Methylene bisphosphonates as the inhibitors of HIV RT phosphorolytic activity

Article abstract of DOI:10.1016/j.biochi.2016.05.012

The structure-function analysis of 36 methylenebisphosphonates (BPs) as inhibitors of the phosphorolytic activity of native and drug-resistant forms of HIV-1 reverse transcriptase (RT) was performed. It was shown that with the increase of the inhibitory potential of BPs towards the phosphorolytic activity raises their ability to inhibit the RT-catalyzed DNA elongation. Herein, we report the impact of the thymidine analog mutations (TAM) on the activity of bisphosphonates, as well as some structural features of the BPs, allowing them to maintain the inhibitory activity on the enzyme resistant to nucleoside analog therapy. We estimated the Mg²⁺-coordinating group structure, the linker and the aromatic pharmacophore influence on the inhibitory potential of the BPs. Based on the 31 BPs SAR, several BPs with improved inhibitory properties were designed and synthesized.

Full text of DOI:10.1016/j.biochi.2016.05.012

Accepted Manuscript
Methylene bisphosphonates as the inhibitors of HIV RT phosphorolytic activity
D.V. Yanvarev, A.N. Korovina, N.N. Usanov, O.A. Khomich, J. Vepsäläinen, E.
Puljula, M.K. Kukhanova, S.N. Kochetkov
PII:
S0300-9084(16)30098-0
10.1016/j.biochi.2016.05.012
BIOCHI 4998
DOI:
Reference:
To appear in: Biochimie
Received Date: 20 March 2016
Accepted Date: 18 May 2016
Please cite this article as: D.V. Yanvarev, A.N. Korovina, N.N. Usanov, O.A. Khomich, J. Vepsäläinen,
E. Puljula, M.K. Kukhanova, S.N. Kochetkov, Methylene bisphosphonates as the inhibitors of HIV RT
phosphorolytic activity, Biochimie (2016), doi: 10.1016/j.biochi.2016.05.012.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Methylene bisphosphonates as the inhibitors of HIV RT phosphorolytic
activity
D.V. Yanvareva,#, A.N. Korovinaa, N.N. Usanova, O.A. Khomicha, J. Vepsäläinenb, E. Puljulab, M. K.
Kukhanovaa, and S.N. Kochetkova
a Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Vavilova st.-32, Moscow,
Russia
b School of Pharmacy, Biocenter Kuopio, University of Eastern Finland, Kuopio, Finland
Keywords: bisphosphonates, reverse transcriptase, HIV, excision, pyrophosphorolysis, resistance,
TAM, SAR
#corresponding author: 119991, Vavilova st.-32, Moscow, Russia. E-mail: JDmitry@hotmail.com, tel.
+7(926)9267025449
Abstract
The structure-function analysis of 36 methylenebisphosphonates (BPs) as inhibitors of the
phosphorolytic activity of native and drug-resistant forms of HIV-1 reverse transcriptase (RT) was
performed. It was shown that with the increase of the inhibitory potential of BPs towards the
phosphorolytic activity raises their ability to inhibit the RT-catalyzed DNA elongation. Herein, we
report the impact of the thymidine analog mutations (TAM) on the activity of bisphosphonates, as well
as some structural features of the BPs, allowing them to maintain the inhibitory activity on the enzyme
resistant to nucleoside analog therapy. We estimated the Mg2+-coordinating group structure, the linker
and the aromatic pharmacophore influence on the inhibitory potential of the BPs. Based on the 31 BPs
SAR, several BPs with improved inhibitory properties were designed and synthesized.
1.
Introduction
Reverse transcriptase (RT) activity in lymphoid cells of the first patients with acquired
immunodeficiency syndrome (AIDS) was detected in the early 1980s. This discovery allowed to assign
HIV to the Retroviridae family and led to intensive studies of HIV infection mechanisms and search
for the specific inhibitors [1, 2]. Further investigations of RT as the HIV key enzyme revealed that RT
has no reparation function and that provides high mutability of viral genome and the emergence of the
drug resistant strains against most antivirals used [3].
There are two main classes of drugs that suppress polymerase activity of HIV RT: non-
nucleoside (NNRTIs) (noncompetitive allosteric inhibitors of RT polymerase reaction) and
nucleoside/nucleotide inhibitors (NRTIs).
NRTIs commonly lack the 3’-OH group of the ribose fragment; after their incorporation into
the growing viral DNA chain, they terminate its biosynthesis and therefore inhibit the viral replication.
According to the generally accepted hypothesis, HIV resistance correlates with the formation of the
specific point mutations in the RT coding pol gene. The resistance to NRTIs emerges through two
main mechanisms. The first one, “the exclusion mechanism”, appears in the presence of amino acid
substitutions M184V/I, which allow the RT significantly reduce the effectiveness of the oxathiolane
nucleosides incorporation into DNA (e.g. β-L-2’,3’-dideoxy-3’-thiacytidine (3TC, Lamivudine®) [4,
5] or K65R substitution resulting in the formation of a molecular bridge to the conservative Arg72,
which allows RT to differentiate tenofovir diphosphate and dATP [6]. The resistance of HIV against
1

ACCEPTED MANUSCRIPT
thymidine-based drugs (azidothymidine (AZT), Zidovudine®) takes place via so-called “excision
mechanism”. As a result of the accumulation of the specific mutations, the enzyme acquires the ability
to remove AZT monophosphate from a terminated primer by the phosphorolysis reaction and thus
resumes viral DNA elongation. Inorganic pyrophosphate (PPi) or ATP/GTP [7, 8] can serve as
substrates for the excision reaction although the GTP is barely utilized in vivo by RT due to its low
intracellular concentrations compared with ATP.
The excision-mediated viral resistance is acquired through the selection of specific amino acid
substitutions in HIV RT, which include M41L, D67N, K70R, L210W, T215Y, T215F, K219Q and
K219E – the so-called thymidine analog mutations (TAMs) [9]. Two of these substitutions K70R and
T215Y appear initially during the therapy and later ones only improve the effect of resistance [10]. A
series of RT mutations e.g. “fingers insertions” can enhance the effectiveness of the excision although
without TAMs they have no effect on the phosphorolysis rates. These mutations are called “excision
enhancing mutations (EEMs)” [11]. As a result, the appearance of viral resistance to the NRTI requires
drug change and its higher therapeutic dose leading to problems with drug toxicity.
The drug resistance occurrence stimulates the development of new antiviral agents. The drug
design based on suppression of several catalytic activities of RT, e.g. inhibitors of the elongation and
phosphorolysis reactions seems very promising and potent to avoid the viral resistance developed. A
number of simple inorganic pyrophosphate analogs such as carbonyldiphosphonic and hypophosphoric
acids as well as phosphonoformate (foscarnet) are already known as good inhibitors of RT catalyzed
DNA polymerization reaction (Fig.1) [12]. The mechanism of their inhibitory activity is apparently
associated with their higher than PPi affinity to the catalytic center of RT, thus, leading to their strong
affinity to the pyrophosphate binding site and, as a result, to the suppression of the polymerization
reaction.
Fig. 1. PPi and its structural analogs, which inhibit DNA-dependent DNA polymerase reaction
catalyzed by HIV RTwt, and their IC50.
However, not every PPi analog with high affinity to RT inhibits pyrophosphorolysis (PPlysis)
as itself may serve as a PPlysis substrate. For example, hypophosphoric acid promotes the 3’-terminal
nucleotide excision of the primer more efficiently than equal concentrations of the natural substrates
(ATP, PPi) [13]. Therefore, the inhibitors to be designed should not only possess high affinity for the
enzyme but also lack the substrate properties.
Despite the potential importance of phosphorolytic excision of NRTIs for clinical practice, only
few examples of the selective small molecule inhibitors based on PPi structure were described [14,
15]. Moreover, SAR studies that estimate the impact of TAM mutations on the activity of these
inhibitors are absent and the mechanisms of its inhibition of phosphorolytic or polymerizing activities
of HIV RT are not elucidated.
Hypothetically, there are two possible ways for the blocking of phosphorolysis reaction. The
first one is the competitive binding of the analog to the enzyme catalytic site instead of PPi or ATP
substrates (path b, Fig.2). The second one is the stabilization of the complex between the enzyme and
the terminated DNA-DNA or RNA-DNA complex. The latter is much alike to the dead-end complex,
appearing when the linking of RT to the first complementary NTP after terminator incorporation
produces a quaternary complex RT/DNA-DNA/AZTMP/dNTP (path a, Fig.2). Currently there are no
synthetic compounds described capable of forming the dead-end complex by binding to RT
nucleotide-binding site (N-site) simultaneously with the P-site being occupied by the “terminator”
monophosphate.
2

ACCEPTED MANUSCRIPT
Fig. 2. Possible mechanisms of PPi analogs action: a) formation of the dead-end complex of AZT-
terminated primer – template-BPs b) competition with PPi or its donors for RT nucleotide-binding site.
PPP-TN3 = AZT triphosphate
2.
Results and discussion
Methylene bisphosphonates (BPs) as PPi analogs are attractive as pyrophosphorolysis
inhibitors. The first example of BP suppressing phosphorolysis was described by Cruchaga et al. [15]
who demonstrated that nitrogen containing bisphosphonate pamidronate was able to completely
suppress the ATP-induced excision (ATPlysis) of AZT monophosphate not being a substrate for this
reaction even at high concentrations. All of the previously described PPi analogs, e.g. foscarnet,
phosphonoacetate [15] and hypophosphate [13] were also substrates of the phosphorolysis reaction that
does not allow them to fully suppress pyrophosphorolysis.
Since well-known drugs etidronic and medronic acids (Fig. 3) don’t suppress RT-inherent
phosphorolytic activity, the inhibiting properties of pamidronate (Aredia®) [15] are likely determined
by the presence of the amino group on the –(CH2)2– linker. Therefore we synthesized and evaluated
structure-activity relationship of a series of BP analogs with different linkers between the NH2 group
and the Mg2+-chelating P-C-P moiety (Fig. 4, 18, 32-36).
Fig. 3. Inorganic pyrophosphate and its therapeutically used mimetics
All of them turn out to be weak inhibitors of HIV RT phosphorolytic activity (IC50> 1 mM) and
only 2-aminoethylidene bisphosphonate (18) bearing the -(CH2)1 – linker, had IC50 about 200 ꢀM
(three times better than pamidronate). Thus, aminoalkyl linker variations show little promise in the
design of the inhibitors of RT-catalyzed phosphorolysis.
Previously Song et al. showed a several-order of magnitude increase of the inhibitory potential
for BPs with aromatic substituents instead of alkyl or aminoalkyl groups [14]. QSAR study revealed
that BPs with electron-deficient aromatic substituents were most perspective among the series.
Furthermore, an increase of inhibitory activity up to 500 times with the introduction of an aromatic
substituent indicates that the mechanism of their action on HIV RT is different from the solely
chelation of Mg2+ cations in the reaction medium [14] .
Taking into account the results of [14], we have modified the most active pamidronate analog
(18) with the benzene ring located at the different distances from the Mg2+-coordinating group (P-C-P)
(Fig. 4, 15, 26, 27). However, the activity of such BPs (26, 27) did not increase if compared with
parent (18) and the most active aniline derivative (15) was less active than its benzyl analog (14).
Fig. 4. Structures of the bisphosphonate inhibitors investigated.
decreasing their inhibition activity of PPlysis (IC50 5- 300 µM).
(PPlysis) for all the compounds of this series with the exception of (21). *IC50 (PPlysis/ATPlysis), µM.
A
. Active species, rank ordered by
B. Low active species. IC50 >500 µM
3

ACCEPTED MANUSCRIPT
We determined IC50s (Table 1) for compounds presented in Fig. 4 in excision inhibition assay using
PAGE.
Table 1. IC50 of the compounds in ATP- and PPi-dependent excision reaction catalyzed by WT and
TAM RTs
*Data are presented as the mean values ±S.E. (P<0.001, three independent experiments)
SAR analysis of the compounds containing different structural changes compared to (18), namely the
introduction of substituted phenyl group, variations either in Mg2+-coordinating group or in linker
length were conducted. The effect of such variations on the inhibitory activity is discussed below.
2.1 Effect of the distance between the aromatic pharmacophore and the P-C-P backbone
The distance between the aromatic substituent and Mg2+-coordinating fragment containing OH-
group at the central carbon atom is a very important factor for the BPs inhibitory activity. In a series of
bisphosphonates (7), (9), (13), (14), (19), (20) with benzene as a pharmacophore and the linker length
from 0 to 5 methylene fragments, IC50 values decreased by more than 20 times for both ATPlysis and
PPlysis (Table 1). The optimal length of linker is equivalent to three or four -(CH2)-fragments for both
reactions (Fig. 5).
Fig. 5. Correlation between the inhibitory activity and linker length. Pi = H2PO3
2.2 Effect of the substituent in the aromatic pharmacophore
Introduction of halogen atoms into the aromatic ring increases the activity of BPs more
efficiently than optimization of the linker length. For example, introduction of the chlorine atom at the
para-position of the benzene ring in (14) (i.e. its conversion to 6) increases the inhibition of
phosphorolytic activity of RTwt by six and eight times for PPlysis and ATPlysis, respectively. Another
chlorine-substituted analogue at the meta- and para positions (2) is fourteen and ten times more
effective than its non-halogenated precursor (14). The iodine substitutions also demonstrate a profound
influence on the activity of aromatic BPs as can be seen from the comparison of (12) and (15).
Previously, the effect of the halogen in the aromatic ring of BPs was ascribed to its electron-
withdrawing nature [14], and thus an electron-deficient benzene ring was proposed as a good
pharmacophore. To validate this strategy, we have synthesized a series of BPs containing pyridine (23)
and its N-alkylated cationic forms (22, 28-31), which are even more electron-deficient than chlorinated
benzenes [16]. However, all of the synthesized pyridine-substituted compounds were unable to inhibit
the phosphorolytic activity of RTwt (IC50>500 µM, fig. 4). Apparently, despite the high electron-
deficiency of the pyridinium analogues, there is no recognition site in the catalytic domain of HIV RT
for cationic aromatic substituents. When compared the inhibitory properties of the bisphosphonate (11)
to (31), it becomes clear that the cationic pyridinium ring of (11) acts only as a carrier for benzene
substituent in the same way as -(CH2)3- and -(CH2)4- act as linkers for BPs (7) and (9) respectively.
Although the enzyme may bind these inhibitors in different locations. A similar result was obtained in
4

ACCEPTED MANUSCRIPT
[17] where the heterocyclic substituent PYPY-BP (Fig. 6) was not active without the benzene ring
presence at the C-6 position.
Fig. 6. Examples of the novel heterocycle based bisphosphonate inhibitors of HIV RTwt polymerizing
activity [17]
Electron-deficient substituents in the benzene ring influence the inhibitory potential of the
compound in ambiguous way, since the loss of activity is found for more electronegatively fluorine-
substituted compounds (compare BPs 2 and 5). Despite the fact that (5) is considerably more active
than the unsubstituted precursor (14), bisphosphonate (5) is 2.1 times less active than dichlorobenzyl
analog in PPlysis and 1.7 times less potent in ATPlysis. It can be assumed that the affinity of the BP to
enzyme is not so dependent on hydrogen bonding, which is known to form well with fluorine atoms
[18], but on the so-called halogen bonding, which is characteristic for halogens with a larger atomic
radius (Cl, Br, I) [19, 20]. Halogen bonding is a noncovalent interaction resembling the hydrogen
bonding in both the properties and the way of formation but formed by the halogen atom and the
nucleophile containing a lone electron pair.
Supporting this, the introduction of electron-donating groups into an aromatic substituent, such
as -OMe, actually leads to a multifold loss of inhibitory activity (compare BPs 24 and 13). Lacbay et
al. observed the similar decrease in activity (2-6 fold) of their series of bisphosphonate inhibitors
when -OMe and -OiPr substituents possessing +M effect were introduced into the benzene moiety
(Fig. 6) [17].
2.3 Effect of the Mg2+-coordinating backbone structure on the inhibitory properties
Since the P-C-P backbone of bisphosphonate inhibitors is responsible for coordination of
magnesium cations in the catalytic center of the HIV RT [11], its structure would determine the
inhibitory potential of these compounds to a great extent. However, this topic is practically negligible
in literature; there were only two examples of P-C-P fragment modifications that lead to the decrease
of inhibitory activity more than 100 times compared with the bisphosphonate (2) (Fig. 7) [14]). It
should be noted that in [14] two modifications were introduced in the same P-C-P fragment: the OH-
group at the central carbon atom was removed and the phosphonate acidic -OH was substituted
to -CH3; therefore, we cannot determine which of the modifications has resulted in the loss of activity.
Fig. 7. BPs with the "defective" chelating group (P-C-P)
We have synthesized the bisphosphonate (10) and found out that hydroxyl group deletion at the
central carbon definitely reduces the activity in 2.8 and 1.9 times compared with (2) for PPlysis and
ATPlysis, respectively. Thus, the total loss of activity observed in [14] happens mainly due to the
replacement of P(O)(OH) to P(O)(CH3).
The introduction of the NH2-group into the P-C-P backbone (8) has practically no effect on
neither the PPlysis nor the ATPlysis when compared with its absence in the bisphosphonate (10). We
expected the amino group to be capable of coordination with Mg2+ the same manner as the hydroxyl
group, and such a moderate difference in the inhibitory properties of the hydroxy-bisphosphonate (2)
and the amino-bisphosphonate (8) required an explanation. We conducted the NMR titration of the
bisphosphonate (8) and found out that its NH2-group is protonated at pH values 7-8, used in enzymatic
+
assays. Fig. 8 demonstrates that most of amino group of (8) are in NH3 form which are unable to
5

ACCEPTED MANUSCRIPT
participate in Mg2+ coordination giving rise to only 10% increase in the activity of the NH2-introduced
P-C-P-based inhibitors (compare 10 vs 8). In general, comparing the data of BPs (13)/(21), (2)/(8) and
(14)/(25) we can conclude that the replacement of the OH group at the central carbon atom of the P-C-
P core leads to a significant loss of the inhibitory activity and that 1-aminomethylenebisphosphonates
have little if any potential for design of “phosphorolytic” RT inhibitors.
We have found some interesting correlations between the pH and the physical properties of 1-
aminobisphosphonate (8) (Fig. 8). The 7 ppm downfield shift in the 31P NMR spectrum shows electron
density change at the phosphorus atoms, that should substantially affect the chelation and, as a result,
the inhibitory potential. However, for the rest of the BPs the effect of pH on the phosphonate group
appeared to be insignificant (Fig. 8).
Fig. 8. 31Р NMR shift dependence on the рН. Dotted vertical lines indicate pH range for the HIV RT
activity studies.
2.4 Rational design on the base of studied inhibitors
SAR analysis of 31 BPs synthesized allowed us to propose the optimal structure with the
potentially effective inhibitory properties. It should be constructed of a methylenebisphosphonate
backbone bearing -OH group at the central carbon atom and an aromatic halogenated pharmacophore
linked through the 3-4 methylene fragments to the Mg2+-coordinating backbone (Fig. 9).
Fig. 9. P-C-P Backbone (green) containing OH-group (blue); halogenated benzene ring (red); linker
(grey). The geometry of the depicted bisphosphonate (Y=Cl) was optimized in vacuum using Gaussian
03 software (B3LYP functional and a 6-31G** (2p, 2d) basis set were applied)
Based on our results, the activity of BPs is more affected by the nature of an aromatic
substituent than the Mg2+-coordinating fragment structure. Therefore, we replaced OH- at the central
P-C-P carbon of of (2) with the second benzyl substituent and synthesized (1). As seen from Table 1,
the introduction of the second aromatic substituent compensated the effect of the removal of the OH-
group at the P-C-P backbone. Although the same transformation of the bisphosphonate (14) to (17)
yielded to the inhibitory activity decrease (Table 1) for both processes. This may be a result of the OH-
group replacement in (14) by the nonchlorinated benzyl substituent that possesses low affinity to RT
and cannot compensate the removal of OH- from P-C-P backbone. Anyway, BPs with two
hydrophobic aromatic groups at the central P-C-P carbon possess significant hydrophobicity and three
times reduced sorption to hydroxyapatite (data will be published later). High affinity of therapeutically
used BPs to bone matrix hydroxyapatites is one of the main reasons for their low concentration in
blood stream; that is why low affinity of the dibenzylated BPs to hydroxyapatites may be
advantageous for antiviral BPs. Thus, high hydrophobicity and low sorption on bone matrix of BPs
like in (1) and (17) should positively affect the efficiency of its penetration into the cell.
Despite the fact that dibenzylated bisphosphonate (1) has superior activity vs
hydroxymethylene one (2) it seems to be advantageous to combine two aromatic substituents and OH-
group in P-C-P backbone. One of the options is to alkylate hydroxyl at the central carbon by benzyl or
a similar aromatic substituent but there is no data describing biochemical properties of BPs bearing
alkylated OH- group. To evaluate the influence of such substitution on the inhibitory activity we
synthesized BPs (3), (4) and found out that inhibitory activity of (4) and (14) was much alike to their
6

ACCEPTED MANUSCRIPT
isosters (2) and (16), respectively. Thus, further modifications of bisphosphonate inhibitors should
combine two halogenated benzene rings and oxygen atom in the α position of the P-C-P backbone in
one structure.
2.5 Structure – activity dependence of BPs inhibition on ATP and PPi induced excision,
DNA/DNA elongation and influence of TAMs on these processes
2.5.1 Peculiarities of BPs inhibition of ATP and PPi induced excision
Most of the studied BPs suppress ATPlysis more effectively than PPlysis (Fig. 10) though the
concentration of ATP in the reaction mixture was 10 times higher than PPi (0.3 mM vs. 3.0 mM).
These concentrations were in accordance with the published data on the concentrations of ATP and
PPi in lymphoid cells [21, 22]. Figure 10 demonstrates the relative efficacy of inhibition of ATPlysis
and PPlysis catalyzed by RTwt.
Fig. 10. The differences in the efficacy of the ATPlysis and PPlysis suppression of some
ꢅꢆꢇ
bisphosphonates. The values were calculated according to the formula: ꢀ1 − ꢁꢂ ꢈ × 100%. Positive
ꢃꢄ
ꢇꢇ
ꢁꢂ
ꢃꢄ
values indicate that a BP suppresses ATPlysis more effectively than PPlysis and vice versa. The
concentration of ATP – 3 mM, PPi – 0.3 mM.
It should be noted that the inhibition efficiency for ATP and PPi induced phosphorolysis is
significantly different for some compounds (3, 6, 8, 9, 12), while the others inhibit both processes
equally well (1, 2, 5, 7, 11, 29). One of the most active inhibitors, (2), and two low-active pyridine
derivatives, (29) and (30), demonstrated unusual behavior, suppressing the PPlysis process more
effectively.
2.5.2 Impact of TAM mutations on BPs affinity to the RT catalytic domain
According to [23], TAM mutations in RT lead to high viral resistance to nucleoside-based
drugs by increasing the rate of phosphorolytic excision (the so-called “rescue activity”). All amino
acid substitutions are located close to the catalytic domain. The most important T215Y replacement
forms the new hydrophobic pocket increasing affinity of RTTAM to ATP as a result of additional π-π
stacking interactions [11]. The stacking interactions of 215Y with the aromatic ring of BPs are possible
too, and it allows us to expect the significant difference in the inhibitory activity of aromatic BPs
towards the WT and TAM-forms of HIV RT. For this reason, we conducted a comparative analysis of
the inhibition activities of the most effective inhibitors of RTwt with those of the TAM enzyme. Since
it is known that TAM do not accelerate the PPlysis [23], IC50 changes were studied only for the
ATPlysis reaction. It should be noted that several reference compounds (1, 2, 7, and 8) slightly reduced
their inhibitory activity in PPlysis, catalyzed by TAM RT (Tab. 1).
As we expected, the increased affinity of the mutant RT to ATP [11] led to a decrease in the
activity of most of the BPs. One can see in Fig. 11, that the most active BPs with aromatic substituents
lost 10-80% in their activity, the greatest loss was observed for alkoxy-bisphosphonates (3) and (4), in
3.7 and 6.4 times, respectively. In contrast, the presence of TAM had practically no impact on the
activity of BPs (1), (12) and (14), increasing the IC50 less than 15% (Fig. 11). The fact that decreasing
of inhibitory activity depends on the structure of the P-C-P backbone suggests that one of TAMs is in a
close proximity to the site where RT binds to the P-C-P backbone and somehow interacts with its
7

ACCEPTED MANUSCRIPT
central carbon surrounding. Noteworthy, inhibitors based on the alkoxy-P-C-P skeleton appeared to be
more sensitive to TAMs and significantly decreased inhibitory activity. At the same time, NH2-
containing P-C-P skeletons (BPs 8 and 21) maintained or even surpassed their inhibitory potential
towards the mutant enzyme, like highly hydrophobic inhibitors bearing two benzene rings (1 and 17).
Despite of the small selection of the compounds, the obtained correlations allow to predict the
structure of bisphosphonate inhibitors targeted to the suppression of RT forms resistant to thymidine
analogues.
Fig. 11. Impact of TAM mutations on the BP inhibitory activity. (A): IC50, of the compounds in
ATPlysis reactions catalyzed by WT/TAM RT. (B): the «prevalence» values of WT RT inhibition over
ꢉꢆ
ꢁꢂ
ꢃꢄ
TAM RT were calculated according to the formula: ꢀ1 − ꢁꢂ ꢈ × 100%. Positive values indicate that
ꢆꢅꢊ
ꢃꢄ
BPs more effectively suppress the WT enzyme than TAM and vice versa.
2.5.3 Inhibition of the DNA polymerase activity of HIV RT by the bisphosphonates
An extra goal was to study the ability of the tested bisphosphonates to inhibit DNA synthesis
catalyzed by HIV RT. Aromatic derivatives of medronic acid have already been described as inhibitors
of the DNA polymerase activity of HIV RT. For example, Song et al. described a series of 42 BPs
suppressing the ATPlysis at nanomolar concentrations but failing to suppress the DNA polymerase
activity at 5 ꢀM – a concentration at which the reference compound foscarnet inhibits the reaction
almost completely [14]. In paper [17], in contrast, bisphosphonate inhibitors were shown to inhibit
DNA polymerase activity (IC50 1-5 ꢀM), comparable with foscarnet.
Fig. 12. Inhibition of the phosphorolytic and DNA-polymerase activity by BPs.
The inhibitory properties of the most active bisphosphonates are summarized in Fig.12. As can
be seen, the ability of PPi analogues to suppress the phosphorolytic excision reaction correlates well
with their inhibition of the DNA polymerase reaction. This behavior is expected for nucleotide-
competitive inhibitors that bind to the target enzyme in the polymerase active site. Thus, represented
bisphosphonates are the binary RT inhibitors that suppress both the NRTI phosphorolytic excision and
the DNA synthesis.
3.
Conclusions
Herein, we have tested a library of 36 bisphosphonate analogues for their ability to inhibit PPi
and ATP-induced phosphorolytic activity as well as elongation reaction catalyzed by both the wild
type- and TAM HIV-1 reverse transcriptases. The four most active compounds were all halogen
containing aromatic species having IC50 values about 5-10 µM for both PPi and ATP induced excision
and 10-30 µM for elongation reactions. We found out what modifications could improve the activity of
bisphosphonates, namely, aromatic ring bearing one or two halogen atoms in the ring should be
located in a distance of three or four methylene groups apart from P-C-P backbone. For most of
bisphosphonates with IC50 <100 µM their inhibitory activity retains upon thymidine analog resistant
enzyme but BPs with α-oxygen atom demonstrated high sensitivity towards TAM that dramatically
decreased their activities.
4.
Materials and Methods
8

ACCEPTED MANUSCRIPT
The structures of 36 bisphosphonates studied are shown in the Figure 4.
All reagents used were purchased from Aldrich or Acros Organics.
Flash chromatography was performed on Kieselgel (40-63 ꢀm, Merck, Germany).
1
The inhibitor concentrations were measured by H NMR using the known concentrations of tert-
1
butanol-d1 in D2O as an internal reference standard (relaxation delay (D1) was set to 60 s for H
measurements). pH measurements was performed using FEP30 Mettler Toledo (Switzerland) pH-
meter with LE409-electrode.
High-resolution electrospray mass spectra were obtained on Applied Biosystems/MDS Sciex
QSTAR XL. NMR spectra were recorded on 400 MHz (1H), 182 MHz (31P), and 92 MHz (13C) AMX
III -400 Bruker spectrometer. Chemical shifts are reported in parts per million (ppm) using
tetramethylsilane (1H), tert-butanol-d1 (13C), and 85% H3PO4 (31P) as external standards. 13C and 31P
NMR spectra were proton-decoupled unless otherwise specified.
[γ-32P]ATP (6000 Ci/mM) was a kind gift from Dr. Ur. S. Skoblov. HIV-1 RTwt (21500 U/ml)
was purchased from “CalBioChem” (USA). Thermostable inorganic pyrophosphatase and T4
polynucleotide kinase were purchased from “New England BioLabs” (USA). All oligonucleotides
were synthesized by the phosphoramidite method on an automatic ABI 3400 DNA synthesizer
(Applied Biosystems, USA) under conditions recommended by manufacturer and purified by
electrophoresis in a 20% polyacrylamide/7 M urea gel.
4.1 Chemistry
All bisphosphonates except (3-5), (16-20), (24), (26), (30) have been synthesized and
characterized in our previous publications and all samples were from the same source [16, 24, 25].
New bisphosphonates were prepared according to following methods: (5), (20) and (24) (method D),
(17) (method E), (3), (4), (16) (method F), (19) (method A), (26) (method C), (30) by method B (for
synthesis, NMR and HRMS characteristics see [26]).
Fig. 13. Chemical methods of synthesis of BPs described in this work.
4.2
31P NMR titration study
31P proton-decoupled NMR spectra were recorded at 300K in 5 mm NMR-tube containing 600 µl of
BPs solution in H2O:D2O (98:2). Capillary with 85% H3PO4 was used as an external reference
standard.
10 ml of BPs solutions in 2% D2O with concentration of free acid BPs 10 mg/ml were thermostated at
20±0.2°C for 10 min and titrated using 1M NaOH. Aliquots (600 µl) after each NMR measurement
were returned to the vessel prior to next titration step. The increase of final volume was no more than
5-7%.
4.3
Biology
[5’-32P]-labeled primer–template complex preparation. Synthetic 42-nt oligonucleotide (5’- CCA GTT
AGC GTA GTC AAG GCT CGA GAC TAC AGG AAT TGA CGG-3’) and 19-nt oligonucleotide
(5’-CCG TCA ATT CCT GTA GTC T-3’) were used to obtain primer-template complex. The reaction
mixture (30 ꢀl) contained 70 mM Tris-HCl (pH 7.6), 10 mM MgCl2, 5mM dithiothreitol, 50 pmol of
9

ACCEPTED MANUSCRIPT
19-nt primer, 5U T4 polynucleotide kinase, and 100 ꢀCi of [γ-32P]rATP. After 1 h of incubation at
37°C, T4 polynucleotide kinase was inactivated by heating at 65°C for 20 min. To obtain the primer–
template complex, [5’-32P]-primer was annealed with a 1.5-excess 42-nt DNA template by heating at
65°C for 5 min and cooling to room temperature. This complex was purified using an illustra™
microspin G-25 column from “GE Healthcare” (UK).
4.3.1. Expression and purification of ТАМ reverse transcriptase enzyme. Plasmid p6HRT encoding
expression cassette determining synthesis of His-p66 subunit of the mutant form of HIV-1 reverse
transcriptase (TAM RT; M41L, D67N, K70R, T215Y, K219Q) and HIV-1 protease was a kind gift
from Professor S.F.J. Le Grice. Mutant p66 RT polypeptide was expressed in Rosetta (DE3) E. coli
strain with consequent processing by protease, yielding p66/p51 heterodimer. TAM RT was purified
using standard Ni-NTA agarose procedure [27]. Enzyme activity was 5500 U/ml.
4.3.2. Inhibition of PP-mediated phosphorolysis. Reaction mixture (10 ꢀl) contained 50 mM Tris-HCl
(pH 8.0), 10 mM MgCl2, 60 mM KCl, 10 nM [5’-32P]-primer–template complexes, 0.5U RT, 300 ꢀM
PPi, and the increasing concentrations of bisphosphonate inhibitors. The reactions proceeded at 37°C
for 30 min.
4.3.3. Inhibition of ATP-mediated phosphorolysis. Reaction mixture (10 ꢀl) contained 50 mM Tris-
HCl (pH 7.5), 10 mM MgCl2, 60 mM KCl, 10 nM [5’-32P]-primer–template complexes, 0.5U RT, 3
mM ATP, inorganic pyrophosphatase (0.1U), and increasing concentrations of bisphosphonate
inhibitors. The reactions proceeded at 37°C for 30 min.
4.3.4. Inhibition of RT polymerase activity. Reaction mixture (10 ꢀl) contained 50 mM Tris-HCl (pH
8.0), 10 mM MgCl2, 60 mM KCl, 10 nM [5’-32P]-primer–template complexes, 0.5U RT, 1 ꢀM of
dATP, dTTP, dGTP, and dCTP and increasing concentrations of bisphosphonate inhibitors. The
reactions proceeded at 37°C for 30 min.
4.3.5. Quantitative analysis of the BPs inhibition. The reactions containing [5’-32P]-primer–template
complexes were quenched by the addition of 5 ꢀl of stop solution, containing formamide, 10% of
EDTA and 1 mg/ml each of bromphenol blue and xylene cyanol, and subjected to electrophoresis in a
15% denaturing PAGE (2 h, 2400 V). Radioactive products were detected using Typhoon FLA 9500
biomolecular imager “GE Healthcare” (UK). The IC50 of the BPs inhibitors were calculated using
Opti-Quant software (Packard Inc., USA).
10

ACCEPTED MANUSCRIPT
Author contributions
DY and SK planned the project and designed the research, JW and EP performed the physical-
chemical experiments, MK, AK and NU performed biochemical measurements and enzymes
purification, DY and OK conducted chemical synthesis and purification, DY, SK and MK wrote the
paper, and all co-authors commented on the paper.
Acknowledgements
We thank Dr. Jukka Leppänen end Dr. Janne Weisell from the University of Eastern Finland for
providing ESI HRMS, Dr Pavel Solyev from the Engelhardt Institute of Molecular Biology RAS
(Moscow, Russia) for his thoughtful comments on the manuscript. Physicochemical measurements
were supported by the Academy of Finland (decision no. 292574) and strategic funding of UEF. All
other studies were supported by the Russian Science Foundation, project № 14-50-00060.
Abbreviations
Bisphosphonate, BP; reverse transcriptase, RT; thymidine analog mutation, TAM; 3'-azido-2', 3'-
dideoxythymidine, AZT; nucleoside analog reverse transcriptase inhibitor, NRTI; non-nucleoside
reverse transcriptase inhibitor, NNRTI; inorganic pyrophosphate, PPi; deoxynucleoside triphosphate,
dNTP; VDPA, vinylidenediphosphonic acid.
Conflict of interest
The authors declare no competing financial interest.
Highlights
•
•
•
•
•
described properties of 36 bisphosphonates as inhibitors of HIV RT;
SAR studies for 31 bisphosphonates;
rational design and synthesis of five inhibitors based on SAR;
the impact of the thymidine analog mutations (TAM) on the activity of the inhibitors;
Mg2+-coordinating group determines inhibitory potential and sensitivity to TAM.
11

ACCEPTED MANUSCRIPT
References
[1] F. Barre-Sinoussi, J. Chermann, F. Rey, M. Nugeyre, S. Chamaret, J. Gruest, C. Dauguet, C. Axler-
Blin, F. Vezinet-Brun, C. Rouzioux, W. Rozenbaum, L. Montagnier, Isolation of a T-lymphotropic
retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS), Science, 220
(1983) 868-871.
[2] R.C. Gallo, S.Z. Salahuddin, M. Popovic, G.M. Shearer, M. Kaplan, B.F. Haynes, T.J. Palker, R.
Redfield, J. Oleske, B. Safai, et al., Frequent detection and isolation of cytopathic retroviruses (HTLV-
III) from patients with AIDS and at risk for AIDS, Science, 224 (1984) 500-503.
[3] J.L. Martinez-Cajas, M.A. Wainberg, Antiretroviral therapy : optimal sequencing of therapy to
avoid resistance, Drugs, 68 (2008) 43-72.
[4] H. Huang, R. Chopra, G.L. Verdine, S.C. Harrison, Structure of a covalently trapped catalytic
complex of HIV-1 reverse transcriptase: implications for drug resistance, Science, 282 (1998) 1669-
1675.
[5] S.G. Sarafianos, K. Das, A.D. Clark, Jr., J. Ding, P.L. Boyer, S.H. Hughes, E. Arnold, Lamivudine
(3TC) resistance in HIV-1 reverse transcriptase involves steric hindrance with beta-branched amino
acids, Proc Natl Acad Sci U S A, 96 (1999) 10027-10032.
[6] K. Das, R.P. Bandwar, K.L. White, J.Y. Feng, S.G. Sarafianos, S. Tuske, X. Tu, A.D. Clark, Jr.,
P.L. Boyer, X. Hou, B.L. Gaffney, R.A. Jones, M.D. Miller, S.H. Hughes, E. Arnold, Structural basis
for the role of the K65R mutation in HIV-1 reverse transcriptase polymerization, excision antagonism,
and tenofovir resistance, J Biol Chem, 284 (2009) 35092-35100.
[7] D. Arion, N. Kaushik, S. McCormick, G. Borkow, M.A. Parniak, Phenotypic Mechanism of HIV-1
Resistance to 3‘-Azido-3‘-deoxythymidine (AZT): Increased Polymerization Processivity and
Enhanced Sensitivity to Pyrophosphate of the Mutant Viral Reverse Transcriptase†, Biochemistry, 37
(1998) 15908-15917.
[8] P.R. Meyer, S.E. Matsuura, A.G. So, W.A. Scott, Unblocking of chain-terminated primer by HIV-1
reverse transcriptase through a nucleotide-dependent mechanism, Proceedings of the National
Academy of Sciences, 95 (1998) 13471-13476.
[9] B.A. Larder, S.D. Kemp, Multiple mutations in HIV-1 reverse transcriptase confer high-level
resistance to zidovudine (AZT), Science, 246 (1989) 1155-1158.
[10] R.E. Jeeninga, W. Keulen, C. Boucher, R.W. Sanders, B. Berkhout, Evolution of AZT Resistance
in HIV-1: The 41-70 Intermediate That Is Not Observed in Vivo Has a Replication Defect, Virology,
283 (2001) 294-305.
[11] X. Tu, K. Das, Q. Han, J.D. Bauman, A.D. Clark, X. Hou, Y.V. Frenkel, B.L. Gaffney, R.A.
Jones, P.L. Boyer, S.H. Hughes, S.G. Sarafianos, E. Arnold, Structural basis of HIV-1 resistance to
AZT by excision, Nature Structural & Molecular Biology, 17 (2010) 1202-1209.
[12] L. Vrang, B. Oberg, PPi analogs as inhibitors of human T-lymphotropic virus type III reverse
transcriptase, Antimicrobial Agents and Chemotherapy, 29 (1986) 867-872.
[13] M.K. Kukhanova, N.F. Zakirova, A.V. Ivanov, L.A. Alexandrova, M.V. Jasco, A.R. Khomutov,
Hypophosphoric acid is a unique substrate of pyrophosphorolysis catalyzed by HIV-1 reverse
transcriptase, Biochemical and Biophysical Research Communications, 338 (2005) 1335-1341.
[14] Y. Song, J.M.W. Chan, Z. Tovian, A. Secrest, E. Nagy, K. Krysiak, K. Bergan, M.A. Parniak, E.
Oldfield, Bisphosphonate inhibitors of ATP-mediated HIV-1 reverse transcriptase catalyzed excision
of chain-terminating 3′-azido, 3′-deoxythymidine: A QSAR investigation, Bioorganic & Medicinal
Chemistry, 16 (2008) 8959-8967.
[15] C. Cruchaga, E. Anso, A. Rouzaut, J.J. Martinez-Irujo, Selective Excision of Chain-terminating
Nucleotides by HIV-1 Reverse Transcriptase with Phosphonoformate as Substrate, Journal of
Biological Chemistry, 281 (2006) 27744-27752.
[16] J. Agapkina, D. Yanvarev, A. Anisenko, S. Korolev, J. Vepsalainen, S. Kochetkov, M. Gottikh,
Specific features of HIV-1 integrase inhibition by bisphosphonate derivatives, Eur J Med Chem, 73
(2014) 73-82.
12

ACCEPTED MANUSCRIPT
[17] C.M. Lacbay, J. Mancuso, Y.-S. Lin, N. Bennett, M. Götte, Y.S. Tsantrizos, Modular Assembly
of Purine-like Bisphosphonates as Inhibitors of HIV-1 Reverse Transcriptase, Journal of Medicinal
Chemistry, 57 (2014) 7435-7449.
[18] W.K. Hagmann, The many roles for fluorine in medicinal chemistry, J Med Chem, 51 (2008)
4359-4369.
[19] L.A. Hardegger, B. Kuhn, B. Spinnler, L. Anselm, R. Ecabert, M. Stihle, B. Gsell, R. Thoma, J.
Diez, J. Benz, J.M. Plancher, G. Hartmann, Y. Isshiki, K. Morikami, N. Shimma, W. Haap, D.W.
Banner, F. Diederich, Halogen bonding at the active sites of human cathepsin L and MEK1 kinase:
efficient interactions in different environments, ChemMedChem, 6 (2011) 2048-2054.
[20] R. Wilcken, M.O. Zimmermann, A. Lange, A.C. Joerger, F.M. Boeckler, Principles and
applications of halogen bonding in medicinal chemistry and chemical biology, J Med Chem, 56 (2013)
1363-1388.
[21] B.A. Barshop, D.T. Adamson, D.C. Vellom, F. Rosen, B.L. Epstein, J.E. Seegmiller, Luminescent
immobilized enzyme test systems for inorganic pyrophosphate: assays using firefly luciferase and
nicotinamide-mononucleotide adenylyl transferase or adenosine-5'-triphosphate sulfurylase, Anal
Biochem, 197 (1991) 266-272.
[22] A. Mas, B.M. Vazquez-Alvarez, E. Domingo, L. Menendez-Arias, Multidrug-resistant HIV-1
reverse transcriptase: involvement of ribonucleotide-dependent phosphorolysis in cross-resistance to
nucleoside analogue inhibitors, J Mol Biol, 323 (2002) 181-197.
[23] A.S. Ray, E. Murakami, A. Basavapathruni, J.A. Vaccaro, D. Ulrich, C.K. Chu, R.F. Schinazi,
K.S. Anderson, Probing the molecular mechanisms of AZT drug resistance mediated by HIV-1 reverse
transcriptase using a transient kinetic analysis, Biochemistry, 42 (2003) 8831-8841.
[24] A. Anisenko, J. Agapkina, T. Zatsepin, D. Yanvarev, M. Gottikh, A new fluorometric assay for
the study of DNA-binding and 3'-processing activities of retroviral integrases and its use for screening
of HIV-1 integrase inhibitors, Biochimie, 94 (2012) 2382-2390.
[25] D.V. Yanvarev, A.N. Korovina, N.N. Usanov, S.N. Kochetkov, Non-hydrolysable analogs of
inorganic pyrophosphate as inhibitors of hepatitis C virus RNA-dependent RNA-polymerase, Bioorg
Khim, 38 (2012) 257-262.
[26] D.V. Yanvarev, A.N. Korovina, N.N. Usanov, O.A. Khomich, J. Vepsäläinen, E. Puljula, M.K.
Kukhanova, S.N. Kochetkov, Data on synthesis of methylene bisphosphonates and screening of their
inhibitory activity towards HIV reverse transcriptase, Data in Brief, in press (2016).
[27] S.F. Le Grice, F. Gruninger-Leitch, Rapid purification of homodimer and heterodimer HIV-1
reverse transcriptase by metal chelate affinity chromatography, Eur J Biochem, 187 (1990) 307-314.
13

ACCEPTED MANUSCRIPT
Table 1
WT
5x TAM WT
BPs
PPlysis, IC50,
µM
ATPlysis,
IC50, µM
4.6±0.1
PPlysis,
IC50, µM
7.2±0.8
17.9±1.1
ATPlysis,
IC50, µM
5.2±0.4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
5.0±0.2
6.0±0.1
11.0±0.9
11.5±1.2
12.8±0.9
12.9±0.4
13.6±3.1
14.2±1.9
16.2±2.5
16.6±2.0
17.8±4.9
26.0±3.1
71.1±6.8
84.6±10.0
102±11
103±6
9.1±0.3
6.5±0.2
5.8±0.2
7.2±0.3
11.0±0.8
7.7±0.7
12.8±1.1
10.2±0.3
9.3±0.9
12.4±1.3
15.4±1.0
19.0±1.9
44.0±2.1
64.7±4.9
53.7±6.0
64.0±5.5
70.4±5.0
165±8
---
---
---
21.7±3.2
46.1±4.8
18.3±1.8
12.0±0.9
20.0±2.9
2.6±0.2
16.7±3.3
18.6±2.0
25.4±4.1
21.2±3.0
59.4±4.7
73.3±5.4
84.9±6.9
720±28
---
18.1±0.9
20.4±2.8
---
---
---
---
---
---
---
---
---
104±9
205±7
228±9
298±11
78.4±3.7
427±24
185±12
---
---
---
150±7
436±20
208±32
14

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT