Selective Toll-like receptor 7 agonists with novel chromeno[3,4-d]imidazol-4(1H)-one and 2-(trifluoromethyl)quinoline/ quinazoline-4-amine scaffolds

Article abstract of DOI:10.1016/j.ejmech.2019.06.030

Toll-like receptors (TLRs) are promising targets for treatment of viral infections, autoimmune diseases, and cancers. Here, two new series of selective small-molecule TLR7 agonists with novel scaffolds and good selectivity over TLR8 are described, some with potencies in the low micromolar range. 8-Hydroxy-1-isobutylchromeno[3,4-d]imidazol-4(1H)-one (26) from the first series was designed and synthesized on the basis of previously described TLR7 antagonist 2, and is shown to be a selective TLR7 agonist (EC₅₀, 1.8 μM). The second series was based on 2-(trifluoromethyl)quinolin-4-amine and 2-(trifluoromethyl)quinazolin-4-amine scaffolds, which were defined according to our in-house ligand-based virtual screening protocol. Further synthesis of a focused library of analogs, biological evaluation, and docking studies provided systematic exploration of the structure?activity relationships, which indicate that a secondary or tertiary amine with smaller flexible alkyl substituents up to three carbon atoms in length, or bulkier rigid aliphatic rings is required at position 4 on 2-(trifluoromethyl)quinoline/quinazoline scaffold for potent TLR7 agonist activity. The influence of selected TLR7 agonists on cytokine production is also reported showing that N-cyclopropyl-2-(trifluoromethyl)quinazolin-4-amine (46) is able to induce increased levels of IL-6 and IL-8. These data demonstrate successful in-silico definition of novel TLR7 versus TLR8-selective compounds as promising chemical probes for further development of potent small-molecule immunomodulators.

Full text of DOI:10.1016/j.ejmech.2019.06.030

Accepted Manuscript
Selective Toll-like receptor 7 agonists with novel chromeno[3,4-d]imidazol-4(1H)-one
and 2-(trifluoromethyl)quinoline/ quinazoline-4-amine scaffolds
Ana Dolšak, Urban Švajger, Samo Lešnik, Janez Konc, Stanislav Gobec, Matej Sova
PII:
S0223-5234(19)30558-6
DOI:
https://doi.org/10.1016/j.ejmech.2019.06.030
EJMECH 11433
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 22 February 2019
Revised Date: 10 June 2019
Accepted Date: 11 June 2019
Please cite this article as: A. Dolšak, U. Švajger, S. Lešnik, J. Konc, S. Gobec, M. Sova, Selective Toll-
like receptor 7 agonists with novel chromeno[3,4-d]imidazol-4(1H)-one and 2-(trifluoromethyl)quinoline/
quinazoline-4-amine scaffolds, European Journal of Medicinal Chemistry (2019), doi: https://
doi.org/10.1016/j.ejmech.2019.06.030.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Selective Toll-like receptor 7 agonists with novel chromeno[3,4-d]imidazol-4(1H)-one and 2-
(trifluoromethyl)quinoline/ quinazoline-4-amine scaffolds
Ana Dolšaka, Urban Švajgera,b, Samo Lešnikc, Janez Koncc, Stanislav Gobeca, Matej Sovaa*
aUniversity of Ljubljana, Faculty of Pharmacy, Aškerčeva 7, SI-1000 Ljubljana, Slovenia
bBlood Transfusion Centre of Slovenia, Šlajmerjeva 6, SI-1000 Ljubljana, Slovenia
cNational Institute of Chemistry, Hajdrihova 19, SI-1001 Ljubljana, Slovenia
*Correspondence to Matej Sova
University of Ljubljana
Faculty of Pharmacy
Aškerčeva 7
SI-1000 Ljubljana, Slovenia
Tel/Fax: +386-1-4769577, +386-1-4258031
E-mail: matej.sova@ffa.uni-lj.si;
Abstract
Toll-like receptors (TLRs) are promising targets for treatment of viral infections, autoimmune
diseases, and cancers. Here, two new series of selective small-molecule TLR7 agonists with novel
scaffolds and good selectivity over TLR8 are described, some with potencies in the low micromolar
range. 8-Hydroxy-1-isobutylchromeno[3,4-d]imidazol-4(1H)-one (26) from the first series was
designed and synthesized on the basis of previously described TLR7 antagonist 2, and is shown to be a
selective TLR7 agonist (EC50, 1.8 µM). The second series was based on 2-(trifluoromethyl)quinolin-4-
amine and 2-(trifluoromethyl)quinazolin-4-amine scaffolds, which were defined according to our in-
house ligand-based virtual screening protocol. Further synthesis of a focused library of analogs,
biological evaluation, and docking studies provided systematic exploration of the structure−activity
relationships, which indicate that a secondary or tertiary amine with smaller flexible alkyl substituents
up to three carbon atoms in length, or bulkier rigid aliphatic rings is required at position 4 on 2-
(trifluoromethyl)quinoline/ quinazoline scaffold for potent TLR7 agonist activity. The influence of
selected TLR7 agonists on cytokine production is also reported showing that N-cyclopropyl-2-
(trifluoromethyl)quinazolin-4-amine (46) is able to induce increased levels of IL-6 and IL-8. These
data demonstrate successful in-silico definition of novel TLR7 versus TLR8-selective compounds as
promising chemical probes for further development of potent small-molecule immunomodulators.
Keywords: Toll-like receptors, agonist, TLR7 selective, chromeno[3,4-d]imidazole-4(1H)-one, 2-
(trifluoromethyl)quinazoline-4-amine, 2-(trifluoromethyl)quinolin-4-amine
Introduction
Toll-like receptors (TLRs) are pattern-recognition receptors that are involved in host cell recognition
and initiation of immune responses against bacteria, viruses, fungi, and parasites [1]. To date, 13
functional TLRs have been identified in vertebrates [2]. TLRs are considered to be emerging

ACCEPTED MANUSCRIPT
therapeutic targets for several life-threatening diseases [3]. In the past few years, there has been
extensive research for small-molecule agonists and antagonists of the TLRs that might be useful in the
battle against microbial infections, cancers, sepsis, and autoimmune diseases [4]. The TLRs that
recognize mainly microbial membranes or cell-wall components (i.e., types 1, 2, 4, 5, 6) are expressed
on cell surfaces. On the other hand, TLR3, 7, 8 and 9, which recognize microbial nucleic acids, are
expressed exclusively in the intracellular compartments, such as endosomes and lysosomes [5–7].
TLR7 and TLR8 specifically recognize single-stranded RNAs, and have thus been shown to be
promising targets for treatment of viral infections, autoimmune diseases, and cancers [3,5,8,9].
Several small-molecule modulators of TLR7 have been developed in recent years [10,11]. Up to 2016,
no crystal structures for TLR7 had been described, and therefore many of the structures of newly
designed potential modulators originated from known ligands, such as imiquimod (1) [12]. This
immunomodulator has been approved by the US Food and Drug Administration for the treatment of
anogenital warts since 1997, and since 2004 for the treatment of actinic keratosis and basal-cell
carcinoma [13,14]. It was only shown later that imiquimod (1) is an immunostimulatory agent that acts
as an agonist on TLR7 [15]. Development of new small-molecule modulators and optimization of the
known ones then led to the discovery of new ligands, which can be structurally divided into several
groups, including imidazoquinolines [16–20], imidazopyridines [11], 8-oxo/hydroxyadenines [21–25],
pyrimidinediamines [26] and quinazolinediamines [27], guanine nucleoside analogs [28], and
tetrahydropyridopyrimidines [29]. Most of these compounds show low micromolar agonist activities
for TLR7 (with some in the nanomolar range) and do not show selectivity for TLR7 over TLR8.
In the present study, we report on two new series of TLR7-selective agonists with low micromolar
potencies, as shown in Figure 1. Compounds from both of these series arose from in-silico design,
which is widely used in the early stages of drug discovery. With no crystal structures of TLR7
available at the time of screening, molecular docking could not be applied. However, several small-
molecule modulators of TLR7 were known and thus used as a starting point for ligand-based virtual
screening. The first series was based on commercially available 1-isobutylchromeno[3,4-d]imidazol-
4(1H)-one, as compound 2, which we described as a hit compound with TLR7 antagonist activity in
our previous study [30], where compound 1 and its analog were used for three-dimensional ligand-
based virtual screening that led to the identification of novel TLR7 antagonist scaffolds from
previously prepared ZINC drug-like database. Herein we show that minor structural changes can
switch antagonism into agonism, which was already demonstrated previously for TLRs [31]. On the
basis of compound 2, we systematically explored the structure−activity relationships of
chromenoimidazolone-based TLR7 modulators. For discovery of novel scaffolds from the second
series, our in-house ligand-based virtual screening protocol, known as LiSiCA [32] was used, with
imiquimod (1) as a reference compound. LiSiCA was set for the two-dimensional screening option
using the ZINC Drugs Now database [33]. The high-ranked hit compounds that were topologically
most similar to the reference compound 1 (according to Tanimoto coefficients) were commercially
available 2-(trifluoromethyl)quinoline/ quinazoline derivatives 32 and 33 (Figure 1), which were
further used as scaffolds for preparation of novel second series of TLR7 agonists.

ACCEPTED MANUSCRIPT
Figure 1. Illustration of the workflow for development of the TLR7 ligands. Here, ligand-based
virtual-screening protocols and further chemical optimization led to the discovery of the novel
selective TLR7 agonists 26 and 46.
Results and discussion
Chromenoimidazolone series
In our previous study [30], ligand-based virtual-screening defined compound 2 as a new antagonist of
TLR7 with an IC50 of 23 µM. Therefore, we explored the chemical space around the
chromenoimidazolone scaffold and synthesized a focused library of analogs. Different substituents
were added onto the main chromenoimidazolone scaffold on three selected positions, denoted as R1,
R2, and R3 in Figure 1. Compounds 17 to 31 were synthesized via two-step synthetic procedure, as
presented in Scheme 1. These syntheses were carried out with slight modifications according to the
procedures published in the literature [34]. Starting from ethyl isocyanoacetate, imidazoles 4 to 16
were formed in a one-pot reaction between the primary amine and aldehyde 3, in the presence of
pyridine, copper(I) iodide, and anhydrous methanol. Further cyclization of imidazole derivatives 4 to
16 in the next step yielded the final chromenoimidazolone products 17 to 31. The cyclization reaction
was performed according to our previously described procedure [30]; however, it was often not
successful, or yielded the product in particularly low amounts, which made the isolation process
difficult. Therefore, a simple, and in particular more sustainable methodology was introduced:
microwave-assisted synthesis. The classical procedures were thus replaced by microwave-assisted
synthesis using 70% H2SO4, which shorten the reaction times (from 3 h to 20 min) and improve the
yields.

ACCEPTED MANUSCRIPT
R1
R1
N
OH
N
H
N
a
b
R3
R2
R3
R2
N
O
O
O
R2
O
OH
O
R3
3
4 - 16
17 - 31
Scheme 1. Reagents and conditions. (a) Ethyl isocyanoacetate, R1-NH2, pyridine, copper(I) iodide,
anhydrous MeOH, 50 °C, 16 h; (b) 70% H2SO4, 80 °C, 3 h or MW, 20 min.
All of these finally synthesized compounds 17 to 31 were fully characterized by spectroscopic and
other methods, and furthermore, prior to any biological evaluation, their cytotoxicities were measured
using the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium
(MTS) cell-proliferation assay. The highest noncytotoxic concentrations for each compound varied
from 62.5 µM to 250 µM (Supplementary Material, section MTS assay results, Table S1).
All of the compounds were then tested for their antagonist activities at their highest noncytotoxic
concentrations. Interestingly, none of them showed any antagonist activities. Therefore, we decided to
test these compounds for their agonist activities, and the data obtained are presented in Table 1.
Indeed, as for TLR7 ligands reported by Shukla et al. [31], a switch from antagonism to agonism was
observed. In line with our previous study, R2 or R3 substitution of hydrogen with polar substituents
resulted in the shift from antagonist to agonist activity. Six compounds (21, 25, 26, 29, 30, 31) showed
promising agonist activities. The most potent chromenoimidazolone derivatives 26 and 29 had EC50
values of 1.8 µM and 29.9 µM, respectively. In our assay, compound 26 exhibited even higher potency
compared to a positive control imiquimod (1). Some structure–activity trends were evident from these
data for optimal substituents R1, R2, and R3 on this chromeno[3,4-d]imidazol-4(1H)-one scaffold.
First, a lipophilic alkyl R1 fragment of medium length (i.e., up to 4 carbon atoms) without any
heteroatoms was obligatory for TLR7 agonist activity in addition to polar substituents on R2 and R3.
The unsubstituted derivative (R2 = R3 = H) with isobutyl group as R1 (compound 2) exhibited
antagonist activity. The optimal substituent for R1 was shown to be isobutyl, which was included in 5
out of the 6 active compounds. The only exception that still showed TLR7 agonist activity was
compound 21, where the bulky isobutyl group on R1 was replaced by a more flexible n-butyl chain.
Furthermore, either R2 or R3 substituent was also indispensable, which was improved by donors or
acceptors of hydrogen bonds. Neither R2 or R3 was inferior, which is shown by the most potent
compounds 26 and 29, which have a hydroxyl group as R3 substituent and a diethylamino group as R2
substituent, respectively. Another outcome which was observed by our in-silico studies was that R2 or
R3 substituents need to be relatively small in size, as there is little space for these substituents in the
active site. This could be seen in the proposed binding mode of 26 from our docking study using the
crystal structure of monkey TLR7 (PDB code 5gmh) and the Gold v5.5. software, as shown in Figure
2. Accordingly, the low micromolar EC50 of 26 appeared to be a consequence of its tight fit within the
active site (Figure 2a). This facilitates the formation of several potential interactions, which were
predicted by the protein-ligand interaction profiler (PLIP) [35] for the best-ranked pose. Among the
proposed interactions, two hydrogen bonds were formed between the oxygen from the 8-hydroxy (R3)
substituent and the amide nitrogens of Gln354 and Val355. The predictions also indicated π–π
stacking of aromatic rings from chromenoimidazolone and Phe408, and a salt bridge between the
carbonyl group of 26 and the amine from Lys432. Moreover, several hydrophobic interactions with
Tyr264, Phe351, Tyr356, Val381, and Phe408 were also proposed. On the contrary, the binding pose
is significantly changed when hydroxyl group is moved to the position 7 (compound 31), which leads

ACCEPTED MANUSCRIPT
to the loss of some key interactions (i.e. salt bridge, some hydrophobic interactions) and two fold
decrease in potency (Figure 3).
Secondly, to determine selectivity toward TLR7 over TLR8, these compounds with agonist activity on
TLR7 (i.e., 21, 25, 26, 29, 30, 31) were also tested on HEK cells expressing TLR8 and the NF-κB/
secreted embryonic alkaline phosphatase (SEAP) gene-reporter system. None of these compounds
showed agonist activity toward TLR8.
Table 1. TLR7 agonist activities of analogs synthesized based on the chromeno[3,4-d]imidazol-4(1H)-
one scaffold.
Compound
Substituents on scaffold
TLR7 agonist
activity
R1
R2
R3
EC50 [µM]a
Imiquimod
8.2
(1)
-H
-H
-F
n.a.
n.a.
17
18
-H
-H
-H
-H
-H
-H
-H
-H
-H
-H
-H
n.a.
n.a.
107
n.a.
n.a.
19
20
21
22
23
-H
-H
-H
-H
-H
-F
n.a.
535
1.8
24
25
26
27
28
29
-Me
-OH
-OMe
-NO2
-H
n.a.
n.a.
29.9
-N(Et)2

ACCEPTED MANUSCRIPT
-OMe
-OH
-H
-H
279
114
30
31
an.a., no activity
(a)
(b)
Figure 2. Predicted binding mode for 26, the most potent chromenoimidazolone derivative, within the
active site of monkey TLR7 (PDB code 5gmh). (a) Compound 26 (sphere representation) almost
perfectly occupies the entrance of the active site. (b) Compound 26 (violet) forms several interactions
with amino acids (yellow) in the active site of TLR7. Blue, hydrogen bonds between OH and
Gln354/Val355; gray, hydrophobic interactions with Tyr264, Phe351, Tyr356, Val381, and Phe408;
green, π–π stacking with Phe408; and yellow, salt bridge with Lys432.
Figure 3. Predicted binding mode for 31 ((a) in sphere representation, (b) in sticks showing the main
interactions with amino acids in the active site) within the active site of monkey TLR7 (PDB code
5gmh). The binding pose is significantly changed when hydroxyl group is moved from position 8 (R3,
compound 26, Figure 2) to 7 (R2, compound 31).
Quinoline/ quinazoline series

ACCEPTED MANUSCRIPT
With no crystal structures for TLR7 available at the time of screening, our in-house ligand-based
virtual screening protocol LiSiCA [32] was used to identify novel compounds that would bind to the
selected target. LiSiCA is based on a graph theoretical approach to find the maximum common
substructure between a reference and a target molecule obtained from a given database. The maximum
common substructure is identified by finding the maximum clique in a product graph using a fast
MCQD algorithm [36]. A base unit of similarity in the product graph in LiSiCA is a single atom type,
which differentiates it from a lot of other ligand-based virtual screening software, where the unit of
similarity is usually a functional or interaction group [37–39]. We therefore obtain very rigorous graph
representations that captured the atom-level details of these compounds.
Screening with LiSiCA was initially performed with imiquimod (1), the well-known TLR7 agonist
that was selected as the reference compound here, and screened against the ZINC Drugs Now database
[33]. LiSiCA was set for the two-dimensional screening option, with zero as the maximum allowed
shortest path difference parameter. This means that a vertex in a product graph is only seen when the
topological distance between two atoms of specific types in a reference molecule is the same as the
distance between the same types of atoms in the target molecule. By setting this value to zero, strict
criteria for screening were imposed, thus scoring highly only the molecules that were very similar to
the reference structure.
The compounds in the database were ranked according to their Tanimoto coefficients. The high-
ranked hit compounds that were topologically most similar to the reference compound were obtained
commercially and evaluated biologically for their activities against TLR7 in the HEK293 cell line. For
the initial cytotoxicity assays, the highest noncytotoxic concentrations were determined to be used in
the preliminary reporter assays. Most of the compounds were noncytotoxic at 250 µM or 125 µM
(Supplementary Material, section MTS assay results, Table S1). All compounds were then tested for
their antagonist and agonist activities against TLR7. While none of them showed any antagonist
activity, 32 and 33 showed promising agonist activities. Further biological evaluation defined 32 and
33 as TLR7 agonists (EC50, 6.3 µM, 4.0 µM, respectively), which incorporated the 2-
(trifluoromethyl)quinolin-4-amine and 2-(trifluoromethyl)quinazolin-4-amine scaffolds, respectively.
In our assay, the potencies of compounds 32 and 33 (Table 2) were in the same range as the positive
control imiquimod (1).
The next goal was to synthesize a focused series of analogs of 32 and 33 to explore the chemical space
around these 2-(trifluoromethyl)quinolin-4-amine and 2-(trifluoromethyl)quinazolin-4-amine
scaffolds, with a focus on the substitution in position 4 (R4) on the basis of synthetic feasibility.
Compounds 40 to 55 were synthesized via the three-step synthetic procedure given in Scheme 2.
Briefly, 2-(trifluoromethyl)quinazoline-4-ol (36) was synthesized from anthranilamide (35) and ethyl
trifluoroacetate according to the known procedure [40]. Compound 36 was then treated with POCl3 at
high temperature, to obtain 4-chloro-2-(trifluoromethyl)quinazoline (37). The last step was
nucleophilic aromatic substitution of chlorine from compound 37 or using commercially available 4-
chloro-2-(trifluoromethyl)quinoline (38) with the appropriate amine. Finally, 18 compounds (39–55)
with modifications in the amine moiety were synthesized according to the described procedures.

ACCEPTED MANUSCRIPT
Scheme 2. Reagents and conditions. (a) Ethyl trifluoroacetate, NaOEt/EtOH, reflux, 24 h; (b) POCl3,
120 °C, 16 h; (c) amine R4, K2CO3, MeCN, rt, 24 h.
The highest noncytotoxic concentrations for the synthesized compounds varied from 7.81 µM to 250
µM for 52 and 45, respectively. The majority of these compounds were noncytotoxic at 62.5 µM or
31.25 µM (Supplementary Material, section MTS assay results, Table S1), and were thus selected for
preliminary evaluation at these concentrations. As our three hit compounds only showed agonism, all
of these synthesized compounds were only tested for agonist activity. Eight compounds (41, 44–50)
were active at these selected concentrations, and so their EC50 values were determined (Table 2). The
only exception was compound 42, which showed some potency in the reporter assay; however, the
EC50 could not be determined due to cytotoxicity issues.
Table 2. TLR7 agonist activities of analogs synthesized based on the 2-(trifluoromethyl)quinolin/
quinazoline-4-amine scaffolds.
Substituents on scaffold
TLR7 agonist
Compound
activity
X
R4
EC50 [µM]a
Imiquimod
6.2
(1)
-CH-
-N-
6.3
4.0
32
33
-N-
-N-
66.6
n.a.
34
36

ACCEPTED MANUSCRIPT
-N-
n.a.
n.a.
37
39
-CH-
-N-
-N-
n.a.
40
41
44.0
-N-
-N-
n.d.
n.a.
42
43
-N-
-N-
-N-
-N-
57.4
4.4
44
45
46
47
16.8
15.0
-N-
-N-
10.9
8.3
48
49
-N-
-N-
15.4
n.a.
50
51
-N-
-N-
n.a.
n.a.
52
53
-N-
n.a.
n.a.
54
-N-
55
an.a., no activity; n.d., not determined

ACCEPTED MANUSCRIPT
With the exception of hit compound 32, all of these active compounds had a quinazoline ring. The
replacement of a carbon atom in aromatic ring with a nitrogen (quinoline to quinazoline) in the
compound 32 resulted in a slight increase in potency, from 6.3 to 4.4 µM for 32 and 45, respectively.
When R4 substituent was changed to 3-methyl-piperidine-1-yl, only quinazoline derivative 33 retained
the activity. The amine moiety varied in size, from the smaller n-, iso- or cyclopropyl, to the bulkier
cyclohexylamine. All of these active compounds were substituted with secondary or tertiary amines on
position 4, whereas compound 40 with a 4-amino group was inactive. In addition, compounds with
flexible substituents that were longer than three carbon atoms and those with additional heteroatoms in
the alkyl chain were inactive, with the exception of a commercially available analog of hit 33 with a 1-
cyclopropyl-piperidine-4-il-amino substituent (compound 34). For the active compounds, their binding
modes in the active site were predicted using the GOLD v5.5 software. The interactions for the best
ranked poses were detected using the PLIP algorithm. The in-silico results of the most potent
compound 33 are presented in Figure 4, and the most potent compounds showed similar binding to 33,
with one hydrogen bond, π–π and several hydrophobic interactions seen. The predicted binding mode
of 34, which was the only active compound that had longer alkyl chains with additional heteroatoms,
is shown in Figure 5. The potency of 34 was weaker compared to the most potent compounds from
this series, which was explained by the different position of 34 within the active site. It showed several
hydrophobic interactions (here also with amino acids Phe349 and Val381, deeper inside the binding
pocket) and halogen bond between fluorine and Gln354. Although PLIP did not predict any hydrogen
bonds, 34 occupied the space inside the active site much more efficiently than the other active 2-
(trifluoromethyl)quinoline/ quinazoline-4-amines. According to these docking studies, and
concomitant with the biological data, it can be concluded that in general smaller flexible alkyl
substituents on the amine group of up to three carbon atoms in length or bulkier rigid ones are required
for agonist activity.
(a)
(b)
Figure 4. Predicted binding mode for 33, the most potent from the library of compounds with the 2-
(trifluoromethyl)quinazoline-4-amine scaffold, within the active site of monkey TLR7 (PDB code
5gmh). (a) Compound 33 (sphere representation, red) occupies the active site with a tight fit. (b)
Predicted binding mode of 33, showing the main interactions with amino acids (yellow) within the
active site. Blue, hydrogen bond to Asp555; gray, several hydrophobic interactions with Phe351,
Ile585, Leu557, and Tyr356; and green, π–π stacking with Phe408.

ACCEPTED MANUSCRIPT
(a)
(b)
Figure 5. Predicted binding mode for 34 within the active site of monkey TLR7 (PDB code 5gmh).
(a) Compound 34 (sphere representation, pink) efficiently occupies the active site. (b) Interactions are
formed within the active site. Gray, hydrophobic interactions with Phe349, Phe351, Tyr356, Val381,
Phe408, Thr532 and Leu557; and green cyan, halogen bond between fluorine and Gln354.
All of these active compounds (i.e., 32–34, 41, 44–50) were then tested for their agonist activities on
TLR8, with none of them seen to be active. Therefore, these selected active compounds from the 2-
(trifluoromethyl)quinolin-4-amine and especially the 2-(trifluoromethyl)quinazolin-4-amine series are
selective TLR7 versus TLR8 agonists with new chemical scaffolds.
The ability of the TLR7 selective compounds 26, 32, 46, 49, which gave the highest response in TLR7
activity assay, and the prototypical TLR7 selective agonist imiquimod (1) to stimulate cytokine
production from human peripheral blood mononuclear cells (PBMCs) were also examined by
measuring the amount of IL-1β, IL-6, IL-8, IL-10, IL-12p70, and TNFα after the addition of 10 µg/mL
of each compound (Figure 6, Supplementary material). The induction of two cytokine production,
namely IL-8 and IL-6, was detected for 2-(trifluoromethyl)quinazoline derivatives 32, 46, 49, whereas
chromenoimidazolone derivative 26 showed no significant induction. The results obtained indicate
that 46 significantly induces IL-8 production (even though a slightly lower compared to 1), whereas 32
and 49 showed only moderate activity (Figure 6). Furthermore, increased production of IL-6 was also
observed 46, while 26, 32, and 49 exhibited only minor increase compared to the negative control. On
the other hand, neither of tested compounds affected IL-12p70 and TNFα production, which was also
observed for 1 and in accordance with previously reported studies for selective TLR7 agonists
[11,18,20].

ACCEPTED MANUSCRIPT
10000
1000
100
10
IL-1β
IL-6
IL-8
IL-10
IL-12p70
TNFα
1
NC
1
26
32
46
49
Compound
Figure 6. The measurement of cytokine secretion in PBMCs after treatment with our selective TLR7
agonists 26, 32, 46, 49, and imiquimod (1). NC: negative control (untreated cells).
Conclusions
In summary, three novel scaffolds with agonist activities against TLR7 are defined here via the two
different ligand-based virtual screening approaches: the chromeno[3,4-d]imidazol-4(1H)-one, 2-
(trifluoromethyl)quinolin-4-amine, and 2-(trifluoromethyl)quinazoline-4-amine scaffolds. Analogs of
active compound 2 from the first series were synthesized according to optimized synthetic procedures,
incorporating microwave-assisted synthesis, and assayed for their antagonist and agonist activities on
TLR7. Substitution of the hydrogen atoms on positions 7 (R2) or 8 (R3) of the chromenoimidazolone
scaffold with small polar substituents resulted in a shift from antagonist to agonist activity. 8-
Hydroxy-1-isobutylchromeno[3,4-d]imidazol-4(1H)-one (26) was the most potent agonist from this
series, with EC50 for TLR7 of 1.8 µM. Biological evaluation of agonist activity on TLR7 of the
commercially available and synthesized hit compounds from the second 2-(trifluoromethyl)quinolin-4-
amine and 2-(trifluoromethyl)quinazoline-4-amine series provided 11 active analogs. Four of these
(i.e., 32, 33, 45, 49) showed agonist activities in the low micromolar range (EC50, 6.3, 4.0, 4.4, 8.3
µM, respectively), similar to the positive control and a known TLR7 agonist imiquimod (1).
Structure−activity relationships indicate that a secondary or tertiary amine on position 4 (R4) is
required for agonist activity. Furthermore, smaller flexible alkyl substituents on the amine group of up
to three carbon atoms in length (such as isobutyl group), or bulkier rigid aliphatic rings (cyclopentyl,
cyclohexyl or 3-methylpiperidine-1-yl) provide improved agonist activitiesAll of these defined TLR7
agonists were selective for TLR7 versus TLR8. Last but not least, N-cyclopropyl-2-
(trifluoromethyl)quinazolin-4-amine (46) also induces IL-6 and IL-8 production in PBMCs. Therefore,
we propose these TLR7-selective compounds as promising chemical probes and an important novel
chemical class of TLR7 agonists for the further development of potent small-molecule
immunomodulators.
EXPERIMENTAL

ACCEPTED MANUSCRIPT
Chemistry and chemical characterization of compounds
The reagents and solvents were from commercial sources (Sigma-Aldrich, Acros Organics, Alfa
Aesar, TCI). The microwave reactions were performed using a Discover (CEM Corporation)
microwave synthesis system with IR and optical temperature control. Reactions were monitored using
thin-layer chromatography on silica-gel plates (Merck DC Fertigplatten Kieselgel 60 GF254) and
visualized under UV light and/or stained with the relevant reagents. Flash column chromatography
was performed on Merck silica gel 60 (mesh size, 70-230), using the indicated solvents. Yields refer to
the purified products and were not optimized. All of the melting points were determined on a Reichert
1
13
hot-stage apparatus, and are uncorrected. H and C NMR spectra were recorded at 295 K in CDCl3,
DMSO-d6, (CD3)2CO or MeOD on a Bruker Avance III NMR spectrometer equipped with a
Broadband decoupling inverse 1H probe. The coupling constants (J) are in Hz, and the splitting
patterns are designated as: s, singlet; bs, broad singlet; d, doublet; dd, double doublet; td, triple
doublet; t, triplet; dt, double triplet; ddd, double of doublet of doublet; and m, multiplet. Mass spectra
and high-resolution mass measurements were performed on a VG-Analytical Autospec Q mass
spectrometer and an Exactive Plus orbitrap mass spectrometer at the Jozef Stefan Institute (Ljubljana,
Slovenia) and at Faculty of Pharmacy, University of Ljubljana, respectively. The analytical purities of
1
the assayed compounds were ≥95%, unless otherwise stated, as determined by H NMR and HPLC
analyses. HPLC analyses were run at 25 °C on an Agilent 1100 system and a ThermoScientific Dionex
UltiMate 3000 system equipped with a quaternary pump and a multiple wavelength detector. An
Agilent Eclipse Plus C18 column was used (4.6 mm × 150 mm, 5 µm), with a flow rate of 1.0
mL/min, detection at 254 nm, and an eluent system of: A, 0.1% aqueous TFA; B, CH3CN. The
following gradients were applied: Method A: 0–15 min, 40%→90% B; 15–17.5 min, 90% B; 17.5–18
min, 90%→40% B; 18–22 min, 40% B. Method B: 0–15 min, 20%→80% B; 15–17.5 min, 80% B;
17.5–18 min, 80%→20% B; 18–22 min, 20% B. Method C: 0–15 min, 10%→90% B; 15–20 min,
90% B. Method D: 0–12 min, 10%→90% B; 12–14 min, 90% B, 14–15 min, 90%→10% B.
General procedure for synthesis of imidazoles (4–16)
Imidazoles (4–16) were synthesized according to a previously described procedure [34]. To a solution
of ethyl isocyanoacetate (1.0 equiv.) in anhydrous MeOH, the appropriate amine (1.0 equiv.),
substituted 3 (1.0 equiv.), pyridine (1.0 equiv.), and copper(I) iodide (1.0 equiv.) were added
sequentially. The reaction mixture was stirred for 16 h at 50 °C. The solvents were evaporated under
reduced pressure. The residue was extracted with EtOAc, H2O, and brine, and purified using flash
column chromatography (DCM/MeOH, 20:1, v/v).
Methyl 1-ethyl-5-(5-fluoro-2-hydroxyphenyl)-1H-imidazole-4-carboxylate (4): Yield: 17%; white
solid; Mp 177–183 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 1.16 (t, J = 7.4 Hz, 3H), 3.43 (q, J = 7.4
Hz, 2H), 3.58 (s, 3H), 6.93 (dd, J1 = 4.8 Hz, J2 = 8.8 Hz, 1H), 7.04 (dd, J1 = 3.2 Hz, J2 = 9.2 Hz, 1H),
7.14 (dt, J1 = 3.2 Hz, J2 = 8.8 Hz, 1H), 7.89 (s, 1H), 9.80 (s, 1H); ESI-HRMS ([M+H]+, m/z): Calcd for
C13H14FN2O3 265.0981, found 265.0988; Rf = 0.42 (DCM/MeOH, 9:1, v/v).
Methyl 5-(2-hydroxyphenyl)-1-isopropyl-1H-imidazole-4-carboxylate (5): Yield: 20%; white solid;
1
Mp 147–149 °C; H NMR (400 MHz, CDCl3): δ (ppm) 1.39 (d, J = 6.8 Hz, 3H), 1.48 (d, J = 6.4 Hz,
3H), 3.83 (s, 3H), 4.20–4.32 (m, 1H), 7.05 (d, J = 7.6 Hz, 1H), 7.14 (d, J = 7.6 Hz, 1H), 7.20 (d, J =
7.6 Hz, 1H), 7.41 (t, J = 7.6 Hz, 1H), 7.89 (s, 1H), 1H from OH is exchanged; ESI-HRMS ([M+H]+,
m/z): Calcd for C14H17N2O3 261.1233, found 261.1239; Rf = 0.43 (DCM/MeOH, 9:1, v/v).
Methyl 5-(2-hydroxyphenyl)-1-propyl-1H-imidazole-4-carboxylate (6): Yield: 8%; white solid; Mp
148–150 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 0.80 (t, J = 7.2 Hz, 3H), 1.61 (sex, J = 7.2 Hz, 2H),

ACCEPTED MANUSCRIPT
3.79–3.85 (m, 2H), 3.86 (s, 3H), 7.05 (t, J = 7.6 Hz, 1H), 7.14 (d, J = 7.6 Hz, 1H), 7.19 (d, J = 8.4 Hz,
1H), 7.41 (t, J = 7.6 Hz, 1H), 7.73 (s, 1H), 1H from OH is exchanged; ESI-HRMS ([M+H]+, m/z):
Calcd for C14H17N2O3 261.1238, found 261.1239; Rf = 0.47 (DCM/MeOH, 9:1, v/v).
Methyl 1-allyl-5-(2-hydroxyphenyl)-1H-imidazole-4-carboxylate (7): Yield: 39%; white solid; Mp
128–131 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 3.82 (s, 3H), 4.37–4.50 (m, 2H), 5.07 (d, J = 16.8
Hz, 1H), 5.24 (d, J = 10.0 Hz, 1H), 5.79–5.89 (m, 1H), 7.02 (t, J = 7.4 Hz, 1H), 7.12–7.18 (m, 2H),
7.38 (t, J = 7.4 Hz, 1H), 7.63 (s, 1H), 1H from OH is exchanged; ESI-HRMS ([M+H]+, m/z): Calcd for
C14H15N2O3 259.1086, found 259.1083; Rf = 0.44 (DCM/MeOH, 20:1, v/v).
Methyl 1-butyl-5-(2-hydroxyphenyl)-1H-imidazole-4-carboxylate (8): Yield: 16%; white solid; Mp
118–122 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 0.77 (t, J = 7.2 Hz, 3H), 1.15 (sex, J = 7.2 Hz, 2H),
1.52 (quin, J = 7.2 Hz, 2H), 3.76 (s, 3H), 3.82 (t, J = 7.2 Hz, 2H), 7.00 (dt, J1 = 0.8 Hz, J2 = 7.6 Hz,
1H), 7.12 (dd, J1 = 1.2 Hz, J2 = 7.6 Hz, 2H), 7.34–7.38 (m, 1H), 7.57 (bs, 1H), 1H from OH is
exchanged; ESI-HRMS ([M+H]+, m/z): Calcd for C15H19N2O3 275.1396, found 275.1398; Rf = 0.41
(DCM/MeOH, 9:1, v/v).
Methyl 5-(2-hydroxyphenyl)-1-(2-methoxyethyl)-1H-imidazole-4-carboxylate (9): Yield: 36%; white
solid; Mp 155–158 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 3.21 (s, 3H), 3.42–3.59 (m, 2H), 3.59 (s,
3H), 3.97–4.02 (m, 2H), 6.94 (dt, J1 = 1.2 Hz, J2 = 7.6 Hz, 1H), 7.00 (dd, J1 = 1.2 Hz, J2 = 8.2 Hz,
1H), 7.22 (dd, J1 = 1.6 Hz, J2 = 8.0 Hz, 1H), 7.31 (ddd, J1 = 1.6 Hz, J2 = 8.0 Hz, J3 = 8.2 Hz, 1H), 7.71
(s, 1H), 8.59 (s, 1H); ESI-HRMS ([M+H]+, m/z): Calcd for C14H17N2O4 277.1192, found 277.1188; Rf
= 0.56 (DCM/MeOH, 9:1, v/v).
Methyl 5-(2-hydroxyphenyl)-1-(3-methoxypropyl)-1H-imidazole-4-carboxylate (10): Yield: 39%;
white solid; Mp 77–79 °C; 1H NMR (400 MHz, acetone-d6): δ (ppm) 1.77–1.86 (m, 2H), 3.17 (s, 3H),
3.21 (t, J = 5.6 Hz, 2H), 3.63 (s, 3H), 3.98–4.02 (m, 2H), 6.96 (t, J = 8.0 Hz, 1H), 7.03 (d, J = 8.0 Hz,
1H), 7.22 (d, J = 8.0 Hz, 1H), 7.31 (t, J = 8.0 Hz, 1H), 7.74 (s, 1H), 8.58 (s, 1H); ESI-HRMS
([M+H]+, m/z): Calcd for C15H19N2O4 291.1339, found 291.1345; Rf = 0.56 (DCM/MeOH, 9:1, v/v).
Methyl 5-(5-fluoro-2-hydroxyphenyl)-1-isobutyl-1h-imidazole-4-carboxylate (11): Yield: 51%; green
1
solid; Mp 86–88 °C; H NMR (400 MHz, CDCl3): δ (ppm) 0.73 (d, J = 5.6 Hz, 6H), 1.68–1.76 (m,
1H), 3.4–3.72 (m, 2H), 3.80 (s, 3H), 6.84 (d, J = 8.0 Hz, 1H), 7.07–7.09 (m, 2H), 7.57 (bs, 2H); ESI-
HRMS ([M+H]+, m/z): Calcd for C15H18FN2O3 293.1301, found 293.1303; Rf = 0.48 (DCM/MeOH,
9:1, v/v).
Methyl 5-(2-hydroxy-5-methylphenyl)-1-isobutyl-1H-imidazole-4-carboxylate (12): Yield: 51%; green
solid; Mp 145–147 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 0.72 (s, 6H), 1.67–1.82 (m, 1H), 2.30 (s,
3H), 3.57–3.98 (m, 5H), 6.85–6.96 (m, 1H), 6.99–7.07 (m, 1H), 7.09–7.18 (m, 1H), 7.61 (s, 1H), 8.30
(bs, 1H); ESI-HRMS ([M+H]+, m/z): Calcd for C16H21N2O3 289.1552, found 289.1554; Rf = 0.68
(DCM/MeOH, 9:1, v/v).
Methyl 5-(2-hydroxy-5-methoxyphenyl)-1-isobutyl-1H-imidazole-4-carboxylate (13): Yield: 41%;
brown oil; 1H NMR (400 MHz, CDCl3): δ (ppm) 0.71 (s, 6H), 1.68–1.80 (m, 1H), 3.58–3.83 (m, 8H),
6.64–6.70 (m, 1H), 6.85–6.91 (m, 1H), 6.99–7.10 (m, 1H), 7.57 (s, 1H), 8.11 (bs, 1H); ESI-HRMS
([M+H]+, m/z): Calcd for C16H21N2O4 305.1501, found 305.1515; Rf = 0.42 (DCM/MeOH, 9:1, v/v).
Methyl 5-(4-(diethylamino)-2-hydroxyphenyl)-1-isobutyl-1H-imidazole-4-carboxylate (14): Yield:
1
31%; brown solid; Mp 94–97 °C; H NMR (400 MHz, acetone-d6): δ (ppm) 0.74 (d, J = 6.6 Hz, 6H),
1.17 (t, J = 7.0 Hz, 6H), 1.81–1.88 (m, 1H), 3.39 (q, J = 7.0 Hz, 4H), 3.63 (s, 3H), 3.73 (d, J = 7.5 Hz,

ACCEPTED MANUSCRIPT
2H), 6.28–6.36 (m, 2H), 6.93–6.98 (m, 1H), 7.63 (s, 1H), 1H from OH is exchanged; ESI-HRMS
([M+H]+, m/z): Calcd for C19H28N3O3 346.2131, found 346.2138; Rf = 0.38 (DCM/MeOH, 20:1, v/v).
Methyl 5-(2-hydroxy-4-methoxyphenyl)-1-isobutyl-1H-imidazole-4-carboxylate (15): Yield: 21%;
green oil; 1H NMR (400 MHz, CDCl3): δ (ppm) 0.72 (m, 6H), 1.70–1.80 (m, 1H), 3.62–3.72 (m, 5H),
3.80 (s, 3H), 6.54 (dd, J1 = 2.4 Hz, J2 = 8.4 Hz, 1H), 6.68 (d, J = 2.4 Hz, 1H), 7.04 (d, J = 8.4 Hz, 1H),
7.49 (s, 1H), 8.80 (bs, 1H); ESI-HRMS ([M+H]+, m/z): Calcd for C16H21N2O4 305.1407, found
305.1411; Rf = 0.47 (DCM/MeOH, 9:1, v/v).
Methyl 5-(5-(benzyloxy)-2-hydroxyphenyl)-1-isobutyl-1h-imidazole-4-carboxylate (16): Yield: 42%;
1
green oil; H NMR (400 MHz, CDCl3): δ (ppm) 0.75 (s, 6H), 1.57–1.63 (m, 1H), 3.59–3.78 (m, 2H),
3.80–3.95 (m, 3H), 5.10 (s, 2H), 6.69 (d, J = 7.6 Hz, 1H), 6.75–6.85 (m, 1H), 6.99–7.08 (m, 1H),
7.42–7.51 (m, 6H), 7.55 (bs, 1H); ESI-HRMS ([M+H]+, m/z): Calcd for C22H25N2O4 381.1814, found
381.1803; Rf = 0.57 (DCM/MeOH, 9:1, v/v).
General procedure for synthesis of chromeno[3,4-d]imidazol-4(1H)-ones (17–31)
Imidazoles (4–16) were cyclized using 2 mL or 20 mL 70% H2SO4, and the reaction mixture was then
stirred either 20 min at 80 °C in a microwave reactor (method A; irradiation was initiated at 50 W,
maximum pressure was set to 10 bar) or 3 h at 80 °C in an oil bath (method B), respectively. After
addition of H2O the mixture was extracted with EtOAc and brine. The combined organic phases were
dried over anhydrous Na2SO4, and the solvent was evaporated under pressure. The product was
subjected to flash column chromatography (DCM/MeOH) to yield pure compound.
1-Ethyl-8-fluorochromeno[3,4-d]imidazol-4(1H)-one (17): Method A; Yield: 6%; white crystals; Mp
1
224–228 °C; H NMR (400 MHz, DMSO-d6): δ (ppm) 1.47 (d, J = 7.2 Hz, 3H), 4.58 (q, J = 7.2 Hz,
2H), 7.45 (dt, J1 = 2.8 Hz, J2 = 8.8 Hz, 1H), 7.58 (dd, J1 = 4.8 Hz, J2 = 9.2 Hz, 1H), 7.82 (dt, J1= 2.8
13
Hz, J2 = 9.2 Hz, 1H), 8.30 (s, 1H); C NMR (100 MHz, CDCl3): δ (ppm) 15.5, 41.7, 108.0 (d, JC,F
=
26.3 Hz), 113.6 (d, JC,F = 9.7 Hz), 116.7 (d, JC,F = 24.3 Hz), 119.2 (d, JC,F = 9.1 Hz), 126.8, 134.0,
144.1, 147.7, 155.8, 158.2 (d, JC,F = 240.7 Hz); ESI-HRMS ([M+H]+, m/z): Calcd for C12H10FN2O2
233.0720, found 233.0726; HPLC: RT 8.067 min (96.2%, Method B); Rf = 0.57 (DCM/MeOH, 9:1,
v/v).
1-isopropylchromeno[3,4-d]imidazol-4(1H)-one (18): Method A; Yield: 46%; white crystals; Mp
145–148 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 1.75 (d, J = 8.0 Hz, 6H), 5.05–5.08 (m, 1H), 7.33–
7.38 (m, 1H), 7.43–7.49 (m, 2H), 7.86 (d, J = 8.0 Hz, 1H), 7.99 (s, 1H); 13C NMR (100 MHz, CDCl3):
δ (ppm) 23.2, 50.0, 113.3, 118.4, 121.0, 124.5, 127.4, 129.6, 134.6, 139.0, 152.1, 157.0; ESI-HRMS
([M+H]+, m/z): Calcd for C13H13N2O2 229.0981, found 229.0977; HPLC: RT 8.369 min (99.4%,
Method B); Rf = 0.57 (DCM/MeOH, 9:1, v/v).
1-propylchromeno[3,4-d]imidazol-4(1H)-one (19): Method A; Yield: 42%; white crystals; Mp 136–
1
138 °C; H NMR (400 MHz, CDCl3): δ (ppm) 1.07 (t, J = 7.2 Hz, 3H), 1.97–2.04 (sex, J = 7.2 Hz,
2H), 4.42 (t, J = 7.2 Hz, 2H), 7.36 (dt, J1 = 1.2 Hz, J2 = 7.6 Hz, 1H), 7.44–7.51 (m, 2H), 7.76 (d, J =
7.6 Hz, 1H), 7.82 (s, 1H); 13C NMR (100 MHz, CDCl3): δ (ppm) 11.1, 23.5, 49.4, 113.3, 118.4, 121.0,
124.5, 128.0, 129.8, 135.5, 143.7, 152.2, 157.0; ESI-HRMS ([M+H]+, m/z): Calcd for C13H13N2O2
229.0976, found 229.0977; HPLC: RT 8.874 min (97.7%, Method B); Rf = 0.55 (DCM/MeOH, 9:1,
v/v).
1-allylchromeno[3,4-d]imidazol-4(1H)-one (20): Method A; Yield: 23%; white crystals; Mp 156–160
°C; 1H NMR (400 MHz, CDCl3): δ (ppm) 4.94 (td, J1 = 1.6 Hz, J2 = 17.2 Hz, 1H), 5.05–5.07 (m, 2H),

ACCEPTED MANUSCRIPT
5.26 (td, J1 = 1.6 Hz, J2 = 10.4 Hz, 1H), 6.08 (tdd, J1 = 4.6 Hz, J2 = 10.4 Hz, J3 = 17.2 Hz, 1H), 7.18–
7.25 (m, 2H), 7.34 (ddd, J1 = 1.2 Hz, J2 = 7.2 Hz, J3 = 8.6 Hz, 1H), 7.63 (dd, J1 = 1.2 Hz, J2 = 8.0 Hz,
1H), 7.80 (s, 1H); 13C NMR (100 MHz, CDCl3): δ (ppm) 49.3, 112.6, 117.7, 118.7, 121.6, 124.4,
127.2, 129.7, 131.0, 135.3, 143.3, 151.7, 156.9; ESI-HRMS ([M+H]+, m/z): Calcd for C13H11N2O2
227.0827, found 227.0821; HPLC: RT 8.215 min (98.4%, Method B); Rf = 0.60 (DCM/MeOH, 9:1,
v/v).
1-butylchromeno[3,4-d]imidazol-4(1H)-one (21): Method A; Yield: 28%; white crystals; Mp 117–121
1
°C; H NMR (400 MHz, CDCl3): δ (ppm) 0.97 (t, J = 7.4 Hz, 3H), 1.37–1.49 (m, 2H), 1.86–1.95 (m,
2H), 4.41 (t, J = 7.3 Hz, 2H), 7.32 (ddd, J1 =1.5 Hz, J2 = 7.1 Hz, J3 = 8.0 Hz, 1H), 7.36–7.41 (m, 1H),
7.41–7.47 (m, 1H), 7.73 (dd, J1 = 1.2 Hz, J2 = 8.0 Hz, 1H), 7.75 (s, 1H); 13C NMR (100 MHz, CDCl3):
δ (ppm) 13.6, 19.8, 32.1, 47.5, 113.2, 118.3, 120.9, 124.5, 127.6, 129.7, 134.9, 143.1, 152.1, 156.9;
ESI-HRMS ([M+H]+, m/z): Calcd for C14H15N2O2 243.1134, found 243.1132; HPLC: RT 10.519 min
(98.0%, Method B); Rf = 0.62 (DCM/MeOH, 9:1, v/v).
1-(2-methoxyethyl)chromeno[3,4-d]imidazol-4(1H)-one (22): Method A; Yield: 54%; white crystals;
Mp 146–150 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 3.37 (s, 3H), 3.88 (t, J = 5.2 Hz, 2H), 4.64 (t, J
= 5.2 Hz, 2H), 7.34–7.38 (m, 1H), 7.49–7.54 (m, 2H), 7.84 (d, J =8.0 Hz, 1H), 7.88 (s, 1H); 13C NMR
(100 MHz, DMSO-d6): δ (ppm) 46.6, 49.2, 58.3, 59.6, 70.1, 113.0, 117.4, 122.2, 124.4, 126.2, 129.6,
135.1, 145.0, 151.4, 156.1; ESI-HRMS ([M+H]+, m/z): Calcd for C13H13N2O3 245.0922, found
245.0926; HPLC: RT 7.153 min (53.7%, Method B); Rf = 0.54 (DCM/MeOH, 9:1, v/v).
1-(3-methoxypropyl)chromeno[3,4-d]imidazol-4(1H)-one (23): Method A; Yield: 3%; colorless oil; 1H
NMR (400 MHz, MeOD): δ (ppm) 2.17–2.23 (m, 2H), 3.36 (s, 3H), 3.46 (t, J = 5.6 Hz, 2H), 4.67 (t, J
= 7.2 Hz, 2H), 7.43–7.49 (m, 2H), 7.59 (ddd, J1 = 1.6 Hz, J2 = 7.2 Hz, J3 = 8.6 Hz, 1H), 8.09 (dd, J1 =
13
1.2 Hz, J2 = 8.0 Hz, 1H), 8.12 (s, 1H); C NMR (100 MHz, MeOD): δ (ppm) 31.2, 45.7, 59.1, 69.7,
114.2, 118.9, 123.2, 126.0, 127.6, 131.2, 136.8, 145.7, 153.4, 158.5; ESI-HRMS ([M+H]+, m/z): Calcd
for C14H15N2O3 259.1082, found 259.1083; HPLC: RT 8.003 (96.9%, Method B); Rf = 0.17
(DCM/MeOH, 50:1, v/v).
8-Fluoro-1-isobutylchromeno[3,4-d]imidazol-4(1H)-one (24): Method B; Yield: 3%; white crystals;
1
Mp 178 °C; H NMR (400 MHz, CDCl3): δ (ppm) 1.03 (d, J = 6.8 Hz, 3H), 2.13–2.27 (m, 1H), 4.21
(d, J = 7.2 Hz, 2H), 7.17 (ddd, J1 = 2.9 Hz, J2 = 7.8 Hz, J3 = 9.1 Hz, 1H), 7.35 (dd, J1 = 2.8 Hz, J2 =
8.9 Hz, 1H), 7.39 (dd, J1 = 4.7 Hz, J2 = 9.1 Hz, 1H), 7.41 (s, 1H); 13C NMR (100 MHz, CDCl3): δ
(ppm) 19.5, 29.0, 54.5, 106.9 (d, JC,F = 26.3 Hz), 113.6 (d, JC,F = 8.8 Hz), 116.7 (d, JC,F = 24.1 Hz),
119.6 (d, JC,F = 8.8 Hz), 127.8, 133.9, 143.9, 148.1 (d, JC,F = 2.2 Hz), 156.2, 158.5 (d, JC,F = 243.6 Hz);
ESI-HRMS ([M+H]+, m/z): Calcd for C14H14FN2O2 261.1039, found 261.1033; HPLC: RT 10.974 min
(98.8%, Method A); Rf = 0.59 (DCM/MeOH, 9:1, v/v).
1-Isobutyl-8-methylchromeno[3,4-d]imidazol-4(1H)-one (25): Method A; Yield: 57%; white crystals;
1
Mp 185–186 °C; H NMR (400 MHz, CDCl3): δ (ppm) 1.07 (d, J = 6.4 Hz, 6H), 2.21–2.32 (m, 1H),
2.48 (s, 3H), 4.21 (d, J = 7.6 Hz, 2H), 7.30 (d, J = 8.4 Hz, 1H), 7.36 (d, J = 8.4 Hz, 1H), 7.48 (s, 1H),
7.73 (s, 1H); 13C NMR (100 MHz, CDCl3): δ (ppm) 19.7, 21.3, 29.0, 54.8, 112.9, 118.0, 120.7, 127.7,
130.5, 133.9, 134.8, 143.3, 150.3, 157.0; ESI-HRMS ([M+H]+, m/z): Calcd for C15H17N2O2 257.1290,
found 257.1295; HPLC: RT 11.770 min (96.5%, Method A); Rf = 0.60 (DCM/MeOH, 9:1, v/v).
8-Hydroxy-1-isobutylchromeno[3,4-d]imidazol-4(1H)-one (26): Product was already cyclized after the
first step. Yield: 9%; orange solid; Mp 225–230 °C; 1H NMR (400 MHz, acetone-d6): δ (ppm) 1.03 (d,
J = 6.7 Hz, 6H), 2.21–2.33 (m, 1H), 4.39 (d, J = 7.6 Hz, 2H), 7.06 (dd, J1 = 2.7 Hz, J2 = 9.0 Hz, 1H),
7.33 (d, J = 9.0 Hz, 1H), 7.38 (d, J = 2.7 Hz, 1H), 8.04 (s, 1H), 8.87 (s, 1H); 13C NMR (100 MHz,

ACCEPTED MANUSCRIPT
DMSO-d6): δ (ppm) 19.3, 28.6, 53.3, 106.4, 113.3, 117.5, 118.5, 126.7, 134.7, 144.8, 144.9, 153.9,
156.3; ESI-HRMS ([M+H]+, m/z): Calcd for C14H15N2O3 259.1083, found 259.1076; HPLC: RT 7.522
min (99.8%, Method B); Rf = 0.60 (DCM/MeOH, 9:1, v/v).
1-Isobutyl-8-methoxychromeno[3,4-d]imidazol-4(1H)-one (27): Method A; Yield: 44%; white
1
crystals; Mp 191–192 °C; H NMR (400 MHz, CDCl3): δ (ppm) 1.07 (d, J = 6.4 Hz, 6H), 2.23–2.33
(m, 1H), 3.90 (s, 3H), 4.21 (d, J = 7.2 Hz, 2H), 7.06 (dd, J1 = 2.4 Hz, J2 = 9.0 Hz, 1H), 7.16 (d, J = 2.8
Hz, 1H), 7.40 (d, J = 9.0 Hz, 1H), 7.81 (s, 1H); 13C NMR (100 MHz, CDCl3): δ (ppm) 19.8, 29.2, 55.0,
55.8, 105.2, 115.7, 119.2, 126.2, 128.2, 130.8, 134.0, 144.5, 146.5, 155.3; ESI-HRMS ([M+H]+, m/z):
Calcd for C15H17N2O3 273.1239, found 273.1245; HPLC: RT 10.901 min (97.8%, Method A); Rf =
0.26 (DCM/MeOH, 20:1, v/v).
1-Isobutyl-8-nitrochromeno[3,4-d]imidazol-4(1H)-one (28): Method A; Yield: 4%; yellow crystals;
1
Mp 212–214 °C; H NMR (400 MHz, CDCl3): δ (ppm) 1.15 (d, J = 6.4 Hz, 6H), 2.23–2.34 (m, 1H),
4.31 (d, J = 7.2 Hz, 2H), 7.66 (d, J = 9.2 Hz, 1H), 7.88 (s, 1H), 8.41 (dd, J1 = 2.8 Hz, J2 = 9.0 Hz, 1H),
8.69 (d, J = 3.2 Hz, 1H); 13C NMR (100 MHz, CDCl3): δ (ppm) 19.8, 29.4, 55.1, 113.6, 117.0, 119.4,
124.5, 128.5, 133.5, 143.8, 144.5, 155.2, 155.7; ESI-HRMS ([M+H]+, m/z): Calcd for C14H14N3O4
288.0984, found 288.0980; HPLC: RT 11.513 min (96.5%, Method A); Rf = 0.57 (DCM/MeOH, 20:1,
v/v).
7-(Diethylamino)-1-isobutylchromeno[3,4-d]imidazol-4(1H)-one (29): Method B; Yield: 17%; yellow
solid; Mp 166–169 °C; 1H NMR (400 MHz, (CD3)2CO): δ (ppm) 1.01 (d, J = 6.7 Hz, 6H), 1.21 (t, J =
7.1 Hz, 6H), 2.18–2.30 (m, 1H), 3.50 (q, J = 7.1 Hz, 4H), 4.31 (d, J = 7.5 Hz, 2H), 6.64 (d, J = 2.7 Hz,
1H), 6.78 (dd, J1= 2.7 Hz, J2 = 9.0 Hz, 1H), 7.73 (d, J = 9.0 Hz, 1H), 7.84 (s, 1H); 13C NMR (100
MHz, DMSO-d6): δ (ppm) 12.4, 19.3, 28.4, 43.9, 53.0, 98.1, 100.4, 108.8, 122.8, 123.0, 136.5, 143.5,
148.6, 154.1, 156.7; ESI-HRMS ([M+H]+, m/z): Calcd for C18H24N3O2 314.1869, found 314.1867;
HPLC: RT 12.145 min (99.9%, Method B); Rf = 0.54 (DCM/MeOH, 9:1, v/v).
1-Isobutyl-7-methoxychromeno[3,4-d]imidazol-4(1H)-one (30): Method A; Yield: 57%; white
1
crystals; Mp 198–200 °C; H NMR (400 MHz, CDCl3): δ (ppm) 1.04 (d, J = 6.4 Hz, 6H), 2.17–2.27
(m, 1H), 3.88 (s, 3H), 4.16 (d, J = 7.6 Hz, 2H), 6.89–6.99 (m, 2H), 7.59 (d, J = 9.2 Hz, 1H), 7.67 (s,
13
1H); C NMR (100 MHz, CDCl3): δ (ppm) 19.6, 28.9, 54.5, 55.6, 102.3, 106.2, 112.3, 121.6, 125.6,
135.4, 142.8, 153.8, 156.9, 160.7; ESI-HRMS ([M+H]+, m/z): Calcd for C15H17N2O3 273.1203, found
273.1209; HPLC: RT 10.901 min (98.8%, Method A); Rf = 0.58 (DCM/MeOH, 9:1, v/v).
7-Hydroxy-1-isobutylchromeno[3,4-d]imidazol-4(1H)-one (31): Method A; Yield: 9%; white crystals;
1
Mp 184–187 °C; H NMR (400 MHz, (CD3)2CO): δ (ppm) 1.03 (d, J = 6.8 Hz, 6H), 2.20–2.34 (m,
1H), 4.40 (d, J = 7.6 Hz, 2H), 6.92 (d, J = 2.4 Hz, 1H), 6.96 (dd, J1 = 2.4 Hz, J2 = 8.8 Hz, 1H), 7.87 (d,
J = 8.4 Hz, 1H), 7.97 (s, 1H), 9.39 (s, 1H); 13C NMR (100 MHz, (CD3)2CO): δ (ppm) 19.8, 29.7, 54.6,
104.6, 106.6, 113.6, 123.9, 126.2, 136.7, 144.4, 154.9, 157.1, 159.9; ESI-HRMS ([M+H]+, m/z): Calcd
for C14H15N2O3 259.1083, found 259.1090; HPLC: RT 7.923 min (96.5%, Method A); Rf = 0.15
(DCM/MeOH, 20:1, v/v).
2-(trifluoromethyl)quinazolin-4-ol (36): To freshly prepared NaOEt from Na (2.48 g, 108.0 mmol, 3.0
equiv.) and abs. EtOH (100 mL) were added 35 (4.90 g, 36.0 mmol, 1.0 equiv.) and ethyl
trifluoroacetate (5.10 mL, 43.2 mmol, 1.2 equiv.). The mixture was refluxed for 24 h. Solvents were
evaporated under pressure and water was added (100 mL). The alkaline solution was then made acidic
(pH 2-3) with the addition of 1 M HCl. The resulting precipitate was collected by filtration and
1
recrystallized from EtOH to yield white crystals. Yield: 32%; white crystals; Mp 218–221 °C; H
NMR (400 MHz, DMSO-d6): δ (ppm) 7.69 (ddd, J1 = 1.2 Hz, J2 = 7.0 Hz, J3 = 8.1 Hz, 1H), 7.83 (dd,

ACCEPTED MANUSCRIPT
J1 = 0.4 Hz, J2 = 8.1 Hz, 1H), 7.93 (ddd, J1 = 1.5 Hz, J2 = 7.2 Hz, J3 = 8.6 Hz, 1H), 8.20 (ddd, J1 = 0.4
Hz, J2 = 1.4 Hz, J3 = 8.0 Hz, 1H), 13.60 (s, 1H); HPLC: RT 6.380 min (99.6%, Method D); Rf = 0.60
(DCM/EtOH, 9:1, v/v).
4-Chloro-2-(trifluoromethyl)quinazoline (37): POCl3 (50 mL) was added to 36 (2.44 g, 11.4 mmol, 1.0
equiv.) in a sealed tube and the mixture was heated at 120 °C for 16 h. After completion of the
reaction, the sealed tube was cooled on ice and then toluene (100 mL) was added to the reaction
mixture. The solution obtained was evaporated under reduced pressure. The oily residue was dissolved
in EtOAc (150 mL), cooled to 0 °C, and saturated solution of NaHCO3 (50 mL) was added slowly
while stirring. Afterwards, this was transferred to a separating funnel, the phases were separated, and
the organic phase was washed again with saturated solution of NaHCO3 (50 mL) and brine (50 mL).
The organic phase was dried over anhydrous Na2SO4, filtered and EtOAc evaporated under reduced
pressure. The residue was subjected to flash column chromatography (Et2O/petrolether, 1:10, v/v) to
1
obtain pure 37 as yellow crystals. Yield: 82%; Mp 56–58 °C; H NMR (400 MHz, DMSO-d6): δ
(ppm) 8.06–8.12 (m, 1H), 8.29–8.32 (m, 2H), 8.44 (m, 1H); HPLC: RT 10.130 min (98.2%, Method
D); Rf = 0.85 (DCM/EtOH, 9:1, v/v).
General procedure for synthesis of quinazoline or quinoline amines (39–55)
To a solution of 37 (1.0 equiv.) or 38 (1.0 equiv.) in MeCN, the appropriate amine (2.0 equiv.) and
K2CO3 (3.0 equiv.) were added. The reaction mixture was stirred at 50 °C for 1 h and then at room
temperature for 16 h. EtOAc was added to the reaction mixture, which was washed with H2O and
brine. The combined organic phases were dried over anhydrous Na2SO4, filtered, and evaporated under
reduced pressure. The product was subjected to flash column chromatography (different solvent
combinations used: EtOAc/n-hexane or Et2O/petrolether) to obtain the pure compound.
4-(3-methylpiperidin-1-yl)-2-(trifluoromethyl)quinoline (39): Yield: 67%; white crystals; Mp 72–73
°C; 1H NMR (400 MHz, DMSO-d6): δ (ppm) 0.93 (d, J = 6.6 Hz, 3H), 1.08–1.18 (m, 1H), 1.75–1.95
(m, 4H), 2.57 (dd, J1 = 10.5 Hz, J2 = 11.7 Hz, 1H), 2.81–2.88 (m, 1H), 3.50–3.60 (m, 2H), 7.21 (s,
1H), 7.67 (ddd, J1 = 1.3 Hz, J2 = 6.9 Hz, J3 = 8.2 Hz, 1H), 7.80 (ddd, J1 = 1.5 Hz, J2 = 6.9 Hz, J3 = 8.4
Hz, 1H), 8.03 (dt, J1 = 0.9 Hz, J2 = 8.5 Hz, 2H); 13C NMR (100 MHz, DMSO-d6): δ (ppm) 18.9, 24.8,
30.7, 32.2, 52.5, 59.2, 103.8 (q, JC-F = 2.1 Hz), 119.0 (q, JC-F = 275.7 Hz), 123.1, 124.0, 127.1, 129.9,
130.4, 147.1 (q, JC-F = 33.1 Hz), 148.0, 159.0; ESI-HRMS ([M+H]+, m/z): Calcd for C16H18F3N2
295.1417, found 295.1417; HPLC: RT 11.670 min (97.6%, Method C); Rf = 0.22 (Et2O/petrolether,
1:10, v/v).
1
2-(trifluoromethyl)quinazolin-4-amine (40): Yield: 58%; white crystals; Mp 208–210 °C; H NMR
(400 MHz, CDCl3): δ (ppm) 6.19 (bs, 2H), 7.62–7.68 (m, 1H), 7.83 (dd, J1 = 0.8 Hz, J2 = 8.3 Hz, 1H),
13
7.87–7.92 (m, 1H), 8.03 (d, J = 8.3 Hz, 1H); C NMR (100 MHz, DMSO-d6): δ (ppm) 114.3, 120.0
(q, JC-F = 276.3 Hz), 123.7, 127.5, 127.9, 134.0, 148.6, 152.1 (q, JC-F = 34.1 Hz), 162.9; ESI-HRMS
([M+H]+, m/z): Calcd for C9H7F3N3 214.0592, found 214.0593; HPLC: RT 6.407 min (100.0%,
Method D); Rf = 0.74 (DCM/MeOH, 9:1, v/v).
N-propyl-2-(trifluoromethyl)quinazolin-4-amine (41): Yield: 13%; yellow crystals; Mp 112–114 °C;
1H NMR (400 MHz, CDCl3): δ (ppm) 1.04 (t, J = 7.4 Hz, 3H), 1.72–1.79 (m, 2H), 3.66–3.72 (m, 2H),
5.99 (t, J = 5.2 Hz, 1H), 7.56 (ddd, J1 = 1.2 Hz, J2 = 7.0 Hz, J3 = 8.3 Hz, 1H), 7.75 (dd, J1 = 0.9 Hz, J2
= 8.4 Hz, 1H), 7.79 (ddd, J1 = 1.4 Hz, J2 = 7.0 Hz, J3 = 8.4 Hz, 1H), 7.96 (dd, J1 = 0.6 Hz, J2 = 8.4 Hz,
1H); 13C NMR (100 MHz, CDCl3): δ (ppm) 11.6, 22.6, 43.4, 114.8, 120.1 (q, JC-F = 276.0 Hz), 120.5,
127.8, 129.6, 133.4, 149.0, 153.3 (q, JC-F = 35.2 Hz), 160.6; ESI-HRMS ([M+H]+, m/z): Calcd for

ACCEPTED MANUSCRIPT
C12H13F3N3 256.1062, found 256.1061; HPLC: RT 9.807 min (96.5%, Method D); Rf = 0.64
(DCM/EtOH, 9:1, v/v).
2-((2-(trifluoromethyl)quinazolin-4-yl)amino)ethan-1-ol (42): Yield: 73%; white crystals; Mp 188–
1
190 °C; H NMR (400 MHz, DMSO-d6): δ (ppm) 3.62–3.70 (m, 4H), 4.86 (t, J = 5.3 Hz, 1H), 7.65
(ddd, J1 = 1.5 Hz, J2 = 6.8 Hz, J3 = 8.2 Hz, 1H), 7.80–7.87 (m, 2H), 8.36 (d, J = 8.2 Hz, 1H), 8.82 (t, J
= 5.0 Hz, 1H); 13C NMR (100 MHz, DMSO-d6): δ (ppm) 43.6, 58.7, 115.0, 120.0 (q, JC-F = 276.0 Hz),
123.1, 127.6, 128.0, 133.6, 148.1, 151.8 (q, JC-F = 34.3 Hz), 160.6; ESI-HRMS ([M+H]+, m/z): Calcd
for C11H11F3N3O 258.0854, found 258.0858; HPLC: RT 6.530 min (100.0%, Method D); Rf = 0.26
(DCM/EtOH, 9:1, v/v).
(2-(trifluoromethyl)quinazolin-4-yl)glycine (43): Methyl (2-(trifluoromethyl)quinazolin-4-yl)glycinate
was prepared according to the general procedure for synthesis of quinazoline-4-amines. The product
was recrystallized from Et2O and then 2 M NaOH was added. The reaction mixture was stirred for 30
min at room temperature. This was followed by extraction with DCM. The water phase was then
acidified with 37% HCl to pH around 2. The resulting precipitate was collected by filtration to provide
1
white crystalline product. Yield: 22%; white crystals; Mp 241–243 °C; H NMR (400 MHz, DMSO-
d6): δ (ppm) 4.22 (d, J = 5.7 Hz, 2H), 7.73 (ddd, J1 = 1.6 Hz, J2 = 6.9 Hz, J3 = 8.3 Hz, 1H), 7.88 (dd, J1
= 1.4 Hz, J2 = 8.6 Hz, 1H), 7.91–7.96 (m, 1H), 8.36 (d, J = 8.4 Hz, 1H), 9.25 (t, J = 6.1 Hz), 12.76 (s,
13
1H); C NMR (100 MHz, DMSO-d6): δ (ppm) 42.4, 114.7, 119.9 (q, JC-F = 275.9 Hz), 122.9, 128.1,
128.2, 134.0, 148.1, 151.5 (q, JC-F = 34.5 Hz), 160.7, 170.8; ESI-HRMS ([M+H]+, m/z): Calcd for
C11H9F3N3O2 272.0641, found 272.0636; HPLC: RT 6.697 min (99.1%, Method D); Rf = 0.26
(DCM/EtOH, 9:1, v/v).
N-isopropyl-2-(trifluoromethyl)quinazolin-4-amine (44): Yield: 70%; white crystals; Mp 107–109 °C;
1H NMR (400 MHz, DMSO-d6): δ (ppm) 1.29 (d, J = 6.6 Hz, 6H), 4.51 (m, 1H), 7.66 (ddd, J1 = 1.4
Hz, J2 = 6.9 Hz, J3 = 8.3 Hz, 1H), 7.82 (dd, J1 = 1.0 Hz, J2 = 8.3 Hz, 1H), 7.88 (ddd, J1 = 1.3 Hz, J2 =
13
6.8 Hz, J3 = 8.2 Hz, 1H), 8.43 (dd, J1 = 0.8 Hz, J2 = 8.4 Hz, 1H), 8.48 (d, J = 7.6 Hz, 1H); C NMR
(100 MHz, DMSO-d6): δ (ppm) 22.3, 43.7, 115.4, 120.7 (q, JC-F = 276.0 Hz), 123.7, 128.6, 128.8,
134.8, 148.6, 152.8 (q, JC-F = 34.0 Hz), 160.4; ESI-HRMS ([M+H]+, m/z): Calcd for C12H13F3N3
256.1062, found 256.1061; HPLC: RT 9.427 min (98.6%, Method C); Rf = 0.43 (EtOAc/n-hexane,
1:1, v/v).
1
N-isobutyl-2-(trifluoromethyl)quinazolin-4-amine (45): Yield: 79%; white crystals; Mp 88–90 °C; H
NMR (400 MHz, DMSO-d6): δ (ppm) 0.91 (d, J = 6.7 Hz, 6H), 1.99–2.09 (m, 1H), 3.39 (dd, J1 = 5.9
Hz, J2 = 6.9 Hz, 2H), 7.63 (ddd, J1 = 1.9 Hz, J2 = 6.4 Hz, J3 = 8.3 Hz, 1H), 7.79–7.86 (m, 2H), 8.38
(dd, J1 = 0.6 Hz, J2 = 8.4 Hz, 1H), 8.79 (t, J = 5.6 Hz, 1H); 13C NMR (100 MHz, DMSO-d6): δ (ppm)
20.2, 27.5, 48.2, 114.9, 120.1 (q, JC-F = 276.0 Hz), 123.0, 127.6, 128.1, 133.5, 148.3, 152.0 (q, JC-F
=
34.2 Hz), 160.7; ESI-HRMS ([M+H]+, m/z): Calcd for C13H15F3N3 270.1218, found 270.1215; HPLC:
RT 10.540 min (100.0%, Method D); Rf = 0.21 (DCM/EtOH, 9:1, v/v).
N-cyclopropyl-2-(trifluoromethyl)quinazolin-4-amine (46): Yield: 86%; yellow crystals; Mp 89–90
°C; 1H NMR (400 MHz, CDCl3): δ (ppm) 0.66–0.70 (m, 2H), 0.94–0.99 (m, 2H), 3.10–3.17 (m, 1H),
6.28 (bs, 1H), 7.52 (ddd, J1 = 1.2 Hz, J2 = 7.0 Hz, J3 = 8.3 Hz, 1H), 7.73–7.79 (m, 2H), 7.92–7.94 (m,
1H); 13C NMR (100 MHz, CDCl3): δ (ppm) 7.2, 24.6, 114.7, 120.0 (q, JC-F = 276.0 Hz), 120.7, 127.7,
129.0, 133.3, 148.6, 153.0 (q, JC-F = 35.2 Hz), 161.7; ESI-HRMS ([M+H]+, m/z): Calcd for C12H11F3N3
254.0905, found 254.0899; HPLC: RT 9.227 min (99.1%, Method D); Rf = 0.51 (EtOAc/n-hexane,
1:1, v/v).

ACCEPTED MANUSCRIPT
N-(cyclopropylmethyl)-2-(trifluoromethyl)quinazolin-4-amine (47): Yield: 68%; white crystals; Mp
1
132–134 °C; H NMR (400 MHz, CDCl3): δ (ppm) 0.35–0.39 (m, 2H), 0.59–0.64 (m, 2H), 1.14–1.22
(m, 1H), 3.56 (dd, J1 = 5.2 Hz, J2 = 7.3 Hz, 2H), 6.12 (bs, 1H), 7.57 (ddd, J1 = 1.2 Hz, J2 = 7.1 Hz, J3 =
13
8.2 Hz, 1H), 7.78–7.82 (m, 2H), 7.95–7.97 (m, 1H); C NMR (100 MHz, CDCl3): δ (ppm) 3.6, 10.4,
46.6, 114.6, 120.2 (q, JC-F = 275.2 Hz), 120.7 127.6, 129.2, 133.3, 148.8, 151.2 (q, JC-F = 33.8 Hz),
160.3; ESI-HRMS ([M+H]+, m/z): Calcd for C13H13F3N3 268.1062, found 268.1060; HPLC: RT
10.057 min (100.0%, Method D); Rf = 0.55 (EtOAc/n-hexane, 1:1, v/v).
N-cyclopentyl-2-(trifluoromethyl)quinazolin-4-amine (48): Yield: 66%; white crystals; Mp 50–52 °C;
1H NMR (400 MHz, DMSO-d6): δ (ppm) 1.55–1.70 (m, 4H), 1.72–1.79 (m, 2H), 2.02–2.07 (m, 2H),
4.51–4.60 (m, 1H), 7.66 (ddd, J1 = 1.5 Hz, J2 = 6.8 Hz, J3 = 8.3 Hz, 1H), 7.82 (dd, J1 = 1.1 Hz, J2 = 8.3
Hz, 1H), 7.88 (ddd, J1 = 1.3 Hz, J2 = 6.8 Hz, J3 = 8.2 Hz, 1H), 8.46 (dd, J1 = 0.8 Hz, J2 = 8.4 Hz, 1H),
13
8.53 (d, J = 6.8 Hz, 1H); C NMR (100 MHz, DMSO-d6): δ (ppm) 23.7, 31.7, 52.6, 114.9, 120.0 (q,
JC-F = 276.3 Hz), 123.4, 127.6, 128.1, 133.6, 148.2, 151.9 (q, JC-F = 33.7 Hz), 160.1; ESI-HRMS
([M+H]+, m/z): Calcd for C14H15F3N3 282.1106, found 282.1098; HPLC: RT 10.743 min (99.0%,
Method C); Rf = 0.61 (EtOAc/n-hexane, 1:1. v/v).
N-cyclohexyl-2-(trifluoromethyl)quinazolin-4-amine (49): Yield: 96%; white crystals; Mp 55–57 °C;
1H NMR (400 MHz, DMSO-d6): δ (ppm) 1.12–1.22 (m, 1H), 1.29–1.48 (m, 4H), 1.61–1.68 (m, 1H),
1.74–1.82 (m, 2H), 1.93–2.00 (m, 2H), 4.11–4.21 (m, 1H), 7.65 (ddd, J1 = 1.4 Hz, J2 = 6.8 Hz, J3 = 8.3
Hz, 1H), 7.80 (dd, J1 = 1.1 Hz, J2 = 8.3 Hz, 1H), 7.86 (ddd, J1 = 1.3 Hz, J2 = 6.8 Hz, J3 = 8.2 Hz, 1H),
13
8.41–8.44 (m, 2H); C NMR (100 MHz, DMSO-d6): δ (ppm) 24.9, 25.3, 31.7, 50.0, 114.9, 120.0 (q,
JC-F = 276.0 Hz), 123.3, 127.5, 128.1, 133.6, 148.3, 152.0 (q, JC-F = 34.3 Hz), 159.6; ESI-HRMS
([M+H]+, m/z): Calcd for C15H17F3N3 296.1369, found 296.1368; HPLC: RT 11.323 min (98.9%,
Method C); Rf = 0.38 (EtOAc/n-hexane, 1:2, v/v).
4-(pyrrolidin-1-yl)-2-(trifluoromethyl)quinazoline (50): Yield: 57%; white crystals; Mp 111–113 °C;
1H NMR (400 MHz, CDCl3): δ (ppm) 2.03–2.06 (m, 4H), 3.93–3.96 (m, 4H), 7.45 (ddd, J1 = 1.3 Hz,
J2 = 7.1 Hz, J3 = 8.4 Hz, 1H), 7.72 (ddd, J1 = 1.3 Hz, J2 = 7.0 Hz, J3 = 8.3 Hz, 1H), 7.90 (dd, J1 = 0.9
Hz, J2 = 8.4 Hz, 1H), 8.16 (dd, J1 = 0.9 Hz, J2 = 8.5 Hz, 1H), 1H from NH is exchanged; 13C NMR
(100 MHz, CDCl3): δ (ppm) 10.6, 46.8, 114.8, 120.2 (q, JC-F = 275.9 Hz), 120.8, 127.8, 129.4, 133.4,
148.9, 153.2 (q, JC-F = 35.1 Hz), 160.4; ESI-HRMS ([M+H]+, m/z): Calcd for C13H13F3N3 268.1062,
found 268.1066; HPLC: RT 10.220 min (95.6%, Method D); Rf = 0.41 (EtOAc/n-hexane, 1:1, v/v).
1
4-(2-(trifluoromethyl)quinazolin-4-yl)morpholine (51): Yield: 87%; white crystals; Mp 98–100 °C; H
(400 MHz, CDCl3): δ (ppm) 3.83–3.86 (m, 4H), 3.88–3.91 (m, 4H), 7.52 (ddd, J1 = 1.3 Hz, J2 = 7.0
Hz, J3 = 8.3 Hz, 1H), 7.77 (ddd, J1 = 1.4 Hz, J2 = 7.0 Hz, J3 = 8.4 Hz, 1H), 7.88 (dd, J1 = 0.9 Hz, J2 =
8.4 Hz, 1H), 7.96 (dd, J1 = 0.7 Hz, J2 = 8.4 Hz, 1H), 1H from NH is exchanged; 13C (100 MHz,
CDCl3): δ (ppm) 50.0, 66.7, 116.0, 120.0 (q, JC-F = 275.9 Hz), 124.8, 127.1, 129.5, 133.4, 151.6, 151.7
(q, JC-F = 35.7 Hz), 164.8; ESI-HRMS ([M+H]+, m/z): Calcd for C13H13F3N3O 284.1011, found
284.1013; HPLC: RT 9.090 min (100.0%, Method D); Rf = 0.74 (DCM/EtOH, 9:1, v/v).
4-(4-methylpiperazin-1-yl)-2-(trifluoromethyl)quinazoline (52): Yield: 91%; yellow crystals; Mp 77–
79 °C; 1H NMR (400 MHz, CDCl3): δ (ppm) 2.38 (s, 3H), 2.60–2.63 (m, 4H), 3.94–3.98 (m, 4H), 7.54
(ddd, J1 = 1.3 Hz, J2 = 7.0 Hz, J3 = 8.3 Hz, 1H), 7.80 (ddd, J1 = 1.4 Hz, J2 = 7.0 Hz, J3 = 8.4 Hz, 1H),
7.92 (dd, J1 = 0.9 Hz, J2 = 8.4 Hz, 1H), 8.01 (dd, J1 = 0.8 Hz, J2 = 8.4 Hz, 1H), 1H from NH is
exchanged; 13C NMR (100 MHz, DMSO-d6): δ (ppm) 45.4, 48.8, 54.3, 115.3, 120.0 (q, JC-F = 276.0
Hz), 125.7, 127.2, 128.4, 133.7, 150.3 (q, JC-F = 34.8 Hz), 150.7, 163.7; ESI-HRMS ([M+H]+, m/z):

ACCEPTED MANUSCRIPT
Calcd for C14H16F3N4 297.1327, found 297.1328; HPLC: RT 5.810 min (100.0%, Method D); Rf =
0.26 (DCM/EtOH, 9:1, v/v).
1-(2-((2-(trifluoromethyl)quinazolin-4-yl)amino)ethyl)piperidin-1-ium chloride (53): Yield: 15%;
1
yellow crystals; Mp 94–96 °C; H NMR (400 MHz, CDCl3): δ (ppm) 1.46–1.53 (m, 2H), 1.60–1.66
(m, 4H), 2.22–2.29 (m, 2H), 2.44–2.52 (m, 3H), 2.64–2.68 (m, 2H), 3.71–3.76 (m, 2H), 7.27 (bs, 1H),
7.56 (ddd, J1 = 1.2 Hz, J2 = 7.1 Hz, J3 = 8.3 Hz, 1H), 7.78 (ddd, J1 = 1.4 Hz, J2 = 7.1 Hz, J3 = 8.4 Hz,
1H), 7.83 (ddd, J1 = 0.5 Hz, J2 = 1.3 Hz, J3 = 8.3 Hz, 1H), 7.94 (dd, J1 = 0.4 Hz, J2 = 1.1 Hz, J3 = 8.4
Hz, 1H); 13C NMR (100 MHz, DMSO-d6): (ppm) 23.9, 25.5, 38.2, 54.0, 56.9, 114.8, 119.9 (q, JC-F
=
276.2 Hz), 122.8, 127.6, 128.0, 133.6, 148.1, 151.8 (q, JC-F = 34.0 Hz), 160.4; ESI-HRMS ([M+H]+,
m/z): Calcd for C16H20F3N4 325.1635, found 325.1640; HPLC: RT 6.347 min (99.4%, Method D); Rf
= 0.28 (DCM/EtOH, 9:1, v/v).
N-(2-(1H-imidazol-5-yl)ethyl)-2-(trifluoromethyl)quinazolin-4-amine (54): Yield: 71%; white crystals;
1
Mp 204–206 °C; H NMR (00 MHz, DMSO-d6): δ (ppm) 2.92 (t, J = 7.3 Hz, 2H), 3.79 (td, J1 = 5.6
Hz, J2 = 7.3 Hz, 2H), 6.86 (bs, 1H), 7.61 (d, J = 1.0 Hz, 1H), 7.67 (ddd, J1 = 1.4 Hz, J2 = 6.8 Hz, J3 =
8.2 Hz, 1H), 7.83 (dd, J1 = 1.1 Hz, J2 = 8.3 Hz, 1H), 7.89 (ddd, J1 = 1.2 Hz, J2 = 6.8 Hz, J3 = 8.2 Hz,
13
1H), 8.31 (dd, J1 = 0.8 Hz, J2 = 8.4 Hz, 1H), 8.92 (t, J = 5.3 Hz, 1H), 12.03 (bs, 1H); C NMR (100
MHz, DMSO-d6): δ (ppm) 26.0, 41.0, 114.9, 116.5, 120.0 (q, JC-F = 276.6 Hz), 122.9, 127.7, 128.1,
129.7, 133.6, 134.7, 151.9 (q, JC-F = 34.2 Hz), 148.1, 160.4; ESI-HRMS ([M+H]+, m/z): Calcd for
C14H13F3N5 308.1123, found 308.1121; HPLC: RT 5.567 min (100.0%, Method D); Rf = 0.28
(DCM/EtOH, 9:1, v/v).
N-(4-aminobenzyl)-2-(trifluoromethyl)quinazolin-4-amine (55): Yield: 97%; brown crystals; Mp 191–
193 °C; 1H NMR (400 MHz, CDCl3): δ (ppm): 3.71 (bs, 2H), 4.75 (d, J = 5.2 Hz, 2H), 6.12 (t, J = 4.6
Hz, 1H), 6.65–6.69 (m, 2H), 7.20–7.24 (m, 2H), 7.53 (ddd, J1 = 1.2 Hz, J2 = 7.0 Hz, J3 = 8.3 Hz, 1H),
7.70 (dd, J1 = 0.9 Hz, J2 = 8.4 Hz, 1H), 7.79 (ddd, J1 = 1.4 Hz, J2 = 7.0 Hz, J3 = 8.4 Hz, 1H), 7.96 (dd,
13
J1 = 0.7 Hz, J2 = 8.4 Hz, 1H); C NMR (100 MHz, DMSO-d6): δ (ppm) 43.8, 113.6, 114.9, 120.0 (q,
J
C-F = 276.3 Hz), 123.0, 125.3, 127.6, 128.0, 128.9, 133.6, 147.8, 148.2, 151.8 (q, JC-F = 34.04 Hz),
160.2; ESI-HRMS ([M+H]+, m/z): Calcd for C16H14F3N4 319.1171, found 317.1173; HPLC: RT 6.423
min (98.4%, Method D); Rf = 0.43 (DCM/EtOH, 9:1, v/v).
Characterization of active commercially available compounds
1
N-isobutyl-2-(trifluoromethyl)quinoline-4-amine (32): Yellow crystals; Mp 85–86 °C; H NMR (400
MHz, CDCl3): δ (ppm) 1.09 (d, J = 6.7 Hz, 6H), 2.08 (sep, J1 = 6.7 Hz, J2 = 13.4 Hz, 1H), 3.20 (dd, J1
= 5.5 Hz, J2 = 6.8 Hz, 2H), 5.29 (bs, 1H), 6.72 (s, 1H), 7.53 (ddd, J1 = 1.0 Hz, J2 = 6.9 Hz, J3 = 8.2 Hz,
1H), 7.71 (ddd, J1 = 1.2 Hz, J2 = 6.9 Hz, J3 = 8.3 Hz, 1H), 7.76 (d, J = 8.4 Hz, 1H), 8.09 (d, J = 8.5
Hz, 1H); 13C NMR (100 MHz, DMSO-d6): δ (ppm) 20.3, 26.8, 50.0, 93.1 (q, JC-F = 2.1 Hz), 118.8,
121.9, 122.1 (q, JC-F = 275.7 Hz), 125.9, 129.5, 130.3, 147.2, 147.4 (q, JC-F = 32.5 Hz), 152.2; ESI-
HRMS ([M+H]+, m/z): Calcd for C14H16F3N2 269.1266, found 269.1269; HPLC: RT 6.453 min
(95.7%, Method C); Rf = 0.53 (EtOAc/n-hexane, 1:1, v/v).
1
4-(3-methylpiperidin-1-yl)-2-(trifluoromethyl)quinazoline (33): White crystals; Mp 50–51 °C; H
NMR (400 MHz, (CD3)2CO): δ (ppm) 0.91 (d, J = 6.5 Hz, 3H), 1.19–1.31 (m, 1H), 1.61–1.70 (m, 1H),
1.77–1.87 (m, 3H), 2.93 (dd, J1 = 10.5 Hz, J2 = 13.0 Hz, 1H), 3.21 (ddd, J1 = 2.8 Hz, J2 = 11.9 Hz, J3 =
13.2 Hz, 1H), 4.26–4.36 (m, 2H), 7.61–7.68 (m, 1H), 7.87–7.90 (m, 2H), 8.01–8.03 (m, 1H); 13C
NMR (100 MHz, DMSO-d6): δ (ppm) 18.9, 24.9, 31.0, 32.5, 49.7, 56.1, 115.5, 120.1 (q, JC-F = 275.9
Hz), 125.8, 127.2, 128.4, 133.7, 150.5 (q, JC-F = 34.5 Hz), 150.9, 163.7; ESI-HRMS ([M+H]+, m/z):

ACCEPTED MANUSCRIPT
Calcd for C15H17F3N3 296.1375, found 296.1377; HPLC: RT 11.987 min (97.6%, Method C); Rf =
0.65 (EtOAC/n-hexane, 1:1, v/v).
N-(1-cyclopropylpiperidin-4-yl)-2-(trifluoromethyl)quinazolin-4-amine (34): Yellow crystals; Mp
1
140–142 °C; H NMR (400 MHz, DMSO-d6): δ (ppm) 0.28–0.32 (m, 2H), 0.39–0.43 (m, 2H), 1.55–
1.68 (m, 3H), 1.86–1.94 (m, 2H), 2.27 (dt, J1 = 2.0 Hz, J2 = 11.7 Hz, 2H), 2.94–3.02 (m, 2H), 4.11–
4.21 (m, 1H), 7.64 (ddd, J1 = 1.4 Hz, J2 = 6.9 Hz, J3 = 8.3 Hz, 1H), 7.81 (dd, J1 = 1.1 Hz, J2 = 8.3 Hz,
1H), 7.87 (ddd, J1 = 1.2 Hz, J2 = 6.9 Hz, J3 = 8.2 Hz, 1H), 8.38–8.42 (m, 1H), 8.46 (d, J = 7.4 Hz, 1H);
13C (100 MHz, DMSO-d6): δ (ppm) 6.0, 30.8, 38.2, 48.9, 52.5, 114.9, 120.1 (q, JC-F = 176.0 Hz),
123.3, 127.8, 128.2, 133.9, 148.3, 152.0 (q, JC-F = 34.4 Hz), 159.9; ESI-HRMS ([M+H]+, m/z): Calcd
for C17H20F3N4 337.1635, found 337.1632; HPLC: RT 6.363 min (92.6%, Method D).
Cytotoxicity
Prior to determination of agonist activities, the cytotoxicities of these compounds were determined on
HEK-Blue hTLR7/8 cells (InVivoGen) using the MTS cell proliferation assay (CellTiter 96 AQueous
One Solution Cell Proliferation Assay, Promega), according to the manufacturer protocol. The cells of
interest were seeded in 96-well plates (3 ×104 cells/well). After 24 h incubation at 37 °C and 5% CO2,
the cells were either left untreated (negative control) or treated with solutions of the tested compounds
in dimethylsulfoxide (DMSO). For positive control, pure DMSO was used. The cells were incubated
again under the same conditions. After 24 h they were treated with MTS, incubated for an additional 2
h, then absorbance at 490 nm was measured.
TLR7 antagonist activity
The antagonist activities of these compounds against TLR7 were determined using the HEK293 cell
line, co-transfected with the hTLR7 gene and an inducible SEAP reporter gene (InvivoGen, San
Diego, CA, USA), as described previously [31]. In the preliminary assays, the cells were seeded into
96-well plates (3 ×104 cell/well). After a 24-h incubation at 37 °C and 5% CO2, the cells were treated
with solutions of the compounds in DMSO at the highest noncytotoxic concentrations, which varied
from 250 µM to 7.8 µM. After 2 h imiquimod was added to the pre-treated cells to 1.0 µg/mL. The
cells were then incubated once again for 24 h under the same conditions. The activity of SEAP in the
cell supernatant was measured using the colorimetric Quanti-Blue reagent, according to the
manufacturer protocol.
TLR7 agonist activity
For determination of the TLR7 agonist activities, the HEK293 cell line was co-transfected with the
hTLR7 gene and an inducible SEAP reporter gene (InvivoGen, San Diego, CA, USA) was used, as
described previously [31]. For the preliminary assays, the cells were seeded in 96-well plates (3 ×104
cell/well). After 24-h incubation at 37 °C and 5% CO2, the cells were treated with solutions of the
compounds in DMSO at their highest noncytotoxic concentrations, which varied from 250 µM to 7.8
µM. For positive control, imiquimod was used at 4 µM and 10 µM. The cells were incubated again for
24 h under the same conditions. The activity of SEAP in the cell supernatant was measured using the
colorimetric Quanti-Blue reagent, according to the manufacturer protocol.
For the active compounds, EC50 was determined in the same manner to that described above. This time
the cells were stimulated with the active compounds at eight different concentrations, starting with the
highest noncytotoxic one. Imiquimod at 20.8, 10.4, 5.2, 2.6, 1.3, 0.65, and 0.325 µM was used as
positive control.

ACCEPTED MANUSCRIPT
TLR8 preliminary agonist activity
For determination of the TLR8 agonist activities, the HEK293 cell line was co-transfected with the
hTLR8 gene and an inducible SEAP reporter gene (InvivoGen, San Diego, CA, USA) was used, as
described previously [31]. For the preliminary assays, the cells were seeded in 96-well plates (3 ×104
cell/well). After a 24-h incubation at 37 °C and 5% CO2, the cells were treated with solutions of the
compounds in DMSO at the highest noncytotoxic concentrations, which varied from 250 µM to 7.8
µM. For the positive control, resiquimod at 3 µM and 15 µM was used. The cells were incubated again
for 24 h under the same conditions. The activity of SEAP in the cell supernatants was measured using
the colorimetric Quanti-Blue reagent, according to the manufacturer protocol.
Quantification of cytokine production
The BD Human Cytometric Bead Array (BD Biosciences) was used to assay the protein levels of IL-
1β, IL-6, IL-8, IL-10, IL-12p70 and TNFα in the cell culture supernatant, according to the
manufacturer’s protocol (BD Life Sciences – Biosciences). Briefly, PBMCs were isolated from buffy
coats obtained at the Blood Transfusion Centre of Slovenia according to institutional guideline. Cells
were seeded into 96-well plates (3 × 105 cells/well) after addition of 300 µL of cRPMI medium and
selected TLR7 agonist (final concentration is 10 µg/mL) into wells. For positive control imiquimod (1)
was used in the same concentration and the negative control were untreated cells. Cells were
centrifuged after 24 h incubation at 37 °C and 5% CO2 and cell culture supernatants were used for
further analysis. In an appropriate assay tube, 50 µL of the Capture Beads, 50 µL of the Detection
Reagent and 50 µL of sample (5-fold dilution) were mixed and incubated for 3 h at room temperature
and protected from light. Samples were then washed with 1 mL of Wash Buffer and centrifuged at
200G for 5 min. The supernatant was carefully aspirated and discarded from each assay tube. The bead
pellet was resuspended by adding 300 µL of Wash Buffer. Flow cytometry was performed using a
FACSCalibur system (Becton Dickinson, Inc.). A standard curve was prepared by serial dilutions of
standards and used for determination of cytokine concentrations in supernatants.
Acknowledgments
The authors acknowledge the financial support from the Slovenian Research Agency (research core
funding No. P1-0208, project grant BI-US/18-20-087 and junior researcher's programme for Ana
Dolšak). We thank Dr. Chris Berrie for critical reading of the manuscript.
Declaration of Interest
The authors declare that they have no conflicts of interest.
References
[1] T. Kawai, S. Akira, Toll-like Receptors and Their Crosstalk with Other Innate Receptors in
Infection and Immunity, Immunity. 34 (2011) 637–650. doi:10.1016/j.immuni.2011.05.006.
[2] M. Smith, E. García-Martínez, M.R. Pitter, J. Fucikova, R. Spisek, L. Zitvogel, G. Kroemer, L.
Galluzzi, Trial Watch: Toll-like receptor agonists in cancer immunotherapy, OncoImmunology.
7 (2018) e1526250. doi:10.1080/2162402X.2018.1526250.
[3] E.J. Hennessy, A.E. Parker, L.A.J. O’Neill, Targeting Toll-like receptors: emerging
therapeutics?, Nat. Rev. Drug Discov. 9 (2010) 293–307. doi:10.1038/nrd3203.
[4] X. Wang, C. Smith, H. Yin, Targeting Toll-like receptors with small molecule agents, Chem.
Soc. Rev. 42 (2013) 4859. doi:10.1039/c3cs60039d.
[5] S. Diebold, Recognition of viral single-stranded RNA by Toll-like receptors☆, Adv. Drug
Deliv. Rev. 60 (2008) 813–823. doi:10.1016/j.addr.2007.11.004.

ACCEPTED MANUSCRIPT
[6] Y. Kumagai, O. Takeuchi, S. Akira, Pathogen recognition by innate receptors, J. Infect.
Chemother. 14 (2008) 86–92. doi:10.1007/s10156-008-0596-1.
[7] T. Kawai, S. Akira, The role of pattern-recognition receptors in innate immunity: update on Toll-
like receptors, Nat. Immunol. 11 (2010) 373–384. doi:10.1038/ni.1863.
[8] M.C. Patel, K.A. Shirey, L.M. Pletneva, M.S. Boukhvalova, A. Garzino-Demo, S.N. Vogel, J.C.
Blanco, Novel drugs targeting Toll-like receptors for antiviral therapy, Future Virol. 9 (2014)
811–829. doi:10.2217/fvl.14.70.
[9] N.J. Horscroft, D.C. Pryde, H. Bright, Antiviral applications of Toll-like receptor agonists, J.
Antimicrob. Chemother. 67 (2012) 789–801. doi:10.1093/jac/dkr588.
[10] M. Czarniecki, Small Molecule Modulators of Toll-like Receptors, J. Med. Chem. 51 (2008)
6621–6626. doi:10.1021/jm800957k.
[11] E. Yoo, B.M. Crall, R. Balakrishna, S.S. Malladi, L.M. Fox, A.R. Hermanson, S.A. David,
Structure–activity relationships in Toll-like receptor 7 agonistic 1H-imidazo[4,5-c]pyridines,
Org. Biomol. Chem. 11 (2013) 6526. doi:10.1039/c3ob40816g.
[12] Miller, R. L., Gerster, J. F., Owens, M. L., Slade, H. B., Tomai, M. A., Imiquimod applied
topically: a novel immune response modifier and new class of drug, Int. J. Immunopharmacol.
21 (1998) 1–14.
[13] E. Vacchelli, L. Galluzzi, A. Eggermont, W.H. Fridman, J. Galon, C. Sautès-Fridman, E.
Tartour, L. Zitvogel, G. Kroemer, Trial watch: FDA-approved Toll-like receptor agonists for
cancer therapy, OncoImmunology. 1 (2012) 894–907. doi:10.4161/onci.20931.
[14] E. Hanna, R. Abadi, O. Abbas, Imiquimod in dermatology: an overview, Int. J. Dermatol. 55
(2016) 831–844. doi:10.1111/ijd.13235.
[15] H. Hemmi, T. Kaisho, O. Takeuchi, S. Sato, H. Sanjo, K. Hoshino, T. Horiuchi, H. Tomizawa,
K. Takeda, S. Akira, Small anti-viral compounds activate immune cells via the TLR7 MyD88–
dependent signaling pathway, Nat. Immunol. 3 (2002) 196–200. doi:10.1038/ni758.
[16] N.M. Shukla, S.S. Malladi, C.A. Mutz, R. Balakrishna, S.A. David, Structure−Activity
Relationships in Human Toll-Like Receptor 7-Active Imidazoquinoline Analogues, J. Med.
Chem. 53 (2010) 4450–4465. doi:10.1021/jm100358c.
[17] N.M. Shukla, D.B. Salunke, R. Balakrishna, C.A. Mutz, S.S. Malladi, S.A. David, Potent
Adjuvanticity of a Pure TLR7-Agonistic Imidazoquinoline Dendrimer, PLoS ONE. 7 (2012)
e43612. doi:10.1371/journal.pone.0043612.
[18] C.E. Schiaffo, C. Shi, Z. Xiong, M. Olin, J.R. Ohlfest, C.C. Aldrich, D.M. Ferguson, Structure–
Activity Relationship Analysis of Imidazoquinolines with Toll-like Receptors 7 and 8 Selectivity
and Enhanced Cytokine Induction, J. Med. Chem. 57 (2014) 339–347. doi:10.1021/jm4004957.
[19] Y. Fujita, K. Hirai, K. Nishida, H. Taguchi, 6-(4-Amino-2-butyl-imidazoquinolyl)-norleucine:
Toll-like receptor 7 and 8 agonist amino acid for self-adjuvanting peptide vaccine, Amino Acids.
48 (2016) 1319–1329. doi:10.1007/s00726-016-2190-7.
[20] P. Larson, T.A. Kucaba, Z. Xiong, M. Olin, T.S. Griffith, D.M. Ferguson, Design and Synthesis
of N1-Modified Imidazoquinoline Agonists for Selective Activation of Toll-like Receptors 7 and
8, ACS Med. Chem. Lett. 8 (2017) 1148–1152. doi:10.1021/acsmedchemlett.7b00256.
[21] Y. Isobe, A. Kurimoto, M. Tobe, K. Hashimoto, T. Nakamura, K. Norimura, H. Ogita, H.
Takaku, Synthesis and Biological Evaluation of Novel 9-Substituted-8-Hydroxyadenine
Derivatives as Potent Interferon Inducers, J. Med. Chem. 49 (2006) 2088–2095.
doi:10.1021/jm051089s.
[22] J.J. Weterings, S. Khan, G.J. van der Heden van Noort, C.J.M. Melief, H.S. Overkleeft, S.H. van
der Burg, F. Ossendorp, G.A. van der Marel, D.V. Filippov, 2-Azidoalkoxy-7-hydro-8-
oxoadenine derivatives as TLR7 agonists inducing dendritic cell maturation, Bioorg. Med.
Chem. Lett. 19 (2009) 2249–2251. doi:10.1016/j.bmcl.2009.02.095.
[23] A. Kurimoto, K. Hashimoto, T. Nakamura, K. Norimura, H. Ogita, H. Takaku, R. Bonnert, T.
McInally, H. Wada, Y. Isobe, Synthesis and Biological Evaluation of 8-Oxoadenine Derivatives
as Toll-like Receptor 7 Agonists Introducing the Antedrug Concept, J. Med. Chem. 53 (2010)
2964–2972. doi:10.1021/jm100070n.
[24] H.G. Bazin, Y. Li, J.K. Khalaf, S. Mwakwari, M.T. Livesay, J.T. Evans, D.A. Johnson,
Structural requirements for TLR7-selective signaling by 9-(4-piperidinylalkyl)-8-oxoadenine
derivatives, Bioorg. Med. Chem. Lett. 25 (2015) 1318–1323. doi:10.1016/j.bmcl.2015.01.037.

ACCEPTED MANUSCRIPT
[25] A.J. Smith, Y. Li, H.G. Bazin, J.R. St-Jean, D. Larocque, J.T. Evans, J.R. Baldridge, Evaluation
of novel synthetic TLR7/8 agonists as vaccine adjuvants, Vaccine. 34 (2016) 4304–4312.
doi:10.1016/j.vaccine.2016.06.080.
[26] M. Beesu, A.C.D. Salyer, M.J.H. Brush, K.L. Trautman, J.K. Hill, S.A. David, Identification of
High-Potency Human TLR8 and Dual TLR7/TLR8 Agonists in Pyrimidine-2,4-diamines, J.
Med. Chem. 60 (2017) 2084–2098. doi:10.1021/acs.jmedchem.6b01860.
[27] W. Embrechts, F. Herschke, F. Pauwels, B. Stoops, S. Last, S. Pieters, V. Pande, G. Pille, K.
Amssoms, I. Smyej, D. Dhuyvetter, A. Scholliers, W. Mostmans, K. Van Dijck, B. Van
Schoubroeck, T. Thone, D. De Pooter, G. Fanning, T.H.M. Jonckers, H. Horton, P. Raboisson,
D. McGowan, 2,4-Diaminoquinazolines as Dual Toll-like Receptor (TLR) 7/8 Modulators for
the Treatment of Hepatitis B Virus, J. Med. Chem. 61 (2018) 6236–6246.
doi:10.1021/acs.jmedchem.8b00643.
[28] J. Lee, T.-H. Chuang, V. Redecke, L. She, P.M. Pitha, D.A. Carson, E. Raz, H.B. Cottam,
Molecular basis for the immunostimulatory activity of guanine nucleoside analogs: activation of
Toll-like receptor 7, Proc. Natl. Acad. Sci. 100 (2003) 6646–6651.
[29] D.C. McGowan, F. Herschke, M.D. Khamlichi, M.L. Rosauro, S.M.P. Benedicto, F. Pauwels, B.
Stoops, V. Pande, A. Scholliers, B. Van Schoubroeck, W. Mostmans, K. Van Dijck, T. Thoné,
H. Horton, G. Fanning, T.H.M. Jonckers, P. Raboisson, Design and synthesis of
tetrahydropyridopyrimidine based Toll-Like Receptor (TLR) 7/8 dual agonists, Bioorg. Med.
Chem. Lett. 28 (2018) 3216–3221. doi:10.1016/j.bmcl.2018.08.015.
[30] U. Švajger, Ž. Horvat, D. Knez, P. Rožman, S. Turk, S. Gobec, New antagonists of toll-like
receptor 7 discovered through 3D ligand-based virtual screening, Med. Chem. Res. 24 (2015)
362–371. doi:10.1007/s00044-014-1127-5.
[31] N.M. Shukla, M.R. Kimbrell, S.S. Malladi, S.A. David, Regioisomerism-dependent TLR7
agonism and antagonism in an imidazoquinoline, Bioorg. Med. Chem. Lett. 19 (2009) 2211–
2214. doi:10.1016/j.bmcl.2009.02.100.
[32] S. Lešnik, T. Štular, B. Brus, D. Knez, S. Gobec, D. Janežič, J. Konc, LiSiCA: A Software for
Ligand-Based Virtual Screening and Its Application for the Discovery of Butyrylcholinesterase
Inhibitors, J. Chem. Inf. Model. 55 (2015) 1521–1528. doi:10.1021/acs.jcim.5b00136.
[33] C.A. Lipinski, Drug-like properties and the causes of poor solubility and poor permeability, J.
Pharmacol. Toxicol. Methods. 44 (2000) 235–249.
[34] T. Meng, Y. Zou, O. Khorev, Y. Jin, H. Zhou, Y. Zhang, D. Hu, L. Ma, X. Wang, J. Shen,
Simple and Efficient Copper(I)-Catalyzed Access to Three Versatile Aminocoumarin-Based
Scaffolds using Isocyanoacetate, Adv. Synth. Catal. 353 (2011) 918–924.
doi:10.1002/adsc.201000895.
[35] S. Salentin, S. Schreiber, V.J. Haupt, M.F. Adasme, M. Schroeder, PLIP: fully automated
protein–ligand interaction profiler, Nucleic Acids Res. 43 (2015) W443–W447.
doi:10.1093/nar/gkv315.
[36] J. Konc, D. Janezic, An improved branch and bound algorithm for the maximum clique problem,
Proteins. 4 (2007).
[37] E.J. Barker, D. Buttar, D.A. Cosgrove, E.J. Gardiner, P. Kitts, P. Willett, V.J. Gillet, Scaffold
Hopping Using Clique Detection Applied to Reduced Graphs, J. Chem. Inf. Model. 46 (2006)
503–511. doi:10.1021/ci050347r.
[38] V.J. Gillet, P. Willett, J. Bradshaw, Similarity Searching Using Reduced Graphs †, J. Chem. Inf.
Comput. Sci. 43 (2003) 338–345. doi:10.1021/ci025592e.
[39] E.J. Barker, E.J. Gardiner, V.J. Gillet, P. Kitts, J. Morris, Further Development of Reduced
Graphs for Identifying Bioactive Compounds, J. Chem. Inf. Comput. Sci. 43 (2003) 346–356.
doi:10.1021/ci0255937.
[40] H.P. Beck, T. Kohn, S. Rubenstein, C. Hedberg, R. Schwandner, K. Hasslinger, K. Dai, C. Li, L.
Liang, H. Wesche, B. Frank, S. An, D. Wickramasinghe, J. Jaen, J. Medina, R. Hungate, W.
Shen, Discovery of potent LPA2 (EDG4) antagonists as potential anticancer agents, Bioorg.
Med. Chem. Lett. 18 (2008) 1037–1041. doi:10.1016/j.bmcl.2007.12.024.

ACCEPTED MANUSCRIPT
HIGHLIGHTS
-
-
-
-
-
Successful in-silico definition of novel TLR7-selective compounds.
Three novel scaffolds possessing TLR7 agonist activity identified.
A focused library of selective TLR7 over TLR8 agonists was synthesized.
Most potent compounds 26 and 45 with EC50s of 1.8 µM and 4.4 µM, respectively.
Systematic exploration of structure–activity relationships, docking studies and cytokine secretion
assay.