Acyclic phosphonate nucleotides and human adenylate kinases: Impact of a borano group on α-P position

Article abstract of DOI:10.1080/15257770801941952

Adenylate kinases are involved in the activation of antiviral drugs such as the acyclic phosphonates analogs PMEA and (R)PMPA. We examine the in vitro phosphorylation of PMEA and PMPA bearing a borano- or a H- group on the phosphorus atom. The α-borano or α-H on PMEA and PMPA were detrimental to the activity of recombinant human AMP kinases 1 and 2. Docking PMEA to the active site of AMP kinase 1 indicated that the borano group may prevent two conserved critical Arg interactions with the α-phosphate, resulting in substrate bad positioning. Copyright Taylor & Francis Group, LLC.

Full text of DOI:10.1080/15257770801941952

This article was downloaded by: [Boston College]
On: 27 September 2013, At: 07:14
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Nucleosides, Nucleotides and Nucleic
Acids
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/lncn20
Acyclic Phosphonate Nucleotides and
Human Adenylate Kinases: Impact of a
Borano Group on α-P Position
D. Topalis a , K. Alvarez b , K. Barral b , H. Munier-Lehmann c , B.
Schneider d , M. Véron d , C. Guerreiro c , L. Mulard c , C. El-Amri e ,
B. Canard b & D. Deville-Bonne a
a Laboratoire d′Enzymologie, Université Paris, 6, Paris, France
b Architecture et Fonction des Macromolecules Biologiques, UMR
CNRS 6098, Marseille, France
c Unité de Chimie Organique, URA CNRS 2128, Institut Pasteur, Paris,
France
d Unité de Régulation Enzymatique des Activités Cellulaires, CNRS
URA 2185, Institut Pasteur, Paris, France
e Peptidome de la Peau des Amphibiens, Université Paris, 6, FRE
CNRS 2852, Paris, France
Published online: 10 Apr 2008.
To cite this article: D. Topalis , K. Alvarez , K. Barral , H. Munier-Lehmann , B. Schneider , M. Véron ,
C. Guerreiro , L. Mulard , C. El-Amri , B. Canard & D. Deville-Bonne (2008) Acyclic Phosphonate
Nucleotides and Human Adenylate Kinases: Impact of a Borano Group on α-P Position, Nucleosides,
Nucleotides and Nucleic Acids, 27:4, 319-331, DOI: 10.1080/15257770801941952
To link to this article: http://dx.doi.org/10.1080/15257770801941952
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or

howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions

Nucleosides, Nucleotides and Nucleic Acids, 27:319–331, 2008
Copyright ꢀC Taylor & Francis Group, LLC
ISSN: 1525-7770 print / 1532-2335 online
DOI: 10.1080/15257770801941952
ACYCLIC PHOSPHONATE NUCLEOTIDES AND HUMAN ADENYLATE
KINASES: IMPACT OF A BORANO GROUP ON α-P POSITION
D. Topalis,1 K. Alvarez,2 K. Barral,2 H. Munier-Lehmann,3 B. Schneider,4
M. Ve´ron,4 C. Guerreiro,3 L. Mulard,3 C. El-Amri,5 B. Canard,2 and
D. Deville-Bonne1
1Laboratoire dꢁEnzymologie, Universite´ Paris 6, Paris, France
2Architecture et Fonction des Macromolecules Biologiques UMR CNRS 6098, Marseille,
France
3Unite´ de Chimie Organique, URA CNRS 2128, Institut Pasteur, Paris, France
4Unite´ de Re´gulation Enzymatique des Activite´s Cellulaires, CNRS URA 2185, Institut
Pasteur, Paris, France
5Peptidome de la Peau des Amphibiens, Universite´ Paris 6, FRE CNRS 2852, Paris, France
Adenylate kinases are involved in the activation of antiviral drugs such as the acyclic
2
phosphonates analogs PMEA and (R)PMPA. We examine the in vitro phosphorylation of PMEA
and PMPA bearing a borano- or a H- group on the phosphorus atom. The α-borano or α-H on
PMEA and PMPA were detrimental to the activity of recombinant human AMP kinases 1 and 2.
Docking PMEA to the active site of AMP kinase 1 indicated that the borano group may prevent two
conserved critical Arg interactions with the α-phosphate, resulting in substrate bad positioning.
Keywords Borano analogues; phosphinate; phosphonate; adenylate kinase
Received 7 February 2007; accepted 11 November 2007.
Present address for B. Schneider: Laboratoire de Diffe´renciation et Prion, CNRS-FRE 2937 Villejuif,
France.
This work is dedicated to the memory of our friend and colleague Dr Simon Robert SARFATI, who
died December 12, 2005.
We thank Pr. Miche`le Reboud (CNRS FRE 2852 - Universite´ Paris6) for helpful discussions. We also
thank J. J. Montagne (Institut Jacques Monod, CNRS & Universite´s Paris 6,7) for mass spectroscopy
experiments and the Service de Mode´lisation et Imagerie Mole´culaire of the Re´seau Fe´de´ratif de
Recherche of Universite´ Paris6 for calculation facilities.
The work was supported by grants from the Agence Nationale pour la Recherche contre le SIDA
(France), Universite´ Pierre-et-Marie-Curie-Paris6, French Centre National de Recherche Scientifique
(FRE 2852 and ANR-05-BLAN-0368-02 for DDB and DT) and ESC-SIDACTION (for BC, KA and KB).
Dr. Karine Barral was supported by a postdoctoral fellowship from Agence Nationale pour la Recherche
contre le SIDA (France).
Address correspondence to D. Deville-Bonne, Laboratoire d’Enzymologie, Universita´ Paris 6,
FRE CNRS 2852, Institut Jacques Monod, 4, place Jussieu, 75251, Paris Cedex 05, France. E-mail:
ddeville@ccr.jussieu.fr
319

320
D. Topalis et al.
INTRODUCTION
Nucleos(t)ide analogues are widely used in the treatment of cancers and
viral infections and require conversion to their tri-phosphate metabolites
before reaching their viral target. Although successful in combination
therapies, their long-term efficacy is limited by the emergence of drug-
resistant strains of virus. The development of new molecules is thus a
permanent challenge. Since a high intracellular concentration of the
triphosphate form of the analogues (or diphosphate form of acyclic nu-
cleotide phosphonates) is correlated with enhanced activity, understanding
the individual pathways of each nucleoside analogue will help to pre-
dict and explain drug potency. Nucleotide analogues with a nonbridging
oxygen on the α-phosphate substituted by a borano group (BH3−) are
promising second-generation molecules.[1] α-Boranophosphates have been
proposed as mimics of natural phosphodiesters in DNA, in antisense
oligonucleotide strategies or for boron neutron-capture therapy.[2−4] They
are also potential antiviral molecules when combined with other chemical
modifications.[2] Mechanistic analysis of reverse transcriptase from resistant
virus indicates that α-borano derivatives are incorporated into DNA by
suppressing reverse-transcriptase mediated resistance when the resistance
is due to a poor catalytic rate.[3] α-Borano-derivatives of AZT- and d4T-
diphosphate also react faster with NDP kinases, allowing the formation of
increasing amounts of the chain terminators.[1] However, the efficacy of
AZT- and d4T- borano monophosphate has not yet been demonstrated on
HIV-infected cells in culture, perhaps because these charged molecules do
not readily enter the cells, or because of problems of phosphorylation in the
cells.
Acyclic nucleoside phosphonates (ANPs) have become a key class of an-
tiviral nucleoside derivatives. Prodrugs of tenofovir, (R)PMPA, and adefovir,
PMEA, were approved for HIV−1 infections in 2001 and for hepatitis B virus
infections in 2002, respectively.[4]. Unlike nucleoside analogues, PMEA and
(R)PMPA (further called PMPA), as dAMP mimics, do not require the
initial phosphorylation catalyzed by a nucleoside kinase (Figure 1). They
rely on cellular adenylate kinases for their activation, resulting in persistent
metabolites.[5−7] The NMP kinases catalyze the phosphotransfer “in-line”
between nucleotides without the formation of a covalent intermediate. Both
the fold and the kinetics are different from NDP kinase, the enzyme involved
in the next step of the salvage pathway.[8] The crystal coordinates of human
AK1 complexed with the bisubstrate inhibitor Ap5A, and of human AK2
with bound Ap4A, were recently deposited in the Protein Data Bank (code
1Z83 and 2C9Y, respectively).[9] AK1 is one of the most abundant cytosolic
proteins in vertebrate muscles with a concentration as high as 0.1 mM.[10]
AK2, located in the mitochondrial intermembrane space, plays a major role
in heart, liver, and kidney energy metabolism.

Reaction of ANP with Human Adenylate Kinases
321
(R)PMPA, tenofovir
PMEA, adefovir
NH2
NH2
N
N
N
N
N
N
N
O
P
N
O
P
-O
O
-O
O
Y
Y
-
Y : H, OH, BH3
FIGURE 1 Structure of analogues of PMEA (adefovir) and (R)PMPA (tenofovir) bearing a borano
group or a hydrogen on the α-phosphorus atom.
We report here the cloning of the gene encoding human AK1 and
AK2, the major cytosolic and mitochondrial adenylate kinases of human
cells, and their expression in E. coli. We used steady-state kinetic analysis
to characterize the specificity of the recombinant enzymes for natural
substrates, acyclic phosphonates analogues, PMEA and PMPA and for their
analogues. The α-boranophosphonate and the H-phosphinates derivated
from PMEA and PMPA have been synthesized recently[11] (Figure 1), as well
as their diphosphate which presented interesting properties by overcoming
resistance of HIV reverse transcriptase.[12] We examined whether the mod-
ifications in α-P affected the rate at which a phosphate from ATP is added
to PMEA and (R)PMPA, a reaction catalysed by human adenylate kinases 1
and 2 (hAK1 and hAK2).
MATERIALS AND METHODS
Chemistry: α-Borano-phosphonates and H-phosphinates derived of
PMEA and PMPA were made according the described procedure.[11]
Cloning of the gene encoding human AMP kinases 1 and 2: The 582 bp and
717 bp fragments corresponding to the AK1 and AK2 genes (coding for
the human AK1 and AK2 proteins) were amplified by PCR[13] using the
human liver Quick-clone cDNA (BD Bio-sciences Clontech) as a template.
The synthetic oligonucleotides used for amplification were
5ꢁ-ggaattccatatgGaagAGAAGCTGAAGAAAAC-3ꢁ and
5ꢁ-ccgctcgagttacttTAGGGCGTCCAGGTGGGTG-3ꢁ for AK1 and
5ꢁ-ggaattccatatgGCTCCCAGCGTGCCAGC-3ꢁ and
5ꢁ-ccgctcgagttaGATAAACATAACCAAGTCTTTACA-3ꢁ for AK2.
The NdeI and XhoI restriction sites (in bold letters in the oligonucleotide
sequences) were created at both ends of the amplified fragments. The

322
D. Topalis et al.
amplified fragments were digested with NdeI and XhoI and inserted into
the pET28a plasmid (Novagen, Inc., Madison, WI, USA) digested with the
same enzymes. The DNA inserts were sequenced using the double-stranded
dideoxynucleotide sequencing technique[14] in order to verify the absence
of any mutation. The plasmids harboring the hAK1 and hAK2 genes were
named pHL20-11 and pHL20-21.
Expression and purification of recombinant hAK1 and hAK2: Recombinant
proteins were produced in E. coli BL21 (DE3) strain (Novagen, Inc.) trans-
formed with the pDIA17[15] and the corresponding expression plasmids.
The same procedure was followed for both hAKs. Cultures were grown
at 37◦C in a rich medium (2YT) containing 30 µg/mL chloramphenicol
and 70 µg/mL kanamycine until an absorbance A600 of 1 was reached.
After induction with 1 mM isopropylthio-β-D-galactoside (IPTG) and con-
tinued incubation for 3 hours at 30◦C, cells were harvested. A cell pellet
corresponding to 1 L of culture was suspended in 40 mL lysis buffer (50
mM phosphate buffer pH 8, 300 mM NaCl, 10 mM imidazol, protease
inhibitors (Complete EDTA-free, Roche Applied Science, Indianapolis,
IN, USA) and 1 mM DTT) and stored at −80◦C. Cells were broken by
sonication and centrifuged at 10,000 rpm for 30 minutes at 4◦C. The
supernatant was placed on a 20 mL Ni-NTA column (Quiagen, Germany)
equilibrated with lysis buffer. The column was washed with lysis buffer
(100 mL) and the protein was eluted with a linear gradient of imidazol
(10-250 mM) pH 8. Fractions containing the enzymatic activity were pooled
and dialysed against 50 mM Tris, 20 mM NaCl, 1 mM DTT, 50% glycerol,
pH 7.5 and stored at –20◦C. The protein appeared to be >95% pure as
judged by SDS–PAGE gels (Figure 2). There was no loss of activity after
3 months storage at –20◦C. The molar extinction coefficients at 278 nm
were calculated from the absorbance of individual amino acids as 9925
M−1 cm−1 for hAK1 and 7250 M−1 cm−1 for hAK2. Mass spectra were
obtained on a Voyager-DE PRO MALDI-TOF mass spectrometer (Applied
Biosystems, Foster City, CA, USA) equipped with a pulsed nitrogen laser
(337 nm, 3 ns pulse). Operating parameters for reflection were: accelerating
voltage (20 kV), grid voltage (75%), guidewire voltage (0.005%) and 100
laser shots per spectrum. The ions of des-Arg1-bradykinin, angiotensin I,
Glu1-fibrinopeptide B and neurotensine were used for external calibration.
Monoisotopic masses were used with deviation for mass assignment within
+/− 0.5 Da.
Kinetic measurements: The catalytic activity of the NMP kinases was fol-
lowed in a spectrophotometer by measuring ADP formation.[16,17] The assay
mix contained 50 mM Tris-HCl pH 7.4, 50 mM KCl, 5 mM MgCl2, 1 mM ATP,
0.2 mM NADH, 10 mM DTT, 1 mM phosphoenolpyruvate and the auxiliary
enzymes: pyruvate kinase (4 U) and lactate dehydrogenase (4 U). The
reaction was started at 37◦C by adding AMP or another phosphate acceptor
and enzyme at the indicated concentration. The magnesium concentration
in the hAK2 experiments was 2 mM, as a slight inhibition was observed for

Reaction of ANP with Human Adenylate Kinases
323
FIGURE 2 Purification of recombinant human adenylate kinases 1 and 2. (A) hAK purifications
SDS/PAGE stained with Coomassie Blue: (1) hAK1 expressing cell extract, (2) pooled active hAK1, (3)
hAK2 expressing cell extract, (4) pooled active hAK2, (M) markers. (B): Absorption spectrum of the
native purified hAK1 enzyme.
Mg2+ concentration above 3 mM. The rates were kept below 0.2 ꢀA340/minute
to avoid limitation by the coupled system. The rates for AMP kinases were
calculated on the basis of two ADPs generated during the reaction with AMP.
When dAMP, PMEA, or PMPA were the phosphate acceptors, only one ADP
was produced as in the AMP kinase assay. One unit of enzyme activity is
defined as 1 µm of NMP transformed per min at 37◦C. Curve-fitting was
performed using Kaleidagraph (Abelbeck Software, Reading, PA, USA) for
a hyperbolic equation.
Structural models: Structural models were built and visualized using
Insight II softwares package (Accelrys, San Diego, CA, USA). The coordinate
of the human adenylate kinase hAK1 (1Z83.pdb) was used for building
our complexes. The pdb file was corrected from some inconstancies, for
example multiple positions for some residues. Lacking atoms were reintro-
duced to be able to accommodate CHARMM force field for minimisation
steps. PMEA was built in SMILES format using JME Molecular Editor
(Novartis, Inc.) and translated in PDB format using a online SMILES
translator (http://cactus.nci.nih.gov/ services/translate). PMEA was then
docked to the active pocket of hAK1, and superimposed to AMP ligand for
comparison of potent interactions expected for these configurations. Since
PMEA minimized structure displayed a certain conformational flexiblity,
rotamers were generated using Biopolymer Insight II module functionalities
to provide a linear version of PMEA phosphonate chain. Linear PMEA
acyclic chain allows a better illustration of interactions network disturbance
compared to AMP.

324
D. Topalis et al.
RESULTS AND DISCUSSION
Characterization of Recombinant Human Adenylate
Kinases 1 and 2
The recombinant enzymes, hAK1 and hAK2, expressed as a His-tag
protein fusion in E. coli, were purified in a single step on a Ni-nitrilotriacetic
acid agarose column yielding 90 mg hAK1 protein and 16 mg hAK2 per
liter of growth medium. The proteins were at least 95% pure by SDS-PAGE
(Figure 2A). The purified AMP kinases were analyzed by mass spectrometry
(MALDI-TOF): The measured mass of the hAK1 was 25,620 (predicted mass
of the recombinant enzyme: 25,929). The measured and expected masses
of hAK2 were 28,704 and 28,640. Both recombinant enzymes had extra 19
amino acid residues, including the 6 His-tag, at the N-terminus. The slight
difference between the masses for hAK1 could be due to some degradation.
The absorbance spectrum of native hAK1 (Figure 2B) is similar to that of
native hAK2, but hAK2 tended to aggregate in absence of glycerol.
The catalytic properties of recombinant hAK1 and hAK2 are presented
on Table 1. At a saturating concentration of ATP (1 mM), AMP phospho-
rylation followed a Michaelis-Menten mechanism, with a maximum rate
of (1240 +/− 50) U/mg for hAK1 and (170 +/− 25) U/mg for hAK2,
and turnover numbers of 500 s−1and 80 s−1, respectively. Excess substrate
AMP
([AMP] > 0.25 mM) was inhibitory with hAK2 (KI
= 0.5 mM). The
magnesium ions (1−10 mM) did not influence the activity of hAK1, but
hAK2 activity was optimum at [Mg2+] = 2 mM. AMP was a better substrate
than dAMP for both kinases (Table 1), and the maximum rates for AMP and
hAK1 gave excellent catalytic efficiencies of 105-106 M−1s−1. The saturation
curves for [ATP] in the presence of 1 mM AMP indicated a KM of 0.060 mM
for both enzymes. Among NTPs, ATP only acted as a phosphodonor for
human AK1 while GTP, CTP and UTP could replaced ATP for hAK2 with
about 15% efficiency (at 0.1 mM NTP and 0.05 mM AMP, results not shown),
as previously observed with cytosolic and mitochondrial extracts from CEM
cells.[7] The values of the catalytic parameters for the two kinases and their
natural substrates (Table 1) are comparable to the kinetic parameters of
hAK1 found in previous studies[18,19]and of enzymes purified from cattle
heart.[20,21] The absence of any post-translational modification and the
presence of the His-tag did not significantly change the enzyme activity, as
expected from the N-acetylation of natural hAK1.
Phosphorylation of PMEA, PMPA and Their Analogues
PMEA was phosphorylated by hAK1 at a maximum turnover of 0.1 s−1
and PMPA at 0.2 s−1, while hAK2 presented higher rates (1.5 and 2.5 s−1
,
respectively) (Figure 3). The KM values for the two enzymes were identical

325

326
D. Topalis et al.
5
4
3
2
1
0
0.5
0.4
0.3
0.2
0.1
0
B
A
(1)
(1)
(2)
(2)
10
0
5
10
15
20
25
0
2
4
6
8
[PMEA or PMPA], mM
[PMEA or PMPA], mM
0.1
0.14
0.12
0.1
(1)
(1)
C
(2)
D
0.08
0.06
0.04
0.02
0
(2)
0.08
0.06
0.04
0.02
0
(3)
(4)
(3)
(4)
0
1
2
3
4
5
0
1
2
3
4
5
[Acyclic phosphonate analogue], mM
[Phosphonate analogue], mM
FIGURE 3 Saturation curves of recombinant human adenylate kinases 1 and 2 with PMEA and PMPA
in 1 mM ATP and 5 mM Mg2+ (hAK1) or 2 mM (hAK2). (A) and (C) hAK1 activity with (1) PMPA
(
)
•
and (2) PMEA
( ); (B) and (D) hAK2 activity with (1) PMPA ( ) and (2) PMEA ( ); (C) hAK1
◦
•
◦
−
−
activity with (3) α-BH3 -PMPA (ꢀ) and (4) α-BH3 -PMEA (ꢀ) derivatives; (D) hAK2 activity with (3)
−
−
α-BH3 -PMPA (ꢀ) and (4) α-BH3 -PMEA (ꢀ) derivatives. (Parameters are shown in Table 2).
(KPMMEA = 6 mM and (KMPMPA = 3 mM, Table 1), revealing lower binding
affinities than for dAMP. As a result, the substrate specificity constants
(kcat/KM) for the phosphonates were 0.2–0.01% of that of dAMP in the hAKs
catalyzed reactions: that is, 800 M−1s−1 and 250 M−1s−1 for PMPA and PMEA
with hAK2, and 75 and 14 M−1s−1 with hAK1 (Table 1). Thus, although
phosphonates are thus poor substrates for both hAK1 and hAK2, they are
phosphorylated with a ten times higher efficiency by the mitochondrial
enzyme hAK2 as observed in previous studies.[7] The KM values determined
in these studies were found slightly higher, which is not surprising for results
obtained with cell extracts, i.e. in the presence of contaminants.[7]
−
Modifications of PMEA and PMPA by replacing an oxygen with a BH3
group resulted in α-boranophosphonates (Figure 1). The modification
drastically reduced phosphorylation by both hAK1 and hAK2: Figures 3C
and 3D show that the α-borano derivatives up 5 mM are phosphorylated
extremely slowly. The resulting catalytic efficiencies (kcat/KM) for hAK2 and
α-BH3-PMPA is 10 M−1s−1 (Table 2). The presence of one hydrogen instead

Reaction of ANP with Human Adenylate Kinases
327
TABLE 2 Catalytic efficiencies of (R)PMPA and PMEA derivatives with hAK1
and hAK2 (Data correspond to at least three independent experiments)
kcat/KM (M−1 s−1
hAK1
)
hAK2
PMPA
PMEA
75
14
5
2
800 50
250 20
α-BH3−-PMPA
α-BH3−-PMEA
α-H-PMPA
α-H-PMEA
1
10
4.3
2
1
<1
<1
<1
<1
<1
of a borano group, resulting in H-phosphinates, is even more deleterious as
these compounds are not substrates at all for AMP kinases 1 and 2. These
analogs are not either substrates for human UMP-CMP kinase (not shown).
Structural Analysis and Molecular Modeling
The structural basis of AMP kinases activity has been investigated in
several studies. AMP kinases, like all NMP kinases, share a highly conserved
structure comprising an AMP binding domain, a central CORE domain
that carries the ATP binding loop (P-loop) and a LID domain providing
catalytic residues necessary for the reaction. The LID domain is larger
in hAK2. The AMP and LID domains are extremely mobile and undergo
closure upon binding of substrates.[22] Protein dynamics studies correlated
the lid-opening rate with the catalytic turnover.[23,24]
In the published structure of hAK2 in a complex with Ap4A, both
adenines were located near the ATP site, providing no information on
the acceptor site. On the contrary, hAK1 complexed to Ap5A (1Z83)
showed the interactions of both ATP and AMP in the active site in a
closed conformation.[9] The structure of PMEA bound to hAK1 was then
modelized using this structure (Figure 4). The adenosine plus the first
phosphate from Ap5A is likely to mimic AMP in the active site of hAK1 in
closed conformation where helices α3 and α4 close up on AMP. As shown
in Figure 4, the 2ꢁ hydroxyl of AMP interacted with the Gln65 main chain
carbonyl belonging to the mobile α4 helix of the NMP binding domain,
locking the nucleotide in the closed binding site. Replacement of AMP by
dAMP suppressed the H-bond between 2ꢁ-OH and the main chain oxygen
of Gln65, resulting in looser binding. AMP was readily replaced by PMEA
because the volume of the acyclic moiety of the analogue is smaller than
a ribose (Figure 4). Thus, both PMPA and PMEA, less bulky than dAMP
and lacking “handles,”[25] fitted snugly into the active site cleft. PMPA is
slightly bulkier with the extra methyl group, and it may fit better into

328
D. Topalis et al.
FIGURE 4 Docking of PMEA in the active site of hAK1. (A) Overall structure of hAK1 (1Z83.pdb)
showing the CORE domain (green), the NMP domain (yellow), the LID domain (dark cyan), the P
loop (orange), the zinc atom (grey) and Ap5A in sticks and CPK colors. (B) Superimposition of PMEA
(blue) and AMP (red) docked in hAK1 AMP binding site. Only the residues interacting with AMP are
shown with the main interactions in dashed lines. (C) Close up of AMP/PMEA binding site of hAK1
(same colors as in A). The superimposition of AMP and PMEA illustrates the disturbance of the AMP
interactions network.
the AMP binding site. The structure of another acyclic analog, ganciclovir-
monophosphate in the active site of GMP kinase from E. coli showed that the
absence of the 2ꢁhydroxyl of the ribose moiety was critical and resulted in
the incomplete closure of the NMP domain.[26] Similarly, we propose that

Reaction of ANP with Human Adenylate Kinases
329
despites the good fit of PMEA in the AMP binding site, the closing of NMP
domain is difficult upon PMEA binding.
In the complex hAK1-Ap5A complex, the phosphate equivalent to the
α-P of AMP is tightly bound to Arg44 from helix α3 and Arg97 belonging
to the CORE domain. These two Arg residues are well conserved in the
other NMP kinases, ensuring an efficient nucleophilic attack on γ -P of
ATP by the equatorial nonbridging oxygen positioned in line. It seems
unlikely that the borano group occupies the equatorial position on the
phosphorus but rather it lies in one of the two other positions, interacting
with the kinase Arg residues. The borano group in boranophosphate is
similar in size to a methyl group in a methylphosphonate and slightly
larger than the oxygen in a natural phosphate.[27] The negative charge
on the natural phosphate oxygen is retained in the boranophosphonate
derivatives but the polarity is affected as the negative charge is polarized
away from the borano.[27] Further, there is no lone pair of electrons in the
–BH3 group, as with oxygen. As a consequence, interaction of the borano
group with the critical Arg residues must be different. A theoretical study
using quantum molecular mechanics calculations proposed a concerted
phosphoryl transfer for the related enzyme UMP-CMP kinase.[28] According
to these authors, the synchronous shift of a proton from NMP to the
γ -phosphate should accompany the phosphate transfer. Such a proton
shift may be difficult (or impossible) with borano-modified NMP or ANP,
resulting in the observed decreased reaction rate.
Similar results were observed with α-boranoP-TMP and TMP kinase. The
reaction of thymidine analogs with TMP kinase was described as a bottleneck
−
in the activation of AZT and d4T.[29] The substitution of a BH3 group for
an oxygen in TMP was found to slow down the phosphotransfer reaction
catalyzed by yeast TMP kinase by nearly two orders of magnitude (result not
shown). Previous data indicated that the 5ꢁ-H-phosphonates and 5ꢁ-fluoro-
phosphonates derived from dAMP were apparently not phosphorylated by
partly purified hAK from placenta, also suggesting that the nature of the
phosphate residue is crucial for the phosphorylation catalyzed by adenylate
kinases.[30]
CONCLUSION
The lack of activity of NMP kinases with the derivatives modified on
the α-P is in complete contrast with the reported improvement (X 10) in
the phosphorylation of α-(Rp)-boranotriphosphate derivatives of AZT and
−
d4T by NDP kinase.[1] The BH3 group is beneficial for the HIV reverse
transcriptase even for enzymes from resistant viruses.[3] The therapeutic
potential of PMEA and PMPA with a α-borano group could be demon-
strated, in future studies, by vectorising or encapsulating the mono or fully

330
D. Topalis et al.
phosphorylated active form: this should facilitate the delivery of the active
derivative to cells. Nanogel formulations of AZT-triphosphate in cross-linked
polymers have already been shown to be particularly cytotoxic for breast
cancer cell lines.[31]
ABBREVIATIONS
ANP, acyclic nucleotide phosphonates; AZT, 2ꢁ,3ꢁ-dideoxy-3ꢁ-azido
thymidine;
cytosine;
9-[2-(phosphonomethoxy)ethyladenine;
cidofovir,
d4T,
(S)-1-(3-hydroxy-2-phosphonylmethoxypropyl)
2ꢁ,3ꢁ-dideoxy-2ꢁ,3ꢁ-didehydrothymidine;
(R)PMPA,
PMEA,
(R)-9-[2-
(phosphonomethoxy)propyl]adenine; hAK1, human adenylate kinase
1 (cytosolic); hAK2, human adenylate kinase 2 (mitochondrial); U, enzyme
activity unit (1 micromole of substrate transformed per minute).
REFERENCES
1. Meyer, P.; Schneider, B.; Sarfati, S.; Deville-Bonne, D.; Guerreiro, C.; Boretto, J.; Janin, J.; Ve´ron,
M.; Canard, B. Structural basis for activation of α-boranophosphate nucleotide analogues targeting
drug-resistant reverse transcriptase. EMBO J. 2000, 19, 3520–3529.
2. Boyle, N.A.; Rajwanshi, V.K.; Prhavc, M.; Wang, G.; Fagan, P.; Chen, F.; Ewing, G.J.; Brooks,
J.L.; Hurd, T.; Leeds, J.M.; Bruice, T.W.; Cook, P.D. Synthesis of 2ꢁ,3ꢁ-dideoxynucleoside 5ꢁ-alpha
-P-borano-beta, gamma (difluoromethylene)triphosphates and their inhibition of HIV-1 reverse
transcriptase. J. Med. Chem. 2005, 48, 2695–2700.
3. Deval, J.; Alvarez, K.; Selmi, B.; Bermond, M.; Boretto, J.; Guerreiro, C.; Mulard, L.; Canard, B.
Mechanistic insights into the suppression of drug resistance by human immunodeficiency virus
type 1 reverse transcriptase using α-boranophosphate nucleoside analogs. J. Biol. Chem. 2005, 280,
3838–3846.
4. De Clercq, E.; Holy, A. Acyclic nucleoside phosphonates: a key class of antiviral drugs. Nat. Rev.
Drug Discov. 2005, 4, 928–940.
5. Votruba, I.; Bernaerts, R.; Sakuma, T.; De Clercq, E.; Merta, A.; Rosenberg, I.; Holy,
A. Intracellular phosphorylation of broad-spectrum anti-DNA virus agent (S)-9-(3-hydroxy-2-
phosphonylmethoxypropyl)adenine and inhibition of viral DNA synthesis. Mol. Pharmacol. 1987,
32, 524–529.
6. Balzarini, J.; Hao, Z.; Herdewijn, P.; Johns, D.G.; De Clercq, E. Intracellular metabolism and
mechanism of anti-retrovirus action of 9-(2-phosphonylmethoxyethyl)adenine, a potent anti-
human immunodeficiency virus compound. Proc. Natl. Acad. Sci. 1991, 88, 1499–1503.
7. Robbins, B.L.; Greenhaw, J.; Connelly, M.C.; Fridland, A. Metabolic pathways for activation of the
antiviral agent 9-(2-phosphonylmethoxyethyl)adenine in human lymphoid cells. Antimicrob. Agents
Chemother. 1995, 39, 2304–2308.
8. Morera, S.; Lacombe, M.-L.; Xu, Y.; LeBras, G.; Janin, J. X-ray structure of nm23 human Nucleoside
Diphosphate kinase B at 1.8A resolution. Structure 1995, 3, 1307–1314.
9. Filippakopoulos, P.; Bunkoczi, G.; Jansson, A.; Schreurs, A.; Knapp, S.; Edwards, A.; Von Delft,
F.; Sundstrom, M. Crystal structure of human AK1A in complex with AP5A. Structural Genomics
Consortium Oxford. PDB code 1Z83, 2005.
10. Maughan, D.W.; Henkin, J.A.; Vigoreaux, J.O. Concentrations of glycolytic enzymes and other
cytosolic proteins in the diffusible fraction of a vertabrate muscle proteome. Mol. Cell Prot. 2005, 4,
1541–1549.
11. Barral, K.; Priet, S.; Sire, J.; Neyts, J.; Balzarini, B.; Canard, B.; Alvarez, K. Synthesis, in vitro
antiviral evaluation and stability studies of novel alpha-borano-nucleotide analogues of 9-(2-

Reaction of ANP with Human Adenylate Kinases
331
(phosphonylmethoxy)ethyl)adenine (PMEA) and (R-9-(2-(phosphonomethoxy)propyl)adenine
(PMPA). J. Med. Chem. 2006, 49, 7799–7806.
12. Frangeul, A.; Barral.; Alvarez, K.; Canard, B. In vitro suppression of K65R reverse transcriptase-
mediated tenofovir- and adefovir-5ꢁ-diphosphate resistance conferred by the boranophosphonate
derivatives. Antimicrob. Agents Chemother. 2007, 51, 3162–3167.
13. Sambrook, J.; Fritsch, E.F.; Manniati, S.T. Molecular cloning: A laboratory manual. Cold Spring
Harbor: NY, 1989.
14. Sanger, F.; Nicklen, S.; Coulson, A.R. DNA sequencing with chain-terminating inhibitors. Proc. Natl.
Acad. Sci. USA 1977, 74, 5463–5467.
15. Munier, H.; Gilles, A.-M.; Glaser, P.; Krin, P.; Danchin, A.; Sarfati, R.; Barzu, O. Isolation and
characterization of catalytic and calmodulin-binding domains of Bordetella pertussis adenylate
cyclase. FEBs. J. 1991, 196, 469–474.
16. Blondin, C.; Serina, L.; Wiesmuller, L.; Gilles, A.-M.; Barzu, O. Improved spectrophotometric assay
of nucleoside monophosphate kinase activity using the pyruvate kinase/lactate dehydrogenase
coupling system. Anal. Biochem. 1994, 220, 219–221.
17. Pasti, C.; Gallois-Montbrun, S.; Munier-Lehmann, H.; Veron, M.; Gilles, A.-M.; Deville-Bonne, D.
Reaction of human UMP-CMP kinase with natural and analog substrates. FEBs. J. 2003, 270, 1784–
1790.
18. Ayabe, T.; Takenaka, H.; Takenaka, O.; Sumida, M.; Maruyama, H.; Onitsuka, T.; Shibata, K.;
Uesugi, S.; Hamada, M. Essential lysine residues in the N-terminal and the C-terminal domain
of hyman adenylate kinase interact with adenine nucleotides as found by site-directed random
mutagenesis. Biochemistry 1997, 36, 4027–4033.
19. Yan, H.; Tsai, M.-D. Nucleoside monophosphate kinases: structure, mechanism and substrate
specificity. Adv. Enzymol. 1999, 73, 103–134.
20. Font, B.; Gautheron, D. General and kinetic properties of pig heart mitochondrial adenylate
kinase. Biochem. Biophys. Acta 1980, 611, 299–308.
21. Walker, E.J.; Dow, J.W. Location and properties of two isoenzymes of cardiac adenylate kinase.
Biochem. J. 1982, 203, 361–369.
22. Vonrheim, C.; Schlauderer, G.J.; Schulz, G.E. Movie of the structural changes during a catalytic
cycle of nucleoside monophosphate kinases. Structure 1995, 3, 483–490.
23. Wolf-Watz, M.; Thai, V.; Henzler-Wildman, K.; Hadjipavlou, G.; Eisenmesser, E.Z.; Kern, D. Linkage
between dynamics and catalysis in a thermophilic-mesophilic enzyme pair. Nat. Struc. Mol. Biol.
2004, 11, 945–949.
24. Bae, E.; Phillips, G.N.Jr. Roles of static and dynamic domains in stability and catalysis of adenylate
kinase. Proc. Natl. Acad. Sci. 2006, 103, 2132–2137.
25. Tuske, S.; Sarafanios, S.G.; Clark Jr, A.D.; Ding, J.; Naeger, L.K.; White, K.L.; Miller, M.D.; Gibbs,
C.S.; Boyer, P.L.; Clark, P.; Wang, G.; Gaffney, B.L.; Jones, R.A.; Jerina, D.M.; Hughes, S.H.; Arnold,
E. Structures of HIV-1 RT-DNA complexes before and after incorporation of the anti-AIDS drug
tenofovir. Nat. Struc. Mol. Biol. 2004, 11, 469–474.
26. Hible, G.; Daalova, P.; Gilles, A.-M.; Cherfils, J. Crystal structures of GMP kinase in complex with
ganciclovir monophosphate and Ap5G. Biochimie 2006, 88, 1157–1164.
27. Summers, J.S.; Shaw, B.R. Boranophosphates as mimics of natural phosphodiesters in DNA. Cur.
Med. Chem. 2001, 8, 1147–1155.
28. Hutter, M.C.; Helms, V. Phosphoryl transfer by a concerted reaction mechanism in UMP-CMP
kinase. Prot. Sci. 2000, 9, 2225–2231.
29. Lavie, A.; Schlichting, I.; Vetter, I.R.; Konrad, M.; Reinstein, J.; Goody, R.S. The bottleneck in AZT
activation. Nat. Med. 1997, 3, 922–924.
30. Skoblov, J.S.; Frank-Kamenetskaya, M.D.; Chernov, D.N.; Krayevsky, A.A. Modified nucleotides as
substrates and inhibitors of adenylate kinase from different sources. FEBS. Let. 1996, 395, 283–285.
31. Vinogradov, S.V.; Kohli, E.; Zeman, A.D. Cross-linked polymeric nanogel formulations of 5ꢁ-
triphosphates of nucleosides analogues: role of a cellular membrane in drug release. Mol.
Pharmacol. 2005, 2, 449–461.