The 6'-fluorinated aristeromycins were designed as dual-target antiviral compounds aimed at inhibiting both the viral RNA-dependent RNA polymerase (RdRp) and the host cell S-adenosyl-l-homocysteine (SAH) hydrolase, which would indirectly target capping of viral RNA. The introduction of a fluorine at the 6'-position enhanced the inhibition of SAH hydrolase and the activity against RNA viruses. The adenosine and N6-methyladenosine analogues 2a-e showed potent inhibition against SAH hydrolase, while only the adenosine derivatives 2a-c exhibited potent antiviral activity against all tested RNA viruses such as Middle East respiratory syndrome-coronavirus (MERS-CoV), severe acute respiratory syndrome-coronavirus, chikungunya virus, and/or Zika virus. 6',6'-Difluoroaristeromycin (2c) showed the strongest antiviral effect for MERS-CoV, with a ∼2.5 log reduction in infectious progeny titer in viral load reduction assay. The phosphoramidate prodrug 3a also demonstrated potent broad-spectrum antiviral activity, possibly by inhibiting the viral RdRp. This study shows that 6'-fluorinated aristeromycins can serve as starting points for the development of broad-spectrum antiviral agents that target RNA viruses.
The 6'-fluorinated aristeromycins were designed as dual-target antiviral compounds aimed at inhibiting both the viral RNA-dependent RNA polymerase (RdRp) and the host cell S-adenosyl-l-homocysteine (SAH) hydrolase, which would indirectly target capping of viral RNA. The introduction of a fluorine at the 6'-position enhanced the inhibition of SAH hydrolase and the activity against RNA viruses. The adenosine and N6-methyladenosine analogues 2a-e showed potent inhibition against SAH hydrolase, while only the adenosine derivatives 2a-c exhibited potent antiviral activity against all tested RNA viruses such as Middle East respiratory syndrome-coronavirus (MERS-CoV), severe acute respiratory syndrome-coronavirus, chikungunya virus, and/or Zika virus. 6',6'-Difluoroaristeromycin (2c) showed the strongest antiviral effect for MERS-CoV, with a ∼2.5 log reduction in infectious progeny titer in viral load reduction assay. The phosphoramidate prodrug 3a also demonstrated potent broad-spectrum antiviral activity, possibly by inhibiting the viral RdRp. This study shows that 6'-fluorinated aristeromycinscan serve as starting points for the development of broad-spectrum antiviral agents that target RNA viruses.
Over
the past 15 years, outbreaks of a number of emerging positive-stranded
RNA (+RNA) viruses,[1] such as the severe
acute respiratory syndrome coronavirus (SARS-CoV),[2] Middle East respiratory syndrome coronavirus (MERS-CoV),[3] chikungunya virus (CHIKV),[4] and Zika virus (ZIKV)[5] have
seriously threatened human health and have had a substantial socio-economic
impact. SARS-CoV and MERS-CoVcause serious respiratory diseases[6] that can be fatal in approximately 10 and 35%
of cases, respectively. CHIKV is transmitted by mosquitoes and causes
a painful arthritis that can persist for months.[7] ZIKV is also transmitted by mosquitoes,[8] although sexual transmission[8] occurs as well. This virus usually causes mild disease, but can
cause neurological complications in adults and fetal death or severe
complications, including microcephaly in infants when women are infected
during pregnancy.[9] CHIKV and ZIKV have
caused massive outbreaks, totaling millions of infections over the
past decade. Currently, there are no effective chemotherapeutic agents
or vaccines that can prevent or cure infections of any of these four
serious pathogens.The aforementioned viruses belong to the
+RNA virus group (Baltimore
class IV),[1] which indicates that their
genomic RNA has the same polarity as mRNA and can be directly translated
by host ribosomes upon release into the cytoplasm of a host cell.
After infection, the genomes of these viruses are translated into
polyproteins that are subsequently cleaved into individual proteins
by viral and/or host proteases. The nonstructural proteins (nsps)
of these viruses harbor a variety of enzymatic activities that are
required for the replication of the viral RNA and invariably include
a RNA-dependent RNA polymerase (RdRp),[10] an enzyme which is not present in uninfected cells. The RdRp transcribes
the genomic RNA into a complementary negative-stranded RNA that subsequently
serves as the template for the synthesis of new positive-stranded
RNA.Many +RNA viruses (including coronaviruses, CHIKV, and
ZIKV) also
encode methyltransferases (MTases)[11] that
are required for methylations of viral mRNA cap structures.[12] Because this capping is crucial for stability
and translation of the viral RNA, and evasion of the host innate immune
response, the viral MTases are considered promising targets for the
development of antiviral therapy.[12] Inhibition
of MTases can be indirectly achieved by the inhibition of S-adenosyl-l-homocysteine (SAH) hydrolase.[13] The SAH hydrolase catalyzes the interconversion
of SAH into adenosine and l-homocysteine. Inhibition of this
enzyme leads to the accumulation of SAH in the cell, which in turn
inhibits S-adenosyl-l-methionine (SAM)-dependent
transmethylase reactions by feedback inhibition.[13,14] Most of the viral MTases are dependent on SAM as the only methyl
donor. Compounds that target cellular proteins might exhibit a broader
spectrum of activity, are less likely to lead to drug-resistance,
but have a higher likelihood of toxicity. Compounds that are specifically
aimed at viral proteins are expected to be less cytotoxic, but might
have a narrower spectrum of antiviral activity and might have a lower
barrier antiviral drug-resistance[14] Thus,
the approach of targeting cellular proteins such as SAH hydrolase
can be considered as a promising strategy for the development of broad-spectrum
antiviral agents.[14] A number of compounds
have been reported to act as SAH hydrolase inhibitors.[14] Type I inhibitors act through inactivation of
the NAD+cofactor, and their inhibitory effect on the catalytic
activity of the enzyme can be reversed by the addition of excess NAD+.[14] Type II inhibitors are irreversible
inhibitors of the SAH hydrolase that form covalent bonds with amino
acid residues in the active site of the enzyme. This irreversible
inhibition cannot be reversed by the addition of NAD+ or
adenosine or by dialysis.[14]Because
both the viral RdRp and host SAH hydrolase are critical
for virus replication, we aimed to design broad-spectrum nucleoside
analogue inhibitors that could directly target RdRp activity and/or
indirectly inhibit the methylation of viral RNA through their effect
on the host SAH hydrolase. Modified nucleosides are usually taken
up by the cell via nucleoside transporters and can be successively
converted into mono-, di-, and triphosphates by cellular kinases.[15] Then, these modified nucleoside triphosphates
(NTPs) can compete with natural NTPs during RNA synthesis or can be
incorporated into the nascent viral RNA, leading to chain termination
or detrimental mutations.[15](−)-Aristeromycin
(1) is a naturally occurring
carbocyclic nucleoside that was originally identified as a metabolite
of Streptomyces citricolor in 1967.[16a] The first synthesis of 1 as racemate
was reported by Clayton and his co-worker,[16b−16d] and its asymmetric syntheses have since been reported.[16e−16h] It is a type I SAH hydrolase inhibitor and exhibits potent antiviral
activity against many viruses.[14a] However,
it could not be further advanced into clinical development because
of its cytotoxicity.[17] Compound 1 was found to be toxic at low concentrations in both adenosine kinase-positive
(AK+) and AK– cells. AK+ cells
were presumably killed by the 5′-phosphorylated form of 1, while the toxicity in AK– cells was caused
by 1 itself.[17] However, this
compound is also metabolized into a triphosphate form and has been
observed to exert a variety of metabolic effects.[17] We aimed to use 1 as a prototype for the design
of dual-target compounds intended at directly inhibiting the viral
RdRp and indirectly inhibiting the capping process through targeting
of cellular SAH hydrolase.Since the introduction of a fluorine
at the 6′-position
of carbocyclic nucleosides has been known to affect biological activities
to a significant extent,[18] we aimed to
synthesize the 6′-fluorinated-aristeromycin analogues 2 by introducing fluorine at the 6′-position of 1 (Figure ). Prisbe and his co-workers[18a] have reported
the synthesis of (±)-6′-α- and (±)-6′-β-fluorinated
aristeromycins and their inhibitory activity on SAH hydrolase, but
the synthesis and biological activity of (±)-6′,6′-difluoroaristeromycin
was not reported, despite the fact that the structure was claimed
in the patent.[18b] Thus, we set out to synthesize
the 6′-fluorinated-aristeromycin analogues 2 in
the optically pure d-forms because biological activity can
generally be attributed to one enantiomer, the d-isomer.
Yin and co-workers[18c] reported the elegant
synthesis of optically pure (−)-6′-β-fluoro-aristeromycin,
but its biological activity was not reported. Their synthetic route
involved the 6-β-fluoroazide as the key intermediate, which
was synthesized by employing SN2 fluorination of the 6-α-triflic
azide with tris(dimethylamino)sulfur(trimethylsilyl)difluoride, whereas
our current approach[19] included the stereoselective
electrophilic fluorination of silyl enol ether with Selectfluor as
the fluorine source. In addition to the adenosine analogues, aimed
at inhibiting SAH hydrolase and/or RdRp, we have also synthesized
6′-fluorinated purine and pyrimidine nucleosides (changes in
B of the structures in Figure ), which could interfere with viral RNA synthesis by targeting
the viral RdRp after their phosphorylation by cellular kinases.[15] To bypass the first and rate-limiting 5′-monophosphorylation
step, we have also synthesized a phosphoramidate prodrug 3 of nucleoside 2, using the McGuigan ProTides.[20] Herein, we report the synthesis of the 6′-fluoro-aristeromycin
analogues 2 and 3 and a preliminary characterization
of their effect on several +RNA viruses, which provided insight into
structure–activity relationships (SARs).
Figure 1
Rationale for the design
of the target nucleosides 2 and 3.
Rationale for the design
of the target nucleosides 2 and 3.
Results and Discussion
Chemistry
For
the synthesis of the target nucleosides 2, the key fluorosugars 8a–c were synthesized
from d-ribose via electrophilic fluorination, as shown in Scheme .
Scheme 1
Synthesis of 6-β-Fluoro-,
6-α-Fluoro-, and 6-Difluorosugar 8a–c
Reagents and conditions: (a)
LiCu(CH2Ot-Bu)2; (b) TESCl,
LiHMDS, THF, −78 °C, 10 min; (c) Selectfluor, DMF, 0 °C,
12 h; (d) NaBH4, MeOH, 0 °C, 30 min. (e) LiBH4, MeOH, 0 °C, 30 min.
Synthesis of 6-β-Fluoro-,
6-α-Fluoro-, and 6-Difluorosugar 8a–c
Reagents and conditions: (a)
LiCu(CH2Ot-Bu)2; (b) TESCl,
LiHMDS, THF, −78 °C, 10 min; (c) Selectfluor, DMF, 0 °C,
12 h; (d) NaBH4, MeOH, 0 °C, 30 min. (e) LiBH4, MeOH, 0 °C, 30 min.d-Ribose was converted to d-cyclopentenone 4 according to our previously published procedure.[21] The 1,4-conjugated addition of 4 with Gilman
reagent yielded the d-cyclopentanone derivative 5.[19,22] Treatment of 5 with
lithium hexamethyldisilazide (LiHMDS) followed by trapping with triethylsilyl
chloride (TESCl) gave silylenol ether 6, which was treated
with (1-chloromethyl-4-fluoro-1,4-diazoniabicyclo[2.2.2]octane bis(tetrafluoroborate):
Selectfluor) in dimethylformamide (DMF) at 0 °C to yield a 5:1
ratio of 6-β-fluorosugar 7a to 6-α-fluorosugar 7b.[19] The stereochemistry of the
fluorine in 7a and 7b was confirmed by 1HNOE experiments. Irradiation of 6-H of 7b gave
NOE effects on its 2-H and 5-H, indicating the 6-α-fluoro configuration,
but no NOE effects were observed on the same experiment in the case
of 7a, confirming the 6-β-fluoro configuration.
The configuration of the fluorine in 7b was further confirmed
by the X-ray crystal structure obtained after it was converted to
the final uracil derivative 2g (Scheme ). Further electrophilic fluorination of
6-β-fluorosugar 7a or 6-α-fluorosugar 7b under the same conditions yielded the 6,6-difluorosugar 7c, which was equilibrated to form a geminal diol because
of the presence of electronegative fluorine atoms. Electrophilic fluorinations
with other electrophilicfluorines such as N-fluorobenzenesulfonimide
(NFSI) or N-fluoro-O-benzenedisulfonimide
(NFOBS) were problematic, resulting in low yields with many side spots.
The reduction of 7a–c with sodium borohydride
(NaBH4) or lithium borohydride (LiBH4) in MeOH
resulted in the production of the 1-hydroxyl derivatives 8a–c.
Scheme 5
Synthesis of Fluorinated
Pyrimidine Nucleoside Analogues 2f–j
Reagents and conditions: (a)
(E)-3-methoxy-2-propenoyl isocyanate, benzene, 4
Å-MS, DMF, −20 °C to rt, 15 h; (b) 2 M H2SO4, dioxane, reflux, 1.5 h; (c) BzCl, pyridine, CH2Cl2, rt, 15 h; (d) (i) 1,2,4-triazole, POCl3, Et3N, CH3CN, rt, 15 h. (ii) NH4OH, dioxane, rt, 15 h. (iii) NH3/MeOH, rt, 15 h.
As the α-fluoro derivative 8b was obtained
as
the minor isomer, as shown in Scheme , we wanted to improve the stereoselective synthesis
of 8b, by using Rubottom[23] oxidation as the key step, as illustrated in Scheme . Rubottom oxidation of silylenol ether 6 with osmium tetroxide (OsO4) and N-methylmorpholine-N-oxide (NMO) followed by trapping
with t-butyldimethylsilyl chloride (TBSCl) produced
6-β-alkoxyketone 9 as a single stereoisomer in
53% yield. The reduction of ketone 9 with NaBH4 gave alcohol 10, which was protected with a benzyl
group to give 11. Removal of the t-butyldimethylsilyl
(TBS) group in 11 with tetra-n-butylammonium
fluoride (TBAF) yielded the 6-β-alcohol 12. To
our disappointment, the treatment of 12 with N,N-diethylaminosulfur trifluoride (DAST)
gave the desired product, 6-α-fluoride13a, but
also the undesired product 1-β-fluoride13b at
a 1:1 ratio. The formation of 13a (route I) resulted
from the direct SN2 reaction of 12a with fluoride,
while 12a was readily converted into the oxonium ion 12b (route II) via its participation of the neighboring benzyl
group, which was attacked exclusively by the fluoride at the less
sterically hindered 1-position to yield the undesired product 13b (route III). However, the product via route IV was not
formed because of the steric effect of the t-butyloxymethyl
substituent.
Scheme 2
Synthetic Approach to 6-α-Fluorosugar 8b via Rubottom
Oxidation
Reagents and conditions: (a)
(i) OsO4, NMO·H2O, THF, rt, 1 h, then NaHCO3, MeOH, rt, 3 h; (ii) TBSCl, imidazole, DMF, rt, 3 h; (b)
NaBH4, MeOH, rt, 1 h; (c) BnBr, NaH, DMF, 0 °C to
rt, 12 h; (d) TBAF, THF, rt, 12 h; (e) DAST, toluene, 0 °C to
rt, 2 h.
Synthetic Approach to 6-α-Fluorosugar 8b via Rubottom
Oxidation
Reagents and conditions: (a)
(i) OsO4, NMO·H2O, THF, rt, 1 h, then NaHCO3, MeOH, rt, 3 h; (ii) TBSCl, imidazole, DMF, rt, 3 h; (b)
NaBH4, MeOH, rt, 1 h; (c) BnBr, NaH, DMF, 0 °C to
rt, 12 h; (d) TBAF, THF, rt, 12 h; (e) DAST, toluene, 0 °C to
rt, 2 h.To avoid the participation of the
neighboring group, we considered
using a cyclic sulfate substrate with electron-withdrawing property
and conformational restraint to be the best choice. Furthermore, cyclic
sulfate has the advantage that it can be utilized as a surrogate for
epoxide during nucleobase condensation, as shown in Scheme . The regioselective cleavage
of the 2,3-acetonide in 10 with trimethylaluminum (AlMe3) followed by treatment of the resulting diol with thionyl
chloride (SOCl2) yielded the 6-β-hydroxyl cyclic
sulfite 14 after the removal of the TBS group. The treatment
of 14 with DAST yielded the desired 6-α-fluoro
cyclic sulfite 15 as a single stereoisomer. The cyclic
sulfite 15 was oxidized to form cyclic sulfate 16, which was subsequently condensed with 6-chloropurine anions;
however, this resulted in decomposition.[19] Thus, we decided to synthesize the 6-α-fluoro derivative 8b according to Scheme .
Scheme 3
Synthetic Approach to 2b via Cyclic Sulfate
Reagents and conditions: (a)
AlMe3, CH2Cl2, −78 °C
to rt, 12 h; (b) SOCl2, Et3N, CH2Cl2, 0 °C, 10 min; (c) TBAF, AcOH, THF, rt, 12 h;
(d) DAST, CH2Cl2, 0 °C to rt, 4 h; (e)
RuCl3, NaIO4, CCl4/CH3CN/H2O (1/1/1.5), rt, 20 min; (f) (i) 6-chloropurine,
18-crown-6, NaH, THF, 65 °C, 15 h; (ii) 20% H2SO4, rt, 1 h.
Synthetic Approach to 2b via Cyclic Sulfate
Reagents and conditions: (a)
AlMe3, CH2Cl2, −78 °C
to rt, 12 h; (b) SOCl2, Et3N, CH2Cl2, 0 °C, 10 min; (c) TBAF, AcOH, THF, rt, 12 h;
(d) DAST, CH2Cl2, 0 °C to rt, 4 h; (e)
RuCl3, NaIO4, CCl4/CH3CN/H2O (1/1/1.5), rt, 20 min; (f) (i) 6-chloropurine,
18-crown-6, NaH, THF, 65 °C, 15 h; (ii) 20% H2SO4, rt, 1 h.Scheme depicts
the synthesis of the aristeromycin analogues 2a–e from the 6-β-fluoro-, 6-α-fluoro-, and 6,6-difluorosugars 8a–c.[19] Compounds 8a–c were treated with triflic anhydride (Tf2O) followed by treatment with sodium azide to give azido derivatives 18a–c. The catalytic hydrogenation of 18a–c yielded the amino derivatives 19a–c, respectively,
which are starting compounds for the base-building process. The treatment
of 19a–c with 5-amino-4,6-dichloropyrimidine[18a−18c,24] in the presence of N,N-diisopropylethylamine (DIPEA) under microwave
radiation conditions yielded 20a–c, which were
cyclized with diethoxymethyl acetate[18a−18c,24] in the presence of microwave radiation to produce the 6-chloropurine
derivatives 21a–c. The treatment of 21a–c with t-butanolicammonia followed by the removal
of protective groups under acidicconditions yielded the 6′-β-fluoro-,
6′-α-fluoro-, and 6′,6′-difluoroaristeromycins 2a–c, respectively. The structure of compound 2c was confirmed by a single-crystal X-ray analysis (see the Supporting Information).[25] The treatment of 21a and 21c with 40%
aqueous methylamine followed by aqueous trifluoroacetic acid (TFA)
resulted in N6-methyl-aristeromycin analogues 2d and 2e, respectively.
Scheme 4
Synthesis of β-Fluoro-,
α-Fluoro-, and Difluoro-Aristeromycin
Analogues 2a–e
Reagents
and conditions: (a)
(i) Tf2O, pyridine, 0 °C, 30 min; (ii) NaN3, DMF, 60–100 °C, 4–15 h; (b) Pd/C, H2, MeOH, rt, 18 h; (c) 5-amino-4,6-dichloropyrimidine, DIPEA, n-BuOH, 170–200 °C, 4–7 h, MW; (d) CH3C(O)OCH(OEt)2, 140 °C, 3 h, MW; (e) NH3/t-BuOH, 120 °C, 15 h; (f) NH2Me/H2O, (40 wt %), EtOH, 30 °C, 2 h; (g) 67% aq TFA,
50 °C, 15 h.
Synthesis of β-Fluoro-,
α-Fluoro-, and Difluoro-Aristeromycin
Analogues 2a–e
Reagents
and conditions: (a)
(i) Tf2O, pyridine, 0 °C, 30 min; (ii) NaN3, DMF, 60–100 °C, 4–15 h; (b) Pd/C, H2, MeOH, rt, 18 h; (c) 5-amino-4,6-dichloropyrimidine, DIPEA, n-BuOH, 170–200 °C, 4–7 h, MW; (d) CH3C(O)OCH(OEt)2, 140 °C, 3 h, MW; (e) NH3/t-BuOH, 120 °C, 15 h; (f) NH2Me/H2O, (40 wt %), EtOH, 30 °C, 2 h; (g) 67% aq TFA,
50 °C, 15 h.The amino derivatives 19a–c were also converted
into the pyrimidine nucleoside derivatives 2f–j, as shown in Scheme . Treatment of 19a–c with
(E)-3-methoxy-2-propenoyl isocyanate, which was prepared
by reacting 3-methoxyacryloyl chloride with silver cyanate,[26] in benzene produced 22a–c, respectively, which were cyclized with 2 M H2SO4 to yield the uridine derivatives 2f–h, respectively. The structures of 2g and 2h were confirmed by the X-ray crystallography (see the Supporting Information) (Scheme ).[27] To synthesize
the cytidine derivatives 2i and 2j, compounds 2f and 2h were benzoylated to give 23a and 23b, respectively, which were converted to the
cytidine derivatives 2i and 2j using conventional
three-step procedures.[28]
Synthesis of Fluorinated
Pyrimidine Nucleoside Analogues 2f–j
Reagents and conditions: (a)
(E)-3-methoxy-2-propenoyl isocyanate, benzene, 4
Å-MS, DMF, −20 °C to rt, 15 h; (b) 2 M H2SO4, dioxane, reflux, 1.5 h; (c) BzCl, pyridine, CH2Cl2, rt, 15 h; (d) (i) 1,2,4-triazole, POCl3, Et3N, CH3CN, rt, 15 h. (ii) NH4OH, dioxane, rt, 15 h. (iii) NH3/MeOH, rt, 15 h.The uracil phosphoramidate analogue Sofosbuvir[20] is used in the clinic as a powerful anti-hepatitis
C virus
agent. Therefore, we have also synthesized the uracil phosphoramidate
prodrugs 3b–c and the adenine phosphoramidate
prodrug 3a derived from the purine and pyrimidine nucleoside
analogues 2a–j by using McGuigan’s ProTide
prodrug methodology,[20] as shown in Scheme . 6′,6′-Difluoro-aristeromycin
(2c) was treated with acetone under acidicconditions
to give 2,3-acetonide 24. The treatment of 24 with di-tert-butyl dicarbonate (Boc2O) yielded a mixture of 25a and 25b in
a 2:1 ratio, which was converted to the phosphoramidate prodrug 26 by treating with phosphoramiditing reagent (A)[29] in the presence of t-butylmagnesium chloride. The treatment of 26 with 50%
formic acid produced the final product, prodrug 3a. The
monofluoro- and difluoropyrimidine derivatives 2f and 2h were similarly converted to the final prodrugs 3b and 3c.
Scheme 6
Synthesis of Phosphoramidate Prodrugs 3a–c
Reagents and conditions: (a)
cH2SO4, acetone, rt, 4 h; (b) (i) TMSOTf, DMAP,
HMDS, 75 °C, 2 h; (ii) Boc2O, THF, rt, 4 h; (iii)
MeOH/Et3N (5:1), 55 °C, 16 h; (c) A, t-BuMgCl, 4 Å-MS, THF, 0 °C to rt, 36 h; (d) 50%
HCOOH, rt, 8 h.
Synthesis of Phosphoramidate Prodrugs 3a–c
Reagents and conditions: (a)
cH2SO4, acetone, rt, 4 h; (b) (i) TMSOTf, DMAP,
HMDS, 75 °C, 2 h; (ii) Boc2O, THF, rt, 4 h; (iii)
MeOH/Et3N (5:1), 55 °C, 16 h; (c) A, t-BuMgCl, 4 Å-MS, THF, 0 °C to rt, 36 h; (d) 50%
HCOOH, rt, 8 h.
Inhibition of SAH Hydrolase
All compounds 1, 2a–j, and 3a–c were assayed
for their ability to inhibit recombinant humanSAH hydrolase protein,
expressed in Escherichia coliJM109,
using a 5,5′-dithiobis-2-nitrobenzoate (DTNB) coupled assay
as described by Lozada-Ramírez et al.[30] As expected, all adenosine derivatives 2a–e potently
inhibited SAH hydrolase, but none of the pyrimidine analogues 2f–j showed any inhibitory activity at concentrations
up to 100 μM. None of the prodrugs 3a–c exhibited
inhibitory activity at concentrations up to 100 μM. This result
is not surprising because adenosine is the substrate for SAH hydrolase.
Among the adenosine analogues, 6′-β-fluoroaristeromycin
(2a) exhibited the most potent inhibitory activity (IC50 = 0.37 μM), which was 3.6-fold more potent than the
control 1 (IC50 = 1.32 μM). However,
6′-α-fluoroaristeromycin (2b, IC50 = 9.70 μM) was 26-fold less potent than the corresponding
6′-β-fluoro analogue 2a and 7.4-fold less
active than the 6′-unsubstituted compound 1. This
indicates that the stereochemistry at the 6′-position is important
for inhibitory activity. Interestingly, the introduction of two fluorines
at the 6′-position resulted in 2c (IC50 = 1.06 μM), which was slightly more potent than the control 1. The inhibitory activity of the 6′-fluoro-aristeromycin
series can be ranked in the following order: 6′-β-F >
6′,6′-F,F > 6′-H > 6′-α-F.
The introduction
of a methyl group at the N6-amino group
of 2a, resulting in 2d, decreased the inhibitory
activity (IC50 = 4.39 μM) by 11.9-fold, while the
addition of a methyl group to the N6-amino
group of 2c, resulting in 2e, increased
the inhibitory activity (IC50 = 0.76 μM) by 1.7-fold.
These results demonstrate that the N6-methyladenine
and the adenine moieties do not lead to a decrease in inhibitory activity.
Antiviral Activity
The novel 6′-fluoro-aristeromycin
analogues 2a–j and 3a–c were
screened for antiviral activity against a variety of +RNA viruses.
The compounds were tested for antiviral activity in cytopathic effect
(CPE) reduction assays at 4 concentrations, that is, 150, 50, 16.7,
and 5.6 μM by preparing 3-fold serial dilutions. Compounds that
demonstrated antiviral activity in this primary screen were further
tested more extensively in dose response experiments at 8 different
concentrations to determine the EC50. Cytotoxicity (CC50) was determined in parallel in uninfected cells (Table ).
Table 1
Inhibition of SAH Hydrolase and the
Replication of Several +RNA Viruses by All Final Nucleoside Analogues 2a–j and 3a–ca,b,c,d
MERS-CoV
SARS-CoV
ZIKV
CHIKV
compound
no.
SAH hydrolase
IC50 (μM)
EC50 (μM)
CC50 (μM)
SI
EC50 (μM)
CC50 (μM)
SI
EC50 (μM)
CC50 (μM)
SI
EC50 (μM)
CC50 (μM)
SI
1
1.32
>50
2
>50
>5
0.64
2.4
3.8
0.8
6.3
7.9
2a
0.37
0.20
0.60
3
ND
ND
ND
ND
>100
>100
2b
9.70
ND
ND
ND
ND
2.54
3.97
1.56
0.53
1.32
2.49
2c
1.06
0.2
3.2
16
0.5
5.9
11.8
0.26
>2.5
>9.6
0.13
>1.25
>9.6
2d
4.39
>50
>50
>100
>100
>100
>100
>100
>100
2e
0.76
>50
12.5
>100
>100
>100
>100
>100
>100
2f
>100
>100
>100
>100
>100
>100
>100
>100
>100
2g
>100
>100
>100
>100
>100
>100
>100
>100
>100
2h
>100
>50
>50
>100
>100
>100
>100
>100
>100
2i
>100
>100
>100
>100
>100
>100
>100
>100
>100
2j
>100
>50
>50
>100
>100
>100
>100
>100
>100
3a
>100
9.3
>50
6.8
>25
>3.7
1.75
>25
>14.3
1.95
>12.5
>6.4
3b
>100
>50
>50
>100
>100
>100
>100
>100
>100
3c
>100
>50
>50
>100
>100
>100
>100
>100
>100
ND: not determined;
SI = CC50/EC50.
EC50: effective concentration
to inhibit the replication of the virus by 50%.
CC50: cytotoxic concentration
to inhibit the replication of normal cells by 50%.
EC50 > 100 indicates
that no antiviral activity was observed at the highest concentration
tested because either there was no protection or the compound was
toxic.
ND: not determined;
SI = CC50/EC50.EC50: effective concentration
to inhibit the replication of the virus by 50%.CC50: cytotoxicconcentration
to inhibit the replication of normal cells by 50%.EC50 > 100 indicates
that no antiviral activity was observed at the highest concentration
tested because either there was no protection or the compound was
toxic.As shown in Table , only the adenosine
derivatives 2a–c exhibited
potent antiviral activities against +RNA viruses, while the other
purineN6-methyladenine derivatives 2d and 2e and pyrimidine derivatives 2f–j did not show significant antiviral activities, not even at 100 μM.
This result suggests that the antiviral activity might be due to an
(indirect) effect on viral MTase activity through the inhibition of
host SAH hydrolase. Inhibition of the viral RdRp appears not to be
important. The mechanism of action of these compounds has been studied
in more detail and results will be published elsewhere.Compound 2a inhibited MERS-CoV replication with an
EC50 of 0.20 μM; however, it was also rather cytotoxic,
resulting in a selectivity index (SI) of 3. Replacement of the remaining
6′-H in 2a with F resulted in compound 2c, which exhibited a > 5-fold reduction in cytotoxicity, while
its
antiviral activity remained unchanged, with an EC50 of
∼0.20 μM and an SI of 15 for MERS-CoV. This compound
was also active against SARS-CoV with an SI of 12.5, suggesting that
it may be a broad-spectrum coronavirus inhibitor. In addition, it
also inhibited ZIKV replication with an EC50 of 0.26 μM
(SI > 10) and was active against CHIKV with an EC50 of
0.13 μM. Compound 2b showed some inhibitory effects
on CHIKV and ZIKV replication, but this was likely due to pleiotropiccytotoxic effects, as the SI was <3. Among the phosphoramidate
prodrugs 3a–c, only the adenosine prodrug 3a exhibited significant broad-spectrum antiviral activities,
demonstrating that it may inhibit the RdRp of RNA viruses after conversion
into the triphosphate form, although it remains to be determined in
biochemical assays whether the triphosphate form affects RdRp activity.[20] Compound 3a had an EC50 of 9.3 μM for MERS-CoV and 6.8 μM for SARS-CoV, but
it also had an SI < 10, and it was therefore not considered a potent
inhibitor of coronavirus replication. However, for CHIKV and ZIKV, 3a had EC50 values of 1.95 and 1.75 μM, respectively,
with good selectivity indices. Interestingly, the prodrug 3a was less potent, but also much less cytotoxic than the parent compound 2c, which is unusual as regularly the phosphoramidate is more
potent than the parent drug.[20] The phosphoramidate 3a might be slowly hydrolyzed to the 5′-monophosphate
by metabolic enzymes, or to the parent drug 2c by a phosphatase,
which could inhibit SAH hydrolase, explaining the observed antiviral
effect. Viral load reduction assays were performed with compound 2c by infecting cells with CHIKV, ZIKV, SARS-CoV, and MERS-CoV,
followed by treatment with different concentrations of 2c. At 30 hpi (CHIKV) or 48 hpi (ZIKV, SARS- and MERS-CoV), infectious
progeny titers in the medium were determined by plaque assay (Figure ). Treatment with
concentrations higher than 1 μM of 2c reduced infectious
CHIKV titers by more than 2 log. The effect on ZIKV infectious progeny
titers was limited and showed a ∼1 log reduction. For SARS-CoV,
the reduction in infectious progeny titer was ∼1.5 log at 2cconcentrations above 0.3 μM. The strongest antiviral effect
was observed for MERS-CoV, with a ∼2.5 log reduction in infectious
progeny titers when infected cells were treated with 2cconcentrations
above 0.3 μM. Follow-up studies to gain more insights into the
mode of action of 2c and 3a and related
compounds are currently ongoing, and results will be published elsewhere.
Figure 2
Effect
of 2c on the infectious progeny of CHIKV, ZIKV,
SARS-CoV, and MERS-CoV. Cells were infected with the virus indicated
on the y-axis of the graph in medium with various concentrations of 2c. Infectious progeny titers were determined by plaque assay
(n = 4) and viability of noninfected cells was monitored
using the CellTiter 96AQueous Non-Radioactive Cell Proliferation Assay
(Promega). Significant differences are indicated by *: *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.
Effect
of 2c on the infectious progeny of CHIKV, ZIKV,
SARS-CoV, and MERS-CoV. Cells were infected with the virus indicated
on the y-axis of the graph in medium with various concentrations of 2c. Infectious progeny titers were determined by plaque assay
(n = 4) and viability of noninfected cells was monitored
using the CellTiter 96AQueous Non-Radioactive Cell Proliferation Assay
(Promega). Significant differences are indicated by *: *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.Finally, we measured the log P of the most active
compound 2c by the pH-metric method, using a T3 Sirius
instrument, because the lipophilicity is a major determinant for compound
absorption, distribution in the body, penetration across biological
barriers, metabolism, and excretion. The measured log P was 0.02, indicating that it is almost equally partitioned between
the lipid and aqueous phases. The relatively low log P of 2c is expected to be overcome by converting it to
the phosphoramidate 3a.
Conclusions
We
have synthesized the 6′-fluorinated aristeromycin analogues 2a–j, which were designed as dual-target antiviral
compounds aimed at inhibiting both the viral RdRp and the host SAH
hydrolase. The electrophilic fluorination of silyl enol ether with
Selectfluor was the key step in the synthesis. We have also synthesized
the phosphoramidate prodrugs 3a–c to determine
whether these would inhibit virus replication through an effect on
the viral RNA polymerase. Figure depicts the summarized SAR of the synthesized 6′-fluorinated
final nucleoside analogues 2a–j and 3a–cconcerning the inhibition of humanSAH hydrolase and the inhibition
of the replication of various +RNA viruses with capped genomes. It
was discovered that the introduction of fluorine at the 6′-position
increases the inhibitory activity on SAH hydrolase and the replication
of selected +RNA viruses. Compared to the 6′-unsubstituted
compound 1, the 6′-fluorinated aristeromycin analogues 2a and 2c more potently inhibited SAH hydrolase
activity and the replication of MERS-CoV, SARS-CoV, ZIKV, and CHIKV.
Among these compounds, 6′-β-fluoroaristeromycin (2a) was the most potent with an IC50 of 0.37 μM
for SAH hydrolase activity and an EC50 of 0.20 μM
for MERS-CoV replication. There was a correlation between the inhibition
of SAH hydrolase and the antiviral activity of the compounds, suggesting
that the latter was mainly due to indirect targeting of viral methylation
reactions. The SAR studies and a lack of antiviral effects of several
purine and pyrimidine analogues suggest that the antiviral effect
of 1, 2a, and 2c is unlikely
due to targeting of the viral RdRp. Compound 2c appears
to be an interesting compound for further development and evaluation
as a broad-spectrum antiviral agent, as it inhibited several coronaviruses,
CHIKV, and ZIKV. More detailed biological studies on the efficacy
of these compounds in virus-infected cells and into their mode of
action are currently ongoing and will be published elsewhere.
Figure 3
Summarized
SAR of 6′-fluorinated aristeromycin analogues 2 and 3.
Summarized
SAR of 6′-fluorinated aristeromycin analogues 2 and 3.
Experimental Section
Chemical Synthesis
General
Methods
Proton (1H) and carbon (13C)
NMR spectra were obtained on a Bruker AV 400 (400/100
MHz), Bruker AMX 500 (500/125 MHz), JEOL JNM-ECA600 (600/150 MHz),
or Bruker AVANCE III 800 (800/200 MHz) spectrometer. Chemical shifts
are reported as parts per million (δ) relative to the solvent
peak. Coupling constants (J) are reported in hertz.
Mass spectra were recorded on a Thermo LCQ XP instrument. Optical
rotations were determined on Jasco III in appropriate solvent. UV
spectra were recorded on U-3000 made by Hitachi in methanol or water.
Infrared spectra were recorded on FT-IR (FTS-135) made by Bio-Rad.
Melting points were determined on a Buchan B-540 instrument and are
uncorrected. The crude compounds were purified by column chromatography
on a silica gel (Kieselgel 60, 70–230 mesh, Merck). Elemental
analyses (C, H, and N) were used to determine the purity of all synthesized
compounds, and the results were within ±0.4% of the calculated
values, confirming ≥95% purity.
To a cooled (−78 °C) solution
of 5 (1568.0 mg, 6.470 mmol) in anhydrous tetrahydrofuran
(THF; 32.0 mL, 0.2 M) was dropwise added chlorotriethylsilane (5.4
mL, 32.355 mmol), followed by addition of LiHMDS (19.0 mL, 1.0 M solution
in THF, 19.0 mmol) under N2. After being stirred at the
same temperature for 10 min, the reaction mixture was quenched with
saturated aqueous NH4Cl (80 mL). The layers were separated,
and the aqueous layer was extracted with ethyl acetate (EtOAc; 150
mL). The combined organic layers were washed successively with H2O and saturated brine, dried over anhydrous MgSO4, filtered, and evaporated. The residue was purified by column chromatography
(silica gel, hexanes/EtOAc, 100/1 to 30/1) to give 6 (2267.0
mg, 98%) as colorless oil: [α]D20 = +36.48 (c 1.23, CHCl3); 1HNMR (400 MHz, CDCl3): δ
4.73 (dd, J = 1.1, 6.0 Hz, 1H), 4.58 (d, J = 2.1 Hz, 1H), 4.36 (d, J = 6.1 Hz, 1H),
3.27 (dd, J = 5.6, 8.6 Hz, 1H), 3.15 (dd, J = 6.6, 8.6 Hz, 1H), 2.72 (dd, J = 5.9,
5.9 Hz, 1H), 1.42 (s, 3H), 1.32 (s, 3H), 1.12 (s, 9H), 0.96 (t, J = 8.0 Hz, 9H), 0.66–0.72 (m, 6H); 13CNMR (100 MHz, CDCl3): δ 154.1, 110.3, 104.4, 82.8,
79.7, 72.5, 63.9, 47.9, 27.4 (3 × CH3-tert-butyl), 27.3, 25.8, 6.5 (3 × triethylsilyl), 4.6 (3 ×
triethylsilyl); IR (neat): 2973, 1648, 1363, 1262, 1204, 1056, 851,
748 cm–1; HRMS (FAB): found, 356.2388 [calcd for
C19H36O4Si+ (M + H)+, 356.2383].
(3aR,5R,6R,6aR)-6-(tert-Butoxymethyl)-5-fluoro-2,2-dimethyldihydro-3aH-cyclopenta[d][1,3]dioxol-4(5H)-one (7a) and (3aR,5S,6R,6aR)-6-(tert-Butoxymethyl)-5-fluoro-2,2-dimethyldihydro-3aH-cyclopenta[d][1,3]dioxol-4(5H)-one (7b)
To a cooled (0 °C) solution
of silyl enol ether 6 (8.75 g, 24.548 mmol) in anhydrous
DMF (123.0 mL, 0.20 M)
was added 1-chloromethyl-4-fluoro-1,4-diazoniabicyclo[2.2.2]octane
bis(tetrafluoroborate) (13.04 g, 36.824 mmol, Selectfluor) in one
portion under N2. After being stirred at the same temperature
for 12 h, the reaction mixture was quenched with saturated aqueous
NH4Cl (130 mL), diluted with EtOAc (130 mL). The layers
were separated and the aqueous layer was extracted with EtOAc (2 ×
100 mL). The combined organic layers were washed successively with
H2O and saturated brine, dried over anhydrous MgSO4, filtered, and evaporated. The residue was purified by column
chromatography (silica gel, hexanes/EtOAc, 40/1 to 20/1) to give 7a and 7b (5.80 g, 91%, total yield, 7a/7b = 5.2:1 by 1HNMR analysis).
To a cooled (0 °C) solution of 7a–c (1 equiv) in MeOH (0.18 M), sodium borohydride
or lithium borohydride
was added in a single portion in a N2 atmosphere. After
stirring for 30 min at the same temperature, the reaction mixture
was neutralized with acetic acid (2 mL) and evaporated. The residue
was diluted with saturated aqueous NH4Cl, and the aqueous
layer was extracted with EtOAc (2 × 100 mL). The combined organic
layers were dried over anhydrous MgSO4, filtered, and evaporated.
The residue was purified by column chromatography (silica gel, hexanes/EtOAc,
20/1) to give 8a–c.
To a cooled (0 °C) solution
of 6 (1275 mg, 3.57 mmol) in anhydrous THF (12 mL, 0.3
M) were added 4-methylmorpholine N-oxide monohydrate
(967 mg, 7.15 mmol, 2 equiv) and osmium tetroxide (1000 mg, 3.93 mmol,
1.1 equiv) under a N2 atmosphere. After stirring for 30
min, to the reaction mixture were added sodium thiosulfate pentahydrate
(300 mg), sodium sulfite (300 mg), and acetone (30 mL) and stirred
for additional 1 h at the same temperature. The layers were separated,
and the aqueous layer was extracted with EtOAc (100 mL). The combined
organic layers were washed with H2O followed by saturated
brine, dried over anhydrous MgSO4, filtered, and evaporated.
The residue was used for the next step without further purification.
To a solution of above generated intermediate in anhydrous DMF (18
mL, 0.19 M) were added TBSCl (1614 mg, 10.71 mmol) and imidazole (729
mg, 10.71 mmol) under a N2 atmosphere. After stirring for
3 h at room temperature, the reaction mixture was quenched with saturated
aqueous NH4Cl (50 mL) and diluted with EtOAc (50 mL). The
layers were separated, and the aqueous layer was extracted with EtOAc
(2 × 50 mL). The combined organic layers were washed successively
with H2O and saturated brine, dried over anhydrous MgSO4, filtered, and evaporated. The residue was purified by column
chromatography (silica gel, hexanes/EtOAc, 40/1 to 20/1) to give 9 (705 mg, 53%) as a colorless syrup: [α]D25 = −103.19
(c 0.30, MeOH); 1HNMR (400 MHz, CDCl3): δ 4.65 (d, J = 6.4 Hz, 1H), 4.53
(d, J = 8.0 Hz, 1H), 4.11 (d, J =
6.3 Hz, 1H), 3.61 (dd, J = 1.6, 8.0 Hz, 1H), 3.30
(dd, J = 2.4, 8.1 Hz, 1H), 2.41–2.46 (m, 1H),
1.42 (s, 3H), 1.30 (s, 3H), 1.03 (s, 9H), 0.88 (s, 9H), 0.13 (s, 3H),
0.05 (s, 3H); 13CNMR (100 MHz, CDCl3): δ
207.2, 110.9, 78.1, 75.8, 73.7, 71.3, 56.9, 42.3, 27.0 (3 × CH3-tert-butyl), 26.4, 25.7 (3 × CH3-tert-butyl), 23.8, 18.3, −4.4, −5.6;
HRMS (FAB+) (m/z): found,
373.2398 [calcd for C19H37O5Si+ (M + H)+, 373.2410]; Anal. Calcd for C19H36O5Si: C, 61.25; H, 9.74. Found: C, 61.26;
H, 9.75.
To a cooled (0 °C) solution of 12 (20
mg, 0.052 mmol) in anhydrous toluene (2.0 mL, 0.026 M) was dropwise
added diethylaminosulfur trifluoride (30 μL, 0.210 mmol, 4.0
equiv) under a N2 atmosphere. After being stirred at room
temperature for 2 h, the reaction mixture was quenched with saturated
aqueous NH4Cl (30 mL) and EtOAc (30 mL). The layers were
separated, and the aqueous layer was extracted with EtOAc (3 ×
50 mL). The combined organic layers were washed successively with
H2O and saturated brine, dried over anhydrous MgSO4, filtered, and evaporated. The residue was purified by column
chromatography (silica gel, hexanes/EtOAc, 30/1) to give 13a (5.6 mg, 30%) and 13b (5.6 mg, 30%) as a colorless
syrup.
To perform regioselective cleavage, to
a cooled (−78 °C) solution of 10 (420 mg,
1.121 mmol) in anhydrous CH2Cl2 (5.6 mL, 0.2
M) was dropwise added trimethylaluminum (3.4 mL, 2.0 M solution in
hexane, 6.727 mmol, 6.0 equiv) under a N2 atmosphere. After
being stirred at room temperature for 12 h, the reaction mixture was
quenched with saturated aqueous NH4Cl (30 mL) and EtOAc
(30 mL). The layers were separated, and the aqueous layer was extracted
with EtOAc (3 × 50 mL). The combined organic layers were washed
successively with H2O and saturated brine, dried over anhydrous
MgSO4, filtered, and evaporated. The residue was purified
by column chromatography (silica gel, hexanes/EtOAc, 10/1) to give
diol intermediate (245 mg, 56%) 10a as a colorless syrup. For the
introduction of cyclic sulfite, to a cooled (0 °C) solution of
diol intermediate 10a (250 mg, 0.639 mmol) in anhydrous
CH2Cl2 (6.4 mL, 0.1 M) was dropwise added triethylamine
(0.3 mL, 2.239 mmol, 3.5 equiv) followed by thionyl chloride (70 μL,
0.959 mmol) under a N2 atmosphere. After being stirred
at room temperature for 30 min, the reaction mixture was quenched
with saturated aqueous NH4Cl (30 mL) and diluted with EtOAc
(30 mL). The layers were separated, and the aqueous layer was extracted
with EtOAc (3 × 50 mL). The combined organic layers were washed
successively with H2O and saturated brine, dried over anhydrous
MgSO4, filtered, and evaporated. The residue was purified
by flash column chromatography (silica gel, hexanes/EtOAc, 10/1) to
give cyclic sulfite intermediate 10b (249 mg, 89%) as
a colorless syrup. For TBS deprotection, to a cooled (0 °C) solution
of 10b (286 mg, 0.654 mmol) in anhydrous THF (6.5 mL,
0.1 M) was added acetic acid (0.13 mL, 0.131 mmol, 0.2 equiv) followed
by TBAF solution (2.6 mL, 1.0 M solution in THF, 2.6 mmol, 4.0 equiv)
under a N2 atmosphere. After being stirred at room temperature
for 12 h, the reaction mixture was quenched with H2O (30
mL) and diluted with EtOAc (30 mL). The layers were separated, and
the aqueous layer was extracted with EtOAc (3 × 50 mL). The combined
organic layers were washed successively with H2O and saturated
brine, dried over anhydrous MgSO4, filtered, and evaporated.
The residue was purified by column chromatography (silica gel, hexanes/EtOAc,
6/1) to give 14 (202 mg, 96%, two diastereomers A and B were generated from sulfoxide stereogeniccenter) as a colorless syrup: for A: 1HNMR
(400 MHz, CDCl3): δ 5.27 (t, J =
5.4 Hz, 1H), 5.02 (d, J = 5.9 Hz, 1H), 4.79 (s, 1H),
4.44 (dd, J = 4.8, 11.4 Hz, 1H), 4.19 (d, J = 3.9 Hz, 1H), 3.80 (dd, J = 2.6, 9.3
Hz, 1H), 1.90–1.94 (m, 1H), 1.27 (s, 9H), 1.21 (s, 9H); 13CNMR (125 MHz, CDCl3): δ 86.9, 82.6, 74.9,
74.5, 74.1, 69.4, 58.2, 43.6, 28.3 (3 × CH3-tert-butyl), 27.2 (3 × CH3-tert-butyl); HRMS (FAB+) (m/z): found, 323.1530 [calcd for C14H27O6S+ (M + H)+, 323.1528]; for B: 1HNMR (500 MHz, CDCl3): δ 4.98–5.07
(m, 2H), 4.79 (d, J = 6.4 Hz, 1H), 4.36 (dd, J = 4.6, 11.5 Hz, 1H), 4.31 (d, J = 4.1
Hz, 1H), 3.84 (d, J = 9.2 Hz, 1H), 3.77 (d, J = 9.3 Hz, 1H), 2.65 (d, J = 10.1 Hz,
1H), 1.25 (s, 9H), 1.21 (s, 9H).
To a solution of cyclic sulfite 15 (13 mg, 0.040 mmol) in CCl4/CH3CN/H2O (1:1:1.5, total 1.75 mL, 0.14 M) was added in one portion sodium
periodate (26 mg, 0.120 mmol), followed by ruthenium(III) chloride
trihydrate (2 mg, 0.008 mmol) at room temperature under a N2 atmosphere. After being stirred at the same temperature for 20 min,
the reaction mixture was quenched with H2O (20 mL) and
diluted with CH2Cl2 (20 mL). The layers were
separated, and the aqueous layer was extracted with CH2Cl2 (2 × 50 mL). The combined organic layers were
washed successively with H2O and saturated brine, dried
over anhydrous MgSO4, filtered, and evaporated. The crude
product 16 was used for the next step without further
purification.
General Procedure for the Synthesis of 18a–c
Triflation
To a cooled (0 °C)
solution of 8a–c (1 equiv) in anhydrous pyridine
(0.32 M), trifluoromethanesulfonic
anhydride (2 equiv) was added dropwise in a N2 atmosphere.
After stirring at the same temperature for 30 min, the reaction mixture
was quenched with H2O (50 mL) and diluted with EtOAc (30
mL). The layers were separated, and the aqueous layer was extracted
with EtOAc (2 × 30 mL). The combined organic layers were washed
with saturated aqueous CuSO4 followed by water, dried over
anhydrous MgSO4, filtered, and evaporated. The residue
was used for the next step without further purification.
Azidation
To a solution of triflate intermediate (1
equiv) in anhydrous DMF (0.19 M), sodium azide (3 equiv) was added
in a single portion at room temperature. After being heated to 60–100
°C and stirred for 4–15 h, the reaction mixture was cooled
to room temperature, quenched with H2O (50 mL), and diluted
with EtOAc (50 mL). The layers were separated, and the aqueous layer
was extracted with EtOAc (2 × 50 mL). The combined organic layers
were washed with H2O followed by saturated brine, dried
over anhydrous MgSO4, filtered, and evaporated. The residue
was purified by column chromatography (silica gel, hexanes/EtOAc,
10/1) to give 18a–c.
To a suspension of 18a–c (1 equiv)
in methanol (0.2 M), 10% palladium on activated carbon (0.03 equiv)
was added and stirred overnight at room temperature in a H2 atmosphere. After filtration, the solvent was removed, and the residue
was used for the next step without further purification.
General
Procedure for the Synthesis of 20a–c
To a solution of 19a–c (1 equiv) in n-butanol (0.38 M), 5-amino-4,6-dichloro pyrimidine (3–10
equiv) and diisopropylamine (10 equiv) were added. The reaction mixture
was placed under microwave irradiation at 170–200 °C for
4–7 h. The solvent was co-evaporated with MeOH, and the residue
was purified with column chromatography (silica gel, hexane/EtOAc,
4/1) to give 20a–c, respectively.
A solution of 20a–c in diethoxymethyl
acetate (0.15 M) was placed under microwave irradiation at 140 °C
for 3 h. The mixture was then co-evaporated with MeOH three times,
and the resulting residue was purified with column chromatography
(silica gel, hexane/EtOAc, 7/1) to give 21a–c.
To
a solution of 21a–c in tert-butanol
(2 mL, 0.27 M) contained in a stainless steel bomb reactor,
saturated ammonia in tert-butanol (15 mL) was added
and the reactor was locked. After being heated to 120 °C with
stirring for 15 h, the mixture was cooled to room temperature and
co-evaporated with MeOH. Without purification, the residue was added
to a TFA/H2O solution (2:1, v/v, total 15 mL) and heated
to 50 °C with stirring for 15 h. After the reaction mixture was
evaporated, the residue was purified by column chromatography (silica
gel, CH2Cl2/MeOH, 9/1) to give 2a–c.
To a solution of 21a and 21c (0.283 mmol) in EtOH (1.5 mL, 0.19 M) in a sealed glass tube, methylamine
(40 wt % in H2O, 10 mL) was added. After being stirred
at room temperature for 2 h, the mixture was concentrated and added
to a TFA/H2O solution (2:1, v/v, total 15 mL) without purification.
After being heated to 50 °C with stirring for 15 h, the reaction
mixture was evaporated. The residue was purified by column chromatography
(silica gel, CH2Cl2/MeOH, 9/1) to give 2d and 2e.
To a cooled (−20 °C) solution of 19a–c (1 equiv) in DMF (0.2 M), 3-methoxyacryloyl isocyanate (2 equiv)
in benzene was added dropwise in a N2 atmosphere. After
the reaction mixture was slowly warmed to room temperature for 15
h with stirring, the reaction mixture was filtered with CH2Cl2 and co-evaporated with toluene and ethanol. The residue
was purified by column chromatography (silica gel, hexane/EtOAc, 1.5/1)
to give 22a–c.
To a stirred solution of 22a–c in 1,4-dioxane
(3 mL, 2.5 M), 2 M sulfuric acid (0.3 mL) was dropwise added. After
refluxing with stirring for 1 h, the reaction mixture was cooled to
room temperature and neutralized with DOWEX 66 ion-exchange resin.
The mixture was filtered and evaporated. The residue was purified
by column chromatography (silica gel, CH2Cl2/MeOH, 9/1) to give 2f–h.
To a cooled (0 °C)
solution of 2f or 2h (1 equiv) in CH2Cl2 (0.07 M), benzoyl chloride (6 equiv) and pyridine
(6.7 equiv) were
added in a N2 atmosphere. After being stirred for 15 h
at room temperature, the reaction mixture was quenched with H2O and extracted with CH2Cl2. The organic
layers were combined and washed with H2O followed by brine,
dried over MgSO4, filtered, and evaporated. The residue
was purified with column chromatography (silica gel, hexane/EtOAc,
1/1) to give the benzoylated intermediate.
Introduction
of Triazole
To a cooled (0 °C) suspension
of 1,2,4-triazole (10 equiv) in anhydrous MeCN (0.6 M), phosphoryl
chloride (10 equiv) was added dropwise in a N2 atmosphere.
After stirring, the benzoylated intermediate (1 equiv) in MeCN (0.14
M) followed by trimethylamine (10 equiv) were added to the reaction
mixture. After additional stirring at room temperature for 15 h, the
reaction mixture was evaporated. The reaction mixture was diluted
with CH2Cl2 and H2O. The layers were
separated, and the organic layers were washed with H2O,
dried over MgSO4, filtered, and evaporated.
Amination
In the sealed glass tube, the above-generated
intermediate in 1,4-dioxane (0.06 M) was added to excess saturated
aqueous ammonia at room temperature. After being stirred at the same
temperature for 2 h, the reaction mixture was evaporated and purified
with flash chromatography (silica gel, CH2Cl2/MeOH, 7/1) to give the benzoyl protected cytosine intermediate.
Benzoyl Deprotection
In a sealed glass tube, the above-generated
benzoyl-protected cytosine intermediate in MeOH (0.2 M) was added
to saturated ammonia in MeOH (0.2 M). After being stirred at the same
temperature for 2 d, the reaction mixture was evaporated and diluted
with H2O and CH2Cl2. The layers were
separated, and the H2O layers were washed with CH2Cl2 10 times and evaporated to give 2i and 2j, respectively.
General Procedure for the Synthesis of 24, 27a and 27b
To a cooled (0 °C)
suspension of 2c, 2f, and 2h (1 equiv) in acetone (0.005 M) were added 1–2 drops of cH2SO4 in N2 (g). After being stirred at
room temperature for 4 h, the reaction mixture was neutralized with
solid NaHCO3, filtered, and evaporated under reduced pressure.
The residue was further purified by silica gelcolumn chromatography
to give 24, 27a, and 27b, respectively.
Synthesis
of tert-Butyl-(9-((3aS,4S,6R,6aR)-5,5-difluoro-6-(hydroxymethyl)-2,2-dimethyltetrahydro-4H-cyclopenta[d][1,3]dioxol-4-yl)-9H-purin-6-yl)carbamate (25a) and Its N6-Di-Boc derivative (25b)
To a suspension of 24 (20 mg, 0.058 mmol) and 4-dimethylaminopyridine
(1 mg, 0.0058 mmol) in hexamethyldisilazane (3 mL), trimethylsilyl
trifluoromethanesulfonate (TMSOTf; 5 μL) was added dropwise
at room temperature in a N2 atmosphere (g). After being
heated to 75 °C with stirring for 2 h, the reaction mixture was
evaporated, and anhydrous THF (7 mL) was added. To a cooled (0 °C)
reaction mixture, di-t-butyl dicarbonate (63 mg,
0.29 mmol) was added. After stirring for 4 h at room temperature,
the reaction mixture was evaporated, and the residue was added to
MeOH/trimethylamine (6 mL, 5:1 (v/v)). After heating to 55 °C
with stirring for 16 h, the reaction mixture was evaporated, and the
residue was purified with column chromatography (silica gel, CH2Cl2/MeOH, 50/1) to give 25a (13 mg,
52%) and 25b (8 mg, 25%) as a colorless syrup.
The gene encoding human
placental SAH hydrolase was cloned
into expression plasmid pPROKcd20. Recombinant SAH hydrolase protein
was produced in E. coliJM109 in 50
mM Tris-HCl (pH 7.5) containing 2 mM ethylenediaminetetraacetic acid
and was purified by DEAE-cellulosecolumn (2.8 cm × 6 cm), ammonium
sulfate fractionation (35–60%), Sephacryl S-300HR (1.0 cm ×
105 cm), and DEAE cellulose (2.8 cm × 24 cm). The protein homogeneity
was confirmed by 10% sodium dodecyl sulfatepolyacrylamide gel electrophoresis.
The protein concentration was determined by using the Bradford method.
Bovine serum albumin was a standard material for protein assay. Enzyme
activity was determined in reaction mixtures (250 μL) that contain
50 mM sodium phosphate (pH 8.0), 2 μM SAH hydrolase (0.5 μM
tetrameric form), and varying concentrations of compounds. The reaction
mixtures were first preincubated with the compounds for 10 min at
37 °C, after which the reaction was initiated by adding 100 μM
SAH. The reaction was allowed to proceed for 20 min, followed by the
addition of 5,5′-dithiobis-2-nitrobenzoate (DNTB) to a final
concentration of 200 μM. The absorbance of the product 5-thio-2-nitrobenzoic
acid (TNB) was measured at 412 nm using a spectrophotometer (Varian,
Cary 100). The molar extinction coefficient for TNB (ε412 = 13 700 M–1 cm–1) was
used in calculations to quantify TNB formation.
Cells, Viruses,
and Compounds
Vero E6 and Vero CCL81
cells were maintained in Dulbecco’s modified Eagle’s
medium (DMEM; Lonza), supplemented with 8% fetal calf serum (FCS;
PAA), 2 mM l-glutamine, 100 IU/mL of penicillin and 100 μg/mL
of streptomycin, and were grown at 37 °C in a humidified incubator
with 5% CO2. Vero cells were maintained in Eagle’s
minimum essential medium (EMEM; Lonza), supplemented with 8% FCS (FCS;
PAA), 100 IU/mL of penicillin and 100 μg/mL of streptomycin,
and were grown at 37 °C in a humidified incubator with 5% CO2. Infections were performed in EMEM with 25 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic
acid (Lonza) supplemented with 2% FCS, l-glutamine, and antibiotics.
Infectious clone-derived CHIKV(CHIKV-LS3) was generated as described
by Scholte et al.[31] The ZIKV strain SL0612
was isolated from an infected traveler returning from Suriname as
described by van Boheemen et al.[32] The
Sindbis virus (SINV) strain HR and Semliki forest virus (SFV) strainSFV4 are part of the LUMC virus collection. The MERS-CoV strain EMC/2012
was isolated from patient material in the Dr. Soliman Fakeeh Hospital,
Jeddah, Saudi Arabia, and was obtained from Erasmus Medical Center,
Rotterdam.[33] The SARS-CoV strain Frankfurt
1 was provided by H. F. Rabenau and H. W. Doerr (Johann Wolfgang Goethe-Universität,
Frankfurt am Main, Germany).[34] The compounds
were dissolved in dimethylsulfoxide to obtain 20 mM stock solutions.
All work with infectious CHIKV, MERS-CoV, SARS-CoV, and ZIKV was performed
inside biosafety cabinets in the BSL-3 facilities of the Leiden University
Medical Center.
Antiviral CPE-Reduction Assays
VeroE6
cells were seeded
at a density of 5000 cells/well (CHIKV) and 10 000 cells/well
(SARS-CoV, SFV and SINV) in a total volume of 100 μL per well
in 96-well plates. Vero cells were seeded at a density of 20 000
cells/well when used for MERS-CoV infections, and Vero CCL81 cells
were seeded at a density of 5000 cells/well for ZIKA infections under
the same conditions as described for Vero E6. The following day, compound
dilutions with concentrations of 150, 50, 16.7, and 5.6 μM were
prepared in the infection medium by 3-fold serial dilution of the
150 μM solution. After replacing the culture medium with the
respective dilutions of the compound, the cells were infected with
CHIKV (MOI 0.005), SFV (MOI 0.025), SINV (MOI 0.025), ZIKV (MOI 0.05),
MERS-CoV (MOI 0.005), or SARS-CoV (MOI 0.01). Viability assays were
conducted in parallel. Each compound was tested at each concentration
in quadruplicate (4 biological replicates per concentration). An MTS
colorimetric assay was conducted 40 hpi for SFV, 76 hpi for SINV,
72 h hpi for MERS- and SARS-CoV, and 96 hpi for CHIKV and ZIKV by
adding 20 μL/well of the CellTiter 96 AQueous One Solution Cell
Proliferation Assay (MTS) reagent (Promega). The assay was stopped
after 2–2.5 h by fixing the cells with 37% formaldehyde. The
absorbance was measured at 495 nm in a Berthold Mithras LB 940 plate
reader, and the values were expressed relative to uninfected (infection)
or untreated (viability) samples. The results represent the average
of quadruplicate samples expressed as the mean ± SD. Compounds
that were found to be protective were further evaluated in CPE reduction
assays by testing 8 different concentrations to determine the EC50 as previously described.[31,34] The cytotoxicity
(CC50) of the compounds was determined in parallel, and
all experiments were performed in quadruplicate. Graph-Pad Prism 8.0.1
was used for EC50 and CC50 determination by
nonlinear regression.
Viral Load Reduction Assays
VeroE6
(CHIKV, ZIKV) cells
were seeded at a density of 7.5 × 104 cells/well in
0.5 mL DMEM/8%FCS in 24-well cell culture plates and allowed to adhere
overnight. For MERS-CoV and SARS-CoV, a cell density of or 6.0 ×
104 cells/well of Vero E6 and Vero cells was used, respectively,
under the same conditions as described above. The next day, compound
dilutions (0–1.5 μM) were prepared in EMEM/2%FCS to which
virus was added to yield inocula for infecting the cells with an MOI
of 0.1 for CHIKV, MOI of 1 for ZIKV, and MOI of 0.01 for SARS- and
MERS-CoV. Cells were incubated at 37 °C with 250 μL/well
of the inoculum for 1 h (CHIKV and SARS- and MERS-CoV) or 2 h (ZIKV).
After the infection, the cells were washed twice with 1 mL/well warm
phosphate-buffered saline and 0.5 mL/well fresh EMEM/2%FCS with different
concentrations of compound 2c (0–1.5 μM)
was added. The cells were incubated for 30 h (CHIKV) or 48 h (ZIKV,
SARS- and MERS-CoV) at 37 °C, after which supernatants were harvested
and stored at −80 °C for determination of the infectious
virus titer by plaque assay. Viability assays were conducted in parallel
as described in the previous paragraph. Plaque assays with CHIKV and
SARS-CoV on VeroE6 cells, MERS-CoV on Vero cells, and ZIKV on Vero
CCL81 cells were performed as described previously.[31,34a,35] Compound 2c was tested at each
concentration in duplicate in two independent experiments (n = 4). Graph-Pad Prism 8.0.1 was used for statistical analysis
with a one-way ANOVA multiple comparison test.
Authors: Sander van Boheemen; Ali Tas; S Yahya Anvar; Rebecca van Grootveld; Irina C Albulescu; Martijn P Bauer; Mariet C Feltkamp; Peter J Bredenbeek; Martijn J van Hemert Journal: Sci Rep Date: 2017-05-24 Impact factor: 4.379
Authors: Stanislaw P Stawicki; Rebecca Jeanmonod; Andrew C Miller; Lorenzo Paladino; David F Gaieski; Anna Q Yaffee; Annelies De Wulf; Joydeep Grover; Thomas J Papadimos; Christina Bloem; Sagar C Galwankar; Vivek Chauhan; Michael S Firstenberg; Salvatore Di Somma; Donald Jeanmonod; Sona M Garg; Veronica Tucci; Harry L Anderson; Lateef Fatimah; Tamara J Worlton; Siddharth P Dubhashi; Krystal S Glaze; Sagar Sinha; Ijeoma Nnodim Opara; Vikas Yellapu; Dhanashree Kelkar; Ayman El-Menyar; Vimal Krishnan; S Venkataramanaiah; Yan Leyfman; Hassan Ali Saoud Al Thani; Prabath Wb Nanayakkara; Sudip Nanda; Eric Cioè-Peña; Indrani Sardesai; Shruti Chandra; Aruna Munasinghe; Vibha Dutta; Silvana Teixeira Dal Ponte; Ricardo Izurieta; Juan A Asensio; Manish Garg Journal: J Glob Infect Dis Date: 2020-05-22
Authors: Kristina Kovacikova; Bas M Morren; Ali Tas; Irina C Albulescu; Robin van Rijswijk; Dnyandev B Jarhad; Young Sup Shin; Min Hwan Jang; Gyudong Kim; Hyuk Woo Lee; Lak Shin Jeong; Eric J Snijder; Martijn J van Hemert Journal: Antimicrob Agents Chemother Date: 2020-03-24 Impact factor: 5.191
Figure 1
Scheme 1
Scheme 5
Scheme 2
Scheme 3
Scheme 4
Scheme 6
Figure 2
Figure 3