
inhibitors of the Pf and/or Pv enzymes. None of these ANPs inhibit the human enzyme. The compounds
12
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
containing 2-imidazoline, 20 and 21 (Scheme 2), or 1,2,3-triazole, 47 (Scheme 4) and 59 (Scheme 6),
attached directly to the C-9 position of the 9-deazahypoxanthine base and with the phosphonopropyl
moiety on the nitrogen next to the C–C bond between the nucleobase and the 5-membered hetrocycle are
low micromolar inhibitors of PfHGXPRT and PvHGPRT. Phosphonate 59 which contains additional
hydroxyethyl moiety has a Ki value of 2-3 M for the two malarial enzymes. Interestingly, compound 58
(Scheme 6), a regioisomer of inhibitor 59, differing only in the position of an attachment of the
phosphonopropyl linker to the triazole ring, does not inhibit any of the HG[X]PRTs tested. Again, this
can be attributed to the inability of derivative 58 to accommodate itself properly in the critical binding
areas in the active site of either enzyme.
Table 1. Ki values of 9-deazahypoxanthine ANPs for human HGPRT, PfHGXPRT, and PvHGPRT.
SI
SI
Ki (M)
5-Membered
heterocycle
Compound
(Pf)* (Pv)*
human
NIa
Pf
6.6
4.7
2
3
NI
Pv
29
6
1
2
2-imidazoline
2-imidazoline
1,2,3-triazole
1,2,3-triazole
1,2,3-triazole
>14
>20
>50
>17
NI
>3
>17
>100
>25
>50
20
21
47
59
82
NIb
NIc
NId
NIe
2
a[inhibitor] = 73 M; b[inhibitor] = 74 M; c[inhibitor] = 71 M; d[inhibitor] = 11.7 M; e[inhibitor] = 61 M.
*Selectivity index. Lowest value likely, given none of the compounds inhibited human HGPRT at the concentrations listed
above, which represents their maximum solubility for 20, 21, 47 and 82. The absorbance of 59 absorbed very strongly at the
wavelength of the experiment. Thus, this was its maximum value that could be measured.
Compound 15 (Scheme 2) did not inhibit either the human or plasmodial enzymes at concentrations of
108 M in contrast to 20 and 21 (Scheme 2) which contain a linker with the phosphonate group. This is
as expected for an ordered sequential reaction mechanism as the nucleobase cannot enter the active site
prior to PRib-PP. The attachment of a phosphonate group to the five-membered ring in 20 and 21,
however, led to the inhibition of the two parasitic enzymes only. Therefore, it can be concluded that this
can simply occur when the 5-phosphate binding site is occupied. Though compounds 20 and 21
containing a non-aromatic imidazoline ring are low M inhibitors of PfHGXPRT and PvHGPRT, their
“oxidized” aromatic forms, compounds 26 and 27 (Scheme 2), respectively, did not inhibit the enzymes
at similar concentrations. This indicates that some, although limited, flexibility of the acyclic linker of
these novel ANPs is beneficial to exert biological activity.
13
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
Compound 82 with a two-carbon linker between the aromatic five-membered ring (triazole) and 9-
deazahypoxanthine did not inhibit either the human or Pf enzymes but does inhibit PvHGPRT which
shows that important differences exist between the active sites of the two plasmodial enzymes.
Conclusions
This study nicely demonstrates how the introduction of a rigid element into the acyclic linker of ANPs
can affect the potency and selectivity of the HG[X]PRT inhibitors. In general, such insertion seems to
lead to the decreased potency of these inhibitors compared to parent ANPs with the fully aliphatic linkers,
but at the same time, to more favorable selectivity and, thus, possibly to less off-target effects of potential
antimalarial treatments. The insertion of different 5-membered heterocyclic rings between the
phosphonoalkyl linker(s) and a purine base influences the Ki values of these compounds for all three
enzymes. Inserting imidazole moiety completely abolished binding whereas the insertion of imidazoline
and suitably substituted triazole (preferably phosphonopropyl attached to the nitrogen next to the C–C
bond between the nucleobase and heterocycle) did result in inhibition of the two plasmodial enzymes
with Ki values between 1-29 while no inhibition of human HGPRT, even at concentrations up to
100 was observed. The reason for this could be due to the flexibility of the mobile loops that
surround the active site of 6-oxopurine PRTs. However, it is clear that the nature of the linker connecting
different groups of the inhibitor can be a crucial factor for conferring both affinity and selectivity between
the human and plasmodial enzymes. The important SAR observations reported herein may be exploited
in future design and development of potent and completely selective inhibitors of plasmodial 6-oxopurine
PRTs.
Experimental section
General
Unless otherwise stated, solvents were evaporated at 40 °C/2 kPa, and compounds were dried at 2 kPa
over P2O5. TLC was performed on TLC aluminium sheets – silica gel 60 F254 (Merck), chromatographic
systems are described in the text. Column chromatography was performed on silica gel 230-400 mesh,
60 Å (Merck). Reverse phase HPLC separation were performed on a Waters Delta 600 instrument with
a Waters 486 Tunable Absorbance Detector using column Phenomenex Gemini C-18 (10 μm, 250 x 21.2
mm, flow 10 ml/min preparative column). 1H and 13C NMR spectra were measured on a Bruker Avance
13
13
II 600 (1H at 600 MHz and C at 151 MHz) and/or Bruker Avance II 500 (1H at 500 MHz and C at
14
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
126 MHz) spectrometers in CDCl3, DMSO or D2O (NaOD additive) and referenced to TMS, 13C
chloroform signal (δ 77.0) or dioxane used as internal standard (δ 3.75 and δ 67.19). Mass spectra were
measured on UPLC–MS (Waters SQD-2). HRMS data were taken on a LTQ Orbitrap XL spectrometer.
Microwave experiments were carried out in 10 mL vial in CEM Discover (Explorer) microwave
apparatus operated at a frequency of 2.45 GHz with continuous irradiation power from 0 to 300W.
Chemicals were purchased from Sigma-Aldrich (Prague, Czech Republic). 6-Chloro-9-deazapurine was
purchased from Santiago (Czech Republic), DEVP from Epsilon Chemie (France). Tetrahydrofuran
(THF) was distilled from lithium aluminium hydride under argon. Other anhydrous solvents are
commercially available. Solvents were dried by standard procedures and stored over molecular sieves (4
Å). Chromatographic system Eluent 1: cyclohexane : choroform : methanol = 6 : 4 : 1.
Detailed synthesis and full compound characterization can be found in the Electronic Supporting
Information.
Determination of Ki values. Recombinant human HGPRT, PfHGXPRT and PvHGPRT were purified
as described previously.10 Human HGPRT and PvHGPRT were stored in 0.1 M Tris-HCl, 0.01 M MgCl2,
pH 7.4, 200 μM PRib-PP, 1 mM dithiothreitol (DTT), −80°C. PfHGXPRT was stored in 0.01 M
phosphate, pH 7.4, 200 μM PRib-PP, 60 μM hypoxanthine, 1 mM dithiothreitol (DTT), −80°C.10
Enzyme assays were conducted in 0.1 M Tris-HCl, 0.01 M MgCl2, pH 7.4, 25oC. The Ki values were
calculated at a fixed concentration of guanine (60 μM) and at variable concentrations of PRib-PP (100
−1000 μM), depending on the Km(app), in the presence of the inhibitor and using the equations, vo =
Vmax.[S]o/Km(app) + [S]o and Km(app) = Km[1+ [I]/Ki].
Acknowledgements
This study is a part of research project RVO61388963 of the Institute of Organic Chemistry and
Biochemistry of the Czech Academy of Sciences and was supported by Czech Science Foundation (Grant
No. 16-06049S), by funds from the National Health and Medical Research Council, Australia (Grant No.
1030353) and by Gilead Sciences (Foster City, CA, USA).
References
[1] J. K. Baird, Clin. Microbiol. Rev. 2009, 22, 508.
15
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
[2] J. de Jersey, A. Holý, D. Hocková, L. Naesens, D. T. Keough, L. W. Guddat, Curr. Top. Med. Chem.
2011, 11, 2085.
[3] P. Bejon, J. Lusingu, A. Olotu, A. Leach, M. Lievens, J. Vekemans, S. Mshamu, T. Lang, J. Gould,
M.-C. Dubois, M.-A. Demoitié, J.-F. Strallaert, P. Vansadia, T. Carter, P. Njuguna, K. O. Awuondo, A.
Malabeja, O. Abdul, S. Gesase, N. Mturi, C. J. Drakeley, B. Savarese, T. Villafana, W. R. Ballou, J.
Cohen, E. M. Riley, M. M. Lemnge, K. Marsh, L. von Seidlein, New Engl. J. Med. 2008, 359, 2521.
[4] D. Saunders, P. Vanachayangkul, C. Lon, N. Engl. J. Med. 2014, 371, 484.
[5] I. B. Muller, J. E. Hyde, Future Microbiol 2010, 5, 1857.
[6] M. J. Downie, K. Kirk, B. C. Mamoun, Eukaryotic Cell 2008, 7, 1231.
[7] E. De Clercq, A. Holý, I. Rosenberg, T. Sakuma, J. Balzarini, P. C. Maudgal, Nature 1986, 323,
464.
[8] a) E. De Clercq, Expert Rev. Anti-infect. Ther. 2003, 1, 21; b) E. De Clercq, A. Holý, Nat. Rev.
Drug Discov. 2005, 4, 928; c) E. De Clercq, Med. Res. Rev. 2013, 33, 1278.
[9] a) D. T. Keough, D. Hocková, A. Holý, L. Naesens, T. Skinner-Adams, J. de Jersey, L. W. Guddat,
J. Med. Chem. 2009, 52, 4391; b) D. Hocková, A. Holý, M. Masojídková, D. T. Keough, J. de Jersey,
L. W. Guddat, Bioorg. Med. Chem. 2009, 17, 6218; c) D. T. Keough, D. Hocková, M. Krečmerová, M.
Česnek, A. Holý, L. Naesens, I. M. Bereton, D. J. Winzor, J. de Jersey, L. W. Guddat, Mol. Biochem.
Parasitol. 2010, 173, 165; d) M. Česnek, D. Hocková, A. Holý, M. Dračínský, O. Baszczyňski, J. De
Jersey, D. T. Keough, and L. W. Guddat, Bioorg. Med. Chem. 2012, 20, 1076–1089. e) M.
Krečmerová, M. Dračínský, D. Hocková, A. Holý, D. T. Keough, L. W. Guddat, Bioorg. Med. Chem.
2012, 20, 1222; f) O. Baszczyňski, D. Hocková, Z. Janeba, A. Holý, P. Jansa, M. Dračínský, D. T.
Keough, L. W. Guddat, Eur. J. Med. Chem. 2013, 67, 81; g) M. Kaiser, D. Hocková, T.-H. Wang, M.
Dračínský, L. Poštová-Slavětínská, E. Procházková, M. D. Edstein, M. Chavchich, D. T. Keough, L.
W. Guddat, Z. Janeba, ChemMedChem 2015, 10, 1707; h) M. Lukáč, D. Hocková, D. T. Keough, L. W.
Guddat, Z. Janeba, Tetrahedron 2017, 73, 692.
[10] D. T. Keough, P. Špaček, D. Hocková, T. Tichý, S. Vrbková, L. Slavětínská, Z. Janeba, L.
Naesens, M. D. Edstein, M. Chavchich, T.-H. Wang, J. de Jersey, L. W. Guddat, J. Med. Chem. 2013,
56, 2513.
[11] a) D. Hocková, D. T. Keough, Z. Janeba, T.-H. Wang, J. de Jersey, L. W. Guddat, J. Med. Chem.
2012, 55, 6209; b) D. T. Keough, D. Hocková, Z. Janeba, T.-H. Wang, L. Naesens, M. D. Edstein, M.
Chavchich, L. W. Guddat, J. Med. Chem. 2015, 58, 827; c) D. Hocková, Z. Janeba, L. Naesens, M. D.
Edstein, M. Chavchich, D. T. Keough, L. W. Guddat, L. W. Bioorg. Med. Chem. 2015, 23, 5502.
16
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
[12] a) D. T. Keough, D. Hocková, D. Rejman, P. Špaček, S. Vrbková, M. Krečmerová, W. Soon Eng,
H. Jans, N. P. West, L. M. J. Naesens, J. de Jersey, L. W. Guddat, L. W., J. Med. Chem. 2013, 56,
6967; b) W. S. Eng, D. Hocková, P. Špaček, Z. Janeba, N. P. West, K. Woods, L. M. J. Naesens, D. T.
Keough, L. W. Guddat, J. Med. Chem. 2015, 58, 4822; c) W. S. Eng, D. Hocková, P. Špaček, O.
Baszczyňski, Z. Janeba, L. Naesens, D. T. Keough, L. W. Guddat, ChemistrySelect 2016, 1, 6267.
[13] a) W. Shi, C. M. Li, P. C. Tyler, R. H. Furneaux, S. M. Cahill, M. E. Girvin, C. Grubmeyer, V. L.
Schramm, S. C. Almo, Biochemistry 1999, 38, 9872; b) C. M. Li, P. C. Tyler, R. H. Furneaux, G.
Kicska, Y. Xu, C. Grubmeyer, M. E. Girvin, V. L. Schramm, Nat. Struct. Mol. Biol. 1999, 6, 582; c)
W. Shi, C. M. Li, P. C. Tyler, R. H. Furneaux, C. Grubmeyer, V. L. Schramm, S. C. Almo, Nat. Struct.
Mol. Biol. 1999, 6, 588.
[14] a) K. Z. Hazleton, M. C. Ho, M. B. Cassera, K. Clinch, D. R. Crump, I. Rosario Jr., E. F. Merino,
S. C. Almo, P. C. Tyler, V. L. Schramm, Chem. Biol. 2012, 19, 721; b) K. Clinch, D. R. Crump, G. B.
Evans, K. Z. Hazleton, J. M. Mason, V. L. Schramm, P. C. Tyler, Bioorg. Med. Chem. 2013, 21, 5629.
[15] J. C. Eads, G. Scapin, Y. Xu, C. Grubmeyer, J. C. Sacchettini, Cell 1994, 78, 325.
[16] R. Pohl, L. Poštová Slavětínská, W. Soon Eng, D. T. Keough, L. W. Guddat, D. Rejman, Org.
Biomol. Chem. 2015, 13, 4693.
[17] E. V. Matveeva, P. V. Petrovskii, I. L. Odinets, Tetrahedron Lett. 2008, 49, 6129.
[18] Patent: WO2009/82247 A1 2009, 32.
[19] a) H. Fujioka, K. Murai, O. Kubo, Y. Ohba, Y. Kita, Tetrahedron 2007, 63, 638; b) M. Ishihara, H.
Togo, Tetrahedron 2007, 63, 1474; c) R. D. Crouch, Tetrahedron 2009, 65, 2387.
[20] J. X. Qiao, S. R. King, K. He, P. C. Wong, A. R. Rendina, J. M. Luettgen, B. Xin, R. M. Knabb, R.
R. Wexler, P. Y. S. Lam, Bioorg. Med. Chem. Lett. 2009, 19, 462.
[21] C. E. McKenna, J. Schmidhuser, J. Chem. Soc., Chem. Commun. 1979, 739.
[22] a) K. Sonogashira, Y. Tohda, N. A. Hagihara, Tetrahedron Lett. 1975, 50, 4467; b) R. Chinchilla,
C. Nájera, Chem. Rev. 2007, 107, 874.
[23] a) J. E. Hein, V. V. Fokin, Chem. Soc. Rev. 2010, 39, 1302; b) M. Meldal, C. W. Tornoe, Chem.
Rev. 2008, 108, 2952; c) H. C. Kolb, K. B. Sharpless, Drug Discovery Today 2003, 8, 1128; d) C. W.
Tornoe, C. Christensen, M. Meldal, J. Org. Chem. 2002, 67, 3057; e) V. V. Rostovtsev, L. G. Green, V.
V. Fokin, K. B. Sharpless, Angew. Chem., Int. Ed. 2002, 41, 2596.
[24] a) R. Huisgen, Angew. Chem., Int. Ed. Eng. 1963, 75, 604; b) R. Huisgen, Angew. Chem., Int. Ed.
Engl. 1963, 75, 742.
17
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201700293
ChemMedChem
[25] a) L. K. Rasmussen, B. C. Boren, V. V. Fokin, Org. Lett. 2007, 9, 5337; b) L. Zhang, X. Chen, P.
Xue, H. H. Y. Sun, I. D. Williams, K. B. Sharpless, V. V. Fokin, G. Jia, J. Am. Chem. Soc. 2005, 127,
15998.
[26] B. C. Boren, S. Narayan, L. K. Rasmussen, L. Zhang, H. Zhao, Z. Lin, G. Jia, V. V. Fokin, J. Am.
Chem. Soc. 2008, 130, 8923.
[27] F. Himo, T. Lovell, R. Hilgraf, V. V. Rostovtsev, L. Noodleman, K. B. Sharpless, V. V. Fokin, J.
Am. Chem. Soc. 2005, 127, 210.
[28] a) O. Baszczyňski, P. Jansa, M. Dračínský, M. M. Kaiser, P. Špaček, Z. Janeba, ACS Adv. 2012, 2,
1282; b) H. J. Cristau, D. Virieux, Tetrahedron Lett. 1999, 40, 703.
[29] J. Storch, M. Bernard, J. Sýkora, J. Karban, J. Čermák, Eur. J. Org. Chem. 2013, 260.
[30] S. Kudret, J. D’Haen, L. Lutsen, D. Vanderzande, W. Maes, Adv. Synth. Catal. 2013, 355, 569.
[31] R. Blaszczyk, T. Gajda, Synth. Commun. 2008, 38, 1110.
18
This article is protected by copyright. All rights reserved.