Structural modulation study of inhibitory compounds for ribonuclease H activity of human immunodeficiency virus type 1 reverse transcriptase

Article abstract of DOI:10.1248/cpb.60.764

Reverse transcriptase of human immunodeficiency virus type 1 (HIV-1) has two enzymatic functions. One of the functions is ribonuclease (RNase) H activity concerning the digestion of only RNA of RNA/DNA hybrid. The RNase H activity is an attractive target for a new class of anti-HIV drugs because no approved inhibitor is available now. In our previous studies, an agent bearing 5-nitro-furan-2-carboxylic acid ester core was found from chemical screening and dozens of the derivatives were synthesized to improve compound potency. In this work, some parts of the chemical structure were modulated to deepen our understanding of the structure-activity relationship of the analogous compounds. Several derivatives having nitro-furan-phenyl-ester skeleton were shown to be potent RNase H inhibitors. Attaching methoxy-carbonyl and methoxy groups to the phenyl ring increased the inhibitory potency. No significant cytotoxicity was observed for these active derivatives. In contrast, the derivatives having nitro-furan-benzyl-ester skeleton showed modest inhibitory activities regardless of attaching diverse kinds of functional groups to the benzyl ring. Both the modulation of the 5-nitro-furan-2-carboxylic moiety and the conversion of the ester linkage resulted in a drastic decrease in inhibitory potency. These findings are informative for designing potent inhibitors of RNase H enzymatic activity of HIV-1.

Full text of DOI:10.1248/cpb.60.764

764
Regular Article
Chem. Pharm. Bull. 60(6) 764–771 (2012)
Vol. 60, No. 6
Structural Modulation Study of Inhibitory Compounds for Ribonuclease
H Activity of Human Immunodeficiency Virus Type 1 Reverse
Transcriptase
Hiroshi Yanagita,a Satoshi Fudo,a Emiko Urano,b Reiko Ichikawa,b Masakazu Ogata,a
Mizuho Yokota,a Tsutomu Murakami,b Honggui Wu,b,c Joe Chiba,c Jun Komano,b and
,a
Tyuji Hoshino*
a Graduate School of Pharmaceutical Sciences, Chiba University; 1–8–1 Inohana, Chuo-ku, Chiba 260–8675, Japan:
b AIDS Research Center, National Institute of Infectious Diseases; 1–23–1 Toyama, Shinjuku-ku, Tokyo 162–8640,
c
Japan: and Faculty of Industrial Science and Technology, Tokyo University of Science; 2641 Yamazaki, Noda, Chiba
278–8510, Japan. Received February 23, 2012; accepted April 4, 2012
Reverse transcriptase of human immunodeficiency virus type 1 (HIV-1) has two enzymatic functions.
One of the functions is ribonuclease (RNase) H activity concerning the digestion of only RNA of RNA/
DNA hybrid. The RNase H activity is an attractive target for a new class of anti-HIV drugs because no ap-
proved inhibitor is available now. In our previous studies, an agent bearing 5-nitro-furan-2-carboxylic acid
ester core was found from chemical screening and dozens of the derivatives were synthesized to improve
compound potency. In this work, some parts of the chemical structure were modulated to deepen our under-
standing of the structure–activity relationship of the analogous compounds. Several derivatives having nitro-
furan-phenyl-ester skeleton were shown to be potent RNase H inhibitors. Attaching methoxy-carbonyl and
methoxy groups to the phenyl ring increased the inhibitory potency. No significant cytotoxicity was observed
for these active derivatives. In contrast, the derivatives having nitro-furan-benzyl-ester skeleton showed
modest inhibitory activities regardless of attaching diverse kinds of functional groups to the benzyl ring.
Both the modulation of the 5-nitro-furan-2-carboxylic moiety and the conversion of the ester linkage resulted
in a drastic decrease in inhibitory potency. These findings are informative for designing potent inhibitors of
RNase H enzymatic activity of HIV-1.
Key words antiviral drug; ribonuclease H enzymatic activity; nitro-furan-phenyl-ester; human immunodefi-
ciency virus type 1 reverse transcriptase; inhibitor
Human immunodeficiency virus type 1 (HIV-1) reverse for catalysis.11–13) The RNase H dual metal mechanism was
transcriptase (RT) is a multi-functional enzyme that facilitates suggested from high resolution co-crystal structures of Ba-
both polymerase and ribonuclease (RNase) H activities and cillus halodurans RNase H with RNA/DNA hybrids at dif-
converts the single-stranded viral RNA into a double-stranded ferent stages along the reaction pathway of phosphodiester
DNA. There exist two active sites in HIV-1 RT responsible cleavage.14,15) The active site contains four carboxyl residues,
for the respective enzymatic functions. Currently, two classes creating an environment capable of holding two metal ions.
of RT inhibitors, nucleoside reverse transcriptase inhibitors It has been assumed that many RNase H inhibitors bind to
(NRTIs) and non-nucleoside reverse transcriptase inhibitors the catalytic center interacting with two divalent metal ions
(NNRTIs), are used clinically. The formers compete with the simultaneously.
natural deoxyribonucleotide triphosphate (dNTP) for nucleo-
Diketo acids are known to work as potent inhibitors for
side incorporation and act as chain terminators after incor- divalent metal-related enzymes.16) Therefore, diketo acid
poration.1) The latter agents are bound to an adjacent location structure has served as a starting point for the design and
from the polymerase active site and block RT polymerase optimization of inhibitors of HIV-1 integrase or influenza
function.2) Both of these inhibitors are targeting polymerase endonuclease. Pyrimidinol is another typical agent bearing
activity of RT. In contrast, no approved inhibitor is available a scaffold called N-hydroxyimide.17) N-Hydroxyimides were
for RNase H activity, although there have been several reports firstly described as inhibitors of influenza endonuclease, but
on the inhibitors that target the RNase H activity of HIV-1 they also show a high potency in biochemical assays of HIV-1
RT.3–5) The role of RNase H activity in the reverse transcrip- RNase H. A natural product β-thujaplicinol is another scaf-
tion process is to remove the viral genomic RNA during the fold and shows a high inhibitory potency for HIV-1 RNase H
synthesis of double-stranded DNA.6,7) Agents targeting RNase activity.18)
H function is expected to be complimentary to the currently
From an in vitro screening using 20000 chemical com-
standard chemotherapy. Hence, RT-associated RNase H activ- pounds, we found chemicals that blocked HIV-1 RT-associated
ity is one of the attractive targets for developing a novel class RNase H activity.19) The agents bearing the 5-nitro-furan-
of antiviral drugs. Furthermore, the potential for dual inhibi- 2-carboxylic acid ester moiety turned out to work as an RNase
tion of RNase H activity and integrase activity of HIV-1 has H inhibitor. Two of the agents were capable of suppressing
been examined because of the structural similarity of their HIV-1 replication in tissue culture. On the basis of the hit
catalytic sites.8–10)
chemicals found in the screening, more than 50 derivatives
HIV-1 RNase H is known to utilize two divalent metals of 5-nitro-furan-2-carboxylic acid were synthesized.20) Inhibi-
tory potency of RNase H enzymatic activity was measured
The authors declare no conflict of interest.
in a biochemical assay. Several derivatives showed higher
*To whom correspondence should be addressed. e-mail: hoshino@chiba-u.jp
© 2012 The Pharmaceutical Society of Japan

June 2012
765
°
inhibitory activities than those of the hit chemicals. Modula- of 10µL at 37 C. Fluorescence at 488nm was monitored every
tion of the 5-nitro-furan-2-carboxylic moiety resulted in a 150s using a multimode detector.
decrease in inhibitory potency. In contrast, many derivatives
HIV-1 RT was expressed in Escherichia coli and puri-
with modulation of other parts maintained inhibitory ac- fied by using a HiTrap Ni affinity column. The purified RT
tivities. These studies indicate that the nitro-furan-carboxylic was dialyzed to reduce the concentration of imidazole from
moiety is one of the potent scaffolds for RNase H inhibitor.
the elution buffer and then incubated with human rino virus
In this study, we further synthesized chemical compounds (HRV) 3C protease to cleave an N-terminal hexahistidine tag.
bearing the 5-nitro-furan-2-carboxylic acid ester moiety and The protein was further purified by nickel-coordinated nitrilo-
examined the potency for anti-HIV drugs blocking RT-associ- triacetic acid (Ni-NTA) to remove the uncleaved protein and
ated RNase H enzymatic activity. The potency of the synthe- HRV 3C protease. The RT was dialyzed against a buffer of
sized compounds was evaluated through the measurement of 50mM Tris–HCl at pH 7.5 and 200mM NaCl and was stored at
°
inhibitory activity with real-time monitoring of fluorescence −20 C with adding 50% (v/v) glycerol.
emission from the digested RNA substrate. In addition, the Assessment of Cytotoxicity 3-(4,5-Dimethylthiazol-
cytotoxicity of these compounds was assessed in 293T cells. 2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was
Computer simulation with molecular dynamics (MD) method carried out with 293T cell line. First, 100µL medium (RPMI-
was also performed to analyze the stability of the binding 1640) supplemented with 10% fetal bovine serum (FBS)
structure of an active compound.
containing 2% dimethylsulfoxide (DMSO) was loaded in a 96-
well plate, and 200µL medium with 10% FBS and 2% DMSO
containing test compounds at a concentration of 200µM was
Experimental
Organic Synthesis Compound 1 was synthesized by cre- added to the wells in the first column of the plate. Different
ating an ester linkage between a nitro-furan carboxylic acid concentrations of compound were prepared for the second,
and an α-chloro-amide bound with benzyl and penthyl groups, third and fourth columns. The final concentrations of these
°
by 3h reaction at 60 C in dimethylformamide (DMF) in the columns were 100, 50, 25 and 12.5µM, respectively. Second,
presence of dimethyl-aminopyridine (DMAP). Chemical mod- 100µL 293T cells at a concentration of 2×105/mL were added
ulation was performed for the nitro-furan moiety, with chang- to the respective wells. The final concentration of DMSO in
°
ing the starting block from furan to thiophene or pyrrole etc. each well was 1%. Third, cells were incubated for 3d at 37 C
2–8. These compounds 2–8 were synthesized in the similar with 5% CO2 atmosphere. A hundred micro liter of superna-
manner to compound 1. The derivatives bearing 5-nitro-furan- tant was removed from the cultured medium and 15µL MTT
2-ester scaffold, compounds 9–27, were prepared by the reac- reagent for dye solution was added to each well and the cells
tion of converting 5-nitro-2-furoic acid into an acid chloride were incubated for 1h. Then, 100µL solution of stop mix was
°
with thionyl chloride, followed by the nucleophilic substitution added, and the cells were incubated overnight at 4 C to suf-
reaction in the presence of NEt3 in tetrahydrofuran (THF) ficiently dissolve the dye. Finally, intensity of OD570/690 was
°
with setting the temperature at 0 C for the initial 30min. and measured by a spectrofluorometer.
elevated it to r.t. afterward. Since a hydroxy group bound to
Molecular Dynamics Simulation A computational model
phenyl ring is more reactive than a hydroxy group bound to of HIV-1 RT domain was constructed from an X-ray crystal
alkane, the substitution reaction dominantly produced phenyl- structure with Protein Data Bank code 3QIO.23) Atom coordi-
ester linkage (9–16) when a nucleophilic reagent contained nates for the missing residues were generated by using Mod-
two hydroxy groups. When a nucleophilic reagent contained eller9.9.24) According to the results of the recent X-ray crystal-
only one hydroxy group, the substitution reaction generated lographic studies on the complex of RNase H domain and its
alkyl-ester linkage (17–27). Compounds 28–30 were produced inhibitors,25–27) the RNase H domain contains two divalent
by generating nitro-furan-carbonyl-alkyl-benzene through the metal ions at the center of the active site. Two Mn2+ ions in
reaction of Weinreb amides containing benzyl group with the crystal structure were replaced with Mg2+ ions. The pro-
alkyl lithium, followed by incorporation of nitro group into tonation states of all of the ionazable residues were predicted
the phenyl ring using white fuming nitric acid and acetic by ProPKa program28) in the presence of two Mg2+ ions at
anhydride. Compounds 29 and 30 were separated by flash the active site. Atom charges of the compounds were deter-
chromatography. Compounds 31–33 were generated by using mined from the electrostatic potential obtained from quantum
hydroxy-amines as nucleophilic reagents.
chemical calculations, followed by the restrained electrostatic
Evaluation of Inhibitory Activity The 50% inhibitory potential (RESP) fitting29) in a similar manner to the previ-
concentration (IC50) of the synthesized compounds for RT- ous studies.30–33) The atom charge for Mg2+ ion was setting to
associated RNase H activity was determined from the chemi- 1.54, which was also determined by the RESP method based
cal concentration reducing the rate for substrate cleavage on the calculated electrostatic potential obtained by QM/MM
reaction to half relative to the control. A real-time monitoring technique carried out in a similar manner to the previous
assay was employed to estimate the IC50.21,22) In short, two work.20) An active compound was combined with HIV-1 RT
oligo-nucleotides were annealed at final concentrations of domain, referring the binding structure predicted in our previ-
2.5 and 0.25µM for substrate. One was oligo-ribonucleotide ous work.20) The compound-bound RT model was placed in a
5′-GAUCUGAGCCUGGGAGCU-3′ with 6-carboxy-fluoros- rectangular box and solvated with TIP3P water molecules,34)
cein (FAM) conjugated at the 3′ end, and the other was oligo- with all of the crystal water molecules remaining. Periodic
deoxyribonucleotide 5′-AGCTCCCAGGCTCAGATC-3′ with boundary conditions were applied to avoid the edge effect in
black hole quencher (BHQ) conjugated at the 5′ end. Enzyme all calculations.
reaction with 100ng RT, 0.025µM oligo-ribonucleotide, and
Minimizations and MD simulations were carried out using
0.25µM oligo-deoxyribonucleotide was carried out in a volume sander module of AMBER9.35) The modified ff03 force field36)

766
Vol. 60, No. 6
was used as the parameters for molecular dynamics. The cut-
off distance for the long range electrostatic and van der Waals
Results
Eight analogues of 5-nitro-furan-2-carboxylic acid ester
energy terms was set to 12.0Å. The expansion and shrinkage were synthesized by converting the 5-nitro-furan moiety
of all covalent bonds connecting to hydrogen atom were con- into other functional groups and examined for their RNase
strained using the SHAKE algorithm.37) Energy minimization H inhibitory activities (Table 1). Compound 1 has a typical
was achieved in three steps. Initially, movement was allowed chemical structure showing an inhibitory potency for HIV-1
only for water molecules. Next, compound and divalent metal RNase H enzymatic activity. This compound bears nitro-furan
ions were allowed to move in addition to the water molecules. ester core connecting to the pentyl- and benzyl-bound amide
Finally, all atoms were allowed to move freely. In each step, group. Replacement of furan with thiophene largely decreased
energy minimization was executed by the steepest descent compound potency (2, 3). Conversion of furan into pyrrole
method for the first 10000 cycles and the conjugated gradient also resulted in complete loss of compound potency (4). At-
method for the subsequent 10000 cycles. After a 0.1ns heating taching a halogen to the 4th position of furan exhibited a
calculation until 310K using the NVT ensemble condition, a slight increase of inhibitory activity (5), while a hydrophobic
20ns equilibrating calculation was executed at 1.0atm and at or aromatic substitute resulted in loss of inhibitory potency (6,
310K under the NPT ensemble condition, with an integration 7). Introduction of even a small hydrophobic group at the 3rd
time step of 2.0fs.
position of furan decreased inhibitory activity (8).
Table 1. RNase H Inhibitory Activity and Cytotoxicity of the Derivatives Modulated at Nitro-Furan Moiety
Compound
Structure
IC50 (µM)
CC50 (µM)
1
8.4
74
2
>50
49
3
4
5
6
30.3
>50
6.6
50
86
3
>50
>100
7
8
>50
>100
25.7
76

June 2012
767
Eight derivatives bearing 5-nitro-furan phenyl ester core with carbonyl group as shown in Table 4. Ester linkage is dis-
were examined as shown in Table 2. A similar degree of advantageous for medicine because esterase digests the link-
inhibitory activity was observed for the analogs having a age and the drug concentration in a body decreases rapidly.
hydroxy group at the meta or para position of phenyl ring (9– Nitro-phenyl group was connected to the carbonyl carbon with
11). While substitution of the hydroxy group with acetyl group changing the length of alkyl chain and the position of nitro
showed a similar degree of inhibitory activity (12), substitu- group (28–30). None of the derivatives showed noticeable in-
tion with ethyl-ester increased compound potency (13). Fur- hibitory activity. The ester linkage was replaced by an amide
ther, introduction of methoxy group at the ortho position also bond (31–33), in which hydroxy-methyl-benzyl, acetyl-methyl-
increased compound potency (14). The compound containing benzyl, or bromo-methyl-benzyl was bound via a carbamoyl
methoxy-carbonyl and methoxy groups at the para or ortho group. This modification also resulted in complete loss of
positions exhibited a fine inhibitory activity (15). Introduction compound potency.
of phenyl-methyl-amine at the para position also maintained
compound potency (16).
Cytotoxicity was little or undetectable for most of the
derivatives modulated at the nitro-furan moiety except for
Eleven derivatives bearing 5-nitro-furan ester core bound compound 5 (Table 1). No detectable cytotoxicity was also ob-
with benzyl-based substitutes were investigated as summa- served for most of the derivatives bearing nitro-furan-phenyl
rized in Table 3. Connection of a benzyl group without any ester core (Table 2). It should be noted that highly active
additional functional substitute showed moderate inhibitory compounds, 14 and 15, showed no noticeable cytotoxicity at a
activity (17). Attaching nitro group, methoxy-ether, or tert- concentration of 100µM. In contrast, the analogs bearing ester
butoxy-ester exhibited slight changes in compound potency core bound with benzyl-based substitute showed some degree
(18–20). Extension of alkyl chain caused no significant differ- of cytotoxicity (Table 3). Further, the analogs converted the
ence in inhibitory potency (21). However, addition of hydroxy ester into amide or ketone showed cytotoxicity in which CC50
group decreased inhibitory activity (22). Introduction of me- ranged from 9 to 32µM (Table 4). Overall, in the measurement
thoxy group further decreased compound potency regardless using 293T cells, half of the synthesized compounds showed
of the position of the group bound to phenyl ring (23–25). noticeable cytotoxicity but almost all of the toxicity-detected
While connection of two methoxy groups improved compound compounds showed little RNase H inhibitory activity. Accord-
potency (26), replacement of two methoxy groups by chlorides ingly, the results of this assay suggest that active compounds
showed low inhibitory activity (27).
have no significant cytotoxicity and the nitro-furan-phenyl
Six derivatives were synthesized by changing the ester bond ester skeleton is the most favorable among them from a cyto-
toxic viewpoint.
In order to examine the stability of the binding structure of
a potent compound inside the active site of the target protein,
MD simulation was performed for the complex of a synthe-
sized derivative 15 and RNase H domain. MD simulation was
carried out for 20ns and root mean square deviation (RMSD)
relative to the structure after heating was calculated as shown
in Fig. S1 of the supplemental information. The RMSD value
showed a gradual increase up to 5ns and scarcely changed
during later 15ns. Accordingly, the binding conformation of
the complex was judged to be equilibrated. In order to extract
the snapshot structure representing a typical binding mode of
compound 15 and RNase H domain, the averaged structure
was obtained using 2500 trajectories from the last 5ns of MD
simulation. The RMSD between each trajectory and the aver-
age structure was calculated, and then one trajectory with the
smallest RMSD value was determined to be the typical com-
plex structure shown in Fig. 1a.
Table 2. RNase H Inhibitory Activity and Cytotoxicity of the Derivatives
Bearing Phenyl Ester Core
Compound
–O–R
IC50 (µM)
CC50 (µM)
>100
9
7.2
10
8.2
24
11
12
9.1
8.7
>100
Two Mg2+ ions were held by the side chains of four acid
residues; Asp443, Glu478, Asp498, Asp549, and the com-
pound was stably bound to the active site without changing
the binding mode during 20ns simulation. In the binding
structure, the nitro-furan-ester core is especially, stably bound
to two Mg2+ ions as shown in Fig. 1a. A stereo view of the
binding structure is presented in Fig. S2 of the supplemental
information. The oxygen atom on the furan ring is oriented
toward the divalent metals. Nitro oxygen and carbonyl oxygen
of the ester are strongly coordinated to the respective Mg2+
ions. Therefore, a large ring-shaped configuration of –Mg–O–
N–C–O–C–C–O–Mg– is formed. The inter-atomic distance
between two Mg2+ ions is 3.8Å. The distance between the ni-
tro oxygen atom and one Mg2+ ion is 1.9Å, and that between
carbonyl oxygen and the other Mg2+ ion is also 1.9Å. Figure
49
13
14
3.6
3.1
>100
>100
15
16
1.4
3.8
>100
>100

768
Vol. 60, No. 6
1b shows the changes of these two distances for the last 5ns of
MD simulation. These graphs clearly indicate that the interac-
tion of oxygen atom and Mg2+ ion is quite strong and that the
oxygen–Mg interaction is essentially important for the binding
of potent compound.
Methoxy-carbonyl and methoxy groups connected to phenyl
stick out from the binding pocket. The distance between me-
thoxy oxygen and hydrogen atom of amino group of Asn474
was monitored as shown in Fig. 1c. The distance largely
fluctuates during the simulation and the interaction is not so
steady. Hence, there is much room for improvement in this
region. The oxygen atom at the ester bond has noticeable
interactions with the amide group of main chain of Gln500.
The change of the distance between the ester oxygen and the
hydrogen atom of the amide group is shown in Fig. 1c. No
abrupt, large change is observed in the distance. Therefore,
the interaction contributes to stabilizing the binding of the
potent compound. The distance between the ester oxygen and
the Cα hydrogen atom of Ser 499 was also monitored. The
distance shows no large change for the last 5ns of simulation,
which suggests the stability of the binding of the ester part
with RNase H domain. Consequently, it is confirmed from the
MD simulation that an active compound 15 is stably bound to
the active site of the RNase H domain with coordinating to
two divalent metal ions and making supportive interaction at
the ester part.
Discussion
According to the data summarized in Table 3, it is sug-
gested that various kinds of functional groups connected to
the benzyl ring have little interaction with the RNase H do-
main and that the functional group-binding region is located
outside the binding pocket and is exposed to the solvent. Fur-
ther, a comparison of inhibitory activity between compounds
17–20 and compounds 21–27 suggests that the length of alkyl
chain connecting to phenyl ring has a significant influence
on the difference in compound potency. That is, the longer
alkyl chain in compounds 21–27 is less favorable in terms of
both inhibitory activity and cytotoxicity. This indicates that a
strategy to increase the inhibitory activity is to position the
substitute closer to the nitro-furan group. The conversion may
allow the aromatic ring or substitute to interact with the target
protein inside the binding pocket.
A comparison of inhibitory activities of compounds in
Tables 2 and 3 indicates that the introduction of phenyl-ester
connected to nitro-furan shows higher compound potency than
that of benzyl-ester. This is consistent with the findings de-
scribed above paragraph and supports the notion that the com-
pound potency increases when the position of the substitute
connected to the ester linkage is closer to the nitro-furan. The
incorporation of polarized substitutes like methoxy, hydroxy,
methoxy-carbonyl, or ethoxy-carbonyl group is effective to
increase the compound potency. In particular, the introduction
of methoxy group at the ortho-position of phenyl ring effec-
tively increases the inhibitory activity.
All the compounds in Tables 2 and 3 have the ester linkage.
The data summarized in Table 4 clearly indicate that conver-
sion of the ester linkage into carbonyl and/or amide bond
results in loss of inhibitory potency. Both bonds will be likely
to make the chemical to be in a straight configuration. If a
compound has a straight form, the side part of the compound
Fig. 1. (a) Binding Structure of an Active Compound to the RNase H
Domain, Obtained by MD Simulation
Compound and several polar residues are shown in stick representation. Two
Mg2+ ions are denoted by spheres. (b) Changes in the distances between nitro
oxygen and Mg2+ ion (d1: red solid line) and between carbonyl oxygen and another
Mg2+ ion (d2: blue dotted line) for the last 5ns of MD simulation. (c) Changes in
the distances between methoxy oxygen and amino hydrogen of Asn474 (d3: green
solid line), between ester oxygen and amide hydrogen of main chain of Gln500 (d4:
cyan broken line), and between ester oxygen and Cα hydrogen of Ser499 (d5: yel-
low dotted line) for the last 5ns of MD simulation. Color images were converted
into gray scale; red solid line, blue dotted line, green solid line, cyan broken line,
yellow dotted line have been converted into gray scale.

June 2012
769
Table 3. RNase H Inhibitory Activity and Cytotoxicity of the Derivatives
Bearing Ester Core Bound with Benzyl-Based Substitute
Table 4. RNase H Inhibitory Activity and Cytotoxicity of the Derivatives
Converted the Ester Bond into Other Kinds of Linkages
Compound
–O–R
IC50 (µM)
CC50 (µM)
–CH2–R or
–NH–R
Compound
IC50 (µM)
>50
CC50 (µM)
17
5.4
38
28
11
18
19
20
21
5.0
5.8
4.5
5.1
>100
75
29
30
>50
>50
17
9
>100
57
31
32
33
>50
>50
>50
29
27
32
22
7.9
39
23
24
25
14.2
18.6
12.5
42
36
36
RNase H reaction site to exert its enzymatic activity.40)
A
theoretical study by De Vivo et al. suggested that the presence
of two divalent metal ions is essential for RNase H activ-
ity and that two metal ions act cooperatively with facilitat-
ing the binding of a substrate and catalyzing the enzymatic
reaction.41) This theoretical finding strongly suggests double
coordination of divalent metal ions at the RNase H domain.
Crystal structures on the complex of the RNase H domain
and its inhibitor were successively reported from three differ-
ent research groups.23,25–27) All of the crystal structures ever
reported showed the presence of two metal ions at the active
site. One of the divalent metal ions was held deep inside the
binding pocket of the RT RNase H domain with making co-
26
27
8.5
51
64
17.4
will collide with the inside wall of the binding pocket of the ordination bonds to three carboxyl groups of Asp443, Glu478
RNase H domain. Accordingly, compounds are hardly com- and Asp498. The other was fixed with making coordination
bined with the binding pocket.
bonds to two carboxyl groups of Asp443 and Asp549. The
The conversion of the nitro-furan group into pyrrole dras- distance between two metal ions was about 4Å. Every inhibi-
tically decreases the inhibitory activity while conversion to tor in crystal structures was revealed to have a similar binding
nitro-thiophen maintains the activity (Table 1). This indicates mode. That is, inhibitors are stabilized with forming coordina-
that a nitro-furan or nitro-thiophene core is indispensable for tion bonds to both metal ions. Accordingly, it is highly proba-
inhibitory potency. The characteristic property of nitro-furan ble that the chemical compounds showing RNase H inhibitory
is its large electric polarity. Oxygen atoms are negatively activity examined in this study are also coordinated to two
charged and these oxygen atoms will be coordinated to diva- divalent metal ions. Hence, negatively charged oxygen atoms
lent metal ions at the RNase H active site. The attachment of of the nitro group, furan, and carbonyl group are aligned in a
non-polar substitute to furan results in decrease of compound straight form. This negatively charged region will be attached
potency. Accordingly, the 3rd and 4th positions of furan be- to the divalent metal ions.
come close to the residues inside the RNase H active site.
The binding structure deduced from MD simulation indi-
RNase H of HIV-1 exerts its enzymatic activity by incor- cates that ether oxygen at the ester bond has an interaction
porating divalent metal ions at the reaction site.38,39) It had with a polar residue, Ser499, which is located at the deep
been controversial how many metal ions were required at the inside of RNase H domain. This residue would have little

770
Vol. 60, No. 6
(1991).
influence on the function of RNase H. Therefore, one of the
designs to improve inhibitory activity is to modify the com-
pound to bear some polar functional group that enables a
strong interaction with Ser499. In order to enhance the bind-
ing affinity of the compounds with RNase H active site, the
incorporation of a polar functional group bound to phenyl
ring is one of the possible conversions of our derivatives. The
distance between methoxy group and the amine of Asn474
largely fluctuated during MD simulation. If the interaction
with Asn474 is enforced, compound will be more stably com-
bined with the RNase H domain.
7) Telesnitsky A., Goff S. P., EMBO J., 12, 4433–4438 (1993).
8) Borkow G., Fletcher R. S., Barnard J., Arion D., Motakis D., Dmit-
rienko G. I., Parniak M. A., Biochemistry, 36, 3179–3185 (1997).
9) Tramontano E., Esposito F., Badas R., Di Santo R., Costi R., La
Colla P., Antiviral Res., 65, 117–124 (2005).
10) Tarrago-Litvak L., Andreola M. L., Fournier M., Nevinsky G.
A., Parissi V., de Soultrait V. R., Litvak S., Curr. Pharm. Des., 8,
595–614 (2002).
11) Cowan J. A., Ohyama T., Howard K., Rausch J. W., Cowan S. M.,
Le Grice S. F., J. Biol. Inorg. Chem., 5, 67–74 (2000).
12) Haren L., Ton-Hoang B., Chandler M., Annu. Rev. Microbiol., 53,
245–281 (1999).
13) Steitz T. A., Smerdon S. J., Jäger J., Joyce C. M., Science, 266,
2022–2025 (1994).
14) Nowotny M., Gaidamakov S. A., Crouch R. J., Yang W., Cell, 121,
1005–1016 (2005).
15) Nowotny M., Gaidamakov S. A., Crouch R. J., Yang W., EMBO J.,
25, 1924–1933 (2006).
16) Shaw-Reid C. A., Munshi V., Graham P., Wolfe A., Witmer M.,
Danzeisen R., Olsen D. B., Carroll S. S., Embrey M., Wai J. S.,
Miller M. D., Cole J. L., Hazuda D. J., J. Biochem., 278, 2777–2780
(2003).
17) Hang J. Q., Rajendran S., Yang Y., Li Y., In P. W. K., Overton H.,
Parkes K. E. B., Cammack N., Martin J. A., Klumpp K., Biochem.
Biophys. Res. Commun., 317, 321–329 (2004).
18) Budihas S. R., Gorshkova I., Gaidamakov S., Wamiru A., Bona M.
K., Parniak M. A., Crouch R. J., McMahon J. B., Beutler J. A., Le
Grice S. F., Nucleic Acids Res., 33, 1249–1256 (2005).
19) Fuji H., Urano E., Futahashi Y., Hamatake M., Tatsumi J., Hoshino
T., Morikawa Y., Yamamoto N., Komano J., J. Med. Chem., 52,
1380–1387 (2009).
20) Yanagita H., Urano E., Matsumoto K., Ichikawa R., Takaesu Y.,
Ogata M., Murakami T., Wu H. G., Chiba J., Komano J., Hoshino
T., Bioorg. Med. Chem., 19, 816–825 (2011).
21) Chan K. C., Budihas S. R., Le Grice S. F., Parniak M. A., Crouch
R. J., Gaidamakov S. A., Isaaq H. J., Wamiru A., McMahon J. B.,
Beutler J. A., Anal. Biochem., 331, 296–302 (2004).
22) Parniak M. A., Min K. L., Budihas S. R., Le Grice S. F., Beutler J.
A., Anal. Biochem., 322, 33–39 (2003).
Conclusion
More than 30 chemical compounds were synthesized for
developing the inhibitors of RNase H activity of HIV-1 re-
verse transcriptase. Inhibitory potency of RNase H enzymatic
activity was measured in a biochemical assay with a real-time
fluorescence monitoring method. The active compounds found
in our previous studies commonly bear nitro-furan ring con-
necting to hydrophobic region via an ester linkage. Conversion
of the nitro-furan group into pyrrole drastically decreased the
inhibitory activity while conversion into nitro-thiophene main-
tained the compound potency. This means that the structural
basis of nitro-furan or nitro-thiophene is indispensable for in-
hibitory activity. An improvement in compound potency was
observed when a phenyl-ester moiety was connected to the
nitro-furan and further methoxy-carbonyl and methoxy groups
were bound to the phenyl ring. No notable change in inhibi-
tory potency was observed when benzyl-ester based substitute
was connected to nitro-furan. Modulation of ester linkage
resulted in complete loss of compound potency. Molecular
dynamics simulation was performed to examine the stability
of the binding structure of a synthesized active compound
to RNase H domain. It was demonstrated that a potent com-
pound was stably bound to the active site with establishing
strong coordinate bonds with divalent metal ions located at the
active site. The present study provides important information
for designing prospective chemical structures inhibiting HIV-1
RNase H activity.
23) Lansdon E. B., Liu Q., Leavitt S. A., Balakrishnan M., Perry J. K.,
Lancaster-Moyer C., Kutty N., Liu X., Squires N. H., Watkins W.
J., Kirschberg T. A., Antimicrob. Agents Chemother., 55, 2905–2915
(2011).
Acknowledgments This work was supported by a Health 24) Martí-Renom M. A., Stuart A. C., Fiser A., Sánchez R., Melo F.,
Sali A., Annu. Rev. Biophys. Biomol. Struct., 29, 291–325 (2000).
25) Kirschberg T. A., Balakrishnan M., Squires N. H., Barnes T., Bren-
and Labor Science Research Grant for Research on Publicly
Essential Drugs and Medical Devices from the Ministry of
Health and Labor of Japan. A part of this work was also sup-
ported by a Grant-in-Aid for Scientific Research (C) from
Japan Society for the Promotion of Science (JSPS). Theo-
retical calculations were performed at the Research Center for
Computational Science, Okazaki, Japan and at the Information
dza K. M., Chen X., Eisenberg E. J., Jin W., Kutty N., Leavitt S.,
Liclican A., Liu Q., Liu X., Mak J., Perry J. K., Wang M., Watkins
W. J., Lansdon E. B., J. Med. Chem., 52, 5781–5784 (2009).
26) Himmel D. M., Maegley K. A., Pauly T. A., Bauman J. D., Das
K., Dharia C., Clark A. D. Jr., Ryan K., Hickey M. J., Love R. A.,
Hughes S. H., Bergqvist S., Arnold E., Structure, 17, 1625–1635
Technology Center of the University of Tokyo and also by the
high-performance computer system at Institute for Media In-
formation Technology in Chiba University.
(2009).
27) Su H. P., Yan Y., Prasad G. S., Smith R. F., Daniels C. L., Abey-
wickrema P. D., Reid J. C., Loughran H. M., Kornienko M., Sharma
S., Grobler J. A., Xu B., Sardana V., Allison T. J., Williams P. D.,
Darke P. L., Hazuda D. J., Munshi S., J. Virol., 84, 7625–7633
(2010).
28) Li H., Robertson A. D., Jensen J. H., Proteins, 61, 704–721 (2005).
29) Cieplak P., Cornell W. D., Bayly C., Kollman P. A., J. Comput.
Chem., 16, 1357–1377 (1995).
References
1) Goody R. S., Müller B., Restle T., FEBS Lett., 291, 1–5 (1991).
2) Ren J., Stammers D. K., Virus Res., 134, 157–170 (2008).
3) Andréola M. L., De Soultrait V. R., Fournier M., Parissi V., Desjob-
ert C., Litvak S., Expert Opin. Ther. Targets, 6, 433–446 (2002).
4) Klumpp K., Mirzadegan T., Curr. Pharm. Des., 12, 1909–1922
(2006).
30) Matsuyama S., Aydan A., Ode H., Hata M., Sugiura W., Hoshino T.,
J. Phys. Chem. B, 114, 521–530 (2010).
31) Sano E., Li W., Yuki H., Liu X., Furihata T., Kobayashi K., Chiba
K., Neya S., Hoshino T., J. Comput. Chem., 31, 2746–2758 (2010).
5) Tramontano E., Mini Rev. Med. Chem., 6, 727–737 (2006).
6) Tanese N., Telesnitsky A., Goff S. P., J. Virol., 65, 4387–4397

June 2012
771
32) Ode H., Neya S., Hata M., Sugiura W., Hoshino T., J. Am. Chem.
Soc., 128, 7887–7895 (2006).
33) Ode H., Matsuyama S., Hata M., Neya S., Kakizawa J., Sugiura W.,
Hoshino T., J. Mol. Biol., 370, 598–607 (2007).
34) Jorgensen W. L., Chandrasekhar J., Madura J. D., Impey R. W.,
Klein M. L., J. Chem. Phys., 79, 926–935 (1983).
1930–1944 (2008).
37) Ryckaert J.-P., Ciccotti G., Berendsen H. J. C., J. Comput. Phys., 23,
327–341 (1977).
38) Sarafianos S. G., Das K., Tantillo C., Clark A. D. Jr., Ding J., Whit-
comb J. M., Boyer P. L., Hughes S. H., Arnold E., EMBO J., 20,
1449–1461 (2001).
35) Case D. A., Darden T. A., Cheatham T. E. III, Simmerling C. L.,
Wang J., Duke R. E., Luo R., Merz K. M., Pearlman D. A., Crow-
ley M., Walker R. C., Zhang W., Wang B., Hayik S., Roitberg A.,
Seabra G., Wong K. F., Paesani F., Wu X., Brozell S., Tsui V.,
Gohlke H., Yang L., Tan C., Mongan J., Hornak V., Cui G., Beroza
P., Mathews D. H., Schafmeister C., Ross W. S., Kollman P. A.,
39) Huang H., Chopra R., Verdine G. L., Harrison S. C., Science, 282,
1669–1675 (1998).
40) Klumpp K., Hang J. Q., Rajendran S., Yang Y., Derosier A., Wong
Kai In P., Overton H., Parkes K. E., Cammack N., Martin J. A.,
Nucleic Acids Res., 31, 6852–6859 (2003).
41) De Vivo M., Dal Peraro M., Klein M. L., J. Am. Chem. Soc., 130,
10955–10962 (2008).
“
”
AMBER 9, University of California, San Francisco, 2006.
36) Katagiri D., Fuji H., Neya S., Hoshino T., J. Comput. Chem., 29,