Synthesis, structure-activity relationship studies and evaluation of a TLR 3/8/9 agonist and its analogues

Article abstract of DOI:10.1007/s00044-021-02736-3

A comprehensive SAR study of a putative TLR 3/8/9 agonist was conducted. Despite the excitement surrounding the potential of the first small molecule TLR3 agonist with a compound that additionally displayed agonist activity for TLR8 and TLR9, compound 1 displayed disappointing activity in our hands, failing to match the potency (EC₅₀) reported and displaying only a low efficacy for the extent of stimulated NF-κB activation and release. The evaluation of >75 analogs of 1, many of which constitute minor modifications in the structure, failed to identify any that displayed significant activity and none that exceeded the modest activity found for 1. [Figure not available: see fulltext.].

Full text of DOI:10.1007/s00044-021-02736-3

MEDICINAL
CHEMISTRY
RESEARCH
Medicinal Chemistry Research
https://doi.org/10.1007/s00044-021-02736-3
ORIGINAL RESEARCH
Synthesis, structure-activity relationship studies and evaluation of a
TLR 3/8/9 agonist and its analogues
Anindya Sarkar1 Anushka C. Galasiti Kankanamalage Qian Zhang Heng Cheng Prasanna Sivaprakasam
1 ●
2 ●
2 ●
3 ●
●
Joseph Naglich3 Chunshan Xie Sanjeev Gangwar Dale L. Boger
3 ●
2 ●
1
●
Received: 7 March 2021 / Accepted: 30 April 2021
©
The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2021
Abstract
A comprehensive SAR study of a putative TLR 3/8/9 agonist was conducted. Despite the excitement surrounding the
potential of the first small molecule TLR3 agonist with a compound that additionally displayed agonist activity for TLR8 and
TLR9, compound 1 displayed disappointing activity in our hands, failing to match the potency (EC ) reported and
5
0
displaying only a low efficacy for the extent of stimulated NF-κB activation and release. The evaluation of >75 analogs of 1,
many of which constitute minor modifications in the structure, failed to identify any that displayed significant activity and
none that exceeded the modest activity found for 1.
Graphical Abstract
●
●
●
Keywords Toll-like receptor (TLR) TLR agonists TLR3 agonist SAR study
Introduction
These authors contributed equally: Anindya Sarkar, Anushka C.
Galasiti Kankanamalage
Toll-like receptors (TLRs) play a critical role in the mam-
Dedication: This article is dedicated to Professor Gary Grunewald, a
treasured longtime friend and early colleague alongside which I had
the pleasure of working for many years.
malian defense against microbial infections [1]. The TLRs
are germline-encoded receptors that recognize pathogen-
associated molecular patterns (PAMPs) or damage-
associated molecular patterns (DAMPs), highly conserved
molecular components present in microorganisms or
Supplementary information The online version contains
supplementary material available at https://doi.org/10.1007/s00044-
0
21-02736-3.
*
Dale L. Boger
boger@scripps.edu
2
Bristol Myers Squibb Research & Development, 700 Bay Road,
Redwood City, CA 94063, USA
1
3
Department of Chemistry and Skaggs Institute for Chemical
Biology, The Scripps Research Institute, 10550 N. Torrey Pines
Road, La Jolla, CA 92037, USA
Bristol Myers Squibb Research & Development,
PO Box 4000, Princeton, NJ 08543, USA

Medicinal Chemistry Research
abnormal cells, respectively [2]. Ligand induced TLR
dimerization and activation marshals both the initial innate
and slower adaptive immune responses to protect the host
from infectious pathogens [3–5] or remove abnormal cells.
TLR activation for the treatment of infectious diseases or
cancer have been widely explored where TLR agonists
serve as attractive and efficacious vaccine adjuvants [6–10].
Despite their potential, there is not yet a wide range of small
molecule TLR agonists available. Most bear poor structural,
physical, or stability properties found in the identified
naturally-occurring agonists that make them unattractive
starting points for direct use or even further optimization
[19]. It represents the first and only small molecule TLR3
agonist for a receptor that ordinarily senses viral double-
stranded RNA [25]. It also acts as a multi-TLR agonist,
activating TLRs 3/8/9 [19]. The compound showed strong
NF-κB activation through hTLR 3, 8 and 9 (EC = 4.8,
50
13.4 and 5.7 μM, respectively) and negligible or no NF-κB
activation via hTLR 2, 4, 5 and 7, enlisting a SEAP reporter
assay and HEK293 cells overexpressing individually each
of the hTLRs. It also exhibited significant cytotoxic activity,
inhibiting the growth of cervical HeLa cancer cells (IC =
50
2.7 μM) by inducing apoptosis.
Based on the structure-activity relationship (SAR) stu-
dies disclosed by the Yin group [19], both benzyl ether
groups were required for the activity. An electron-
withdrawing 4-nitro group was required, as its removal or
even its relocation to the ortho or meta positions resulted in
loss of activity (Fig. 1). The chromene precursors, a chro-
manone or chromanol, did not display activity, under-
scoring the importance of the olefin. The benzyl ether
linkage was also found to be important, as a decreased
activity was observed when it was replaced with ester.
Substitution, steric bulk, and optimally a spirocyclic ring,
incorporating a heteroatom (the tetrahydropyran group) on
the right side of the molecule was also essential to the
activity. By building on the limited SAR reported by the
Yin group, we hoped to discover compounds with improved
potency and properties (solubility, selectivity, metabolic
stability, toxicity, pharmacokinetics, and in vivo efficacy).
Our initial objective was to confirm the activity and the
mechanism of action of 1, followed by optimization of the
compound potency.
[
11–17].
Humans possess several TLRs that are expressed het-
erogeneously in macrophages, dendritic cells, and other
compartments. TLRs 1, 2, 4, 5 and 6 are located on the cell
membrane and bind bacterial lipids, peptidoglycans, lipo-
proteins and proteins, whereas TLRs 3, 7, 8 and 9 are
situated in the endosome membrane within the cell and are
activated by viral or bacterial nucleic acids. Complementing
the selective targeting of a single TLR, multi-TLR-based
therapies are an emerging area of research in oncology, as
the simultaneous activation of multiple TLRs has potential
merits. It can boost the immune response by triggering
synergy between different TLR pathways [18]. TLRs are
distributed heterogeneously and the effectiveness of a single
TLR activation may be diminished if the target cells do not
express a particular TLR. In contrast, if multiple TLRs are
activated, they can complement each other for presentation
and cross-presentation of antigens to T cells to enhance
immunity. Pertinent to the work detailed herein, maturation
of macrophage dendritic cells is promoted by TLR3 ago-
nists in the tumor microenvironment, thus stimulating pre-
sentation/cross-presentation to
T
cells, resulting in
Results and discussion
expansion of tumor-specific cytotoxic T lymphocytes. On
the other hand, TLR9 agonists target only plasmacytoid
dendritic cells. As noted below, the first small molecule
TLR3 agonist was recently disclosed that also acts on TLR9
and TLR8 [19].
In the recent years, a still limited number of attractive
small molecule TLR agonists have been described [17]. We
reported the discovery of neoseptins [20, 21], the first
structurally characterized class of murine TLR4 agonists
that bear no structural similarity to bacterial lipopoly-
saccharide (LPS) or its active core Lipid A (LPA). More
recently, we disclosed the discovery of the diprovocims
Chemistry and SAR study
To establish insights into the binding of 1 to hTLR8, its
structure was compared with resiquimod (2, R848), a
known TLR8 agonist used as a control in TLR8 and TLR7
assays (Fig. 2a), assuming it binds directly to TLR8 at the
[
22, 23], a class of exceptionally potent, synthetic small
molecule TLR2/TLR1 agonists that show efficacious adju-
vant activity in an anticancer intramuscular vaccination
against murine melanoma and protect mice from tumor
rechallenge in an antigen specific manner [24]. In our
continuing efforts, we were intrigued with the disclosure of
compound 1, derived from a HTS screening lead by Yin
Fig. 1 Structure of reported TLR 3/8/9 agonist and summary of prior
SAR (Ref. [19])

Medicinal Chemistry Research
Fig. 2 a Structure of resiquimod (2, R848). b 2D Binding overlay of 1 and resiquimod (pose 1). c 2D Binding overlay of 1 and resiquimod (pose
2
). d Partial structure of 5’-xCx DNA, H-bonding to D534 in TLR9 agonist binding site
Fig. 3 Structures of compounds 9−22, 26−54
canonical small molecule agonist site. Resiquimod forms a
salt bridge with Asp543 present in the agonist binding site
4) [26], where the cytosine group forms a salt bridge with
Asp534 in the binding site (Fig. 2d). We also inferred that
the nitrophenyl groups in 1 are likely involved in pi-pi
stacking interactions with the active site residues in TLR8.
No small molecule bound structure is known for TLR3,
making it difficult to assign putative binding site(s) [27].
Upon digesting the results of the Yin SAR studies, we
elected to independently conduct (1) modifications of the
benzyl groups with introduction of a variety of substituents
that include nitro group bioisosteres and heterocyclic
replacements of the phenyl ring to manipulate its electronic
(
Supporting Information Fig. 1). Docking studies and a
simple overlay of 1 with R848 suggested that (1) the oxy-
gen in the tetrahydropyran ring in 1 H-bonds to T574
(
TLR8)/G565 (TLR9) backbone NHs, and (2) the two nitro
groups are involved in polar interactions with multiple
arginines and lysines (Supporting Information Fig. 2a, b).
Similar conclusions were drawn from a comparison of the
putative binding of 1 to hTLR9 with that of 5’-xCx DNA,
known TLR9 agonists (Supporting Information Figs 3 and

Medicinal Chemistry Research
Fig. 4 Structures of 23−25, disubstituted benzyl derivatives containing a 4-nitro group, and 4-nitrobenzoyl derivative 32
Scheme 1 Synthesis of compounds 1 and 9−67
Fig. 5 Structures of compounds 55−64
and lipophilic character, diminish its symmetrical nature,
and introduce more sp3 character to improve solubility
while enhancing activity. Independently we pursued (2)
modification of the tetrahydropyran (THP) ring, and (3)
examination of alternative substituted chromene scaffolds
with potential introduction of more basic counterparts to
enhance receptor binding and endosomal localization.
The synthesis of compounds 9−67 largely followed that
detailed by Yin and is outlined in Scheme 1. Friedel-Crafts
acylation of 3,4-dimethoxyphenol with acetic anhydride in
alkylated with the corresponding benzylic bromides
(K CO , acetone, reflux, 12 h) to furnish compounds 1 and
2
3
9−67.
We replaced the 4-nitro group with bioisosteres,
including a nitrile and oxadiazole (compounds 12 and 13,
respectively). In addition, a series of heterocyclic, haloge-
nated and other substituted derivatives (compounds 14−22,
31−34, 39−48) were prepared in efforts to replace the 4-
nitro group on the benzyl substituent (Fig. 3). Several dis-
ubstituted benzyl derivatives containing the 4-nitro group
were also synthesized (Fig. 4, 23−25), and the p-nitro-
phenyl group was replaced with a variety of additional
bioisosteres (compounds 26−28). In order to promote
possible pi-pi stacking, other electron-deficient aryls and
indoles were also examined (compounds 9–11, 17, 18, 38,
39, 54) [28]. In addition, the 4-nitrobenzyl ether was also
replaced with a 4-nitrobenzoyl group (32). However, all
compounds described above displayed little or no activation
of TLR 8 and 9 compared to lead compound 1. As expected,
substituents at the meta-position did not result in an
improvement in activity (e.g., compound 29). Introduction
of alkyl substituents (49−52) also did not enhance
activation.
.
presence of BF Et O (60 °C, 3 h, 90%) yielded the acet-
3
2
ophenone 3, which was condensed with tetrahydro-4H-
pyran-4-one (methanol, 23 °C, 12 h, 83%) to obtain chro-
manone 4. Demethylation of 4 with boron tribromide in
anhydrous CH Cl (0 °C to 23 °C, 12 h, 85%) furnished
2
2
catechol 5, and subsequent reduction with lithium borohy-
dride (LiBH , 2 M in THF, 23 °C, 12 h, 78%) afforded
4
chromanol 6. The latter reduction with LiBH4 proved
cleaner, easier to implement, and provided 6 in higher yield
than that conducted with LiAlH as reported. Dehydration
4
of 6 with 4 N HCl (ethyl acetate, 23 °C, 12 h, 74%) yielded
chromene 7 and was much cleaner than the reported
dehydration conducted in acetone. Compound 7 was

Medicinal Chemistry Research
Fig. 6 Preparation and structure of unsymmetrical derivatives
Scheme 2 Preparation of candidate agonists with a piperidine subunit
5
-Nitrofuranyl (55, cLogP = 4.0) and 5-nitrothienyl (56,
possible only the one nitro group of the 4-nitrophenyl group
that potentially interacts with a tyrosine residue (Y348) in
the resiquimod binding site is needed (Supplementary Fig.
2c). Therefore, analogs of 1 that possess a single 4-nitro
group on each of the benzyl substituents were prepared and
examined (Fig. 6, compounds 65 and 66). The synthesis of
these unsymmetrical derivatives is shown in Fig. 6. In
addition, we replaced the 4-nitro group of one of the benzyl
groups with a fused oxadiazole as a potential bioisostere for
the nitro group while maintaining the one putative key 4-
nitrophenyl group (compound 67). However, compounds
65−67 were inactive, indicating that each 4-nitro group
plays an important role in lead compound 1.
cLogP = 5.2) replacements of the 4-nitrophenyl ring with
cLogP values (cLogP values were calculated with the
module within ChemDraw Professional 16.0) more favor-
able or close to that of the lead molecule (1, clogP = 5.5)
were synthesized (Fig. 5). Additionally, a series of sulfonyl
derivatives 59−64 (cLogP ca. 3) were prepared to address
the high lipophilicity of the lead molecule (cLogP of
compound 1 is 5.5) while simultaneously maintaining
potential binding interactions of the nitro group. In order to
3
introduce sp character, and potentially improve solubility,
derivatives were prepared that contain each enantiomer of
the pyrrolidin-2-one substituent (57 and 58). Unfortunately,
none of these compounds resulted in activation of TLR8
and 9.
The necessity of each individual 4-nitro substituent on
the two benzyl groups was also defined. Based on docking
studies and overlay models of 1 with resiquimod, it is
A final set of modifications of the lead candidate 1
targeted the spirocyclic tetrahydropyran (THP) ring. In a
subset of these studies, the tetrahydropyran was replaced
with two isomeric piperidine rings (Scheme 2 and Fig. 7,
68−71) with the goal to (1) improve solubility, (2)

Medicinal Chemistry Research
8
8
8 (Scheme 3). None of the three compounds (86, 87 and
8), including 88 possessing a single added methyl group
relative to 1, exhibited any TLR activation.
Biological evaluation
The Yin group reported that compound 1 showed strong
NF-κB activation in a reporter assay through activation of
hTLR 3, 8, and 9 (EC = 4.8, 13.4 and 5.7 μM, respec-
5
0
tively) [19]. In contrast, compound 1 prepared herein at two
different sites and by two different procedures and assessed
multiple times only showed weak activity for NF-ĸB acti-
vation in our hands, with EC s of 22, 27 and 58 μM for
5
0
TLRs 3, 8 and 9, respectively. Because of expectations of
improvements in the modest activity reported for 1, the
analogues were examined at concentrations up to 5 μM.
However, none of the analogs of 1 prepared showed an
improvement in activity, with essentially all compounds
being inactive and with all EC50 values being greater than
5
μM (Supporting Information Tables 1 and 2). Selected
members of the full set (13, 18, 24, 55, 56 and 69) were
rescreened at concentrations up to 100 μM and, with the
exception of 69, were inactive (EC > 100 μM) and dis-
Fig. 7 Structures of compounds 68−82
5
0
potentially provide endosomal delivery, localization, or
trap, (3) improve receptor interactions directly or indir-
ectly through attached side chains, and (4) provide a site
for conjugation to neo-antigens and proteins or antibodies
for targeted delivery. The introduction of a com-
plementary basic atom could also form a salt bridge with
D543 in TLR8 and D534 in TLR9. A 2-hydroxyethyl
substituent was appended to the nitrogen to attenuate the
amine basicity and potentially mimic the five-atom chain
in R848, binding in a narrow pocket in TLR8 (compounds
played no NF-ĸB activation. Compound 69 (with a piper-
idine ring replacing the tetrahydropyran ring) exhibited a
weak EC50 of 111 μM in TLR9, but no activity (>125 μM)
in TLR8. Consistent with reported results with 1 [19],
neither compound 1 nor any of its analogs showed any
activity with TLR7.
We also measured the stimulated release of the cytokine
tumor necrosis factor α (TNF-α) by the action of 1 in both
mouse macrophages and in partially differentiated human
THP-1 cells as described in our prior studies [20–24]. It
showed no activity in mouse macrophages and very weak
activity in human THP-1 cells, where the weak stimulated
TNF-α release continues to rise slowly up to 100 µM, the
highest concentration examined (EC50 > 100 μM). The
extent of this stimulated release was not at a level that we
would consider active in this sensitive assay using native
cells. This contrasts the reported strong production of
TNF-α in THP-1 cells by the Yin group (EC50 5−10 μM)
[19]. Thus, although the differences in the potency of 1 in
the reporter assays for hTLR 3, 8, and 9 (2- to 10-fold)
might be attributable to subtle differences in the assay
implementation, those observed with human THP-1 cells
cannot.
7
2 and 73).
The six-membered 4-THP spirocyclic ring was replaced
with spirocyclic rings of altered sizes and modified het-
eroatoms (Fig. 7), including the 4-thiopyran dioxide (74), 4-
thiopyran (75) and its methyl sulfonium salt (76), qua-
ternary dimethyl substitution (77, original screening lead
[
19]), regioisomers of 4-THP (78), and ring contracted
versions of the 4-THP (79 and 80, tetrahydrofurans 81 and
2). However, none of the modifications improved TLR
8
activation. Compounds 74−82 or their precursors were
prepared by the route outlined in Scheme 1, introducing the
modifications in the condensation of 3 with the corre-
sponding ketone.
With the aim of restricting the spatial orientation of the
nitrophenyl groups and in efforts to explore a site for
additional receptor interactions of the chromene, compound
Conclusions
8
8 was prepared and contains a C8-methyl substituent
attached to the core benzene ring. The known compound 84
was converted to chromanone 86, which was reduced to
chromanol 87, followed by dehydration to afford chromene
A comprehensive SAR study of a putative TLR 3/8/9
agonist was conducted. Despite our excitement with the
potential of the first small molecule TLR3 agonist with a

Medicinal Chemistry Research
Scheme 3 Synthesis of compound 88
compound that additionally displayed agonist activity at
TLR8 and TLR9, compound 1 displayed disappointing
activity in our hands, failing to match the potency (EC )
Candidate agonist synthesis
1-(2-Hydroxy-4,5-dimethoxyphenyl)ethan-1-one (3) A solu-
5
0
reported and displaying only a low efficacy for the extent of
stimulated NF-κB release. The evaluation of >75 analogs of
tion of 3,4-dimethoxyphenol (2 g, 12.97 mmol) in acetic
.
anhydride (5 mL) under N was treated with neat BF Et O
2
3
2
1
, many of which constitute minor modifications in its
(0.16 mL, 1.3 mmol) at 0 °C for 10 min, and the reaction
mixture was stirred at 60 °C for 3 h. After completion of the
reaction, the reaction mixture was cooled to room tem-
perature and neutralized with the addition of saturated
aqueous sodium bicarbonate solution, followed by extrac-
tion with ethyl acetate. The crude product was purified by
structure, failed to identify any that displayed significant
activity and none that exceeded the modest activity we
found for 1 or that was initially reported. Although the
authors of the original work [19] have not responded to our
requests for guidance on their report of more promising
observations with 1, it is possible that it could be related to
the low solubility of the lead candidate compound and
potential that compound aggregates may have been assessed
in the original work.
flash column chromatography (SiO , EtOAc/hexane 2:3) to
2
o
1
give 3 (2.3 g, 90%) as a white solid: mp 103–104 C;
H
NMR (600 MHz, CDCl ) δ 12.66 (s, 1H), 7.06 (s, 1H), 6.46
3
1
3
(s, 1H), 3.92 (s, 3H), 3.88 (s, 3H), 2.57 (s, 3H). C NMR
151 MHz, CDCl ) δ 202.1, 160.1, 156.8, 111.7, 111.6,
(
3
1
00.5, 56.7, 6.2, 26.4. MS (ESI) calcd for C H O [M +
15 19 5
+
Methods and materials
H] , 197.1; found, 197.1.
Experimental
6,7-Dimethoxy-2’,3’,5’,6’-tetrahydrospiro[chromane-2,4’-
pyran]-4-one (4) A solution of 3 (2.0 g, 10.2 mmol) in
methanol was treated sequentially with pyrrolidine
(0.84 mL, 10.2 mmol) and tetrahydro-4H-pyran-4-one
(3.06 g, 30.6 mmol), and the reaction mixture was warmed
at reflux for 12 h. Purification by flash column chromato-
All reagents and solvents were used as supplied without
further purification unless otherwise noted. All reactions
were performed in anhydrous solvents unless otherwise
noted. CHCl was pre-treated with alumina for at least 24 h
3
prior to use. Preparative TLC (PTLC) and column chro-
matography were conducted using Millipore SiO 60 F
graphy (SiO , EtOAc/hexane 1:4) gave 4 (2.35 g, 83%) as a
2
1
yellow solid: mp 124−125 °C: H NMR (500 MHz, CDCl )
2
254
3
PTLC (0.5 mm) and Zeochem ZEOprep 60 ECO SiO2
40–63 μm), respectively. Analytical TLC was conducted
δ 7.29 (s, 1H), 6.49 (s, 1H), 3.96 (s, 3H), 3.89 (s, 3H),
3.89–3.77 (m, 4H), 2.71 (s, 2H), 2.05–1.99 (m, 2H), 1.79
(ddd, J = 14.2, 11.0, 5.0 Hz, 2H). MS (ESI) calcd for
(
1
using Millipore SiO 60 F254 TLC (0.250 mm) plates. H
2
1
3
+
and C NMR spectra were obtained using a Bruker
Avance III HD 600 MHz spectrometer equipped with
either a 5 mm QCI or 5 mm CPDCH probe or a Bruker
Avance III 500 MHz spectrometer equipped with a 5 mm
BBFO probe. IR spectra were obtained using a Thermo
Nicolet 380 FT-IR with a SmartOrbit Diamond ATR
accessory. Mass spectrometry analysis was performed by
direct sample injection on an Agilent G1969A ESI-TOF
mass spectrometer.
C H O [M + H] , 279.1; found, 279.1.
1
5
19
5
6,7-Dihydroxy-2’,3’,5’,6’-tetrahydrospiro[chromane-2,4’-
pyran]-4-one (5) A solution of 4 (0.5 g, 1.8 mmol) in
CH Cl (5 mL) was slowly treated with boron tribromide
2
2
(1 M in CH Cl , 5.4 mL, 5.4 mmol) at 0 °C under nitrogen,
2
2
and the reaction mixture was stirred at 23 °C for 12 h. The
reaction mixture was poured into ice water and extracted
with EtOAc, (3x). The combined organic layers were dried

Medicinal Chemistry Research
with Na SO , filtered, concentrated, and the residue was
The characterization for 1, 9–64 is provided in the
Supporting Information section. Experimental procedures
for adaptations of the general procedure and characteriza-
tion data for the synthesis 65–67 (Fig. 6), 68–73 (Scheme 2),
74–82 (Fig. 7), and 86–88 (Scheme 3) are also provided in
Supporting Information.
2
4
purified by flash column chromatography (SiO , EtOAc/
2
hexane 4:1) to give 5 (383 mg, 85%) as an off-white solid:
1
H NMR (500 MHz, DMSO-d ) δ 10.22 (s, 1H), 9.05 (s,
6
1
2
9
2
H), 7.03 (s, 1H), 6.36 (s, 1H), 3.68–3.61 (m, 4H), 2.65 (s,
H), 1.81 (dt, J = 13.8, 3.5 Hz, 2H), 1.68 (ddd, J = 14.4,
+
.3, 6.1 Hz, 2H). MS (ESI) calcd for C H O [M + H] ,
13
15
5
51.1; found, 251.1.
In vitro human TLR cell reporter assays
2
’,3’,5’,6’-Tetrahydrospiro[chromane-2,4’-pyran]-4,6,7-triol
Human HEK-blue reporter cells (Invivogen, San Diego, CA)
containing a TLR transgene are plated on day 0 at an optimal
cell concentration per well of a 384-well plate (20 μL cells/
(
6) A mixture of 5 (500 mg, 2 mmol) and 2 M LiBH in
4
THF (2 mL, 4 mmol) in THF (10 mL) was stirred at 23 °C
for 12 h. The solution was quenched with the addition of
methanol, and the reaction mixture was adjusted to pH 1−2
with the addition of aqueous 2 N HCl prior to removal of
solvents under reduced pressure. The residue was extracted
with EtOAc (10 mL) and washed sequentially with water
and saturated aqueous NaCl. The organic phase was dried
with Na SO , filtered, concentrated, and the residue was
well) and incubated for 16 h at 37 °C, 5% CO . Compounds
2
are added (100 nL/well) to each well using an ATS liquid
handling instrument (Wagner Life Science, Middleton, MA)
and the mixtures are incubated for 24 h at 37 °C, 5% CO2.
Freshly prepared Q-Blue reagent (Invivogen, San Diego, CA;
suspended in water) is added to each well (10 μL/well) and
the mixtures are incubated at 37 °C, 5% CO for 30 min
2
4
2
purified by flash column chromatography (SiO , EtOAc) to
afford 6 (393 mg, 78%) as a yellow solid: H NMR
before measuring the absorbance (620 nm) using an Envision
plate reader (PerkinElmer, Hopkinton, MA). Assay baselines
are set using relative light unit (RLU) values from DMSO-
treated control cells and 100% is set using reference agonist
RLU values from cells treated at the highest compound
concentration [reference agonists: poly I:C HMW (TLR3),
resiquimod (TLR7, TLR8) and ODN-2006 (TLR9)]. Effec-
tive concentration values (EC ; compound concentration
2
1
(
500 MHz, DMSO-d ) δ 10.12 (s, 1H), 9.05 (s, 1H), 7.03 (s,
6
1
3
2
H), 6.46 (br s, 1H), 6.36 (s, 1H), 4.78–4.75 (m, 1H),
.68–3.61 (m, 4H), 2.65 (s, 2H), 1.81 (dt, J = 13.8, 3.5 Hz,
H), 1.68 (ddd, J = 14.4, 9.3, 6.1 Hz, 2H). MS (ESI) calcd
+
for C H O [M + H] , 253.1; found, 253.1.
1
3
17
5
5
0
2
(
’,3’,5’,6’-Tetrahydrospiro[chromene-2,4’-pyran]-6,7-diol
7) Compound 6 (200 mg, 0.79 mmol) was dissolved
in ethyl acetate (1 mL), and aqueous 4 N HCl
0.4 mL,1.58 mmol) was added. The reaction mixture was
yielding half-maximal response) are calculated using BMS
proprietary data analyses software.
Acknowledgements We gratefully acknowledge the financial support
of NIH (CA042056, DLB) and Bristol Myers Squibb.
(
stirred at 23 °C for 12 h, and then basified with the addition
of saturated aqueous NaHCO solution (3 mL) followed by
extraction with EtOAc (10 mL). The combined organic
3
Compliance with ethical standards
layers were dried with Na SO , filtered, concentrated, and
2
4
Conflict of interest The authors declare no competing interests.
purified by flash column chromatography (SiO , EtOAc/
2
hexane 1:3) to afford 7 (137 mg, 74%) as an off white solid:
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
1
H NMR (400 MHz, CDCl ) δ 10.15 (s, 1H), 9.15 (s, 1H),
3
6
.56 (s, 1H), 6.48 (s, 1H), 6.29 (d, J = 9.7 Hz, 1H), 5.51 (d,
J = 9.7 Hz, 1H), 3.92 (td, J = 11.2, 2.6 Hz, 2H), 3.77 (dt,
J = 11.5, 4.0 Hz, 2H), 1.96 (d, J = 14.0 Hz, 2H), 1.77 (td,
J = 10.2, 5.4 Hz, 2H). MS (ESI) calcd for C H O [M +
References
13
15
4
1
2
3
. Ferrandon D, Imler JL, Hoffmann JA. Sensing infection in Dro-
sophila: Toll and beyond. Semin Immunol. 2004;16:43–53.
. Janeway CA Jr, Medzhitov R. Innate immune recognition. Annu
Rev Immunol. 2002;20:197–216.
. Akira S, Uematsu S, Takuechi O. Pathogen recognition and innate
immunity. Cell. 2006;124:783–801.
+
H] , 235.1; found, 235.1.
Synthesis of compounds 9–64, general procedure
A mixture of 7 (10 mg, 0.043 mmol) and K CO (35.4 mg,
2
3
4. Moresco EM, LaVine D, Beutler B. Toll-like receptors. Curr Biol.
0
(
.26 mmol) in acetone (1 mL) was treated with the bromide
0.093 mmol) and stirred at 60 °C under nitrogen for 12 h.
The reaction mixture was filtered, concentrated, and purified
2011;21:R488–R493.
. Kang JY, Lee JO. Structural biology of the Toll-like receptor
family. Annu Rev Biochem. 2011;809:917–941.
. Beutler B. Neo-ligands for innate immune receptors and the
etiology of sterile inflammatory disease. Immunol Rev.
2007;220:113–128.
5
6
by flash column chromatography (SiO , EtOAc/hexane) to
afford compounds 1 and 9–64.
2

Medicinal Chemistry Research
7
8
9
. Kanzler H, Barrat FJ, Hessel EM, Coffman RL. Therapeutic tar-
geting of innate immunity with Toll-like receptor agonists and
antagonists. Nat Med. 2007;13:552–559.
. Hennessy EJ, Parker AE, O’Neill LA. Targeting Toll-like receptors:
emerging therapeutics? Nat Rev Drug Discovery. 2010;
Moresco EM, Berger M, Zhan X, Zhang H, Boger DL,
Beutler B. TLR4/MD-2 activation by a synthetic agonist with no
similarity to LPS. Proc Natl Acad Sci USA. 2016;113:
E884–E893.
21. Morin MD, Wang Y, Jones BT, Su L, Surakattula MM, Berger M,
Huang H, Beutler EK, Zhang H, Beutler B, Boger DL. Discovery
and structure-activity relationships of the neoseptins: a new class
of Toll-like receptor-4 (TLR4) agonists. J Med Chem. 2016;59:
4812–4830.
22. Morin MD, Wang Y, Jones BT, Mifune Y, Su L, Shi H, Moresco
EMY, Zhang H, Beutler B, Boger DL. Diprovocims: a new and
exceptionally potent class of toll-like receptor agonists. J Am
Chem Soc. 2018;140:14440–14454.
9
:293–307.
. Connolly DJ, O’Neill LA. New developments in Toll-like receptor
targeted therapeutics. Curr Opin Pharmacol. 2012;12:510–518.
1
1
0. Rakoff-Nahoum S, Medzhitov R. Toll-like receptors and cancer.
Nat Rev Cancer. 2009;9:57–63.
1. Hoebe K, Jiang Z, Georgel P, Tabeta K, Janssen E, Du X,
Beutler B. TLR signaling pathways: opportunities for activation
and blockade in pursuit of therapy. Curr Pharmaceut Des.
2
006;12:4123–4134.
23. Su L, Wang Y, Wang J, Mifune Y, Morin MD, Jones BT, Mor-
esco EMY, Boger DL, Beutler B, Zhang H. Structural basis of
TLR2/TLR1 activation by the synthetic agonist diprovocim.
J Med Chem. 2019;62:2938–2949.
24. Wang Y, Su L, Morin MD, Jones BT, Mifune Y, Shi H,
Wang K, Zhan X, Liu A, Wang J, Li X, Tang M, Ludwig S,
Hildebrand S, Zhou K, Siegwart D, Moresco EMY, Zhang H,
Boger DL, Beutler B. Adjuvant effect of the novel TLR1/2
agonist Diprovocim synergizes with anti-PD-L1 to eliminate
melanoma in mice. Proc Natl Acad Sci USA. 2018;115:
E8698–E8706.
25. Alexopoulou L, Holt AC, Medzhitov R, Flavell RA. Recognition
of double-stranded RNA and activation of NF-kappaB by Toll-
like receptor 3. Nature. 2001;413:732–738.
26. Ohto U, Ishida H, Shibata T, Sato R, Miyake K, Shimizu T. Toll-
like receptor 9 contains two DNA binding sites that function
cooperatively to promote receptor dimerization and activation.
Immunity. 2018;48:649–658.e4.
1
1
1
1
2. Czarniecki M. Small molecule modulators of Toll-like receptors.
J Med Chem. 2008;51:6621–6626.
3. Meyer T, Stockfleth E. Clinical investigations of Toll-like receptor
agonists. Expert Opin Invest Drugs. 2008;17:1051–1065.
4. Wang X, Smith C, Yin H. Targeting Toll-like receptors with small
molecule agents. Chem Soc Rev. 2013;42:4859–4866.
5. Peri F, Calabrese V. Toll-like receptor 4 (TLR4) modulation by
synthetic and natural compounds: An update. J Med Chem.
2
014;57:3612–3622.
1
6. Lu BL, Williams GM, Brimble MA. TLR2 Agonists and their
structure–activity relationships. Org Biomol Chem. 2020;18:5073–5094.
7. Federico S, Pozzetti L, Papa A, Carullo G, Gemma S, Butini S,
Campiani G, Relitti N. Modulation of the innate immune response
by targeting Toll-like receptors: a perspective on their agonists
and antagonists. J Med Chem. 2020;63:13466–13513.
8. Hajishangallis G, Lambris JD. More than complementing Tolls:
complement–Toll-like receptor synergy and crosstalk in innate
immunity and inflammation. Immunol Rev. 2016;274:233–244.
9. Zhang L, Dewan V, Yin H. Discovery of small molecules as
multi-Toll-like receptor agonists with proinflammatory and
anticancer activities. J Med Chem. 2017;60:5029–5044.
1
1
1
2
27. Choe J, Kelker MS, Wilson IA. Crystal structure of human toll-
like receptor
3 (TLR3) ectodomain. Science. 2005;309:
581–585.
28. Bissantz C, Kuhn B, Stahl M.
guide to molecular interactions.
53:5061–5084.
A
J
medicinal chemist’s
Med Chem. 2010;
0. Wang Y, Su L, Morin MD, Jones BT, Whitby LR,
Surakattula MM, Huang H, Shi H, Choi JH, Wang KW,