Novel anti-HIV agents are still needed to overcome resistance issues, in particular inhibitors acting against novel viral targets. The ribonuclease H (RNase H) function of the reverse transcriptase (RT) represents a validated and promising target, and no inhibitor has reached the clinical pipeline yet. Here, we present rationally designed non-diketo acid selective RNase H inhibitors (RHIs) based on the quinolinone scaffold starting from former dual integrase (IN)/RNase H quinolinonyl diketo acids. Several derivatives were synthesized and tested against RNase H and viral replication and found active at micromolar concentrations. Docking studies within the RNase H catalytic site, coupled with site-directed mutagenesis, and Mg2+ titration experiments demonstrated that our compounds coordinate the Mg2+ cofactor and interact with amino acids of the RNase H domain that are highly conserved among naïve and treatment-experienced patients. In general, the new inhibitors influenced also the polymerase activity of RT but were selective against RNase H vs the IN enzyme.
Novel anti-HIV agents are still needed to overcome resistance issues, in particular inhibitors acting against novel viral targets. The ribonuclease H (RNase H) function of the reverse transcriptase (RT) represents a validated and promising target, and no inhibitor has reached the clinical pipeline yet. Here, we present rationally designed non-diketo acid selective RNase H inhibitors (RHIs) based on the quinolinone scaffold starting from former dual integrase (IN)/RNase Hquinolinonyl diketo acids. Several derivatives were synthesized and tested against RNase H and viral replication and found active at micromolar concentrations. Docking studies within the RNase H catalytic site, coupled with site-directed mutagenesis, and Mg2+ titration experiments demonstrated that our compounds coordinate the Mg2+ cofactor and interact with amino acids of the RNase H domain that are highly conserved among naïve and treatment-experienced patients. In general, the new inhibitors influenced also the polymerase activity of RT but were selective against RNase H vs the IN enzyme.
The human immunodeficiency
virus type 1 (HIV-1) is the agent responsible
for the acquired immunodeficiency syndrome (AIDS). According to the
last estimates by the World Health Organization (WHO) and the Joint
United Nations Programme on HIV and AIDS (UNAIDS), globally, there
were 38 million people living with HIV in 2018 and only 62% of them
were receiving antiretroviral treatment by the end of 2018.[1]In total, 44 Food and Drug Administration
(FDA)-approved medicines
can be used in the treatment of HIV, including multiclass combination
products, nucleoside reverse transcriptase (RT) inhibitors (NRTIs),
non-nucleoside RT inhibitors (NNRTIs), protease inhibitors (PIs),
integrase (IN) inhibitors (INSTIs), fusion inhibitors, CCR5 antagonists,
postattachment inhibitors, and pharmacokinetic enhancers.[2] Treatment with HIV medicines is called antiretroviral
therapy (ART), which involves taking a combination of drugs as a single
pill or in various pill combinations and which generally comprehends
combinations of at least three drugs from different HIV drug classes
(usually NRTIs, NNRTIs, and INSTIs).[2,3] These approaches
have resulted in suppression of viral replication, with decreased
death rates[4] and morbidity.[5] Still, therapy suspension or lack of adherence is associated
with a rapid viral rebound because such therapies do not affect the
viral reservoir of latently infected cells, being the main obstacle
to viral eradication.Despite the undisputed advantage of ART,
this therapy still has
several drawbacks, which include long-term toxicity and drug–drug
interactions.[6] Moreover, life-long treatment
strongly impairs the adherence, drastically promoting the selection
of variants of the virus resistant to current therapies.[7] This resistance phenomenon represents the major
clinical challenge in the fight against AIDS. Therefore, new anti-HIV
agents are still urgently needed, in particular, inhibitors acting
against novel viral targets that can contribute overcoming the resistance
issue.[8−10]Since the discovery of HIV, RT has been the
first exploited therapeutic
target. RT is an RNA-dependent DNA polymerase that utilizes a strand
of RNA to synthesize double-stranded viral DNA that can eventually
integrate into the genome of the infected cell.[11] It is a multifunctional enzyme with DNA polymerase RNA-
and DNA-dependent (RDDP and DDDP, respectively) and endonuclease (ribonuclease
H, RNase H) activities. RNase Hfunction is essential for virus replication
since it specifically cleaves the RNA moiety of the RNA/DNA hybrid
to generate a DNA duplex to be integrated into the host cell. The
RNase H active site contains a highly conserved DEDD motif consisting
offour carboxylate amino acid residues in close proximity (D443,
E478, D498, and D549) that interact with two Mg2+ ions.[11] It is worthy of note that a similar arrangement
is observed in the active site ofHIV-1 IN, another metalloenzyme
that plays critical roles in viral infection. Indeed, three highly
conserved residues in the catalytic core domain of this enzyme (D64,
D116, and E152; DDE motif) coordinate the two Mg2+ ions
necessary for its trans-esterase activity.[12]Despite being a valid and promising drug target, RNase H inhibitors
have not reached the clinical pipeline yet. Indeed, all of the RT-targeting
drugs approved thus far are inhibitors of the RDDP activity and the
development ofRNase H inhibitors (RHIs) has lagged behind so that
no drug targeting RNase H has been approved yet. This can be attributed
to two reasons: (i) the availability of expertise on inhibitors of
other DNA polymerases[13] that encouraged
the development of drugs targeting the RT-associated RDDPfunction,
and (ii) the open morphology of the RNase Hfunction that is hard
to target and showing a strong competition with the substrate for
access to the catalytic core.[14]However,
RNase H plays a key role in the viral life cycle and shows
a high degree of conservation of the entire domain upon naïve
and treatment-experienced patients.[15] Thus,
more recently, efforts were boosted in the development of new RHIs
as relevant to enhance the antiretroviral armory and potentially able
to counteract circulating HIV-1 strains resistant to the approved
drugs.[15−17] In recent years, the development of more effective
screening techniques[18,19] and the availability of more
and more detailed structural data helped design and identify new inhibitors
that can be grouped into two main categories: active-site and allosteric
inhibitors. The first ones are small molecules that showed RNase H
inhibitory activity at low micromolar or submicromolar ranges. These
inhibitors mainly contain a hydrophobic moiety linked to a two-metal-cation
chelating core, an element reminiscent of that observed for canonical
HIV-1 INSTIs.[20−23] This moiety plays a key role as a driving force for the binding,
conferring a high potency of inhibition, but leads to limited selectivity
towards RT-associated RDDP and/or IN activities.[20−22,24]In this field on class="Chemical">f research, our group has previously
reported DKA
derivatives that proved to be dual inhibitors of IN and RH[25−27] or INSTIs endowed with marginal RH inhibition activities.[10,28]
Among the dual inhibitors, an example is the pyrrolyl diketo
hexenoic
ester (RDS 1643, 1), which was the first DKA derivative
reported as RHI to have an antiviral effect[29−32] (Figure ), further developed on RDS 1759,[25] with a selective mode-of-action and the ability
to target conserved residues within the RT RNase H domain. Conversely,
quinolinonyl diketo butanoic derivative 2 (Figure ) showed a more prominent IN
inhibitory activity with respect to that against RNase H, reporting
IC50 values of 0.028 and 5.1 μM, respectively.[33]
Figure 1
Inhibitors of HIV-1 RNase H function of RT and/or IN enzymes.
Inhibitors ofHIV-1RNase Hfunction of RT and/or IN enzymes.Despite the relevance of the DKA branch in the
inhibitory activity,
it is well-known that the DKA chain suffers from several limits related
to the pharmacokinetic and pharmacodynamic profiles. Indeed, the DKA
chain is responsible for the poor solubility, high metabolic turnover,
and low permeability through the cell membrane of molecules containing
such a chain. Furthermore, a time-dependent decrease in the activity
in solution at room temperature has been proven, even during short
periods.[30−34] Therefore, to overcome the limits of the DKA moiety, a variety of
compounds were developed by transferring the DKA chain to scaffolds
characterized by improved druglike qualities that dialed out the undesirable
DKA properties. A notable example is the INSTI elvitegravir (3, Figure ) approved by the FDA as a successful anti-HIV drug. In this compound,
the DKA chain was shortened into a carboxylic acidfunction that,
together with the ketone group in the 4-position of the quinolinone
ring, chelates the two Mg2+ ions within the IN catalytic
site.[35]Recently, we successfully
obtained a new selective RHIs by design
ofpyrrolyl pyrazole carboxylic acids.[36] This new scaffold has been achieved converting the diketo group
of our previously reported dual IN/RH inhibitors pyrrolyl DKA derivatives[25−27] into a pyrazole moiety (Figure A). In this way, we obtained compounds active at micromolar/submicromolar
concentrations against RNase H and selective for RNase H vs INfrom
5 to >18 times. Also, these non-DKA inhibitors blocked the viral
replication
and proved to interact with conserved residues within the RNase H
active site domain.[36]
Figure 2
Design of pyrrolyl pyrazole
carboxylates as RH inhibitors (A), quinolinones as INIs
(B), and the new quinolinonyl
non-DKA derivatives as RHIs (C).
Design of pyrrolyl n class="Chemical">pyrazole
carboxylates as RH inhibitors (A), quinolinones as INIs
(B), and the new quinolinonyl
non-DKA derivatives as RHIs (C).
Besides pyrrolyl DKA inhibitors, we also reported quinolinonyl
DKA derivatives. We recently designed a few series ofquinolinones
and defined the structural elements that were vital in influencing
the activity. Indeed, the bifunctional DKA derivatives were IN inhibitors
nonselective against 3′-processing vs strand
transfer steps,[37] while the quinolinones
with small substituents on 6- and 7-positions gave potent INSTIs,[28] and finally, the basic quinolinones bearing
an amino substituent in the 7-position resulted in IN inhibitors with
marginal activity against RH (nM against ST and >10 μM against
RH) (Figure B).[10]In this paper, we applied an isosteric
approach to those quinolonyl
DKAs, with the aim to obtain a new class of compounds endowed with
selective inhibitory activity against RNase H. We referred to integrase
inhibitor 3, conceived as a DKA isoster capable of chelating
ions within the catalytic site. Thus, starting from quinolonyl DKA,
we designed non-DKA quinolinonyl derivatives in which the DKA unit
is replaced by the carboxylic unit in the 3-position and the carbonyl
group in the 4-position of the quinolone moiety. In this way, we retained
the chelating unit capable of binding the ions within the catalytic
site. At the same time, we changed the distance between the benzyl
moiety and the chelating group, critical to achieving optimal interaction
with the IN catalytic site, with the aim to produce selectivity toward
the RH function of the RT (Figure C).All of the newly designed compounds 4a–t and 5a–t are characterized
by the introduction of different
substituents in the 6-position of the quinolinone ring (Chart ). In detail, we introduced
(i) a hydrogen atom or a hydroxyl group; (ii) various ether groups
characterized by different degrees offreedom or steric hindrance
as methoxy, phenoxy, aminoalkyloxy, and arylmethyloxy; (iii) acetyl,
cyano, nitro, trifluoromethyl, and methylsulfonyl groups; and (iv)
a phenylpropenone moiety.
Chart 1
Structures of the Newly Designed Quinolinonyl
Derivatives 4a–t and 5a–t
The newly synthesized compounds have been evaluated in
vitro for their ability to inhibit the specific enzymatic
activity of recombinant RNase H, for theircytotoxicity and antiviral
activity against HIV-1 inhuman cells. Besides, a rationalization
of the interaction with the biological target has been proposed, based
on docking studies using the crystal structure ofRNase H, and validated
by site-directed mutagenesis on the residues indicated as being crucial
for the binding. Finally, selected derivatives have been tested for
their activity against IN and RDDPfunctions of the RT to evaluate
their selectivity.
Results and Discussion
Chemistry
Compounds 4d and 5d were obtained as already reported.[38] The
synthesis of derivatives 4a–c,e–i,k and 5a–c,e–i,k is outlined in Scheme . Condensation of(E)-3-(4-aminophenyl)-1-phenylprop-2-en-1-one[39] (6) with diethyl ethoxymethylenemalonate
(EMME) gave intermediate 7, which was submitted to thermal
ring closure to give 8k (Gould–Jacobs reaction).[40] Derivatives 5a–c,e–i,k were obtained by alkylation in position 1 of the proper quinolinonyl
derivative 8a–c,e–i(41−49) or 8k with p-fluorobenzyl bromide
in the alkaline medium. The subsequent base-catalyzed hydrolysis ofester derivatives 5a–c,e–i,k afforded the
corresponding acids 4a–c,e–i,k.
Scheme 1
Synthetic
Route to 4a–c,e–i,k and 5a–c,e–i,k Derivatives
Reagents and conditions: (i)
EMME, 90 °C, 3 h, 90% yield; (ii) diphenyl ether, reflux, 2 h,
100% yield; (iii) 4-fluorobenzyl bromide, K2CO3, N,N-dimethylformamide (DMF),
100 °C, 2–3 h, 70–93% yield; (iv) proper base,
1:1 tetrahydrofuran (THF)/EtOH, reflux or room temp, 1–4 h,
80–100% yield.
Synthetic
Route to 4a–c,e–i,k and 5a–c,e–i,k Derivatives
Reagents and conditions: (i)
EMME, 90 °C, 3 h, 90% yield; (ii) n class="Chemical">diphenyl ether, reflux, 2 h,
100% yield; (iii) 4-fluorobenzyl bromide, K2CO3, N,N-dimethylformamide (DMF),
100 °C, 2–3 h, 70–93% yield; (iv) proper base,
1:1 tetrahydrofuran (THF)/EtOH, reflux or room temp, 1–4 h,
80–100% yield.
The synthesis of compounds 4j,l–t and 5j,l–t was performed
as reported in Scheme . The synthetic approach resembles
the one described above for compounds 4a–c,e–i,k and 5a–c,e–i,k. Noteworthy, the synthetic
pathway starts with an O-alkylation ofdiethyl 2-(((4-hydroxyphenyl)amino)methylene)malonate[50] (9) with the appropriate alkyl
halide to obtain intermediates 10a–j.
Scheme 2
Synthetic
Route to 4j,l–t and 5j,l–t Derivatives
Reagents and conditions: (i)
appropriate halide, t-BuOK, DMF, 0 °C to room
temp, 2–3 h, 60–90% yield; (ii) diphenyl ether, reflux,
2–3 h, 90–100% yield; (iii) 4-fluorobenzyl bromide,
KCO, DMF, 100 °C,
2–3 h, 40–90% yield; (iv) NaOH 20%, 1:1 THF/EtOH, reflux,
1–2 h, 50–100% yield.
Synthetic
Route to 4j,l–t and 5j,l–t Derivatives
Reagents and conditions: (i)
appropriate halide, t-BuOK, n class="Chemical">DMF, 0 °C to room
temp, 2–3 h, 60–90% yield; (ii) diphenyl ether, reflux,
2–3 h, 90–100% yield; (iii) 4-fluorobenzyl bromide,
KCO, DMF, 100 °C,
2–3 h, 40–90% yield; (iv) NaOH 20%, 1:1 THF/EtOH, reflux,
1–2 h, 50–100% yield.
Evaluation
of Biological Activities
In Vitro Screening for RNase
H Inhibitory Activity
All compounds 4a–t and 5a–t were tested in vitro in enzyme inhibition assays
against recombinant RNase H (Table ), using the known RNase H inhibitors RDS 1759[25] (11) and β-thujaplicinol
(BTP) used as positive controls. The assays were performed in conditions
of competition with the substrate, without preincubation, starting
the reaction by adding the enzyme, to avoid overestimation of compound
potency and proving that it was not affected by substrate displacement,
as reported for other active compounds.[14,51]
Table 1
Enzymatic Activities on RNase H of
the Newly Synthesized Compounds 4a–t and 5a–t
cpd
X
R
anti-RH activity (IC50 ± SD)a
4a
H
H
>100
4b
H
CN
>100
4c
H
CF3
>100
4d
H
COCH3
16.3 ± 1.42
4e
H
NO2
56.0 ± 4
4f
H
SO2CH3
100
4g
H
OH
74 ± 9
4h
H
OCH3
>100
4i
H
OPh
32.0 ± 11.0
4j
H
O(CH2)3N(CH3) 2
15.4 ± 3.0
4k
H
5.9 ± 0.6
4l
H
H
34.0 ± 0.1
4m
H
2,3-Cl2
8.19 ± 0.05
4n
H
3,4-Cl2
29.5 ± 0.5
4o
H
1-yl
1.51 ± 0.21
4p
H
2-yl
30.3 ± 1.7
4q
H
13.5 ± 1.7
4r
H
8.27 ± 0.45
4s
H
Ph
7.47 ± 1.55
4t
H
OBnb
7.48 ± 0.28
5a
Et
H
ntc
5b
Et
CN
>100
5c
Et
CF3
59.0 ± 5.0
5d
Et
COCH3
11.0 ± 1.0
5e
Et
NO2
47.8 ± 10
5f
Et
SO2CH3
>100
5g
Et
OH
>100
5h
Et
OCH3
>100
5i
Et
OPh
45.5 ± 1.5
5j
Et
O(CH2)3N(CH3)2
>100
5k
Et
15.3 ± 1.6
5l
Et
H
8.0 ± 1.6
5m
Et
2,3-Cl2
24.0 ± 4.9
5n
Et
3,4-Cl2
37.2 ± 7.6
5o
Et
1-yl
1.49 ± 0.33
5p
Et
2-yl
19.6 ± 0.05
5q
Et
55.3 ± 1.7
5r
Et
28.4 ± 4.3
5s
Et
Ph
>100
5t
Et
OBnb
>100
11(25)
7.50 ± 1.32
BTPd
0.19 ± 0.03
Inhibitory concentration
50% (μM)
determined from dose-response curves: experiments performed against
HIV-1 RT-associated RNase H activity.
Bn, benzyl.
nt, not tested.
BTP, β-thujaplicinol.
Inhibitory concentration
50% (μM)
determined from dose-response curves: experiments performed against
HIV-1 RT-associated RNase H activity.Bn, benzyl.nt, not tested.BTP, β-thujaplicinol.In general, the newly designed
quinolinones were proven active
against RH, with 27 out of 39 tested compounds showing measurable
IC50 under 100 μM concentration. Moreover, 20 compounds
were active at concentrations up to 34 μM, and 8 compounds were
active in the low micromolar range. The most active compounds of the
series were 4o and 5o having comparable
IC50 values (about 1.5 μM).The acid derivatives
were generally more active than the ester
counterparts, although notable exceptions can be cited like the equipotent
couples 4d–5d, 4o–5o, and 4e–5e and the case ofester 5l being more
active than the acid counterpart 4l.In general,
compounds with small substituents (4a–h and 5a–h) were found to be active in the high
micromolar range (IC50 ≥ 50 μM) or were inactive,
with the sole exception of the couple 4d and 5d showing IC50 values 16.3 and 11.0 μM, respectively.Indeed, within this subseries, the best acting compound, 5e, showed a decrease of 3–4-fold in activity with
respect to the acetyl counterparts 4d and 5d.The removal of the substituent in position 6 of the quinolinonyl
ring led to a loss of activity, as observed for the acid 4a. Similarly, by replacing the acetyl group proper of compounds 4d and 5d with a hydroxyl one, a decrease in
inhibitory potency was observed (5g, IC50 >
100 μM; 4g, IC50 = 74 μM). It
is also worthy of note that the methylation of compounds 4g and 5g led to derivatives 4h and 5h, which reported no activity.An increase in the dimension
of the substituent gave compounds
endowed with better activity. Indeed, within the ether subseries 4h–j and 5h–j, the methoxy compounds 4h and 5h were inactive, phenyl ethers4i and 5i reported a moderate inhibition with
IC50 values of 32.0 and 45.5 μM, respectively, while
the dimethylaminopropyl derivative 4j showed good efficacy
(IC50 = 15.4 μM). Definitively, among the acid 4h–j, it is possible to notice that the activity increases
in the following order: 4h < 4i < 4j.A further increase in the moiety placed in the 6-position
of the
quinolinone ring, like the one with the phenylpropenone substituent
(4k and 5k), gave good inhibitory potencies.
In particular, the ester 5k showed comparable activity
with respect to that of compounds 4d and 5d, while its acid counterpart 4k reported a twofold gain
in activity shifting to low micromolar activity (5k,
IC50 = 15.3 μM; 4k, IC50 =
5.9 μM).Following this trend, arylmethyloxy ether derivatives 4l–t and 5l–t reported the most
promising inhibitory
profile. Indeed, although the esters 5s,t containing
two aromatic rings in the 6-position of the quinolinone core were
found inactive, the acid counterparts 4s,t were found
highly active (IC50 values 7.47 and 7.48 μM, respectively).
This trend of activity could be ascribable to the coexistence of the
carboxylic acidfunction in the 3-position, along with substituents
characterized by both moderate degrees offreedom and steric hindrance
in the 6-position of the quinolinonyl ring (a benzyloxybenzyl group
in the case of derivative 4t and a biphenylmethyl moiety
for derivative 4s).Within this subseries, 7 derivatives
(4m,o,r–t and 5l,o) out of 18 tested
showed high inhibitory activities,
with IC50 values lower than 10 μM, 5 compounds (n class="Chemical">4n,q and 5m,p,r) proved to be active with 10
μM < IC50 < 30, and only 6 ethers (4l,p and 5n,q,s,t) reported inhibitory activity with IC50 values >30 μM.
Compounds 4q and 5p reported inhibitory
potencies comparable to those of compounds 4d and 5d (4q, IC50 = 13.48 μM; 5p, IC50 = 19.59 μM).The ester5o and its acid counterpart 4o, characterized
by a methylnapht-1-yl group in position 6 of the
quinolinonyl ring, proved to be the best acting compounds among the
newly synthesized quinolinonyl derivatives showing IC50 values of 1.49 and 1.51 μM, respectively.Interestingly,
the regioisomers 4p and 5p of the compounds
described above, obtained by replacing the methylnapht-1-yl
group with a methylnapht-2-yl one, showed a decrease in activity by
about 13–20 times.A similar decrease in potency was
obtained by substituting the
methylnapht-1-yl group with a benzyl ring (4l and 5l, IC50 = 34.0 and 8.0 μM, respectively),
thus suggesting that for both acid and ester compounds, the methylnapht-1-yl
group is advisable for enzymatic inhibition.The dichloro derivatives 4m and 5m, isosters
of compounds 4o and 5o, reported lower inhibitory
potencies, as well. Indeed, the ester 5m showed an IC50 value 16 times lower than its isoster 5o (5m, IC50 = 24.03 μM; 5o, IC50 = 1.49 μM). The decrease in activity was less evident
for the corresponding acid 4m, resulting in a 5-fold
loss in activity with respect to its isosteric counterpart 4o (4m, IC50 = 8.19 μM; 4o, IC50 = 1.51 μM).Difn class="Chemical">ferently, this trend
of activity is not detected for the methylnapht-2-yl
derivatives 6q, 7q, 4p, and 5p and theirdichloro isosters4n and 5n,
which showed comparable activity.
Likewise, no big difference
in inhibitory potencies can be outlined
between the dichloro derivatives4n and 5n and the corresponding isomers 4m and 5m, with the sole exception of the acid 3,4-dichloro derivative 4m, which is active at the micromolar concentration level
(4m, IC50 = 8.19 μM). Finally, collectively,
the arylmethyloxy acid derivatives resulted in more potent RHIs than
the ester counterparts. Indeed, we obtained five acids active in the
micromolar range, one compound around 10 μM, and three derivatives
active around 30 μM concentrations. Only two of the corresponding
esters showed IC50 values in the micromolar range, five
compounds were active at concentrations higher that 20 μM, and
two compounds were inactive up to 100 μM.
Cell-Based
Assays
To determine the effect of compounds
on viral replication, compounds 4d,k–t and 5d,k–p,r were chosen to test the antiviral activity
in HeLa-CD4-LTR-β-gal cells (Table ). In general, the acid derivatives gave
measurable EC50 values that ranged from 1.73 to 16.1 μM
with only compounds 4d,p,r (out of 11 tested) inactive
up to 50 μM. Among them, compound 4t reported the
lowest EC50 of the series, proving not to be cytotoxic
up to high concentrations (100 μM), thus showing the best antiviral
profile (SI > 57.8). In general, the ester derivatives showed a
weaker
activity when compared to the ones of the acid series, and only two
compounds out of the eight tested were active in the low micromolar
range (5n, IC50 = 14.6 μM; 5m, IC50 = 17.8 μM).
Table 2
Biological
Effects on RT-RDDP and
IN Activities, Cytotoxicity, and Antiviral Activities of Compounds 4d,k–t and 5d,k–p,r
activity in the enzyme assay IC50 (μM)a
antiviral
activity (μM)
cpd
RDDPb
INc
EC50d
CC50e
SIf
4d
ntg
>100
>50
>200
4k
>100
3.38 ± 0.42
5.4 ± 3.1
17.0 ± 4.0
3.1
4l
38.5 ± 7.1
0.41 ± 0.03
11.7 ± 2.5
>100
>8.6
4m
5.6 ± 0.6
>100
13.3 ± 4.5
>200
>15
4n
2.0 ± 0.8
3.25 ± 0.85
16.1 ± 5.5
>200
>12.4
4o
11.4 ± 2.6
>100
8.4 ± 0.7
51 ± 11
6.1
4p
nt
nt
>50
>200
4q
5.1 ± 1.8
>100
11.3 ± 3.5
>200
>17.7
4r
nt
nt
>50
>200
4s
11.6 ± 5.5
>250
2.5 ± 1.03
>100
>40
4t
2.2 ± 0.1
>100
1.73 ± 0.47
>100
>57.8
5d
nt
>100
>50
>200
5k
nt
nt
>50
>100
5l
nt
nt
>100
>100
5m
1.88 ± 0.04
0.05 ± 0.01
17.8 ± 2.3
21.6 ± 2.8
1.2
5n
24.1 ± 8.6
9.45 ± 0.55
14.6 ± 3.1
20.6 ± 4.4
1.4
5o
nt
nt
>100
>100
5p
nt
nt
>50
73 ± 24
5r
nt
nt
>16
>200
11
>100
2.9 ± 0.5
68 ± 10
>13.7
RALh
0.019 ± 0.01
EFVi
0.035 ± 0.011
0.53 ± 0.04
Inhibitory concentration 50% (μM)
determined from dose-response curves.
Experiments performed against HIV-1
RT-RDDP activity.
Experiments
performed against HIV-1
IN activity.
Effective concentration
50% (μM).
Cytotoxic
concentration 50% (μM).
Selectivity index = CC50/EC50.
nt, not tested.
RAL, raltegravir.
EFV, efavirenz.
Inhibitory concentration 50% (μM)
determined from dose-response curves.Experiments performed against n class="Species">HIV-1
RT-RDDP activity.
Experiments
performed against n class="Species">HIV-1
IN activity.
Efn class="Chemical">fective concentration
50% (μM).
Cytotoxic
concentration 50% (μM).Selectivity index = CC50/EC50.nt, not tested.RAL, n class="Chemical">raltegravir.
EFV, n class="Chemical">efavirenz.
Counter-Assays against IN and RDDP HIV-1 Activities
Since
severalRNase H active site inhibitors were reported to inhibit
also other related viral targets;[22] compounds 4e,k,l–o,r–t and 5m,n active against
viral replication were tested for their ability to inhibit the other
HIV-RT enzymatic function RDDP and the activity of the HIV-1 IN, structurally
related to HIV-1RNase H (Table ).The results showed that seven out of the 12
compared derivatives were selective for HIV-1RNase H inhibition over
HIV-1 IN inhibition, showing IC50 > 100 μM against
the IN enzyme, three compounds showed comparable activities, and only
two derivatives were substantially more active against IN than against
RH.Conversely, we did not observe a selectivity against the
RT-RDDP
activity since most of the tested compounds inhibited the polymerase
function with IC50 values on the same order of magnitude
ofRNase H IC50 values. Here, we cannot exclude the possibility
that the RDDP inhibition might result from an allosteric modulation
of this RT function rather than a direct binding to the NNRTI binding
site as already demonstrated for other RHIs.[52] Nevertheless, this additional inhibition may contribute to the antiviral
activity displayed in cell-based assays. Only two compounds showed
selectivity for RH, namely, derivative 4k that did not
influence the RDDP activity up to 100 μM and 4o. Interestingly, the last one was the best performing RNase H inhibitor,
with an increase of 7.5-fold against RDDP (1.51 vs 11.4 μM)
and was totally inactive against HIV-1 IN (IC50 > 100
μM).
It was chosen to be further characterized for its binding mode, together
with compound 4t, which showed the most promising antiviral
activity.
Molecular Modeling
To clarify the reasons behind the
inhibitory activity displayed by the novel compounds, molecular docking
studies were performed on 4o and 4t. In
particular, to predict the binding poses of compounds 4o and 4t in the RNase H binding site (X-ray crystal structure
with the PDB code 3QIP)[53] and to adequately probe the possible
conformational changes in the active site induced by the rather bulky
side chains at the 6-position of the quinolinonyl core, we elected
to employ the Induced Fit routine of Glide docking software[54,55] (see Molecular Modeling methods). In general, we expect that the
whole set of newly designed inhibitors would bind to the viral enzyme
with similar poses to those that we are reporting for 4o (Figure A) and 4t (Figure B).
Figure 3
Predicted binding poses of 4o (A) and 4t (B) in the HIV RNase H binding site (PDB 3QIP). Important residues are labeled. 4o is represented as yellow sticks, 4t as orange
sticks, the magnesium ions are depicted in purple, and their coordination
with the nearby atoms is also represented in purple. H-bonds are depicted
as dashed green lines. Charge-transfer interactions are represented
as dashed red lines. The protein is depicted as blue ribbons and sticks.
Predicted binding poses ofn class="Chemical">4o (A) and 4t (B) in the HIV RNase H binding site (PDB 3QIP). Important residues are labeled. 4o is represented as yellow sticks, 4t as orange
sticks, the magnesium ions are depicted in purple, and their coordination
with the nearby atoms is also represented in purple. H-bonds are depicted
as dashed green lines. Charge-transfer interactions are represented
as dashed red lines. The protein is depicted as blue ribbons and sticks.
Our model suggests for both n class="Chemical">4o and 4t that the oxygen atoms of the ketone and the position-4
carboxyl/ester
moiety tightly chelate the Mg2+ atoms in the active site,
in a geometry that is consistent with other cocrystallized HIV-1 RNase
H inhibitors.[53,56]
Interestingly, the 4-carboxyl
group is localized in a highly polar
section of the binding site, which comprises H539, D549, D443, D498,
and E478 and two conserved water molecules that take part in the chelation
of one of the magnesium ions. While we infer that the ester derivatives 5a–t should engage in the same interaction pattern
that characterizes their acidic counterparts, our theoretical model
would place the esters’ lipophilic ethyl chain toward the above-described
highly hydrophilic area. In some cases, this should unfavorably impact
the binding affinity.The quinolinonyl core of4oforms van der Waals contacts
with the side chains of W535 and A538. In the 4t case,
the said core adopts a slightly different pose, possibly diminishing
the strength of the van der Waals contacts with the two residues.
The 4o pendant p-fluorobenzyl ring is
lodged in a cleft lined by the side chains of the polar residues H539
and K540. Here, the phenyl ring establishes a T-shaped charge-transfer
interaction with H539, which is intensified by the electron-withdrawing
(EWG) fluorine substituent and a H-bonding interaction between its p-fluorine atom and the K540 side chain.Regarding
the 4t pendant p-fluorobenzyl
ring, while it is predicted to point toward the same cleft, it would
be oriented away from H539 and closer to the side chains of P537 and
W535, possibly giving rise to a parallel-displaced π–π
interaction with the latter residue, also enhanced by the p-fluorine atom. Besides, the K540 side chain would be oriented
to form a cation-π with the ligand pendant phenyl ring, although
this interaction should be weakened by the presence of an EWG substituent
such as fluorine.[57,58]As for the 6-position substituent,
several of the newly designed
compounds feature an oxygen atom, which should be well-positioned
to accept a H-bond from either the backbone NH of Q500, as exemplified
by the 4o binding pose, or the side chain of the nearby
S499, as in the 4t case. Furthermore, the 4o naphthylmethoxy group would be positioned in a polar pocket lined
by the residues R448, N474, Q475, S499, and Y501. Here, like the other
arylmethyloxy ether derivatives, 4o naphthyl can form
a parallel-displaced charge-transfer interaction with the aromatic
side chain ofY501, similar to what we already demonstrated for another
class ofRNase H inhibitors.[36] On the other
hand, our in silico analysis also revealed that the
compound 4t methylphenyl-4-oxybenzyl group should reach
an enzyme region that is only partially overlapping with the binding
site area contacted by the naphthylmethoxy moiety of4o. While the 4t extended chain should still engage in
a π–π interaction with Y501 through its aromatic
portions, it should also be able to access an adjacent cavity lined
by the residues Q475, L479, Y501, and I505, providing further anchoring
points for the ligand. It could also be inferred that the other compounds
with a comparably long 6-position substituent, such as 4r and 4s, might lodge their chain in the same pocket
contacted by 4t. To better elucidate the interaction
pattern established by derivatives devoid of aromatic moieties at
the 6-position (4a–h,j and 5a–h,j), we also
docked compounds 4c and 4d in the RNase
H binding site. The resulting docking solutions are presented in Figure . Indeed, with the
exception of the missing contacts with Y501, the predicted binding
poses for the two ligands are largely congruent with what we detailed
for 4o. Here, the 4d carbonyl oxygen of
the acetyl group at the 6-position seems to be favorably positioned
to accept a H-bond from the backbone NH of Q500. However, in 4c, the trifluoride substituent would establish a weak H-bond
with the Q500 backbone NH. Thus, it could be inferred that only some
substituents, such as the acetyl and the nitro group in 4d and 4e, respectively, possess a spatial arrangement
that can tightly engage in this specific contact.
Figure 4
Predicted binding poses
of 4c (A) and 4d (B) in the HIV RNase H
binding site (PDB 3QIP). Important residues are labeled. 4c is represented
as salmon sticks, 4t as magenta
sticks, the magnesium ions are depicted in purple, and their coordination
with the nearby atoms is also represented in purple. H-bonds are depicted
as dashed green lines. Charge-transfer interactions are represented
as dashed red lines. The protein is depicted as blue ribbons and sticks.
Predicted binding poses
ofn class="Chemical">4c (A) and 4d (B) in the HIV RNase H
binding site (PDB 3QIP). Important residues are labeled. 4c is represented
as salmon sticks, 4t as magenta
sticks, the magnesium ions are depicted in purple, and their coordination
with the nearby atoms is also represented in purple. H-bonds are depicted
as dashed green lines. Charge-transfer interactions are represented
as dashed red lines. The protein is depicted as blue ribbons and sticks.
Arguably, the described interacting points (i.e., a strong H-bond-accepting atom and/or an aromatic
ring) in position
6 of the quinolone core should provide anchoring points that help
in stabilizing the chelation geometry. Thus, these considerations
could account for the weaker activity, or complete lack thereof, of
the derivatives 4a–c,f–h and 5a–c,f–h.
Site-Directed Mutagenesis
To experimentally verify
the binding model suggested by computational studies, compounds 4o and 4t were tested against the RNase H activity
of several point-mutants ofHIV-1RNase H, generated by independently
introducing an alanine substitution at residues R448, K451, N474,
Q475, Y501, W535, and K540 (Figure ). An activity curve was performed for all of the enzymes
(Figure S1), and a concentration was chosen
in the linear dose-response range to perform the enzymatic-inhibition
assays. In agreement with the proposed binding pose, results showed
that the potency of inhibition of compound 4o was significantly
affected when tested against all of the mutated enzymes (Figure A) with a loss of
potency of 3.7-fold against R448A, 2.2-fold against K451A, 6.0-fold
against K540A, 16.3-fold against N474A, and completely losing inhibitory
activity against the mutants Q475A, Y501A, and W535A, with a loss
of potency greater than 19.1-fold (see the Supporting Information Table S1). According to the binding model, results
showed that compound 4t binds in a slightly different
orientation, establishing a less extended network of interactions:
its inhibitory activity was not affected by the presence ofR448A,
K451A, N474A, or K540A substitutions (Figure A) (p value > 0.05; see
the Supporting Information Table S2), while
it lost 8.2-fold potency against Q475A RT and was totally inactive
against the RNase Hfunction ofY501A and W535A RTs, with a decrease
in potency greater than >52.4-fold. Interestingly, the analyzed
residues
are among the most conserved toward naïve and treated patients,[15] with a degree of conservation up to 99.9% for
Y501, N474, and Y501 and greater than 100% for W535 and Q475, thereby
increasing the evidence that the design of inhibitors targeting conserved
regions within the RNase H active site is a possible path for lead
development.
Figure 5
Inhibition of HIV-1 RT-associated RNase H activity of
mutated HIV-1
RTs by quinolinonyl non-diketo acid derivatives. Panel A: 4o; panel B: 4t.
Inhibition ofn class="Species">HIV-1 RT-associated RNase H activity of
mutated HIV-1
RTs by quinolinonyl non-diketo acid derivatives. Panel A: 4o; panel B: 4t.
Investigation of Magnesium Complexation
To investigate
the potential importance of the interaction between the active compounds
and Mg2+ ions, spectrophotometric complexation studies
were carried out on the best active derivatives 4o and 5o. Titration of compound 5o with MgCl2 produced an increase in absorbance at 246 nm (Figure panel B), thereby indicating coordination
of this compound with Mg2+. Moreover, as observed in Job’s
plot (Figure , panel
C), the stoichiometry of the complex 5o–Mg2+ is 1:1. Similarly, for acid 4o, an increase
in absorbance was observed at 256 nm (see the Supporting Information Figure S2 panel B), supporting our hypothesis
according to which the shortening of the DKA branch into a carboxylic
acid function could, together with the ketone group of the quinolinone
ring, interact with the Mg2+ ions. Also for derivative 4o, we observed a stoichiometry of 1:1 for the complex 4o–Mg2+(see the Supporting Information Figure S1 panel C).
Figure 6
[A] UV spectra of 5o in EtOH 3.81 10–5 M (black trace) and 5o (3.81 10–5 M) + MgCl2 (3.81
10–3 M) (red trace).
[B] Increments of A at 246 nm obtained during the titration of 5o with MgCl2. [C] Job’s plot obtained for 5o and MgCl2. ΔA at 246
nm was plotted vs the molar ratio of 5o. The maximum ΔA was observed at X = 0.54,
which corresponds to a stoichiometry of 1:1 for the complex 5o–Mg2+.
[A] UV spectra of5o in EtOH 3.81 10–5 M (black trace) and 5o (3.81 10–5 M) + MgCl2 (3.81
10–3 M) (red trace).
[B] Increments of A at 246 nm obtained during the titration of5o with MgCl2. [C] Job’s plot obtained for 5o and MgCl2. ΔA at 246
nm was plotted vs the molar ratio of5o. The maximum ΔA was observed at X = 0.54,
which corresponds to a stoichiometry of 1:1 for the complex 5o–Mg2+.
Conclusions
In this work, we reported a new series of quinolonyl
non-DKA derivatives
as RHIs. This new class of compounds was conceived by shortening the
quinolonyl DKA chain typical of INSTI into a carboxylic acidfunction
that, together with the ketone group in the 4-position of the quinolinone
ring, could still chelate the two Mg2+ ions. The newly
designed compounds showed activity against RH and were also confirmed
to be able to chelate the magnesium ions by spectrophotometric complexation
studies.Among the newly synthesized derivatives, arylmethyloxy
acid quinolonyl
derivatives demonstrated inhibitory activities within the micromolar/low
micromolar range, resulting in IC50 values lower than that
of the ester counterparts, with compound 4o being the
most potent acid derivative (IC50 = 1.51 μM). Docking
studies within the RNase H catalytic site highlighted a possible reason
for this trend of activity. Indeed, the 4-carboxyl group is localized
in a highly polar section of the binding site. As a result, the esters
would place their lipophilic ethyl chain toward this highly hydrophilic
area and this should unfavorably impact the binding affinity. Interestingly,
this trend was also observed in acutely infected cells, with derivative 4t being the best acting compound (EC50 = 1.73
μM) with no cytotoxic activity. Site-directed mutagenesis experiments
confirmed the docking calculations, demonstrating also that our compounds
are capable of interacting with amino acids highly conserved among
naïve and treated patients. Overall, these results confirmed
the effectiveness of this class of quinolonyl non-DKA derivatives
as new RHIs.It is worth noting that the quinolonyl DKAs were
in general active
at low nanomolar concentrations against IN, showing marginal activity
against RH. Conversely, the new quinolonyl non-DKAs were good RH inhibitors,
with marginal activity against IN, as hypothesized in the rationale.
This series was nonselective against RDDP of the RT so that the antiviral
activity that resulted can be ascribed to inhibiting effects of both
RT functions.
Experimental Section
General
Melting points were determined on a Bobby Stuart
Scientific SMP1 melting point apparatus and are uncorrected. Compound
purity was always >95% as determined by combustion analysis. Analytical
results agreed to within ±0.40% of the theoretical values. IR
spectra were recorded on a PerkinElmer Spectrum-One spectrophotometer. 1HNMR spectra were recorded at 400 MHz on a Bruker AC 400
Ultrashield 10 spectrophotometer (400 MHz). Dimethyl sulfoxide-d6 99.9% (CAS 2206–27–1), deuterochloroform
98.8% (CAS 865–49–6), and acetone-d6 99.9% (CAS 666–52–4) of isotopic purity
(Aldrich) were used. Column chromatographies were performed on silica
gel (Merck; 70–230 mesh). All compounds were routinely checked
on thin-layer chromatography (TLC) using aluminum-baked silica gel
plates (Fluka DC-Alufolien Kieselgel 60 F254). Developed
plates were visualized by UV light. Solvents were of reagent grade
and, when necessary, were purified and dried by standard methods.
Concentration of solutions after reactions and extractions involved
the use of a rotary evaporator (Büchi) operating at a reduced
pressure (ca. 20 Torr). Organic solutions were dried over anhydrous
sodium sulfate (Merck). All solvents were freshly distilled under
nitrogen and stored over molecular sieves for at least 3 h prior to
use. Analytical results agreed to within ±0.40% of the theoretical
values.
General Experimental Procedures
General Procedure A (GP-A)
to Obtain O-Alkyl
Derivatives 10a–j
To a mixture of diethyl
2-(((4-hydroxyphenyl)amino)methylene)malonate (17.9 mmol) in anhydrous
DMF (80 mL), t-BuOK (26.9 mmol) and the proper halide
(26.9 mmol) were added at 0 °C. Then, the resulting mixture was
stirred at room temperature for the proper time and monitored by TLC.The reaction was quenched with water (30 mL) and extracted with
ethyl acetate (3 × 50 mL). The combined organic layers were washed
with brine (200 mL), dried over Na2SO4, concentrated,
and purified by column chromatography on silica gel. For derivate 10a, the alkylating agent was obtained as a free base, upon
the reaction between the commercially available 3-dimethylamino-1-propyl
chloride hydrochloride and triethylamine in anhydrous THF. For each
compound, alkylating agent; reaction time; chromatography eluent;
recrystallization solvent; yield (%); melting point (°C); IR; 1HNMR; and elemental analysis are reported.
General
Procedure B (GP-B) to Obtain Quinolinonyl Ester Derivatives
(8b,k,j,l–t)
The proper substituted anilidomethylenemalonic
ester (19.0 mmol) was suspended in diphenyl ether (0.304 mol, 48 mL),
stirred under reflux for the proper time, and monitored by TLC. Upon
completion of the reaction, the mixture was cooled down and poured
into n-hexane (50 mL). The resulting precipitate
was filtered, washed three times with n-hexane (10
mL) and petroleum ether, and dried under an IR lamp to afford the
pure product. For each compound, reaction time; recrystallization
solvent; yield (%); melting point (°C); IR; 1HNMR;
and elemental analysis are reported.
General Procedure C (GP-C)
to Obtain N-Alkyl
Quinolinonyl Ester Derivatives (5a–t)
A mixture of the appropriate quinolinonyl ester (15.0 mmol), 4-fluorobenzyl
bromide (45.0 mol), and K2CO3 anhydrous (21.0
mmol) in DMF anhydrous (130 mL) was stirred at 100 °C for the
proper time and monitored by TLC. The mixture was cooled, treated
with water (40 mL), and extracted with ethyl acetate (3 × 100
mL). The organic layer was washed with brine (200 mL), dried over
anhydrous sodium sulfate, and concentrated under vacuum. The crude
product was purified by chromatography (SiO2) to afford
the pure product. For each compound, reaction time; chromatography
eluent; recrystallization solvent; yield (%); melting point (°C);
IR; 1HNMR; and elemental analysis are reported.
General
Procedure D (GP-D) to Obtain Quinolinonyl Carboxylic
Acid Derivatives (4a–t)
A solution ofNaOH 20% (0.172 mol) in distilled water was added to a suspension
of the appropriate ester (0.010 mol) in 1:1 THF/ethanol (50 mL), and
the reaction was stirred vigorously under reflux for the proper time.
The reaction was monitored by TLC. Upon completion of the reaction,
the organic phase was removed under vacuum and the resulting suspension
was acidified with 1 NHCl (pH 4–5). The resulting solid was
filtered, washed with water, and dried under an IR lamp to afford
the product of interest. For 4b, 0.4 M LiOH (50.0 mmol)
was used as a base instead ofNaOH and the reaction was performed
at room temperature. For each compound, reaction time; recrystallization
solvent; yield (%); melting point (°C); IR; 1HNMR;
and elemental analysis are reported.
The X-ray crystal structure of the
HIV-1RNase H protein (PDB code 3QIP)[53] was downloaded
from the Protein Data Bank.[59,60] The “Protein
Preparation Wizard” tool of Maestro[61] was employed to prepare the protein structure for the subsequent
docking calculations. Specifically, hydrogens atoms were added according
to their appropriate protonation states (pH = 7.4) and minimized,
bond orders and disulfide bonds were calculated, and every water molecule
was removed except for two conserved water molecules that participate
in the magnesium ion chelation. The derivatives 4c,d,o,t were virtually created through
the Maestro 2D Sketcher tool. Then, with Ligprep, another utility
present in the Schrödinger suite, also the ligands were prepared
for the docking experiments, specifying that the ligands would participate
in a chelation process. In particular, every hydrogen atom was added,
and the appropriate tautomeric and ionization states were generated.
Subsequently, the obtained ligands underwent a minimization with the
OPLS3e force field.[62] Then, the ligands
were docked using Glide, part of the Schrödinger suite,[54] which is a grid-based docking software that
takes advantage of an energetic approach to find favorable interactions
in the ligand–receptor complex. Multiple sets offields are
precalculated to render on a grid the structure of the receptor, along
with its properties, to provide an accurate score of the ligand binding
pose. For the grid generation, the grid box was placed at the mass
center of the cocrystal ligand. This box provides a precise measure
of the actual size of the search space. Nevertheless, ligands are
allowed to leave the box boundaries in the course of grid minimization.
The length of the outer box was fixed at 22 Å for the X, Y, and Z coordinates,
while for the inner box, the default values of 10 Å were left.
In the grid generation settings, it was specified that the ligand
would preferentially chelate the magnesium ions. To dock compounds 4o and 4t, we employed Glide Induced Fit routine,[55] which allows one to predict possible conformational
changes in the binding site residues induced by the ligand binding
process. Thus, during the docking process, an increased Coulomb–vdW
cutoff and a reduced van der Waals radii are employed, along with
a temporary removal of some of the most flexible side chains, to finally
yield a diverse ensemble of ligand poses. For each of the generated
poses, the residues’ side chains in the ligand proximity are
reoriented to accommodate the ligand. Then, the ligand and the nearby
residues undergo a minimization process. Finally, each ligand pose
is redocked into its own minimized macromolecule, with each complex
ranked according to its GlideScore. To dock compounds 4c and 4d, the standard Glide protocol with a rigid treatment
of the protein was employed. On the whole, standard settings were
used. The best-scoring complexes in terms of GlideScore were selected.
All of the figures were rendered with the UCSF Chimera package.[63]
Magnesium Complexation Study
Complexation
studies with
MgCl2 were carried out on compounds 4o and 5o. The effects of the magnesium ion were evaluated by a spectrophotometric
method, using a Perkin Elmer Lambda 40 UV–vis spectrophotometer
and a Hellma quartz cuvette with a 1 cm optical path. Magnesium chloride
(1 M solution) was purchased from Sigma-Aldrich (Milano, Italy) and
was diluted with absolute ethanol to obtain stock solutions ranging
from 4 10–5 to 0.2 M.For the titration experiments,
each studied compound was dissolved in 50–100 mL of absolute
ethanol to the final concentration of 3.8 × 10–5 M for compound 4o and 4.1 × 10–5 M for compound 5o. Each solution (3 mL) was placed
in a cuvette, and the UV–vis spectrum was recorded between
230 and 370 or 600 nm using ethanol as the reference. Thereafter,
small volumes (10–15 μL) of the appropriate MgCl2ethanolic stock solution were added both in the sample and
in the reference cuvettes and the UV–vis spectra were recorded;
the Mg2+ concentration in the solutions was increased from
0 to 100–200-fold with respect to the studied compound, in
about 20 consecutive increments. Each experiment was conducted in
triplicate.To determine the stoichiometric coefficients of
the complexes,
Job’s method[64] was used, which requires
mixing, in appropriate proportions, equimolar solutions ofmetal ion
and ligand, so that the final volume and the total moles present in
the cuvette are equal for each measurement. The absorbance values
were recorded at the wavelength where the higher difference in absorbance
was observed in the UV–Vis titration experiments (320 nm for
compound 4o and 246 nm for compound 5o).
For each obtained absorbance value, the nominal absorbance values
of equimolar solutions of the metal and ligand were subtracted, to
obtain the ΔA due exclusively to the complex
formation.The resulting ΔA were reported
in a graph
as a function of the mole fraction of the ligand; the mole fraction X, which caused the maximum variation in absorbance, was
determined by linear regression analysis and used to calculate the
value of the coefficient n, which corresponds to
the number of ligand molecules per cation, applying the following
equation
Expression
and Purification of Recombinant HIV-1 RTs
His-tagged p66/n class="Gene">p51
HIV-1 RT group M subtype B coded in a p6HRT-prot
plasmid was expressed in the Escherichia coli strain M15.[65] Heterodimeric RTs were
expressed essentially and purified as described.[66]
Expression and Purification of Recombinant
HIV-1 IN
His-tagged NL4-3 IN was expressed from a pET15b
plasmid in the E. coli BL21 pLys strain.
Recombinant protein was
purified as previously described[67] following
a batch preparation on Ni-NTA beads (Qiagen, Paris, France).
Site-Directed
Mutagenesis
Amino acid substitutions
were introduced into the p66HIV-1 RT subunit of a HIV-1 RT using
a QuikChange mutagenesis kit, following the manufacturer’s
instructions (Agilent Technologies Inc., Santa Clara, CA).
RNase
H Polymerase-Independent Cleavage Assay
The HIV-1
RT-associated RNase H activity was measured as described.[68] Briefly, the reaction was carried out in a black
96-well plate in a total volume of 100 μL. Serial dilutions
of compounds were added to the reaction mix containing 50 mM TrisHCl pH 7.8, 6 mM MgCl2, 1 mM dithiothreitol (DTT), 80 mM
KCl, 250 nM hybrid RNA/DNA (50-GTTTTCTTTTCCCCCCTGAC-30-Fluorescein,
50-CAAAAG AAAAGGGGGGACUG-30-Dabcyl). The reaction was started by the
addition of 20 ng ofHIV-1 wt RT, 20 ng ofR448A RT, 20 ng ofK451A
RT, 40 ng of K540 RT, 60 ng ofQ475A RT, 500 ng ofN474A RT, 500 ng
ofY501A RT, and 500 ng of W535A RT and incubated for 1 h at 37 °C.
Products were quantified with a Perkin–Elmer Victor 3 multilabel
counter plate reader at an excitation–emission wavelength of
490/528 nm. Experiments were performed in duplicate and replicated
at least two times. Data were analyzed as described.[36] Mean ± standard deviation of IC50 values
were determined, and p values were calculated between the IC50 value against the wt and IC50 value against the mutants
by paired, two-tailed t tests using GraphPad Prism
6.01 software (GraphPad Software, Inc.; San Diego, CA). Figures were
made with GraphPad Prism 6 version 6.01.
Polymerase Assay
The HIV-1 RT-associated RNA-dependent
DNA polymerase activity was measured as described.[65] Briefly, the reaction was carried out in a black 96-well
plate in a total volume of 25 μL. Serial dilutions of compounds
were added to the reaction mix containing 60 mM Tris HCl buffer pH
8.1, 8 mM MgCl2, 60 mM KCl, 13 mM DTT, 2.5 mM poly(A)-oligo(dT),
and 100 mM dTTP. The reaction was started by the addition of 20 ng
ofHIV-1 wt RT and incubated for 30 min at 37 °C, followed by
addition of 2 mL of 200 mM ethylenediamine tetraacetic acid (EDTA).
Reaction products were detected by addition of 170 mL of revealing
solution containing Picogreen in 10 mM Tris HCl pH 7.5, 1 mM EDTA,
and measured with a multilabel counter plate reader Victor 3, equipped
with filters 502/523 nm (excitation/emission wavelength).
IN Assay
The DNA substrate was generated by annealing
an equimolar amount of 19T (GTGTGGAAAATCTCTAGCA) and 21B (ACTGCTAGAGATTTTCCACAC).
Both oligonucleotides were purchased from Integrated DNA Technologies,
Inc. (Coralville, IA), and the gel was purified in-house. ST reactions
were performed by adding molecules or an equivalent volume of 100%
dimethyl sulfoxide (DMSO, used as the drug solvent) to a mixture of
20 nM duplex DNA substrate and 400 nM IN in 50 mM MOPS pH 7.2, 7.5
mM MgCl2, and 14 mM 2-mercaptoethanol. Mixtures were incubated
at 37 °C for 2 h, and the reaction was quenched by addition of
an equal volume of loading buffer [formamide containing 1% sodium
dodecyl sulfate (SDS), 0.25% bromophenol blue, and xylene cyanol].
Reaction products were separated in 16% polyacrylamide denaturing
sequencing gels. Dried gels were visualized using a FLA5000 (Fuji
Photo Film, Tokyo, Japan). Densitometric analyses were performed using
ImageQuant 5.1 software from GE Healthcare. Data analyses (linear
regression, IC50 determination, and standard deviation)
were performed using Prism 6.07 software from GraphPad (San Diego,
CA).
Cell-Based Assays
HIV-1 replication was monitored using
HeLa-CD4-LTR-β-gal reporter cells as previously described.[69] Briefly, β-galactosidase activity was
monitored 24 h post infection using a replication-competent HIV produced
in-house (H9-laï coculture) and a serial dilution of molecules
or an equivalent volume ofDMSO. In parallel, compounds’ toxicity
was determined at 24 h in the same cell line using CellTiter 96 AQueous
One Solution Cell Proliferation Assay (Promega, France).
Authors: Silvia Salerno; Aída Nelly García-Argáez; Elisabetta Barresi; Sabrina Taliani; Francesca Simorini; Concettina La Motta; Giorgio Amendola; Stefano Tomassi; Sandro Cosconati; Ettore Novellino; Federico Da Settimo; Anna Maria Marini; Lisa Dalla Via Journal: Eur J Med Chem Date: 2018-03-07 Impact factor: 6.514
Authors: Yingzhi Bi; Patrick Stoy; Leonard Adam; Bin He; John Krupinski; Diane Normandin; Ron Pongrac; Laurie Seliger; Andrew Watson; John E Macor Journal: Bioorg Med Chem Lett Date: 2004-03-22 Impact factor: 2.823
Authors: Roberta Costi; Roberto Di Santo; Marino Artico; Alessandra Roux; Rino Ragno; Silvio Massa; Enzo Tramontano; Massimiliano La Colla; Roberta Loddo; M Elena Marongiu; Alessandra Pani; Paolo La Colla Journal: Bioorg Med Chem Lett Date: 2004-04-05 Impact factor: 2.823
Figure 1
Figure 2
Chart 1
Scheme 1
Scheme 2
Figure 3
Figure 4
Figure 5
Figure 6