Enzymes involved in RNA capping of SARS-CoV-2 are essential for the stability of viral RNA, translation of mRNAs, and virus evasion from innate immunity, making them attractive targets for antiviral agents. In this work, we focused on the design and synthesis of nucleoside-derived inhibitors against the SARS-CoV-2 nsp14 (N7-guanine)-methyltransferase (N7-MTase) that catalyzes the transfer of the methyl group from the S-adenosyl-l-methionine (SAM) cofactor to the N7-guanosine cap. Seven compounds out of 39 SAM analogues showed remarkable double-digit nanomolar inhibitory activity against the N7-MTase nsp14. Molecular docking supported the structure-activity relationships of these inhibitors and a bisubstrate-based mechanism of action. The three most potent inhibitors significantly stabilized nsp14 (ΔTm ≈ 11 °C), and the best inhibitor demonstrated high selectivity for nsp14 over human RNA N7-MTase.
Enzymes involved in RNA capping of SARS-CoV-2 are essential for the stability of viral RNA, translation of mRNAs, and virus evasion from innate immunity, making them attractive targets for antiviral agents. In this work, we focused on the design and synthesis of nucleoside-derived inhibitors against the SARS-CoV-2 nsp14 (N7-guanine)-methyltransferase (N7-MTase) that catalyzes the transfer of the methyl group from the S-adenosyl-l-methionine (SAM) cofactor to the N7-guanosine cap. Seven compounds out of 39 SAM analogues showed remarkable double-digit nanomolar inhibitory activity against the N7-MTase nsp14. Molecular docking supported the structure-activity relationships of these inhibitors and a bisubstrate-based mechanism of action. The three most potent inhibitors significantly stabilized nsp14 (ΔTm ≈ 11 °C), and the best inhibitor demonstrated high selectivity for nsp14 over human RNA N7-MTase.
Severe
acute respiratory syndrome coronavirus 2 (SARS-CoV-2)
is the third highly pathogenic coronavirus, emerging in the
human population in 2019, after SARS-CoV and Middle East respiratory
syndrome coronavirus (MERS-CoV) emerged in 2003[1] and 2012,[2] respectively. Currently,
treatments for diseases caused by CoVs are still limited. Therefore,
this health emergency highlights the crucial need to identify effective
treatments for SARS-CoV-2 and its variants. Until now, the promising
drug repurposing strategy has failed to find an effective treatment
for COVID-19. The alternative strategy is to develop new direct-acting
antiviral drugs with a rational design.[3,4] Although more
time-consuming, this second approach of exploring the structure–activity
relationships (SARs) of compounds in an attempt to understand their
mode of action is more compelling for medicinal chemists.CoVs
have a genome composed of a large, single-stranded, positive-sense
RNA with a cap structure at its 5′ end that ensures mRNA stability
by protecting it from cellular 5′-exo-ribonucleases. This structure
consists of an N7-methylguanosine linked by
a 5′-5′-triphosphate bridge to the 5′-terminal
nucleotide (adenosine in CoVs) (cap 0: 7mGpppA), which
can be further methylated at its 2′-O position
(cap 1: 7mGpppAm).[5,6] Specifically, N7-methylation of the viral RNA cap plays a key role in
the translation of viral RNA into proteins. Furthermore, inhibition
of the SARS-CoV-2 nsp14 N7-methyltransferase
(MTase) blocks the enzymatic cascade of viral RNA methylations, as
the 2′-O-MTase (nsp16) only recognizes N7-methylated cap substrates.[7,8] Nsp14 is also
considered as an antiviral target because the replication of N7-MTase catalytic mutants is strongly impaired.[9] Therefore, this crucial yet uncommon and under-explored
enzyme seemed an enticing target to us for the development of antiviral
therapies.[9−11]Curiously, until recently, few inhibitors of
the viral MTase nsp14
have been developed, and the lack of selective inhibitors is an exciting
challenge not only for new antiviral therapies but also for functional
studies of this enzyme.[12] It is noteworthy
that nsp14 has an original fold,[13] which
is not the canonical Rossmann fold, and that the N7-MTase domains of CoVs are highly conserved. This particular structural
organization and sequence conservation could facilitate the development
of specific inhibitors. Recently, several studies have reported high-throughput
screening of existing libraries of small molecules against nsp14 activity.
However, very few compounds have been identified as potential inhibitors
of SARS-CoV-2.[14−18] In addition to the drug-repurposing approach, the method of de novo drug discovery with structure-guided design of nsp14
substrates yielded potent inhibitors with IC50 values in
the nanomolar to submicromolar range.[19,20] However, both
articles did not report studies on the inhibitory activity of these
nsp14 inhibitors in SARS-CoV-2-infected cells. Before the emergence
of SARS-CoV-2, our group had pioneered the synthesis of selective
inhibitors targeting the SARS-CoV nsp14 N7-MTase
using dinucleosides as mimetics of the S-adenosyl-l-methionine (SAM) that is the methyl donor in the N7-methylation of the cap (Figure ).[21] Initially, these compounds
were designed to interact with 2′-O-MTases
in a selective manner by mimicking the structure of the transition
state of viral cap-mRNA during the 2′-O-methylation
of 7mGpppN1-RNA. Unexpectedly, while all of
the synthesized compounds were barely active against 2′-O-MTases, some of them (D1–D5, Figure , Table ) exhibited inhibition of the
SARS-CoV nsp14 N7-MTase in the submicromolar range.
These dinucleosides consisted of two adenosines, linked in 2′-5′
by a substituted benzenesulfonamide ethyl linker. Molecular
docking studies suggested that the phenyl ring binds to the cap-binding
pocket of SARS-CoV nsp14 and establishes π–π stacking
interactions with the Phe426 amino acid that naturally stacks the
guanosine of the viral mRNA cap structure (Supporting
Information, Figures S1 and S2).[22] Thus, this work demonstrated that dinucleosides could act as competitive
bisubstrate inhibitors by occupying both the SAM-binding pocket and
the cap-binding pocket (Supporting Information, Figure S2). Unfortunately, none of the developed compounds exhibited
antiviral activity in SARS-CoV-infected VeroE6 cells (unpublished
results). This lack of efficacy could be explained by poor cellular
internalization of these large, high-molecular-weight dinucleosides,
which could impair targeting of the nsp14 enzyme in the cytoplasm.[23]
Figure 1
Design of SARS-CoV-2 nsp14 inhibitors 1–39 derived from initial dinucleoside inhibitors D1–D5 of SARS-CoV nsp14.[21]
Table 1
Comparison of IC50 Values
of Sinefungin, Dinucleosides D1–D5, and Compounds 2, 3, 5, 6, and 7 on SARS-CoV nsp14 and SARS-CoV-2 nsp14
IC50a (μM)
compound
SARS-CoV nsp14
SARS-CoV-2 nsp14
sinefungin
0.36b
0.278 ± 0.008
D1c
2.6 ± 0.3
n.d.
D2c
70.4 ± 4.9
n.d.
D3c
3.9 ± 0.4
n.d.
D4c
5.7 ± 0.5
n.d.
D5c
0.6 ± 0.1
n.d.
2
7.6 ± 1.2
14.1 ± 1.0
3
25 ± 6.1
5.1 ± 2.0
5
2.6 ± 0.3
1.4 ± 0.2
6
3.3 ± 0.2
3.5 ± 0.2
7
1.4 ± 0.2
2.1 ± 0.2
Concentration inhibiting N7-MTase activity by 50%; mean value from three independent
experiments.
Values from
the literature.[10]
Values from the literature.[21] n.d.: not determined.
Design of SARS-CoV-2 nsp14 inhibitors 1–39 derived from initial dinucleoside inhibitors D1–D5 of SARS-CoV nsp14.[21]Concentration inhibiting N7-MTase activity by 50%; mean value from three independent
experiments.Values from
the literature.[10]Values from the literature.[21] n.d.: not determined.Here, with the aim to obtain smaller and presumably more efficient
molecules to enter cells, we designed and developed SAM nucleoside
analogues 1–39 of a reduced size
and molecular weight relative to our initial dinucleoside nsp14 inhibitors
(Figure ). These nucleosides,
designed as bisubstrates, consist of an adenosine interacting with
the SAM-binding pocket and various benzenesulfonamide
moieties which occupy the cap-binding pocket. Indeed, the arylsulfonamide
moiety was crucial for the inhibitory activity of dinucleosides D1–D5 on SARS-CoV nsp14.[21] The synthesized compounds 1–39 were
screened for their ability to inhibit the N7-MTase
activity of SARS-CoV-2 in vitro.
Results and Discussion
Design
In the search for effective inhibitors, we started
to optimize the structure of novel nucleoside analogues 1–9 with the goal in mind to improve interactions
with the SARS-CoV-2 N7-MTase nsp14. A thorough structural
analysis highlighted that three parts of interest in these bisubstrate
molecules could be modified to explore SARs: the linker between the
5′-deoxyadenosine and the phenyl ring, the substituents
on the phenyl ring, and the functionalization of the 5′-nitrogen
atom of the nucleoside (Figure ). In addition, molecular docking performed prior to synthesis
supported the choice of chemical modifications envisioned in the adenosine
scaffold of 7, the nucleoside counterpart of the most
active inhibitor, D5 (Supporting Information, Figure S3). Currently, since the crystal structure of the N7-MTase domain of SARS-CoV-2 nsp14 (highly conserved among
CoVs) bound to the SAM cofactor is still uncharacterized, the high
structural similarity to SARS-CoV nsp14 (95% amino acid sequence homology)
enables modeling with the crystallized nsp14-SAM complex (PDB: 5C8T).[24]First, nucleoside 1 without a substituent
in the phenyl ring and nucleosides 2–9 with a nitro (NO2) group at different positions (ortho (o), meta (m), or para (p)) of the aromatic moiety were designed to evaluate whether the monoadenosine
structure was not detrimental to the inhibitory activity compared
with previous dinucleoside inhibitors D1–D5 (Figure A).[21] Furthermore, nucleosides 8 and 9 were functionalized with an ethyl (Et) group on the N-sulfonamide moiety. This N-substitution has
been proposed to improve hydrophobicity and therefore cellular penetration.
It could also improve affinity with nsp14, as it could reduce the
total nonpolar surface area exposed to water and, therefore, provide
beneficial association entropy. Supported by our previous results,
the NO2 group is well meta-oriented in
the benzenesulfonamide group to provide a double hydrogen
bond with Arg310 in nsp14. This amino acid naturally interacts with
the triphosphate bridge of the cap structure through two hydrogen
bonds.[22] Thus, nucleosides 10–19 were designed with the m-NO2 group and bearing various para substituents
with N-H- or N-Et-sulfonamide
motifs (Figure B).
Their inhibitory activity was compared with that of compound 7 (m-NO2, p-Cl).
Due to the narrow cavity surrounding the para position
in the phenyl ring, only small substituents were inserted (H, F, Br,
methyl (Me), Et). In compounds 20–25, the m-nitro group was replaced by an m-cyano (CN) group to anticipate the mutagenic potential of the NO2 group in vivo (Figure C).[25] In addition,
upstream docking studies showed a strong similarity between NO2 and CN overlays facing Arg310 (Figures and 4).
Figure 2
Rational design
of nucleoside analogues 1–39 as inhibitors
of SARS-CoV-2 nsp14. (A) Initial compounds 1–9 mainly derived from D1–D5. (B) Compounds 10–19 derived from m-NO2 compound 7 in which p-Cl was
replaced by various substituents. (C) Compounds 20–25 containing a m-CN
group replacing a m-NO2 group. (D) Compounds 26–32 with hydrophobic substituents (Cl,
Me) in the phenyl ring. (E) Compounds 33–35 with an N-butyl chain terminated with
various groups. (F) Compounds 36–39 with an amide linkage replacing a sulfonamide linkage.
Figure 3
Overlay of m-nitro (dO–H Arg310 = 2.3, 2.1 Å) and m-cyano (dN–H Arg310 = 2.5, 2.0 Å) derivatives 10 and 20. The π–π stacking
interaction with Phe426 is shown in yellow.
Figure 4
2D representation
of the SARS-CoV nsp14 sites targeted by compounds 33–35. Amino acids surrounding the cap-binding
pocket and SAM-binding pocket are shown in black. Amino acids surrounding
the targeted pocket are shown in blue. Nucleoside is shown in green.
All residues shown here are conserved in the SARS-CoV-2 nsp14.[24]
Rational design
of nucleoside analogues 1–39 as inhibitors
of SARS-CoV-2 nsp14. (A) Initial compounds 1–9 mainly derived from D1–D5. (B) Compounds 10–19 derived from m-NO2 compound 7 in which p-Cl was
replaced by various substituents. (C) Compounds 20–25 containing a m-CN
group replacing a m-NO2 group. (D) Compounds 26–32 with hydrophobic substituents (Cl,
Me) in the phenyl ring. (E) Compounds 33–35 with an N-butyl chain terminated with
various groups. (F) Compounds 36–39 with an amide linkage replacing a sulfonamide linkage.Overlay of m-nitro (dO–H Arg310 = 2.3, 2.1 Å) and m-cyano (dN–H Arg310 = 2.5, 2.0 Å) derivatives 10 and 20. The π–π stacking
interaction with Phe426 is shown in yellow.2D representation
of the SARS-CoV nsp14 sites targeted by compounds 33–35. Amino acids surrounding the cap-binding
pocket and SAM-binding pocket are shown in black. Amino acids surrounding
the targeted pocket are shown in blue. Nucleoside is shown in green.
All residues shown here are conserved in the SARS-CoV-2 nsp14.[24]The cap-binding pocket
is surrounded by aromatic hydrophobic residues
(Phe401, Phe506), forming a cavity near the meta and para sites of the benzenesulfonamide ring (Figure ). To take advantage
of this, nucleosides 26–32 were designed
with hydrophobic substituents (Cl, Me) at these positions (Figure D). Note that docking
studies show a pivot of the benzenesulfonamide ring: m-Cl does not face Arg310 but occupies the hydrophobic cavity,
validating our approach to improve affinity (Supporting
Information, Figure S4). In addition to the SAM-binding pocket
and the cap-binding pocket, a side cavity close to the SAM-binding
pocket and surrounded by Arg289, Val290, Asn388, and His427 (Figure ) was targeted by
compounds 33–35 (Figure E). Their N-sulfonamide linker was functionalized with a butyl chain terminated
with various groups (CO2Et, OAc, NPht) that could interact
with these residues. In particular, docking studies suggest that a
phthalimide (Pht) group could interact via a π*–cation
interaction with Arg289 (Figure , and Supporting Information, Figure S5). The contribution of this R5-butyl chain
was evaluated by comparing the inhibitory activities of 33–35 with that of 15 with a N-H-sulfonamide moiety (Figure F).At the same time, Otava et al.
reported the synthesis of SARS-CoV-2
inhibitors designed to target the same pocket in nsp14.[20] Because this cavity is also adjacent to the
N7 position of the SAM adenine, the authors functionalized the C7
position of the 7-deazaadenosine SAH analogues with aromatic
systems. This rational design led to the identification of several
compounds with an inhibitory effect against SARS-CoV-2 nsp14 in the
low micromolar to nanomolar range, supporting the relevance of our
target. Finally, to demonstrate the key role of the sulfonamide
linker for significant inhibitory activity, it was replaced by an
amide bond in nucleosides 36–39.
Chemical Synthesis
The synthesis of compounds 1–32 began with the preparation of the
readily accessible 5′-amino-2′,3′-isopropylideneadenosine 40 (Scheme ).[21] Coupling of 40 with
the corresponding commercially available benzenesulfonyl chloride
reagents afforded the protected nucleosides 41a–u in 24–81% yield.[21,26] The low yield
of 24% associated with the synthesis of the p-F-m-NO2-benzenesulfonamide derivative 41i can be explained by a side reaction of the aromatic nucleophilic
substitution at the p-F site by 40.
The aromatic nucleophilic substitution to introduce a methoxy (OMe)
group at the para position was carried out from the p-Cl (41g) and p-F (41o) derivatives with sodium methanolate at 50 °C for
48 h, giving the nucleosides 41m and 41q in 76% and 98% yield, respectively.[27,28] Fluorinated
derivative 41o was engaged for the SNAr reaction
in place of the p-Cl-m-CN-benzenesulfonamide
derivative 41p to give the compound 41q in
higher yield (98% instead of 24%). In nucleosides 42a–k, the N-sulfonamide
linker was functionalized with an ethyl chain using ethyl p-toluenesulfonate and potassium iodide in DMF after
18 h at 50 °C in 40–86% yield. Moreover, the N-sulfonamide linker was substituted with an R5-butyl
chain in compounds 43a–c with the
corresponding alkyl bromide reagent under basic conditions in DMF
after 18 h at 50 °C in 55–62% yield (Scheme ).[29] Finally, the amide bond in nucleosides 44a–d was formed by coupling 5′-NH2 adenosine 40 with a suitable carboxylic acid (37–92% yield) (Scheme ). Here, several
conditions were screened using various peptide coupling agents (HBTU,
EDC, PyBOP) in the presence of various bases (iPr2NH, DIEA,
DMAP) to test the coupling of 40 with benzoic acid.[30,31] It was found that coupling with EDC in the presence of DMAP at 0
°C was the most efficient and had the shortest reaction time
(2 h). Finally, the 2′,3′-O-isopropylidene-protecting
group was removed in all intermediates 41a–u, 42a–k, 43a–c, and 44a–d under the same acidic conditions with a mixture of formic acid and
water (1:1) during 24–48 h, giving nucleosides 1–39 after purification.[32,33]
Scheme 1
Synthetic Route for Compounds 1–32
Reagents
and conditions: (a)
appropriate substituted benzenesulfonyl chloride, Et3N, DMF, 0 °C, 24–81%; (b) MeONa 0.5 M/MeOH, 50 °C,
48 h, 80–98%; (c) HCO2H/H2O 1:1 v/v,
25 °C, 24–48 h, 45–84%; (d) EtOTs, KI, K2CO3, DMF, 50 °C, 18 h, 40–81%.
Scheme 2
Synthetic Route for Compounds 33–35
Reagents and conditions: (a)
appropriate alkyl bromide, K2CO3, DMF, 50 °C,
18 h, 55–62%; (b) HCO2H/H2O 1:1 v/v,
25 °C, 58–89%.
Scheme 3
Synthetic Route for
Compounds 36–39
Reagents and conditions:
(a)
appropriate carboxylic acid, EDC, DMAP, 0 °C, 2 h, 37–92%;
(b) HCO2H/H2O 1:1 v/v, 25 °C, 62–90%.
Synthetic Route for Compounds 1–32
Reagents
and conditions: (a)
appropriate substituted benzenesulfonyl chloride, Et3N, DMF, 0 °C, 24–81%; (b) MeONa 0.5 M/MeOH, 50 °C,
48 h, 80–98%; (c) HCO2H/H2O 1:1 v/v,
25 °C, 24–48 h, 45–84%; (d) EtOTs, KI, K2CO3, DMF, 50 °C, 18 h, 40–81%.
Synthetic Route for Compounds 33–35
Reagents and conditions: (a)
appropriate alkyl bromide, K2CO3, DMF, 50 °C,
18 h, 55–62%; (b) HCO2H/H2O 1:1 v/v,
25 °C, 58–89%.
Synthetic Route for
Compounds 36–39
Reagents and conditions:
(a)
appropriate carboxylic acid, EDC, DMAP, 0 °C, 2 h, 37–92%;
(b) HCO2H/H2O 1:1 v/v, 25 °C, 62–90%.
SARS-CoV-2 N7-MTase nsp14
Inhibition Studies
Compounds 1–39 were tested for N7-MTase inhibitory activity
using a radioactive MTase assay
(filter-binding assay) that involves measuring the [3H]-radiolabeled
methyl transferred from the SAM methyl donor onto the cap structure
of an RNA substrate (GpppAC4).[34] It should be noted that the first synthesized nucleosides, 2, 3, and 5–7, were initially tested against SARS-CoV nsp14. Then, after the emergence
of SARS-CoV-2, their inhibitory ability was measured against the SARS-CoV-2
nsp14 protein for comparison (Table ). Similar IC50 values in the single-digit
micromolar range (except for compound 3) were obtained
with both nsp14 enzymes, in agreement with the high sequence homology
between these two viral N7-MTases. Moreover, compounds 2, 3, and 5–7 are derived from the initial dinucleosides D1–D5,[21] respectively, and their IC50 values are comparable in the same micromolar range. Compound D5, with a p-Cl-m-NO2-phenylsulfonamide moiety, was the best dinucleoside
inhibitor of SARS-CoV nsp14 (IC50 = 0.6 μM), and
its nucleoside derivative 7 showed similar low IC50 values of 1.4 μM against SARS-CoV nsp14 and 2.1 μM
against SARS-CoV-2 nsp14.The next series of synthesized compounds
was then evaluated only against the SARS-CoV-2 nsp14 N7-MTase at a fixed concentration of 5 μM. With the exception
of eight compounds (1, 4, 8, 27, 36–39) that showed
no or low inhibitory activity against nsp14, most of the bisubstrate
nucleoside analogues displayed at least 65% inhibition. They were
then tested in a dose–response assay with increasing compound
concentration (Supporting Information, Figure
S6), and MTase activity was measured using a filter-binding assay
to determine the corresponding IC50 (Table ). Among the 32 potential inhibitors, 11
compounds—14, 15, 17–19, 25, 31–35—displayed higher inhibitory activity than the broad-spectrum
inhibitor, sinefungin (IC50 = 278 nM). Remarkably, seven
of these compounds—17–19, 25, 32, 34, 35—exhibited
high activity, with IC50 values between 19 and 80 nM. All
these potent nsp14 inhibitors bear a similar scaffold: an N-alkylsulfonamide linker between adenosine and the
phenyl ring that contains a substituent in the para position and a substituent in the meta position.
The N-atom of the linker is substituted with either an ethyl group
(R4) in compounds 17–19 and 25, 32, an acetyl-ended butyl chain
(34), or a phthalimide group (35)
(Figure ). Finally,
it is worth noting that, except for compound 17, the para substituents are electron-donating groups (EDGs: Me
or OMe) while the meta substituents are rather electron-withdrawing
groups (EWGs: such as NO2, CN, or Cl). Among the seven
best inhibitors 17–19, 25, 32, 34, 35, the most potent
one, 25, is para-OMe and meta-CN substituted (IC50 = 19 nM). Further investigations
on the mechanism of action showed that 25 is a SAM-competitive
nsp14 inhibitor (Supporting Information, Figure
S7).
Table 2
Screening for Inhibitory Activity
of Sinefungin and Compounds 1–39 at 5 μM
on SARS-CoV-2 N7-MTase nsp14a
compound
inhibition
of SARS CoV-2 nsp14 at 5 μM (%)
SARS CoV-2 nsp14 IC50 (μM)
compound
inhibition
of SARS CoV-2 nsp14 at 5 μM (%)
SARS CoV-2 nsp14 IC50 (μM)
sinefungin
100
0.278 ± 0.008
20
80
5.3 ± 0.9
1
NI
n.d.
21
90
0.899 ± 0.1
2
25
14.12 ± 0.9
22
100
1.01 ± 0.2
3
77
n.d.
23
100
0.514 ± 0.05
4
21
n.d.
24
65
0.960 ± 0.1
5
70
1.44 ± 0.2
25
100
0.019 ± 0.02
6
60
3.49 ± 0.1
26
25
13.02 ± 0.93
7
75
2.07 ± 0.2
27
20
n.d.
8
NI
n.d.
28
87
0.482 ± 0.06
9
84
5.0 ± 1.0
29
90
0.344 ± 0.04
10
80
3.589 ± 0.9
30
65
0.442 ± 0.08
11
87
4.4 ± 0.5
31
100
0.100 ± 0.01
12
85
2.9 ± 0.1
32
95
0.056 ± 0.01
13
100
0.342 ± 0.04
33
100
0.114 ± 0.01
14
98
0.244 ± 0.01
34
100
0.080 ± 0.006
15
90
0.146 ± 0.01
35
100
0.030 ± 0.001
16
75
0.516 ± 0.08
36
NI
n.d.
17
100
0.080 ± 0.01
37
NI
n.d.
18
100
0.038 ± 0.002
38
NI
n.d.
19
100
0.044 ± 0.003
39
NI
n.d.
Values are the mean of three independent
experiments. The N7-MTase activity was measured using
a filter-binding assay. Assays were carried out in reaction mixture
[40 mM Tris-HCl (pH 8.0), 1 mM DTT, 1 mM MgCl2, 2 μM
SAM, and 0.1 μM 3H-SAM] in the presence of 0.7 μM
GpppAC4 synthetic RNA and incubated at 30 °C and SARS-CoV-2
nsp14 (50 nM). Compounds were dissolved in 100% DMSO. NI: no inhibition
detected at 5 μM. n.d.: IC50 not determined.
Values are the mean of three independent
experiments. The N7-MTase activity was measured using
a filter-binding assay. Assays were carried out in reaction mixture
[40 mM Tris-HCl (pH 8.0), 1 mM DTT, 1 mM MgCl2, 2 μM
SAM, and 0.1 μM 3H-SAM] in the presence of 0.7 μM
GpppAC4 synthetic RNA and incubated at 30 °C and SARS-CoV-2
nsp14 (50 nM). Compounds were dissolved in 100% DMSO. NI: no inhibition
detected at 5 μM. n.d.: IC50 not determined.
Structure–Activity Relationships (SARs)
As previously
mentioned, we aimed to optimize the structure of nucleoside analogues
by investigating the SARs upon modification of different parts of
the bisubstrate molecules.
From Adenine Dinucleoside Precursors to Monoadenosine
SAM Mimics
Inhibition data for compounds 2, 3, and 5–7 compared with
those of D1–D5 clearly indicate that removal of
adenosine 2′-O-connected
at the arylsulfonamide ethyl linker in D1–D5 is not detrimental to the inhibitory activity of the monoadenosine
compounds against the SARS-CoV N7-MTase nsp14 (Table ).
Arylsulfonamide
Moieties
Three N,N-dimethyl-N-arylsulfonamide derivatives, corresponding
to the arylsulfonamide moieties of compounds 6, 7, and 13, were prepared and evaluated for their
potential to inhibit the SARS-CoV-2 N7-MTase nsp14
at 5 and 50 μM. None of them induces detectable inhibition,
demonstrating the necessity of the 5′-linked nucleoside structure
on the phenylsulfonamide core to achieve methylation inhibition.
2′,3′-O-Isopropylidene Adenosine
Two intermediate compounds, 41f and 41k, corresponding to the 2′,3′-O-protected
precursors of 6 and 13, display no inhibition
of nsp14 at 50 μM, showing that the 2′-OH and 3′-OH
of ribose may be involved in the inhibitor interactions with SARS-CoV-2
nsp14. In fact, it was previously shown that the 2′-OH and
3′-OH of ribose interact through an H-bond with Asp352 of SARS-CoV
nsp14.[22]
Substituents on the Phenyl
Ring
The impact of substituents
on the phenyl ring was demonstrated with compounds 1 and 8 (without substituents) barely inhibiting nsp14 at 5 μM,
compared to most other prepared nucleosides in which the aromatic
moiety was decorated with one or two substituents and exhibited 65–100%
inhibition of methylation. In addition, the positioning of the EWG
seems to be directly correlated with activity, with the efficiency
of the compounds increasing from the ortho position
to the para position and then to the meta position. In the case of compounds 2 and 26, the introduction of a nitro group or chlorine atom at the para position leads to moderate inhibition (IC50 = 14.12 and 13.02 μM, respectively), while in compounds 10 and 20, with a nitro group or a cyano group
located at the meta position, the activity increases
to 3.6 and 5.3 μM, respectively. Surprisingly, however, the m-Cl-substituted analogue 27 does not show
inhibition at 5 μM. A clear improvement was observed when the
phenyl ring was doubly substituted in the para and meta positions. When a m-NO2 group is associated with an EWG such as Cl, F, or Br located at
the para position in compounds 7, 11, and 12, respectively, IC50 values
are in the same micromolar range (from 2.07 to 4.4 μM). Moreover,
replacement of m-NO2 by m-CN in the p-F or p-Cl analogues 21 or 22 slightly increased the inhibitory effect
(IC50 = 0.899 and 1.01 μM). Remarkably, the analogue 28, doubly substituted with chlorine in para and meta positions, showed better inhibition. Of
special interest, an EDG (Me, Et, OMe) in the para position significantly enhanced the inhibitory activity of compounds 13–15 and 23 (146 nM <
IC50 < 514 nM). In conclusion, a meta EWG substituent (NO2, CN or Cl) and a para EDG substituent (Me or OMe) on the phenyl ring appear to be the
best combination to obtain nsp14 inhibitors with an IC50 < 0.5 μM (13–15, 23, and 29).
Substituents of the Nitrogen
Atom of the Sulfonamide Linker
After identifying several
optimal substituted aromatic rings for
effective inhibition, we next studied the functionalization of the
nitrogen atom in the sulfonamide linker with an ethyl group
to give the meta-nitro compounds 17–19, a meta-CN compound 25, and
the chlorine-containing nucleosides 31 and 32. Compared to the N-H-sulfonamide linker (0.146
μM < IC50 < 14 μM), without any exception,
the corresponding N-ethylsulfonamide linker
confers a higher level of N7-MTase nsp14 inhibition,
in the range of IC50 < 100 nM. As an example, the IC50 values of the p-Me-m-NO2-N-Et-sulfonamide derivative 18 and the corresponding NH- derivative 13 were
38 and 342 nM, respectively. The best inhibition was observed for
the m-CN-p-OMe compound 25, with IC50 = 19 nM. Nevertheless, the corresponding m-NO2-p-OMe analogue 19 and m-Cl-p-Me analogue 32 exhibited comparable high potency, with IC50 = 44 and
56 nM, respectively. Furthermore, the crucial N-substitution for high
nsp14 inhibition was demonstrated with compounds 33–35 containing butyl chains with various end groups (CO2Et, OAc, N-phthalimide). Complete inhibition
of SARS-CoV-2 nsp14 was observed at 5 μM of 33–35. Notably, nucleoside 35, carrying the phthalimide
moiety, appeared to be the best inhibitor in the series (IC50 = 30 nM). This result can be compared to recent work by Otava et
al., who identified several compounds with a similar aromatic ring
attached to the N7 position of adenine that inhibited SARS-CoV-2 nsp14
in the same low nanomolar range.[20]
Linker
Modification
To support the crucial role of
the sulfonamide linker in bisubstrate molecules, a series of
analogues of compounds 1, 7, 13, and 6 were designed with an amide linkage to give
compounds 36–39, respectively. One
of the most striking results was the lack of inhibition of the N7-MTase nsp14 by the amide-linked compounds 36–39, thus endorsing the sulfonamide linkage
in the scaffold.
Molecular Docking Studies of SARS-CoV nsp14
in Complex with 25
In this work, molecular modeling
experiments were
performed both before and after the nucleoside synthesis. Before,
these experiments allowed us to design the aromatic moiety combining
various EWG and EDG substituents at positions suitable to strongly
interact with the viral protein. To corroborate our results in the
enzymatic assays, we performed computational docking studies with
the most potent inhibitor 25, using Autodock Vina.[35] The docking was based on the structure of the
SARS-CoV nsp14-nsp10 complex solved in the presence of SAM (PDB ID: 5C8T).[22] Compound 25 was modeled in the SAM- and cap-binding
pockets of the SARS-CoV nsp14 structure, which shares 95% identity
in the sequence with SARS-CoV-2 nsp14. At first sight, nucleoside 25 perfectly overlays with the adenosine of the SAM-bound
structure (Supporting Information, Figure
S8). As with previous dinucleoside bisubstrates,[21] the benzenesulfonamide ring interacts with
Phe426 through π–π stacking interactions (d = 3.6 Å) (Figure ). Formation of a double hydrogen-bond interaction
was observed between the cyano group and Arg310 (d = 2.5, 2.6 Å), which normally interacts with the second phosphate
group of the triphosphate bond in the cap structure, also equivalent
to what had been observed with dinucleosides bearing a nitro group
in the meta position. Specifically for compound 25, a π–alkyl interaction occurs between the
CH3 of the OMe group and two aromatic residues, Tyr420
and Phe506, that naturally hold the purine portion of guanosine in
the cap structure.[22]
Figure 5
Modeling results of docking
compound 25 with the cap-binding
pocket of SARS-CoV nsp14 (PDB ID: 5C8T, resolution 3.2 Å). Contribution
of the cyanobenzenesulfonamide core of 25. Tyr420 and Phe506 (both in green), hydrogen bonds (yellow), and
the π–π stacking interaction (cyan) are shown.
Modeling results of docking
compound 25 with the cap-binding
pocket of SARS-CoV nsp14 (PDB ID: 5C8T, resolution 3.2 Å). Contribution
of the cyanobenzenesulfonamide core of 25. Tyr420 and Phe506 (both in green), hydrogen bonds (yellow), and
the π–π stacking interaction (cyan) are shown.To support the hypothesis of a bisubstrate mechanism,
we also performed
docking studies with 25 and the structure of the SARS-CoV
nsp14-nsp10 complex in the presence of SAH and GpppA. Here, we show
that the 3-cyano-4-methoxybenzenesulfonamide ring of compound 25 clearly occupies the cap-guanosine binding pocket surrounded
by Phe401, Tyr420, Phe426, Thr428, and Phe506 residues, while the
overlay with SAH is correct (Figure ).
Figure 6
Modeling results in the cap-binding (left) and SAM/SAH-binding
pocket (right) of SARS-CoV nsp14 (PDB ID: 5C8S, resolution 3.3 Å) with 25. GpppA and SAH are shown in orange and cyan, respectively. Phe426
is shown in green. π–π stacking interactions (yellow)
are shown.
Modeling results in the cap-binding (left) and SAM/SAH-binding
pocket (right) of SARS-CoV nsp14 (PDB ID: 5C8S, resolution 3.3 Å) with 25. GpppA and SAH are shown in orange and cyan, respectively. Phe426
is shown in green. π–π stacking interactions (yellow)
are shown.
Thermal Shift Assays for
SARS-CoV-2 nsp14
To further
confirm the direct interaction between compounds 25, 32, and 35 and the N7-MTase
nsp14, we performed thermal shift assays (TSAs) (Supporting Information, Figure S9). We observed a significant
shift in melting temperature (Tm) of nsp14
with high concentrations of the compounds (>2 μM). Data in
Table
S1 (Supporting Information) show that the
ΔTm of approximately +11 °C
at 0.5 mM 25, 32, and 35 indicates
remarkable stabilization of nsp14 with these three inhibitors. Dose–response
curves showing the Tm’s of 25, 32, and 35 as a function of
each compound’s concentration indicate that these three inhibitors
increased the stability of the protein more efficiently than the pan-inhibitor
sinefungin or the natural co-substrate SAM (Supporting
Information, Figure S9). This suggests that these bisubstrates
have potent interactions with the nsp14 protein, leading to a substantial
inhibition (19 nM < IC50 < 56 nM). Moreover, it is
noteworthy that all three nucleoside analogues increased the stability
of nsp14 more effectively at 0.5 mM than the dinucleoside D5 that was used at 1 mM in the previous study (Tm shift was 10.8 °C).[21] Comparing
the structures of the newly designed compounds with D5, this demonstrates that the second adenosine in D5 is
not required in the structure of the sulfonamide-containing
bisubstrates to interact with SARS-CoV-2 nsp14 effectively.
Dose–Response
Testing of Compound 25 against
SARS-CoV-2 nsp14 and Other MTases
To support
the observed inhibition of compound 25 against SARS-CoV-2
nsp14, we also tested the compound in a dose–response assay.
After pre-incubation of nsp14 with increasing concentrations of 25, MTase activity was measured by a filter-binding assay
(FBA). The IC50 of compound 25, deduced from
the Hill slope equation (Y = 100/[1 + ((X/IC50)^Hillslope)]) curve-fitting, was 11.81 ± 1.80
nM, confirming its inhibitory activity in the low nanomolar range
(Figure ). In addition,
we evaluated the inhibition of compound 25 in the presence
of other viral N7- and 2′-O-MTases: N7-MTase from vaccinia virus (D1/D12 complex)
and 2′-O-MTases from Dengue virus (NS5 MTase),
vaccinia virus (VP39), and SARS-CoV-2 (nsp10/nsp16 complex). The dose–response
assay showed no inhibition of these enzymes by compound 25 up to 50 μM (Supporting Information, Figure S10). Moreover, human RNA N7-MTase (hRNMT),
which exhibit N7 activity, was also tested in a dose–response
assay. The results showed some inhibition of N7-MTase
activity at a high concentration (50 μM), IC50 =
52.8 ± 8.31 μM. Compound 25 specifically inhibits
SARS-CoV-2 N7-MTase nsp14 with a high selectivity
of ∼2000-fold, in comparison with dinucleoside D5, which showed a selectivity of 413-fold.[21] This evidences that compound 25 has high specificity
to target SARS-CoV-2 N7-MTase nsp14.
Figure 7
IC50 curve
monitored by FBA. Increasing concentrations
of compound 25 were incubated with 50 nM SARS-CoV-2 nsp14
in the reaction mixture [40 mM Tris-HCl (pH 8.0), 1 mM DTT, 1 mM MgCl2, 2 μM SAM, and 0.1 μM 3H-SAM (Perkin
Elmer)] in the presence of 0.7 μM GpppAC4 synthetic
RNA. Reactions were incubated at 30 °C during 30 min, and the
enzymatic activity was determined by FBA. Values were normalized and
fitted with Prism (GraphPad) using the following equation: Y = 100/[1 + ((X/IC50)^Hillslope)]
(n = 3; mean value ± SD).
IC50 curve
monitored by FBA. Increasing concentrations
of compound 25 were incubated with 50 nM SARS-CoV-2 nsp14
in the reaction mixture [40 mM Tris-HCl (pH 8.0), 1 mM DTT, 1 mM MgCl2, 2 μM SAM, and 0.1 μM 3H-SAM (Perkin
Elmer)] in the presence of 0.7 μM GpppAC4 synthetic
RNA. Reactions were incubated at 30 °C during 30 min, and the
enzymatic activity was determined by FBA. Values were normalized and
fitted with Prism (GraphPad) using the following equation: Y = 100/[1 + ((X/IC50)^Hillslope)]
(n = 3; mean value ± SD).
Conclusions
Using a rational structure-guided design consistent
with a bisubstrate
strategy targeting the SARS-CoV-2 N7-MTase nsp14,
we designed and synthesized 39 SAM-derived compounds with a similar
scaffold containing a 5′-aminoadenosine linked to a disubstituted
phenyl ring through either a sulfonamide linker or an amide
linker. In a biochemical assay, seven out of the 39 compounds tested
were found to strongly inhibit N7-MTase nsp14, with
IC50 values ranging from 19 to 80 nM. The three most potent
inhibitors, 25, 32, and 35,
are high-affinity ligands for nsp14, as judged by TSA. Compound 25 is selective for the N7-MTase nsp14 over
other viral N7- or 2′-O-MTases
and, interestingly, over human N7-MTase (SI >
2000).
SAR studies consistent with our modeling results reveal an effective
bisubstrate structure: adenosine (occupying SAM-binding site) linked
to a p-EDG-m-EWG-substituted phenyl
group (occupying the cap RNA substrate binding site) through an N-ethylsulfonamide motif. These promising results
pave the way to develop new N-arylsulfonamide-containing
bisubstrate SAM analogues. Indeed, based on docking studies, our particular
scaffold interacts with two key conserved residues—Arg310 and
Phe426 in SARS-CoV nsp14—of the catalytic pocket that have
been identified as critical for N7-MTase nsp14 activity
and consequently for SARS-CoV-2 replication.[9] Our results strengthen the emerging status of this enzyme as a valid
target for antiviral rational-designed inhibitors. Further optimizations
are underway to increase the cellular permeability of this series
of potent nsp14 inhibitors with physicochemical properties tailored
for cellular activity, which will enable the characterization of their
mode of action.
Experimental Section
Chemistry
General Procedures
All dry solvents and reagents
were purchased from commercial suppliers and were used without further
purification. DIEA was distilled over calcium hydride. Thin-layer
chromatography (TLC) analyses were carried out on silica plate 60
F254. Purifications by column chromatography were performed
using Biotage Isolera 1 system with FlashPure cartridges (Buchi).
NMR experiments were recorded on Bruker 400, 500, or 600 MHz spectrometers
at 20 °C. HRMS analyses were obtained with electrospray ionization
(ESI) in positive mode on a Q-TOF Micromass spectrometer. Analytical
HPLC was performed on a UHPLC Thermoscientific Ultimate 3000 system
equipped with a LPG-3400RS pump, a DAD 3000 detector, and a WPS-3000TBRS
autosampler, Column Oven TCC-3000SD. Compounds were analyzed by RP-HPLC
on a Column Nucleodur C18 ec 100-3, 4.6 × 75 mm (Macherey
Nagel) at 30 °C. The following HPLC solvent systems were used:
1% CH3CN in 12.5 mM TEAAc (buffer A), 80% CH3CN in 12.5 mM TEEAc (buffer B). Flow rate was 1 mL/min. UV detection
was performed at 260 nm. Solid compounds 1–39 were stored at −20 °C for several months without
any degradation. Compounds 1–32 were
analyzed by HPLC and are >95% pure.
General Procedure A for
the Synthesis of Compounds 41a, 41d, 41h–l, 41n–p, and 41r–u
To a solution
at 0 °C under argon of 5′-deoxy-5′-amino-2′,3′-isopropylideneadenosine 40 (1.00 equiv) in anhydrous DMF (C = 0.05
M) were added Et3N (2.00 equiv) and the corresponding benzenesulfonyl
chloride reactant (1.25 equiv) in three portions. After stirring at
0 °C (−10 °C for 41i) for 1.5–3
h, the reaction mixture was diluted with AcOEt and brine. The aqueous
layer was extracted with AcOEt and the combined organic extracts were
washed with brine, dried over Na2SO4, and concentrated
under vacuum. The residue was purified by flash column chromatography
(dry sample, silica gel, linear gradient 0–4% MeOH in CH2Cl2) to give desired compounds as colorless solids.
General Method B for the Synthesis of Compounds 42a–k
A suspension of synthesis intermediates 41 (1.00 equiv), ethyl p-toluenesulfonate (1.50
equiv), KI (0.10 equiv) and K2CO3 (3.00 equiv)
in anhydrous DMF (C = 0.1 M) was stirred under argon at 50 °C
for 16 h. After cooling to room temperature, the reaction mixture
was diluted with AcOEt and brine. The aqueous layer was extracted
with AcOEt and the combined organic extracts were washed with brine,
dried over Na2SO4 and concentrated under vacuum.
The residue was purified by flash column chromatography (dry sample,
silica gel, linear gradient 0–4% MeOH in CH2Cl2) to give the desired compound as a colorless solid.
General Method C for the Synthesis of Compounds 43a-c
A suspension of 41m (1.00 equiv), the corresponding
alkyl bromide reagent (1.50 equiv), and K2CO3 (3.00 equiv) in anhydrous DMF was stirred under argon at 50 °C
for 16 h. After cooling to room temperature, the reaction mixture
was diluted with AcOEt and brine. The aqueous layer was extracted
with AcOEt and the combined organic extracts were washed with brine,
dried over Na2SO4 and concentrated under vacuum.
The residue was purified by flash column chromatography (dry sample,
silica gel, linear gradient 0–3% MeOH in CH2Cl2) to give the desired compound as a colorless solid.
General Method D for the
Synthesis of Compounds 44a–d
To a solution
at 0 °C under argon of 5′-amino-5′-deoxy-2′,3′-isopropylideneadenosine 40 (1.00 equiv) were successively added corresponding benzoic
acid (1.80 equiv), EDC (2.00 equiv) and DMAP (0.30 equiv) in anhydrous
DMF (C = 0.35 M). After stirring at 0 °C for
2 h, the reaction mixture was diluted with AcOEt and saturated NH4Cl solution. The aqueous layer was extracted three times with
AcOEt and the combined organic extracts were washed with brine, dried
over Na2SO4 and concentrated under vacuum. The
residue was purified by flash column chromatography (dry sample, silica
gel, linear gradient 0–5% MeOH in CH2Cl2) to give the desired compound as a foam.
General Method E for the Synthesis of Final
Compounds 1–39
Synthesis intermediates
were treated
with a formic acid/water (1/1 v:v, C = 0.05 M) solution.
After stirring at 25 °C for 24–48 h until completion of
the reaction, solvents were removed under vacuum and the crude was
co-evaporated three times with absolute EtOH. The residues were purified
by flash column chromatography (dry sample, silica gel, linear gradient
0–10% MeOH in CH2Cl2). Fractions containing
pure product were concentrated and trituration in Et2O
afforded desired compounds as solids.
All calculations
were performed using
Autodock Vina (The Scripps Research Institute, La Jolla, CA) on an
MSI computer with a 2.30 GHz Intel Core i5-8300H. The solved X-ray
crystal structure of SARS-CoV nsp14 (PDB 5C8T) was used as a static receptor for docking.
Both the co-crystalized ligand SAM and ions were removed from the
SARS-CoV nsp14 protein using VMD 1.9.3 software. The ligand structures
were drawn and minimized using MarvinSketch (ChemAxon). Targeted protein
and ligand structures with polar hydrogens were converted to the required
PDBQT format using MGL Tools (version 1.5.6). The docking was performed
with a search box located at x = −11.155, y = −40.77, z = −3.688 coordinates,
with a search box size of 25 × 25 × 25 Å3. After calculations, PDB files were analyzed using Pymol (version
2.3).
Expression and Purification of Recombinant Proteins
Dengue virus serotype 2 methyltransferase (NS5 MTase) and human RNA
N7-methyltransferase (hRNMT) coding sequences were cloned in fusion
with a N-terminus hexa-histidine tag in Gateway plasmids. The proteins
were expressed in E. coli and purified following
previously described protocols.[34,36] Vaccinia virus capping
enzyme (D1/D12 complex) and mRNA Cap 2′-O-methyltransferase
(VP39) were purchased (New England Biolabs). SARS-CoV-2 nsp14, nsp10,
and nsp16 coding sequences were cloned in fusion with a N-terminus
hexa-histidine tag in pET28 plasmids. The proteins were expressed
in E. coli C2566 and purified in a two-step IMAC
using cobalt beads. Cells were lysed by sonication in a buffer containing
50 mM Tris pH 6.8, 300 mM NaCl, 10 mM imidazole, 5 mM MgCl2, and 1 mM BME, supplemented with 0.25 mg/mL lysozyme, 10 μg/mL
DNase, and 1 mM PMSF. The protein was purified through affinity chromatography
with HisPur Cobalt resin 480 (Thermo Scientific), washing with an
increased concentration of salt (1 M NaCl) and imidazole (20 mM),
prior to elution in buffer supplemented with 250 mM imidazole. The
protein was further purified by a size exclusion chomatography (GE
Superdex S200) in a final buffer of 50 mM Tris pH 6.8, 300 mM NaCl,
5 mM MgCl2, and 1 mM βME.
MTase Filter-Binding Assay
(FBA)
The transfer of tritiated
methyl from [3H] SAM onto RNA substrate was monitored by
filter-binding assay, performed according to the method described
previously.[37] For hRNMT and SARS CoV-2
nsp14, assays were carried out in reaction mixture [40 mM Tris-HCl
(pH 8.0), 1 mM DTT, 1 mM MgCl2, 2 μM SAM, and 0.1
μM 3H-SAM (Perkin Elmer)] in the presence of 0.7
μM GpppAC4 synthetic RNA and human RNA N7 MTase (hRNMT)
(50 nM) and SARS-CoV-2 nsp14 (50 nM). For SARS CoV-2 nsp10/nsp16 (1.2
μM/0.2 μM) the reaction was performed in the presence
of 0.7 μM mGpppAC4 synthetic RNA. For NS5 MTase (500
nM) the reaction buffer does not contain MgCl2 and the
reaction was performed in the presence of 0.7 μM mGpppAC4 synthetic RNA. For vaccinia virus capping enzyme (D1–D12)
(41 U), the commercial buffer (New England Biolabs) at 1× concentration
was used and the reaction was performed in the presence of 0.7 μM
GpppAC4 synthetic RNA. For vaccinia virus VP39 (24 U) the
commercial buffer (New England Biolabs) at 1× concentration was
used and the reaction was performed in the presence of 0.7 μM
mGpppAC4 synthetic RNA.The enzymes were first mixed
with the compound suspended in 100% DMSO (5% final DMSO) before the
addition of RNA substrate and SAM and then incubated at 30 °C.
For hRNMT, 3% DMSO final concentration was used. Control reactions
were performed in the presence of 5% DMSO or 3% DMSO for hRNMT. Reactions
mixtures were stopped after 30 min by their 10-fold dilution in ice-cold
water. Samples were transferred to diethylaminoethyl (DEAE) filtermat
(Perkin Elmer) using a Filtermat Harvester (Packard Instruments).
The RNA-retaining mats were washed twice with 10 mM ammonium formate
pH 8.0, twice with water and once with ethanol. They were soaked with
scintillation fluid (Perkin Elmer), and 3H-methyl transfer
to the RNA substrates was determined using a Wallac MicroBeta TriLux
liquid scintillation counter (Perkin Elmer). For IC50 measurements,
values were normalized and fitted with Prism (GraphPad software) using
the following equation: Y = 100/[1 + ((X/IC50)^Hillslope)]. IC50 is defined as the
inhibitory compound concentration that causes 50% reduction in enzyme
activity.
Thermal Shift Assays
TSA experiments
were performed
on a Bio-Rad C1000 thermal cycler CFX96 real-time system. Briefly,
freshly purified nsp14 protein (5 μM 20 mM Hepes pH 7.5, 150
mM NaCl and do not exceed 5% of DMSO) was incubated with compounds
at a concentration ranging from 500 μM to 8.4 nM (3-to-3 serial
dilution) in the presence of Sypro Orange dye (SIGMA) used at 0.00056×.
The TSA was performed in 96-well plates (4titude FrameStar 96 Well
Skirked PCR plate) with a melt temperature increment of 1 °C
each minute from 25 to 95 °C. The Tm was determined using the Boltzmann nonlinear regression formula
(Graph-Pad PRISM 9) and ΔTm was
calculated by subtracting the compound Tm with that of negative control DMSO. The experiments were performed
in duplicate and Kd was determined from
the inflection point of the melting curve (ΔTm in function of the concentration).
Authors: Renata Kasprzyk; Tomasz J Spiewla; Miroslaw Smietanski; Sebastian Golojuch; Laura Vangeel; Steven De Jonghe; Dirk Jochmans; Johan Neyts; Joanna Kowalska; Jacek Jemielity Journal: Antiviral Res Date: 2021-07-23 Impact factor: 5.970
Authors: Fadi Soukarieh; Matthew W Nowicki; Amandine Bastide; Tuija Pöyry; Carolyn Jones; Kate Dudek; Geetanjali Patwardhan; François Meullenet; Neil J Oldham; Malcolm D Walkinshaw; Anne E Willis; Peter M Fischer Journal: Eur J Med Chem Date: 2016-08-24 Impact factor: 6.514
Authors: François Ferron; Lorenzo Subissi; Ana Theresa Silveira De Morais; Nhung Thi Tuyet Le; Marion Sevajol; Laure Gluais; Etienne Decroly; Clemens Vonrhein; Gérard Bricogne; Bruno Canard; Isabelle Imbert Journal: Proc Natl Acad Sci U S A Date: 2017-12-26 Impact factor: 11.205
Authors: Souradeep Basu; Tiffany Mak; Rachel Ulferts; Mary Wu; Tom Deegan; Ryo Fujisawa; Kang Wei Tan; Chew Theng Lim; Clovis Basier; Berta Canal; Joseph F Curran; Lucy S Drury; Allison W McClure; Emma L Roberts; Florian Weissmann; Theresa U Zeisner; Rupert Beale; Victoria H Cowling; Michael Howell; Karim Labib; John F X Diffley Journal: Biochem J Date: 2021-07-16 Impact factor: 3.857
Figure 1
Figure 2
Figure 3
Figure 4
Scheme 1
Scheme 2
Scheme 3
Figure 5
Figure 6
Figure 7