There are currently three HIV-1 integrase (IN) strand transfer inhibitors (INSTIs) approved by the FDA for the treatment of AIDS. However, the emergence of drug-resistant mutants emphasizes the need to develop additional agents that have improved efficacies against the existent resistant mutants. As reported herein, we modified our recently disclosed 1-hydroxy-2-oxo-1,2-dihydro-1,8-naphthyridine-3-carboxamides IN inhibitors to develop compounds that have improved efficacies against recombinant IN in biochemical assays. These new compounds show single-digit nanomolar antiviral potencies against HIV vectors that carry wild-type (WT) IN in a single round replication assay and have improved potency against vectors harboring the major forms of drug resistant IN mutants. These compounds also have low toxicity for cultured cells, which in several cases, results in selectivity indices (CC50/EC50) of greater than 10000. The compounds have the potential, with additional structural modifications, to yield clinical agents that are effective against the known strains of resistant viruses.
There are currently three HIV-1 integrase (IN) strand transfer inhibitors (INSTIs) approved by the FDA for the treatment of AIDS. However, the emergence of drug-resistant mutants emphasizes the need to develop additional agents that have improved efficacies against the existent resistant mutants. As reported herein, we modified our recently disclosed 1-hydroxy-2-oxo-1,2-dihydro-1,8-naphthyridine-3-carboxamides IN inhibitors to develop compounds that have improved efficacies against recombinant IN in biochemical assays. These new compounds show single-digit nanomolar antiviral potencies against HIV vectors that carry wild-type (WT) IN in a single round replication assay and have improved potency against vectors harboring the major forms of drug resistant IN mutants. These compounds also have low toxicity for cultured cells, which in several cases, results in selectivity indices (CC50/EC50) of greater than 10000. The compounds have the potential, with additional structural modifications, to yield clinical agents that are effective against the known strains of resistant viruses.
Acquired immunodeficiency
syndrome (AIDS) is an infectious disease
caused by the human immunodeficiency virus (HIV). Reverse transcriptase
(RT), protease (PR), and integrase (IN) are the three viral enzymes
that are required for viral replication and all three have been targeted
by anti-AIDS therapeutics. IN catalyzes the insertion of viral DNA
into the host genome in two sequential steps, termed “3′-processing”
(3′-P) and “strand transfer” (ST). The 3′-P
reaction cleaves two nucleotides from the 3′ end of the viral
DNA, exposing a deoxycytosine residue that is used in a nucleophilic
attack on the host DNA in the ST reaction. Both of these reactions
involve two Mg2+ ions held in place by three acidic residues
Asp64, Asp116, and Glu152 that collectively constitute the “DDE”
motif.[1] IN inhibitors are the most recently
developed class of anti-AIDS drugs. Merck’s raltegravir (RAL, 1, Figure 1) (October 2007)[2] and Gilead’s elvitegravir (EVG) (August
2012)[3] were the first two IN inhibitors
to be approved by the FDA. The approved IN inhibitors selectively
block the ST step, and members of this class of drugs are called “IN
strand transfer inhibitors” (INSTIs) because of their ability
to preferentially block the enzyme’s ST reaction relative to
the 3′-P reaction.[4] All of the known
INSTIs share important structural features, which include a coplanar
arrangement of three heteroatoms that chelate the two catalytic Mg2+ ions, and a halobenzyl ring that binds to the penultimate
base (a deoxycytidine) adjacent to the deoxyadenosine that lies at
the 3′ end of the viral DNA after the 3′-P reaction.
Binding of INSTIs blocks the ST reaction by displacing the viral 3′-terminal
deoxyadenosine from the catalytic Mg2+ ions. Treatment
with 1 and EVG selects for resistant forms of HIV, and
there is considerable cross-resistance to these two drugs. GlaxoSmithKline’s
dolutegravir (DTG, 2, Figure 1)[5,6] is a recently FDA-approved second-generation INSTI
(August 2013), which shows improved efficacies against RAL and EVG-resistant
strains of HIV.[7,8] However, 2 also selects
for resistant strains of HIV.[7] This emphasizes
the need to develop agents that can overcome resistant strains of
IN, including the emerging strains resistant to 2. We
recently reported that 1-hydroxy-1,8-naphthyridin-2H-one-3-carboxamides, which include both 4-unsubstituted and 4-hydroxyl-containing
analogues (3 and 4, respectively, Figure 1), potently inhibit wild-type (WT) IN in biochemical
assays and show good antiviral efficacies in single-round infection
assays of HIV-1 infectivity (Figure 1).[9] Importantly, members of this series retain good
antiviral potency against a set of mutants resistant to 1 in these latter assays. Here we describe structural variation at
the 4-position of compound 3, which yielded agents of
type 5 (Figure 1) that enhance
their efficacy against additional mutant forms of IN that are resistant
to 1.
Figure 1
Structures of HIV-1 integrase inhibitors described in
the text.
Mg2+-chelating heteroatoms are shown in red, with the 4-position
of the 1-hydroxy-1,8-naphthyridin-2(1H)-one ring
being indicated in blue.
Structures of HIV-1 integrase inhibitors described in
the text.
Mg2+-chelating heteroatoms are shown in red, with the 4-position
of the 1-hydroxy-1,8-naphthyridin-2(1H)-one ring
being indicated in blue.
Results and Discussion
Inhibitor Design
IN is a member
of the polynucleotidyl
transferase class of enzymes that share similar catalytic mechanisms.[10] There is also a large body of data that describes
the known INSTIs, their efficacy against WT and drug-resistant forms
of HIV, and their interactions with prototype foamy virus (PFV) IN.
Accordingly, the structures of previously described inhibitors can
be used to aid the design of new anti-IN compounds. This is exemplified
by the development of 1 by Merck, which can be traced
to dihydroxypyrimidine carboxamide inhibitors of the hepatitis C virus
NS5b RNA polymerase (RNAP).[10,11] Inhibitors of HIV-1
ribonuclease H (RNase H) have been reported, which, like INSTIs, inhibit
their target enzyme by chelating two Mg2+ ions in the enzyme
active site. In RNase H, the Mg2+ ions are held in place
by a “DEDD motif” (D443, E478, D498, and D549).[12−14] Compounds that inhibit both IN and RNase H have been reported.[15,16] By analogy to the RNAP example, the known RNase H inhibitors might
provide insights that can be used to design improved INSTI inhibitors.
Accordingly, the design of our bicyclic 1-hydroxy-1,8-naphthyridin-2H-one INSTIs (3 and 4)[9] was guided by the report that compounds such
as the biaryl-containing 6 (Figure 2), which have submicromolar inhibitory potency against RNase H (IC50 = 0.64 μM), also have low micromolar inhibitory potency
against HIV-1 IN (ST IC50 = 2.4 μM).[14] In the case of 6, the reported EC50 value (half-maximal concentration providing protection against viral-induced
cell death) in a HIV-1 HXB2 single-cycle viral replication assay in
HeLa P4-2 cells was 34 μM.[14]
Figure 2
HIV-1 ribonuclease
H (RNase H) inhibitors 6 and integrase
inhibitor 5a with the core metal-chelating 1-hydroxy-1,8-naphthyridin-2(1H)-one system shown in red and the key 4-amino group indicated
in blue.
HIV-1 ribonuclease
H (RNase H) inhibitors 6 and integrase
inhibitor 5a with the core metal-chelating 1-hydroxy-1,8-naphthyridin-2(1H)-one system shown in red and the key 4-amino group indicated
in blue.Metal chelation by the 1-hydroxy-1,8-naphthyridin-2H-one nucleus, which is common to 3, 4,
and 6, can theoretically be achieved via the heteroatom
triad formed by the N-hydroxyl group, the 2-oxo group,
and the 8-naphthyridine nitrogen. However, an important component
of 3 and 4, not found in 6,
is a halobenzyl group, which is known to be important for binding
to IN by interacting with the penultimate deoxycytosine in the 3′-end
of enzyme-bound viral DNA.[1] In the case
of 3 and 4, this binding function is served
by a 2′,4′-difluorobenzyl carboxamide group, which is
appended at the 3-position of the bicyclic nucleus. Previous work
has shown that the nature and pattern of halogen phenyl substitution
can significantly affect the potency of INSTIs.[17] In developing 3 and 4, we found
that a 2′,4′-difluorobenzyl moiety, which is present
in 2, was superior to the other halobenzyl rings we tested.[9] An important feature of 3 and 4 is that the carbonyl oxygen of the 2′,4′-difluorobenzyl
amide group may not be an obligatory component of the metal-chelating
triad. As a consequence, there may be greater flexibility in this
region of the molecule than is found with inhibitors, such as 1, where the halobenzyl amide carbonyl participates in Mg2+ chelation. This flexibility is reminiscent of what is seen
with 2, where the flexibility of the haloamide component
is thought to contribute to the ability of 2 to maintain
efficacy against certain forms of IN that are resistant to 1.[18−20]In our current work, we further modified the 1-hydroxy-1,8-naphthyridin-2H-one nucleus by incorporating new functionalities at the
4-position. In undertaking these efforts, we noted that for RNase
H inhibitors such as 6, an extended aryl functionality
increased their inhibitory potency.[14] Therefore,
we began by preparing inhibitor 5a (Figure 2). In contrast to 6, where the aryl functionality
is attached through a methylene unit, for reasons of synthetic simplicity,
we employed a 4-amine group in 5a. Subsequently, we prepared
a series of analogues (5a–5v) using
an iterative process of design, synthesis, biological evaluation,
and redesign.
Synthesis
Amidation of methyl ester 7 (obtained
in three steps from commercially available methyl 2-fluoronicotinate)[9,21] using 2,4-difluorobenzylamine gave the known amide 8 in 70% yield (Scheme 1).[9,21] Subsequent
reaction with toluenesulfonyl chloride produced the tosylated analogue 9 (93% yield), which was treated with a variety of amines
to provide 10a–10v (Scheme 1). A subset of these amines (10a–10u) was converted to final products (5a–5u) by hydrogenolysis of the N-benzyloxy
group (H2/10% Pd·C). In the case of final product 5p, acetylation of intermediate 10o to yield 10p was done prior to debenzylation. For final product 5v, treatment of intermediate 10v with TFA yielded
the free amine 11 prior to debenzylation (Scheme 1).
Scheme 1
Synthesis of 4-Amino-N-(2,4-difluorobenzyl)-1-hydroxy-2-oxo-1,2-dihydro-1,8-naphthyridine-3-carboxamides 5a–5v
Reagents and conditions: (i)
2,4-diFBnNH2; (ii) TsCl, TEA, MeCN; (iii) R1R2NH or R1R2NH-HCl, DIEA, DMF; (iv)
Ac2O, TEA, DCM; (v) TFA, DCM; (vi) H2, 10% Pd·C,
MeOH.
Synthesis of 4-Amino-N-(2,4-difluorobenzyl)-1-hydroxy-2-oxo-1,2-dihydro-1,8-naphthyridine-3-carboxamides 5a–5v
Reagents and conditions: (i)
2,4-diFBnNH2; (ii) TsCl, TEA, MeCN; (iii) R1R2NH or R1R2NH-HCl, DIEA, DMF; (iv)
Ac2O, TEA, DCM; (v) TFA, DCM; (vi) H2, 10% Pd·C,
MeOH.
Biological Evaluation
Compounds
were evaluated in biochemical
assays using radiolabeled oligonucleotides to measure their inhibitory
potential in the 3′-P and ST reactions.[9,22] The
initial series of compounds was designed to examine the role of aromatic
functionality at the 4-position. The IN ST inhibitory potency of 5a (IC50 = 0.34 ± 0.08 μM, Table 1) was approximately 10-fold better than the value
previously reported for 6, which contains a similar 1,4-phenyl
group.[14] The conformationally constrained
biphenyl amine analogue 5b was slightly less potent than 5a. In contrast, introducing a 4′-nitrile or a 4′-amino
group onto 5a (giving 5c and 5d, respectively) slightly increased the potency of the compound in
the ST reaction relative to 5a. Importantly, shortening
the 4-substituent by removal of one phenyl ring gave an approximate
19-fold enhancement in potency relative to 5a (5e, ST IC50 = 0.018 ± 0.006 μM, Table 1).
Table 1
Inhibitory Potencies
of Carboxamides 5a–5e Obtained Using
an in Vitro IN Assaya
Assays were performed using a gel-based
protocol with Mg2+ cofactor as describe in ref (22).
Assays were performed using a gel-based
protocol with Mg2+ cofactor as describe in ref (22).Antiviral potencies were evaluated in a cell-based
assay using
lentiviral vectors carrying WT IN as well as mutant forms of IN that
are resistant to 1, Y143R, N155H, and the double mutant,
G140S/Q148H.[23−25] In these assays, amine 5a was approximately
two orders-of-magnitude more potent against the WT enzyme (EC50 = 372 ± 63 nM, Table 2) than
what has been reported for 6 (32 μM).[14] All members of the series (5a–5e) showed nanomolar ST inhibitory potencies against WT enzyme,
with 5d and 5e being significantly more
potent (EC50 = 6.3 ± 2.4 and 14 ± 1.9 nM, respectively)
than other members of the series (EC50 values >100 nM).
Because all compounds showed no cytotoxicity up to 250 μM),
selectivity indices (SI = CC50/EC50) were from
at least 500 to greater than approximately 40000 (Table 2). Of particular note, while compound 5e was
only slightly less potent against the WT vector than what has been
reported for 1 (EC50 = 4 ± 2 nM), it
was significantly less susceptible to loss of efficacy against the
mutants: Y143R (2-fold loss versus a reported 54-fold loss for 1), N155H (8-fold loss versus a reported 39-fold loss for 1), and G140S/Q148H (32-fold versus a reported 425-fold loss
for 1).[26] The large loss of
potency incurred by 1 against the Y143R mutant derives
from a loss of π–π stacking of the inhibitor with
the Y143 phenyl ring. The ability of compounds in the current series
to retain good efficacy against the Y143R mutant indicates that they
do not have a similar interaction as 1 with the aryl
ring of Y143.
Table 2
Antiviral Potencies of Carboxamides 5a–5e in Cells Infected with HIV-1 Constructs
Containing WT or Mutant IN
EC50 (FC, IN mutantsc)
compd
CC50 (μM)a
EC50 (nM, WT)b
Y143R
N155H
G140S/Q148H
SId
5a
>250
372 ± 63
1×
N/Ae
N/Ae
>672
5b
>250
171 ± 57
3×
N/Ae
N/Ae
>1462
5c
>250
123 ± 21
3×
N/Ae
N/Ae
>2033
5d
>250
6.3 ± 2.4
16×
67×
N/Ae
>39683
5e
>250
14 ± 1.9
2×
8×
32×
>17857
Cytotoxic concentration resulting
in 50% reduction in the level of ATP in human osteosarcoma (HOS) cells.
Values obtained from cells
infected
with lentiviral vector harboring WT IN.
Cells were infected with viral constructs
carrying IN mutations and indicated values correspond to the fold-change
(FC) in EC50 relative to WT.
Selectivity index calculated as
the ratio of CC50 to EC50.
Not available.
Cytotoxic concentration resulting
in 50% reduction in the level of ATP in humanosteosarcoma (HOS) cells.Values obtained from cells
infected
with lentiviral vector harboring WT IN.Cells were infected with viral constructs
carrying IN mutations and indicated values correspond to the fold-change
(FC) in EC50 relative to WT.Selectivity index calculated as
the ratio of CC50 to EC50.Not available.We prepared an additional series of analogues (5f–5v) in which several alkylamines were
introduced at the 4-position
(Table 3). In biochemical assays in vitro,
most of these analogues exhibited low nanomolar inhibitory potencies
in the ST reaction. However, compounds 5i and 5k, which contained cycloheptyl and n-butylphenyl
substituents, respectively, had ST IC50 values of 0.46
± 0.18 and 0.28 ± 0.11 μM, respectively, which were
markedly elevated relative to other members of the series. A third
member of the series, having an (S)-ethyl N-prolinate group, was also significantly less potent [(S)-5u, ST IC50 = 0.31 ± 0.04
μM] (Table 3).
Table 3
Inhibitory
Potencies of Carboxamides 5f–5v Obtained
Using an in Vitro IN Assaya
IC50 (μM)
compd
R1
R2
3′-processing
strand transfer
5f
–CH3
–H
3.7 ± 0.4
0.027 ± 0.004
5g
–CH3
–CH3
21 ± 2
0.087 ± 0.012
5h
–morpholino
77 ± 12
0.079 ± 0.013
5i
–cycloheptyl
–H
13 ± 1.1
0.46 ± 0.18
5j
–CH2CH2Ph
–H
8.0 ± 1.5
0.050 ± 0.012
5k
–CH2(CH2)3Ph
–H
12 ± 2.0
0.28 ± 0.11
5l
–CH2(CH2)3CH3
–H
8.2 ± 1.6
0.024 ± 0.009
5m
–CH(CH3)2
–H
1.8 ± 0.2
0.016 ± 0.004
5n
–CH2CH2NH2
–H
4.5 ± 0.2
0.039 ± 0.006
5o
–CH2CH2OH
–H
0.55 ± 0.07
0.010 ± 0.009
5p
–CH2CH2OAc
–H
5.3 ± 0.5
0.027 ± 0.006
5q
–CH2CO2CH3
–H
0.71 ± 0.10
0.021 ± 0.011
(S)-5r
–NHCH(CH3)CO2CH3
–H
7.4 ± 0.8
0.017 ± 0.011
(R)-5r
–NHCH(CH3)CO2CH3
–H
5.8 ± 0.6
0.027 ± 0.005
(S)-5s
–NHCH(Ph)CO2CH3
–H
16.7 ± 1.4
0.010 ± 0.002
(R)-5s
–NHCH(Ph)CO2CH3
–H
13.5 ± 1.0
0.0082 ± 0.0015
(S)-5t
–NHCH(CH2OH)CO2CH3
–H
5.8 ± 0.5
0.0084 ± 0.0032
(R)-5t
–NHCH(CH2OH)CO2CH3
–H
4.4 ± 0.5
0.013 ± 0.04
(S)-5u
–Pro-OEt
86 ± 6
0.31 ± 0.04
5v
–H
–H
2.5 ± 0.3
0.019 ± 0.002
Assays were performed
using a gel-based
protocol with Mg2+ cofactor as describe in refs (9) and (22).
Assays were performed
using a gel-based
protocol with Mg2+ cofactor as describe in refs (9) and (22).Antiviral potencies were determined for 5f–5v in cells infected with viral vectors harboring
WT and mutant
forms of IN (Table 4). Most compounds of the
series displayed EC50 values in the low nanomolar range
against the WT vector, with a majority of the compounds showing single-digit
nanomolar potencies. These compounds also showed low cytotoxicity,
which resulted in good SI values, with several compounds showing SI
> 10000. Noteworthy exceptions were compounds 5g, 5h, and (S)-5u, which not only
had significantly reduced antiviral potencies (EC50 values
of 268 ± 8, 1200 ± 260, and 590 ± 72 nM, respectively)
but also showed greater cytotoxicity (CC50 values of 13
± 1.8, 8.4 ± 3.2, and 18 ± 7 μM, respectively)
(Table 4). These latter compounds are the only
members of the series having tertiary amines at the 4-position. As
such, these analogues would not be able to form internal hydrogen
bonds between their 4-amino groups and the 3-carboxamide carbonyl
oxygen. In spite of their poor antiviral efficacies, the in vitro
ST IC50 values for 5g and 5h (87
and 79 nM, respectively, Table 3) were only
modestly elevated relative to most other members of the series (typically
30 nM or lower). In some cases, the in vitro IC50 values
for 5g and 5h were better than compounds
such as 5i and 5k (460 and 280 nM, respectively,
Table 3), which paradoxically exhibited good
EC50 values against the WT vector (12 ± 3 and 50 ±
13 nM, respectively, Table 4). These data could
indicate that being able to form an intramolecular hydrogen bond between
the 4-amino group and the 3-carboxamide carbonyl oxygen has a more
important role for the activity of the compounds in an antiviral assay
done in cultured cells than in the biochemical assay done in vitro.
Table 4
Antiviral Potencies of Carboxamides 5f–5v in Cells Infected with HIV-1 Constructs
Containing WT or Mutant IN
EC50 (nM, IN mutantsc)
compd
CC50 (μM)a
EC50 (nM, WT)b
Y143R
N155H
G140S/Q148H
SId
5f
>250
3.1 ± 0.6
6.1 ± 2.5 (2×)
18 ± 5.2 (6×)
87 ± 11 (28×)
>80645
5g
13 ± 1.8
268 ± 8
273 ± 52 (1×)
2300 ± 700 (9×)
6800 ± 400 (25×)
49
5h
8.4 ± 3.2
1200 ± 260
1800 ± 870 (1.5×)
3730 ± 920 (3×)
6920 ± 2000 (6×)
7
5i
68 ± 8.7
12 ± 3
6.6 ± 1.7 (0.55×)
14 ± 4 (1×)
35 ± 12 (3×)
5667
5j
>250
14 ± 8
18 ± 2.6 (1×)
79 ± 5.2 (6×)
119 ± 3.4 (9×)
>17857
5k
102 ± 8
50 ± 13
40 ± 15 (0.80×)
116 ± 22 (2×)
243 ± 39 (5×)
2040
5l
94 ± 24
7.4 ± 1.4
8.8 ± 2.7 (1×)
12 ± 4.5 (2×)
71 ± 0.14 (10×)
12703
5m
>250
7.2 ± 3.0
7.4 ± 0.5 (1×)
44 ± 6.7 (6×)
154 ± 16 (21×)
>34722
5n
9.6 ± 3.7
35 ± 11
57 ± 13 (2×)
N/Ae
N/Ae
277
5o
24 ± 3
5.2 ± 0.6
4.6 ± 1.8 (0.88×)
25 ± 4 (5×)
43 ± 15 (8×)
4615
5p
>250
4.5 ± 1.5
4.8 ± 2.9 (1×)
3.1 ± 0.3 (0.69×)
35 ± 14 (8×)
>55556
5q
>250
3.8 ± 1.2
4.6 ± 2.2 (1×)
19 ± 7 (5×)
36 ± 16 (9×)
>65789
(S)-5r
>250
4.2 ± 1.6
4.8 ± 1.4 (1×)
15.3 ± 3.3 (4×)
141 ± 20 (33×)
>59 524
(R)-5r
>250
11.3 ± 4.7
8.1 ± 1.7 (0.7×)
27 ± 9.9 (2×)
89 ± 26 (8×)
>22 124
(S)-5s
>250
9 ± 2.4
8.5 ± 1.9 (0.9×)
17.6 ± 6.4 (2×)
55.6 ± 6.2 (6×)
>27 778
(R)-5s
>250
7.4 ± 3.3
8.2 ± 1.8 (1 × )
22 ± 4 (3 × )
48 ± 12 (6 × )
>33 784
(S)-5t
>250
9.7 ± 3
11 ± 4 (1 × )
29 ± 8 (3 × )
122 ± 33 (13 × )
>25 773
(R)-5t
>250
14.2 ± 3.7
10.2 ± 1.4 (0.7 × )
71 ± 5 (5 × )
284 ± 120 (20 × )
>17 606
(S)-5u
18 ± 7
590 ± 72
1660 ± 920 (3 × )
N/Ae
N/Ae
31
5v
>250
1.1 ± 0.7
2.5 ± 0.6 (2 × )
5.3 ± 2.3 (5 × )
35 ± 9 (32 × )
>227 273
Cytotoxic
concentration resulting
in 50% reduction in the level of ATP in human osteosarcoma (HOS) cells.
Values obtained from cells
infected
with lentiviral vector carrying WT IN.
Cells were infected with viral constructs
carrying IN mutations and indicated values correspond to the fold-change
(FC) in EC50 relative to WT.
Selectivity index (SI) calculated
as the ratio of CC50 to EC50.
Not available.
Cytotoxic
concentration resulting
in 50% reduction in the level of ATP in humanosteosarcoma (HOS) cells.Values obtained from cells
infected
with lentiviral vector carrying WT IN.Cells were infected with viral constructs
carrying IN mutations and indicated values correspond to the fold-change
(FC) in EC50 relative to WT.Selectivity index (SI) calculated
as the ratio of CC50 to EC50.Not available.The main objective of the current study was to derive
minimally
cytotoxic inhibitors having good antiviral potency against cells infected
with WT virus, which also retained their efficacy against viruses
harboring mutant forms of IN that are resistant to 1.
As shown in Table 4, relative to their potencies
against WT, almost all members of the current series maintained complete
or nearly complete efficacy against virus having the Y143R mutant.
In addition, most members of the series showed good retention of efficacy
against virus having the N155H and G140S/Q148H mutants (with a few
exceptions, 10-fold or less loss of potency) (Table 4). Members of the series also commonly exhibited high SI values,
in the range of four orders-of-magnitude.On the basis of these
data, we examined selected members of the
series (5o–5q and 5v) against a more extensive panel of INSTI-resistant mutants that
included R263 K and G118R mutants, which have recently been identified
through in vitro selection studies with second-generation INSTIs.[5] For reference, we also included 1 and 2 as well as the parent 1-hydroxy-1,8-naphthyridin-2H-ones (3 and 4), which formed
the starting points for the current series. Although 1 is potent against viral vectors that carry WT IN (EC50 = 4 ± 2 nM), it shows extensive loss of antiviral efficacy
against the mutants, Y143R (EC50 = 162 ± 16 nM; 41-fold
loss), N155H (EC50 = 154 ± 33 nM; 38-fold loss), and
G140S/Q148H (EC50 = 1900 ± 300 nM; 475-fold loss).
Compound 1 is more tolerant of the G118R mutant (EC50 = 36 ± 5 nM; 9-fold loss) and even less affected by
the R263 K mutant (EC50 = 9 ± 4 nM; 2-fold loss) (Table 5). In contrast, the recently FDA-approved second-generation
inhibitor 2, while showing similar potency to 1 against WT vector (EC50 = 1.6 ± 0.9 nM), exhibits
a significantly smaller loss of potency against vectors having the
mutants Y143R (EC50 = 4.3 ± 1.2 nM; 3-fold loss),
N155H (EC50 = 3.6 ± 1.3 nM; 2-fold loss), and G140S/Q148H
(EC50 = 5.8 ± 0.5 nM; 4-fold loss) (Table 5). However, as expected, 2 shows some
loss of potency against virus having the R263 K (EC50 =
11 ± 3 nM; 7-fold loss) and G118R (EC50 = 13 ±
5 nM; 8-fold loss) mutants.
Table 5
Antiviral Potencies
of 5o–5q and 5v Compared
with 1, 2, and Previously Reported Carboxamides 3 and 4(9) in Cells
Infected
with HIV-1 Constructs Containing WT or Mutant IN
EC50 [nM/(FC), IN mutantsb]
compd
EC50 (nM, WT)a
Y143R
N155H
R263 K
G118R
G140S/Q148H
1
4 ± 2
162 ± 16 (41×)
154 ± 33 (38×)
9 ± 4 (2×)
36 ± 5 (9×)
1900 ± 300 (475×)
2
1.6 ± 0.9
4.3 ± 1.2 (3×)
3.6 ± 1.3 (2×)
11 ± 3 (7×)
13 ± 5 (8×)
5.8 ± 0.5 (4×)
3
5.1 ± 1.9
4.9 ± 0.8 (1×)
134 ± 23 (26×)
N/Ac
N/Ac
438 ± 121 (86×)
4
6.2 ± 2.9
11 ± 2 (2×)
31 ± 8 (5×)
36 ± 8 (6×)
107 ± 8 (17×)
308 ± 125 (50×)
5o
5.2 ± 0.6
4.6 ± 1.8 (0.88×)
25 ± 4 (5×)
26 (5×)
27 ± 1 (5×)
43 ± 15 (8×)
5p
4.5 ± 1.5
4.8 ± 2.9 (1×)
3.1 ± 0.3 (0.69×)
16 ± 5 (4×)
44 ± 11 (10×)
35 ± 14 (8×)
5q
3.8 ± 1.2
4.6 ± 2.2 (1×)
19 ± 7 (5×)
26 ± 8 (7×)
41 ± 18 (11×)
36 ± 16 (9×)
5v
1.1 ± 0.66
2.5 ± 0.6 (2×)
5.3 ± 2.3 (5×)
6.4 ± 2.3 (6×)
16 ± 5 (15×)
35 ± 9 (32×)
Values obtained
from cells infected
with lentiviral vector harboring WT IN.
Cells were infected with viral constructs
carrying IN mutations and indicated values correspond to the fold-change
(FC) in EC50 relative to WT.
Not available.
Values obtained
from cells infected
with lentiviral vector harboring WT IN.Cells were infected with viral constructs
carrying IN mutations and indicated values correspond to the fold-change
(FC) in EC50 relative to WT.Not available.Among the current series, the antiviral potencies of 5o–5q are approximately equivalent to 1 against WT enzyme, with 5v being slightly more potent
(equivalent to 2) (Table 5). Compounds 5p, 5q, and 5v also show fold-loss
of potencies that are approximately equivalent to 2 against
the panel of mutants, with the exception of 5v, which
shows greater fold-loss of potency against the mutant G140S/Q138H
(32-fold). However, because 5v has greater potency against
the WT virus than either 5p or 5q, the effective
antiviral potencies against virus having the G140S/Q138H mutant are
approximately equal for the three compounds (EC50 = 35
nM). A summary analysis of Table 5 shows that 5p exhibits a profile against the panel of mutants that is
similar to 2. Because of its slightly better potency
against WT IN, compound 5v exhibits the best overall
performance among the new inhibitors, with both absolute potencies
and fold-loss of potencies against the various mutant vectors that
are similar to 2. The sole exception is the G140S/Q148H
mutant, where 5v suffers an approximate 7-fold loss of
potency relative to 2. Importantly, 5v shows
from 5- to 10-fold enhanced performance across the entire panel relative
to the starting compound 4. The standard assays are performed
in the presence of 5% fetal bovine serum (FBS). To examine the potential
effects of serum protein binding on their antiviral potencies, compounds 5p, 5q, and 5v were evaluated against
the WT vector in the presence of increasing amounts FBS (5%, 10%,
and 15%). These experiments showed that the antiviral potencies of
the compounds were essentially unchanged relative to values shown
in Tables 4 and 5, which
were obtained in the presence of 5% fetal bovine serum protein (data
not shown).The title compounds were originally designed to
chelate the two
Mg2+ ions at the IN active site. However, HIV-1 reverse
transcriptase (RT) also contains two active sites, the polymerase
active site and the RNase H active site, each of which has two bound
Mg2+ ions. To examine whether representative compounds
of the series could also bind Mg2+ ions at one or both
of the active sites in RT, we tested 5p, 5q, and 5v, to see whether they inhibited either the RNase
H or the polymerase activity of RT. Nevirapine, a well-known nonnucleoside
reverse transcriptase inhibitor (NNRTI), was included as a positive
control in the polymerase inhibition assays. The polymerase inhibition
assays show that all three compounds are able to inhibit the DNA-dependent
DNA polymerase activity of HIV-1 RT (Figure 3A). However, the data also show that the compounds differ in their
potency. All of the compounds are less potent than nevirapine in the
polymerase assay, Compound 5v the most potent, followed
by 5p, then 5q. It is not clear at this
point whether the compounds are binding the Mg2+ ions at
the polymerase active site or binding within the NNRTI-binding pocket
near the polymerase active site, similar to nevirapine. It is also
possible that they may be binding at some other site in HIV-1 RT.
Figure 3
Polymerase
and RNase H inhibition assays (A): The effects of 5p, 5q, and 5v on the DNA-dependent
DNA polymerase activity of RT are shown. Reactions were performed
with WT RT in the presence of various concentrations of compounds
(0, 0.02, 0.1, 0.5, 1, or 10 μM). Samples were precipitated
with EtOH and then fractionated on a 15% polyacrylamide sequencing
gel. After electrophoresis, the gel was exposed to X-ray film. RNase
H inhibition assay (B,C,D): The effects of 5p, 5q, and 5v on the RNase H activity of RT are
shown. The reactions were incubated for the amount of time indicated
(1, 5, 10, and 15 min) in the presence of the indicated concentration
of the individual compounds (0, 0.1, 0.5, 1, and 10 μM). The
size of intact RNA (full length) is 60 NT, as shown in the “No-RT”
lane. The RNA fragments derived from the −17 and −8
families of cleavages are shown.
Polymerase
and RNase H inhibition assays (A): The effects of 5p, 5q, and 5v on the DNA-dependent
DNA polymerase activity of RT are shown. Reactions were performed
with WT RT in the presence of various concentrations of compounds
(0, 0.02, 0.1, 0.5, 1, or 10 μM). Samples were precipitated
with EtOH and then fractionated on a 15% polyacrylamide sequencing
gel. After electrophoresis, the gel was exposed to X-ray film. RNase
H inhibition assay (B,C,D): The effects of 5p, 5q, and 5v on the RNase H activity of RT are
shown. The reactions were incubated for the amount of time indicated
(1, 5, 10, and 15 min) in the presence of the indicated concentration
of the individual compounds (0, 0.1, 0.5, 1, and 10 μM). The
size of intact RNA (full length) is 60 NT, as shown in the “No-RT”
lane. The RNA fragments derived from the −17 and −8
families of cleavages are shown.The compounds were also tested for their ability to inhibit
the
RNase H activity of RT. When RT binds an RNA–DNA template/primer
(T/P), the 3′ end of the DNA primer is preferentially located
at the polymerase active site; the RNase H active site contacts the
RNA template approximately 17 to 18 nucleotides (NT) from the polymerase
active site.[27] The initial RNase H cleavage
occurs approximately 17 NT from the polymerase active site. These
cleavages have been designated as the −17 family of cleavages.
The RT then alters its interactions with the T/P, so that RNase H
can make additional secondary cleavages approximately 8 NT from the
3′ end of the primer (−8 cleavages). The full length
RNA is 60 NT in length. We found that 5p, 5q, and 5v varied in their abilities to inhibit RNase
H activity (Figure 3B–D). However, the
ranking of the potencies of the compounds in the RNase H assay is
different from the polymerase assay (Figure 3A), which suggests that the compounds are interacting with the RT
in different places in the two assays. The data show that 5q is the most potent inhibitor of RNase H (it was the least potent
compound in the polymerase assay). Compound 5q did not
completely block RNase H activity, even at the highest concentration;
a large amount of full length RNA was present in the reaction that
contained 5q at the 10 μM concentration. It is
also apparent that more of the product of the −17 cleavages
remained after 10 and 15 min as compared to the “no compound”
controls. This results from partial blocking the −8 cleavages.
Similar results were obtained with 5v, while 5p had very little effect on RNase H activity. At the highest concentration,
there were subtle effects on the −8 cleavages, but the compound
is obviously less potent, in the RNase H assay, than the other two.
It is quite difficult to use the results of the in vitro assays to
estimate IC50 values. In the polymerase assay, it is clear
that all of the compounds are less potent than nevirapine. We did
not have a potent RNase H inhibitor that would allow us to make a
similar comparison in the RNase H inhibitor assay. Although it is
clear that the current compounds are primarily IN inhibitors, because
these compounds can inhibit both the RNase H and polymerase activity
of HIV-1 RT, they could serve as the starting point for the synthesis
of additional compounds that would be specifically designed to inhibit
these alternate targets. In that regard, because there are no potent
RNase H inhibitors, and because many of the known RNase H inhibitors
are relatively toxic, using these compounds as leads to develop new
RNase H inhibitors is a potentially attractive option.
Conclusion
Our current study examines the effects of introducing an amine
functionality at the 4-position of our previously reported 1-hydroxy-1,8-naphthyridin-2H-ones (3 and 4).[9] The focus of the work was to enhance antiviral potency,
with particular emphasis on retaining efficacy against viruses harboring
mutant forms of IN that have been shown to be resistant to the first-generation
INSTI, 1. For reference we employed 2, which
is a recently FDA-approved second-generation INSTI with improved performance
against the known resistant mutants. Most members of the series of
new inhibitors (5) show single-digit nanomolar antiviral
potency against WT enzyme, with SI values greater than 10000. As a
whole, the family of new inhibitors exhibited a smaller fold-change
than 1 in antiviral assays that employed IN mutants Y143R,
N155H, and the double mutant G140S/Q148H. Among the new inhibitors,
compound 5p showed a profile against the panel of mutants
that was comparable to 2, with 5v exhibiting
the best overall absolute performance among the new inhibitors, approximately
5- to 10-fold enhancement relative to the starting compound 4. Although 2 is more effective than first-generation
INSTIs in its ability to retain efficacy against resistant forms of
IN, it has been shown that 2 can select for resistant
forms of the enzyme. Therefore, there is a continuing need for the
development of new agents as potential alternatives to the currently
approved panel of three FDA-approved INSTIs. The structural class
of agents presented herein may represent an attractive platform for
developing such next-generation INSTIs.
Experimental
Section
General Synthetic
1H and 13C
NMR data were obtained on a Varian 400 MHz spectrometer or a Varian
500 MHz spectrometer and are reported in ppm relative to TMS and referenced
to the solvent in which the spectra were collected. Solvent was removed
by rotary evaporation under reduced pressure, and anhydrous solvents
were obtained commercially and used without further drying. Purification
by silica gel chromatography was performed using CombiFlash Rf 200i with EtOAc–hexanes solvent systems. Preparative
high pressure liquid chromatography (HPLC) was conducted using a Waters
Prep LC4000 system having photodiode array detection and Phenomenex
C18 columns (catalogue no. 00G-4436-P0-AX, 250 mm ×
21.2 mm 10 μm particle size, 110 Å pore) at a flow rate
of 10 mL/min. Binary solvent systems consisting of A = 0.1% aqueous
TFA and B = 0.1% TFA in acetonitrile were employed with gradients
as indicated. Products were obtained as amorphous solids following
lyophilization. Electrospray ionization–mass spectra (ESI-MS)
were acquired with an Agilent LC/MSD system equipped with a multimode
ion source. Purities of samples subjected to biological testing were
assessed using this system and shown to be ≥95%. High-resolution
mass spectra (HRMS) were acquired by LC/MS-ESI using an LTQ-Orbitrap-XL
at 30K resolution.
General Procedure A
Preparation of 1-(Benzyloxy)-N-(2,4-difluorobenzyl)-2-oxo-1,2-dihydro-1,8-naphthyridine-3-carboxamide
(10a–10v)
A solution of p-methylbenzenesulfonate (9) (1.0 mmol), N-ethyl-N-isopropylpropan-2-amine (10 mmol)
and amines R1R2NH or R1R2NH·HCl (5.0 mmol) in DMF (2.0 mL) was heated to 50 °C (1
h). The crude mixture was purified by CombiFlash silica gel chromatography
(hexanes and ethyl acetate) to provide amides (10a–10v).
General Procedure B
Preparation of N-(2,4-Difluorobenzyl)-1-hydroxy-2-oxo-1,2-dihydro-1,8-naphthyridine-3-carboxamides
(5a–5v)
Amides (10a–10u or 11; 0.2 mmol) were dissolved
in MeOH (15 mL) and EtOAc (5 mL) and the solution degassed and stirred
at room temperature under H2 over Pd·C (10%, 0.2 mmol)
(1 h). The mixture was filtered and the filtrate was concentrated,
and the resulting residue was purified by HPLC to provide amides (5a–5v).
Treatment of 10v as described
under general procedure B and purification by preparative HPLC (linear
gradient of 30% B to 50% B over 30 min; retention time = 19.5 min)
provided 5v as a white solid in 47% yield. 1H NMR (400 MHz, DMSO-d6) δ 10.59 (t, J = 5.8 Hz, 1H), 8.66 (dd, J = 4.5, 1.4 Hz, 1H),
8.61 (dd, J = 8.1, 1.6 Hz, 1H), 7.36 (d, J = 6.6 Hz, 1H), 7.30 (dd, J = 8.1, 4.6
Hz, 1H), 7.18 (d, J = 9.2 Hz, 1H), 7.00 (d, J = 2.4 Hz, 1H), 4.46 (d, J = 5.8 Hz, 2H).
ESI-MS m/z: 347.1 (MH+). HRMS calcd C16H13F2N4O3 [MH+], 347.0950; found, 347.0953.
Integrase Biochemical Assays
Determination of IN 3′-
P and ST inhibitory values using an in vitro assay IN reactions were
carried out using [γ-32P]-labeled DNA as previously
described.[9,22]
Cellular Cytotoxicity Assays
Cytotoxicity
was measured
using the humanosteosarcoma cell line, HOS (Dr. Richard Schwartz,
Michigan State University, East Lansing, MI), by monitoring ATP levels
using a luciferase reporter assay as previously reported.[9]
Single-Round HIV-1 Infectivity Assays
As previously
described,[9] the humanembryonic kidney
cell line 293T was transfected with the pNLNgoMIVR–ΔLUC vector, which was made from pNLNgoMIVR–ΔEnv.HSA by removing the HSA reporter gene
and replacing it with a luciferase reporter gene between the NotI
and XhoI restriction sites.[28] VSV-g-pseudotyped HIV was produced transfecting 293T cells as
described previously.[29] On the day prior
to transfection, 293T cells were plated on 100 mm diameter dishes
at a density of 1.5 × 106 cells per plate. 293T cells
were transfected with 16 μg of pNLNgoMIVR–ΔLUC and 4 μg of pHCMV-g (obtained from Dr. Jane Burns,
University of California, San Diego) using the calcium phosphate method.
At approximately 6 h after the calcium phosphate precipitate was added,
the 293T cells were washed twice with phosphate-buffered saline (PBS)
and incubated with fresh media (48 h). The virus-containing supernatants
were then harvested, clarified by low-speed centrifugation, filtered,
and diluted for preparation in infection assays. On the day prior
to the screen, HOS cells were seeded in a 96-well luminescence cell
culture plate at a density of 4000 cells in 100 μL per well.
On the day of the screen, cells were treated with the compounds from
a concentration range of 10 μM to 0.0005 μM using 11 serial
dilutions and then incubated at 37 °C (3 h). After this incubation,
100 μL of virus stock diluted to achieve a maximum luciferase
signal between 0.2 and 1.5 RLUs was added to each well and the plates
were incubated at 37 °C (48 h). Infectivity was measured by using
the Steady-lite plus luminescence reporter gene assay system (PerkinElmer,
Waltham, MA). Luciferase activity was measured by adding 100 μL
of Steady-lite plus buffer (PerkinElmer) to the cells, incubating
at room temperature (20 min), and measuring luminescence using a microplate
reader. Activity was normalized to infectivity in the absence of target
compounds. KaleidaGraph (Synergy Software, Reading, PA) was used to
perform regression analysis on the data. EC50 values were
determined from the fit model.
Vector Constructs
pNLNgoMIVR–ΔEnv.LUC has
been described previously.[28] The IN coding
region was removed from pNLNgoMIVR–ΔEnv.LUC (between KpnI and SalI sites) and inserted between the KpnI and SalI sites of pBluescript II KS+. Using this
construct as the wild-type template, the following IN-resistant mutants
were prepared via the QuikChange II XL (Stratagene, La Jolla, CA)
site-directed mutagenesis protocol: G118R, Y143R, Q148H, Q148 K, N155H,
G140S + Q148H, G140A + Q148 K, and E138 K + Q148 K. The following
sense with cognate antisense (not shown) oligonucleotides (Integrated
DNA Technologies, Coralville, IA) were used in the mutagenesis: G118R,
5′-GTACATACAGACAATCGCAGCAATTTCACCAGTAC-3′; E138 K, 5′-GGCGGGGATCAAGCAGAAATTTGGCATTCCCTA-3′;
G140A, 5′-GGGGATCAAGCAGGAATTTGCCATTCCCTACAATC-3′; G140S,
5′-GGGGATCAAGCAGGAATTTAGCATTCCCTACAATC-3′; Y143R, 5′-GCAGGAATTTGGCATTCCCCGCAATCCCCAAAGTCAAGGA-3′;
Q148H, 5′-CATTCCCTACAATCCCCAAAGTCATGGAGTAATAGAATCTA-3′;
Q148 K, 5′-CATTCCCTACAATCCCCAAAGTAAAGGAGTAATAGAATCTATGAA-3′;
N155H, 5′-CCAAAGTCAAGGAGTAATAGAATCTATGCATAAAGAATTAAAGAAAATTATAGGACA-3′.
The double mutation G140S + Q148H was constructed by using the previously
generated Q148H mutant and the appropriate oligonucleotide for the
second mutation, G140S. The double mutation G140A + Q148 K was made
by using the Q148 K mutant and the appropriate oligonucleotide for
the second mutation, G140A. The double mutation E138 K + Q148 K was
made by using the Q148 K mutant and the appropriate oligonucleotide
for the second mutation, E138 K. The DNA sequence of each construct
was verified independently by DNA sequence determination. The mutant
IN coding sequences from pBluescript II KS+ were then subcloned into
pNLNgoMIVR–ΔEnv.LUC (between
the KpnI and SalI sites) to produce
the full-length mutant HIV-1 IN constructs. These DNA sequences were
checked independently by DNA sequence determination.
Reverse Transcriptase
(RT) Assays
The assay used to
determine whether the compounds inhibit the of polymerase activity
of RT has been previously described.[30] Briefly,
the −47 sequencing primer (5′-CGCCAGGGTTTTCCCAGTCACGAC-3′;
New England Biolabs) was 5′ end-labeled with [γ-32P] ATP and T4 polynucleotide kinase. After purification,
the labeled primer was annealed to single-stranded M13mp18 DNA (1
μg of DNA for each sample to be assayed) by heating and slow
cooling. For each sample, 0.1 μg of wild-type RT (approximately
17 nM final) added to the labeled primer template (approximately 9.0
nM) in 25 mM Tris–HCl (pH 8.0), 75 mM KCl, 10 mM MgCl2, 100 μg of BSA per mL, and 10 mM CHAPS. The reaction mixture
was supplemented with 0.5 μM (each) of dATP, dCTP, dGTP, and
TTP. The compounds were added to give final concentrations of 0, 0.02,
0.1, 0.5, 1, or 10 μM. The reactions were allowed to proceed
at 37 °C for 60 min and were stopped by the addition of EDTA.
The samples were precipitated by the addition of two volumes of EtOH,
fractionated by electrophoresis on a 6.0% polyacrylamide gel, and
the gel was autoradiographed. Reactions were carried out in duplicate.
RNaseH Assays
This procedure has been previously described.[30] Briefly, RNA oligonucleotide (5′-GGGGCCACUUUUUAAAAGAAAAGGGGGGACUGGAAGGGCUAAUUCACUCAC-3′)
was obtained from Dharmacon Research, Inc. The RNA oligonucleotide
was 5′-end labeled and was then annealed to a DNA oligonucleotide
(5′-GAGTGAATTAGCCCTTCCAGTCCC-3′) by heating and slow
cooling. A 0.2 μM concentration of the RNA/DNA hybrid was suspended
in 25 mM Tris (pH 8.0), 50 mM NaCl, 5.0 mM MgCl2, 100 μg
of bovine serum albumin/mL, 10 mM CHAPS, and 1 U of Superasin (Ambion).
The compounds were added to the reactions to give the following final
concentrations (0, 0.1, 0.5, 1, and 10 μM). The reaction volume
was 12 μL. The reactions were initiated by the addition of 50.0
ng of RT and were incubated at 37 °C. Aliquots were removed at
the indicated time points, and the reactions halted by addition of
2× gel loading buffer (Ambion). The reaction products were fractionated
on a 15% polyacrylamide sequencing gel. Products were visualized by
exposure to X-ray film.
Authors: A Jacobo-Molina; J Ding; R G Nanni; A D Clark; X Lu; C Tantillo; R L Williams; G Kamer; A L Ferris; P Clark Journal: Proc Natl Acad Sci U S A Date: 1993-07-01 Impact factor: 11.205
Authors: Xue Zhi Zhao; Elena A Semenova; B Christie Vu; Kasthuraiah Maddali; Christophe Marchand; Stephen H Hughes; Yves Pommier; Terrence R Burke Journal: J Med Chem Date: 2007-12-21 Impact factor: 7.446
Authors: Xue Zhi Zhao; Kasthuraiah Maddali; B Christie Vu; Christophe Marchand; Stephen H Hughes; Yves Pommier; Terrence R Burke Journal: Bioorg Med Chem Lett Date: 2009-03-28 Impact factor: 2.823
Authors: John K Pratt; Pamela Donner; Keith F McDaniel; Clarence J Maring; Warren M Kati; Hongmei Mo; Tim Middleton; Yaya Liu; Teresa Ng; Qinghua Xie; Rong Zhang; Debra Montgomery; Akhteruzzaman Molla; Dale J Kempf; William Kohlbrenner Journal: Bioorg Med Chem Lett Date: 2005-03-15 Impact factor: 2.823
Authors: Vincenzo Summa; Alessia Petrocchi; Victor G Matassa; Marina Taliani; Ralph Laufer; Raffaele De Francesco; Sergio Altamura; Paola Pace Journal: J Med Chem Date: 2004-10-21 Impact factor: 7.446
Authors: Paul L Boyer; Steven J Smith; Xue Zhi Zhao; Kalyan Das; Kevin Gruber; Eddy Arnold; Terrence R Burke; Stephen H Hughes Journal: J Virol Date: 2018-06-13 Impact factor: 5.103
Authors: Mathieu Métifiot; Barry C Johnson; Evgeny Kiselev; Laura Marler; Xue Zhi Zhao; Terrence R Burke; Christophe Marchand; Stephen H Hughes; Yves Pommier Journal: Nucleic Acids Res Date: 2016-07-01 Impact factor: 16.971
Authors: Xue Zhi Zhao; Steven J Smith; Daniel P Maskell; Mathieu Métifiot; Valerie E Pye; Katherine Fesen; Christophe Marchand; Yves Pommier; Peter Cherepanov; Stephen H Hughes; Terrence R Burke Journal: J Med Chem Date: 2017-08-10 Impact factor: 7.446
Authors: Steven J Smith; Andrea Ferris; Xuezhi Zhao; Gary Pauly; Joel P Schneider; Terrence R Burke; Stephen H Hughes Journal: Viruses Date: 2021-12-14 Impact factor: 5.048
Authors: Xue Zhi Zhao; Steven J Smith; Daniel P Maskell; Mathieu Metifiot; Valerie E Pye; Katherine Fesen; Christophe Marchand; Yves Pommier; Peter Cherepanov; Stephen H Hughes; Terrence R Burke Journal: ACS Chem Biol Date: 2016-02-05 Impact factor: 5.100