Antifungal activity, mode of action variability, and subcellular distribution of coumarin-based antifungal azoles

Article abstract of DOI:10.1016/j.ejmech.2019.07.003

Azole antifungals inhibit the biosynthesis of ergosterol, the fungal equivalent of cholesterol in mammalian cells. Here we report an investigation of the activity of coumarin-substituted azole antifungals. Screening against a panel of Candida pathogens, including a mutant lacking CYP51, the target of antifungal azoles, revealed that this enzyme is inhibited by triazole-based antifungals, whereas imidazole-based derivatives have more than one mode of action. The imidazole-bearing antifungals more effectively reduced trailing growth associated with persistence and/or recurrence of fungal infections than triazole-based derivatives. The imidazole derivatives were more toxic to mammalian cells and more potently inhibited the activity of CYP3A4, which is one of the main causes of azole toxicity. Using live cell imaging, we showed that regardless of the type of azole ring fluorescent 7-diethylaminocoumarin-based azoles localized to the endoplasmic reticulum, the organelle that harbors CYP51. This study suggests that the coumarin is a promising scaffold for development of novel azole-based antifungals that effectively localize to the fungal cell endoplasmic reticulum.

Full text of DOI:10.1016/j.ejmech.2019.07.003

Accepted Manuscript
Antifungal activity, mode of action variability, and subcellular distribution of coumarin-
based antifungal azoles
Rebecca Elias, Raphael I. Benhamou, Qais Z. Jaber, Orly Dorot, Sivan Louzoun
Zada, Keren Oved, Edward Pichinuk, Micha Fridman
PII:
S0223-5234(19)30624-5
DOI:
https://doi.org/10.1016/j.ejmech.2019.07.003
EJMECH 11500
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 14 May 2019
Revised Date: 30 June 2019
Accepted Date: 1 July 2019
Please cite this article as: R. Elias, R.I. Benhamou, Q.Z. Jaber, O. Dorot, S.L. Zada, K. Oved, E.
Pichinuk, M. Fridman, Antifungal activity, mode of action variability, and subcellular distribution of
coumarin-based antifungal azoles, European Journal of Medicinal Chemistry (2019), doi: https://
doi.org/10.1016/j.ejmech.2019.07.003.
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
Antifungal Activity, Mode of Action Variability,
and Subcellular Distribution of Coumarin-based
Antifungal Azoles
Rebecca Elias,a Raphael I. Benhamou, a Qais Z. Jaber,a Orly Dorot,b Sivan Louzoun Zada a
Keren Oved,b Edward Pichinuk,b and Micha Fridman a*.
a School of Chemistry, Raymond & Beverly Sackler Faculty of Exact Sciences, Tel Aviv
University, Tel Aviv, 6997801, Israel. b Blavatnik Center for Drug Discovery, Tel Aviv
University, Tel Aviv, 6997801, Israel.
*Correspondence to: mfridman@tauex.tau.ac.il
KEY WORDS: Antifungals; Fluorescent Imaging, Antifungal Azoles, Organelle targeting
ABBREVIATIONS
CYP: Cytochrome P450; CYP3A4: Cytochrome P450 3A4; CYP51: cytochrome P450
lanosterol 14α-demethylase; ER: endoplasmic reticulum; IC50: concentration toxic to 50% of the
enzyme; MIC: minimal inhibitory concentration; MTT: 3-(4,5- dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide; SAR: activity relationship study; ZOI: zone of inhibition.
1

ACCEPTED MANUSCRIPT
2

ACCEPTED MANUSCRIPT
ABSTRACT
Azole antifungals inhibit the biosynthesis of ergosterol, the fungal equivalent of cholesterol in
mammalian cells. Here we report an investigation of the activity of coumarin-substituted azole
antifungals. Screening against a panel of Candida pathogens, including a mutant lacking CYP51,
the target of antifungal azoles, revealed that this enzyme is inhibited by triazole-based
antifungals, whereas imidazole-based derivatives have more than one mode of action. The
imidazole-bearing antifungals more effectively reduced trailing growth associated with
persistence and/or recurrence of fungal infections than triazole-based derivatives. The
imidazole derivatives were more toxic to mammalian cells and more potently inhibited the
activity of CYP3A4, which is one of the main causes of azole toxicity. Using live cell imaging,
we showed that regardless of the type of azole ring fluorescent 7-diethylaminocoumarin-based
azoles localized to the endoplasmic reticulum, the organelle that harbors CYP51. This study
suggests that the coumarin is a promising scaffold for development of novel azole-based
antifungals that effectively localize to the fungal cell endoplasmic reticulum.
3

ACCEPTED MANUSCRIPT
INTRODUCTION
The incidence of severe fungal infections has risen sharply in recent decades, due in large
part to a considerable increase in the number of immunosuppressed and/or immunocompromised
patients.[1,2] Moreover, fungal infections are increasingly being caused by drug-resistant strains
with members of Candida genus, in particular Candida albicans and Candida glabrata, the most
frequently encountered fungal pathogens.[3–7] Thus, antifungal drug development is a high
priority.
Alarmingly, prevention and treatment of fungal infections relies on a very limited number of
drugs belonging to three main classes: the azoles, the polyenes, and 5-fluorocytosine.[8–10] The
azole antifungals are commonly used as first line treatment of both topical and invasive fungal
infections worldwide.[11,12] Members of the azole class inhibit lanosterol 14α-demethylase
(CYP51), a cytochrome P450 that catalyzes a demethylation step in the biosynthesis of the
fungal plasma-membrane sterol ergosterol.[13–15] Notably, an orthologue of the fungal CYP51
is found in mammals, and humans express many cytochrome P450 enzymes.[14] First-
generation azole drugs such as ketoconazole (1), miconazole (2), and clotrimazole (3) can cause
hepatotoxic effects, impairment of gastrointestinal and endocrine functions, and skin irritation.
The second-generation azoles (i.e., itraconazole (4) and fluconazole (5)) and third-generation
drugs (i.e., voriconazole (6) and posaconazole (7)) display markedly improved safety profiles
and better oral bioavailability.[16] Unfortunately, however, all three generations of azole
antifungals inhibit mammalian cytochrome P450s.[17,18] Global increase in cases of resistance
to antifungal azoles and side effects associated with the systemic use of azole antifungals have
driven the exploration for new types of antifungal azoles in recent years.[19–24]
4

ACCEPTED MANUSCRIPT
Figure 1. Structures of clinically used antifungal azoles 1-7, which were used as controls in this
study, and of coumarin- and quinolone-based antifungal azole derivatives 8–21.
In designing novel antifungal azoles, perturbations in the activities of heme proteins,
especially P450 isoforms present in mammalian cells, must be reduced or, ideally, avoided. In
5

ACCEPTED MANUSCRIPT
humans, CYP3A4 makes up the largest fraction of the total CYP content in the liver and
intestines; this enzyme is responsible for the metabolism of multiple small molecules. CYP3A4
is unique from other major CYPs in its exceptionally relaxed substrate selectivity. As a result, it
is estimated that this enzyme is responsible for metabolizing over 50% of the marketed drugs. As
a result, CYP3A4 inhibition is one of the main causes of adverse drug/drug and drug/food
interactions. Unfortunately, all clinically used antifungal azoles inhibit CYP3A4. Second- and
third-generation azole antifungals vary in CYP3A4 inhibition potency: 4 and 7 are more potent
inhibitors than are 5 and 6.[25–27]
CYP51 and a number of other enzymes involved in the biosynthesis of ergosterol localize
primarily to the endoplasmic reticulum (ER).[9,28] By designing inherently fluorescent
antifungal azole derivatives we recently discovered that certain azole antifungals localize mainly
to the mitochondria.[29] We therefore reasoned that these antifungal azoles do not reach their
target in the ER. Based on this hypothesis, we developed antifungal azole 8 by attaching a 7-
(diethyl)-aminocoumarin fluorescent segment to the pharmacophore of the antifungal azole
fluconazole (5) (Figure 1).[30] Azole 8 localizes primarily to the ER in Candida cells, and its
antifungal activity was 4- to 64-fold more potent than that of fluconazole against a panel of
Candida strains.
To explore the potential of coumarin-based azole antifungals we conducted a structure
activity relationship (SAR) investigation of a collection of coumarin-based and 2-quinolone-
based azoles that were systematically altered at four structural motifs. These compounds were
evaluated with respect to antifungal activity, mode of action, subcellular localization, target
selectivity, and toxicity to mammalian cells.
6

ACCEPTED MANUSCRIPT
RESULTS AND DISCUSSION
Synthesis of coumarin- and quinolone-based azole antifungals
To study the SAR of ER-localizing antifungal azole 8, we generated derivatives 9–21 by
altering one or more of four different chemical functionalities of this azole scaffold: the two
halogen atom substituents on the phenyl ring, the azole ring, the hetero-atom (oxygen in the
coumarin scaffold or an NH in the 2-quinolone scaffold), and the substituent at the C7 position
of the coumarin or 2-quinolone (R1-R4, respectively, Scheme 1). Several of the marketed azole
antifungals including 1, 2, and 4 (Figure 1) are racemic mixtures. Computer simulated docking
of the R and S enantiomers of aminocoumarin-based azole 8 revealed that the interactions with
cytochrome P450DM take place through a similar binding mode;[29] therefore, the azole
antifungals in this study were synthesized and evaluated as racemic mixtures.
Briefly, nucleophilic displacement of the chloride atom of commercially available α-chloro-
2,4-di-halo-acetophenones 1a and 1b by an imidazole or a triazole ring gave the corresponding
azole intermediate compounds 2a-d (Scheme 1). Conversion of the ketones of 2a-d to the
corresponding racemic mixtures of epoxides using trimethylsulfoxonium iodide in the presence
of a base gave intermediates 3a-d, which were then reacted with sodium azide to afford
intermediates 4a-d. Catalytic hydrogenation of 4a-d gave free amine intermediates 5a-d, which
were coupled to the free carboxylic acid functionality of the coumarins and 2-quinolones to yield
antifungal azoles 8-21 (Figure 1). Antifungal azoles 9-21 were characterized by 1H and 13C NMR
19
as well as by F NMR when applicable and by HRMS; purities were ≥95% as determined by
HPLC (see Experimental section and Supplementary Material for NMR spectra and HPLC
traces).
7

ACCEPTED MANUSCRIPT
Scheme 1. Synthesis of coumarin- and 2-quinolone-based antifungal azoles. Reagents and
conditions: a) NaHCO3, toluene, reflux, 4 h; b) trimethylsulfoxonium iodide, 20% NaOH,
toluene, 60 °C, 4 h; c) NaN3, DMF, 65 °C, 12 h; d) H2, palladium on carbon, isopropanol, room
temperature overnight; e) HATU, DIPEA, DMF, room temperature overnight.
Coumarin-based azoles display potent anti-Candida activity
To test antifungal activity we chose 10 strains representing different species of Candida, the
most common human fungal pathogenic genus, which cause both superficial and systemic
infections (Table 1).[31] The chosen panel included C. albicans (Strains A-C), C. glabrata
(Strains D-F), C. parapsilosis (Strains G, H), C. guilliermondii (Strain I) and C. dubliniensis
(Strain J). To evaluate antifungal activity we determined the minimal inhibitory concentration
(MIC) values for each of the azoles using the broth microdilution method over a concentration
range of 0.007–64 µg/mL. The antifungal activities of azoles 8-21 and of seven clinically used
azoles including first-generation (1, 2, and 3), second-generation (4 and 5), and third-generation (6
and 7) were evaluated (Table 1).
8

ACCEPTED MANUSCRIPT
Table 1. Minimal inhibitory concentration (MIC) values [µg/mL].
Yeast Strains a
A
B
C
D
E
F
G
H
I
J
Comp. #
0.007
0.007
0.007
0.007
0.5
0.007
0.007
0.007
0.007
0.25
0.007
0.007
0.06
0.06
0.03
0.03
0.125
0.015
0.03
0.06
0.125
1
0.007
0.007
0.007
0.007
0.25
0.007
0.007
0.03
0.06
0.03
0.03
0.25
0.03
0.06
0.06
0.25
0.5
1
1
0.06
0.06
0.06
0.125
8
0.007
0.06
0.007
0.007
1
0.007
0.003
0.007
0.03
2
0.007
0.06
0.007
0.03
4
0.007
0.06
0.007
0.06
4
1
2
1
2
4
0.5
1
3
1
4
64
0.5
1
64
2
5
0.007
0.007
0.06
0.015
0.03
0.015
0.125
0.03
0.06
0.03
0.25
2
0.125
0.125
8
0.015
0.015
0.125
0.015
0.06
0.015
0.5
0.03
0.03
1
0.06
0.03
1
0.06
0.03
1
6
1
7
4
16
8
8
8
2
0.125
0.5
0.06
2
0.5
0.25
0.125
2
0.5
0.5
0.25
2
9
4
16
2
4
10
11
12
13
14
15
16
17
18
19
20
21
4
0.25
4
>64
2
32
4
1
0.125
0.25
0.125
0.25
2
0.25
0.5
0.5
2
0.125
0.5
0.25
1
0.25
0.5
0.25
0.5
8
>64
2
>64
2
2
0.5
1
1
4
>64
32
>64
64
>64
64
8
16
8
8
0.25
1
0.125
2
0.5
4
0.25
2
2
1
4
4
>64
>64
64
>64
16
>64
8
>64
8
>64
8
1
1
0.5
1
64
64
>64
>64
>64
>64
>64
>64
[a] A = C. albicans 90028; B = C. albicans P-87; C = C. albicans SN152; D = C. glabrata 66032; E = C.
glabrata 2001; F = C. glabrata 192; G = C. parapsilosis 90018; H = C. parapsilosis 22019; I = C.
guilliermondii T-47; J = C. dubliniensis T-99. MIC values were determined using the broth double-
dilution method starting from a concentration of 64 µg/mL. Cells were grown in Casitone medium at 37 °C
under CO2 for 24 h. Each concentration was tested in triplicate, and results were confirmed in two
independent sets of experiments.
9

ACCEPTED MANUSCRIPT
Displacement of the 1,2,4-triazole ring with an imidazole enhanced antifungal activity.
Compared to the 1,2,4-triazole-based antifungal azoles 8, 10, 12, and 14, the MIC values of the
corresponding imidazole-based compounds 9, 11, 13, and 15, respectively, were 2- to 32-fold
lower against the majority of the tested strains and were equally potent against the rest. The
coumarin-based azoles 8-11, which contain a 7-diethylaminocoumarin, displayed the most potent
and broad spectrum antifungal activity.
Comparison between the MIC values of azole 12 and azole 18, differing solely by the C7
substituent (R4, Scheme 1) of the coumarin segment, revealed that the alkyl group on the ether
did not significantly impact either spectrum of activity or potency. Interestingly, the 7-hydroxy
coumarin-based azole 21 did not inhibit the growth of any strain in the panel (MIC ≥ 64 µg/mL,
Table 1). Although certain differences in the structure of the alkyl ether at the C7-position of the
coumarin are tolerated, the presence of an alcohol at this position completely abrogates
antifungal activity. This may be explained by the incompatibility of the hydrophilic alcohol in
the highly hydrophobic pocket at the catalytic domain of CYP51.[32,33] Of the 14 coumarin-
and 2-quinolone-based azoles, 7-diethylaminocoumarin-based azoles 10 and 11 and 7-(1-
ethylpropoxy)-coumarin based azoles 13 and 15 displayed the best activity against the non-
albicans Candida strains tested.
Finally, similar to the third-generation azoles 6 and 7, 7-diethylaminocoumarin-based azoles
8-11 and 7-(1-ethylpropoxy)-coumarin-based azoles 12-15 displayed potent activity against the
tested C. albicans strains with MIC values of 0.007-0.125 µg/mL. These derivatives had better
antifungal activity than the second-generation antifungal azole 5, which, due to low cost, is most
commonly used for treatment of fungal infections worldwide.
10

ACCEPTED MANUSCRIPT
Imidazole-based derivatives inhibit CYP51 as well as other critical fungal processes
It was previously reported that, in addition to inhibition of CYP51-mediated ergosterol
biosynthesis, clinically used azole antifungals such as 1 and 2 (Figure 1) inhibit the activity of a
variety of additional targets including the activities of several membrane-bound enzymes.[8] To
identify the structural motifs associated with selective inhibition of CYP51 and those that
contribute to other antifungal effects, we compared the antifungal activities of representative
coumarin-based azoles in a mutant strain lacking copies of both the ERG11 and ERG3 genes.
ERG11 encodes CYP51, the target of antifungal azoles, which is essential for fungal cell growth
under aerobic conditions when the C-5 sterol desaturase-encoding ERG3 is expressed.[34,35]
The erg11∆∆ erg3∆∆ mutant strain is therefore viable despite the absence of CYP51.
The effects of 1,2,4-triazole-based azoles 8 and 10 and imidazole-based azoles 9 and 13 and
control drugs were analyzed using the disk diffusion assay against the double knockout strain
and the parental strain SN152 (Figure 2). A zone of growth inhibition was visible after 24 hours
in plates seeded with the parental strain and treated with disks containing 1,2,4-triazole-based
azoles 5, 6, 8, or 10; these compounds did not inhibit growth of the erg11∆∆ erg3∆∆ mutant,
however (Figure 2). These results indicate that CYP51 is the main target of the 1,2,4-triazole-
based azole antifungals tested. The imidazole-based azoles (1, 2, 9, and 13) inhibited growth of
both the parental strain and of the mutant (Figure 2). Thus, the imidazole-based azoles affect
more than one process in the fungal cells.
11

ACCEPTED MANUSCRIPT
Figure 2. Activity of azole antifungals against CYP51-deficient yeast. C. albicans strains
SN152 (parental) and the erg11∆∆ erg3∆∆ mutant were plated onto casitone plates and allowed
to grow in the presence of a disk containing 25 µg of the indicated compound at 30 °C for 24 h.
Photographs of representative plates are shown.
12

ACCEPTED MANUSCRIPT
Imidazole-based compounds reduce drug tolerance
Fungal susceptibility is commonly determined quantitatively in vitro by analysis of MIC
values after 24ꢀh of growth in the presence of the tested antimicrobial agent. However, MIC
measurements do not measure residual growth, also termed tolerance or trailing.[36,37]
Tolerance occurs when subpopulations of fungal cells multiply despite prolonged exposure to
an antifungal drug, and this property can contribute to the persistence and/or recurrence of
fungal infections. An increase in MIC value after exposure to a drug is indicative of tolerance.
To evaluate the level of tolerance, we compared MIC values after 24 hours of incubation to those
measured after 48 hours against two C. albicans and two C. glabrata strains (Table 2).
Table 2. MIC values [µg/mL] at 24 and 48 hours.a
Yeast strain
Compound
C. albicans 90028
C. albicans SN152 C. glabrata 66032 C. glabrata 192
24 hours 48 hours 24 hours 48 hours 24 hours 48 hours 24 hours 48 hours
0.007
0.007
0.007
0.007
0.5
0.007
0.015
0.007
0.007
0.5
0.007
0.007
0.007
0.007
0.25
0.007
0.015
0.007
0.007
1
1
4
0.06
0.06
0.06
0.125
8
0.125
0.125
0.25
0.125
8
1
2
4
2
4
4
3
1
>64
>64
4
4
64
0.5
1
5
0.007
0.007
0.06
0.007
0.007
0.125
0.06
0.06
0.03
0.5
0.007
0.007
0.03
0.007
0.007
0.125
0.125
0.06
0.06
0.5
0.125
0.125
8
0.25
0.125
>64
4
6
>64
>64
32
7
4
8
0.015
0.03
0.06
8
2
9
0.03
4
>64
>64
>64
4
4
>64
1
10
11
12
13
14
15
0.015
0.125
0.03
0.03
4
0.25
4
0.25
>64
2
8
0.06
0.06
0.06
0.03
0.06
0.06
0.06
1
2
0.06
0.06
>64
2
>64
>64
2
>64
1
0.03
0.06
0.5
[a] MIC values were determined using the broth double-dilution method staring from a concentration of
64 g/mL. Cells were grown in Casitone medium at 37 °C under CO2. Each concentration was tested in
triplicate, and results were confirmed in two independent sets of experiments.
µ
13

ACCEPTED MANUSCRIPT
Tolerance was more significant for the C. glabrata strains than for the C. albicans strains.
We observed higher MIC values for all tested azoles for C. glabrata ATCC 66032 at 48 hours
than at 24 hours (Table 2). Interestingly, of the clinically used azoles, the imidazole-based drugs
1, 2, and 3 maintained better efficacy over time with 48-hour MIC values no more than two
double dilutions higher than 24-hour MIC values against C. glabrata ATCC 66032 (Table 2). In
contrast, significant trailing growth of C. glabrata ATCC 66032 was observed for the triazole-
based compounds 4, 5, 6, and 7 (Table 2). For example, the 48-hour MIC of 7 was 64-fold higher
than the 24-hour MIC value against this strain.
The coumarin-based azoles 8-15 are pairs of compounds that vary solely in the azole ring.
Although MIC values measured after 48 hours were generally higher than those at 24 hours,
there was less tolerance in samples treated with imidazoles 9, 11, 13, and 15 than in samples
treated with triazoles 8, 10, 12, and 14. Since we showed that the antifungal activity of the
imidazoles is not based solely on the inhibition of CYP51, the reduction in trailing growth likely
results from the non-CYP51-targeted activity of the imidazole-based compounds.
Imidazole ring leads to more potent inhibition of human CYP3A4
Inhibition of human CYP3A4 is an undesired activity that is associated with the toxicity of
antifungal azoles.[18] The ability of the coumarin-based azoles to inhibit CYP3A4 was
determined in an assay with recombinant human CYP3A4.[38] The results are summarized in
Table 3. Of the clinically used azoles tested, imidazoles 1 and 3 potently inhibited CYP3A4 with
3 displaying the most potent enzymatic inhibition among all antifungals tested in this study
(Table 3). Of the clinically used triazoles tested, 5 was the poorest CYP3A4 inhibitor, whereas
the third-generation azole 7 potently inhibited the enzyme. Similar to the effects of the imidazole
14

ACCEPTED MANUSCRIPT
ring on antifungal activity, comparison between the inhibition of CYP3A4 by the triazole
derivatives of coumarin-based azoles 8, 10, 12 and 14 and their corresponding imidazole-based
compounds 9, 11, 13 and 15 revealed that the imidazole ring leads to more potent inhibition of
this CYP. An exception was observed for aminocoumarin-based triazole 8 and its imidazole
derivative 9, which had similar IC50 values (~1
µM, Table 3).
Table 3. CYP3A4 inhibition by azole derivatives.a
Compound
IC50 [µM]
0.1826 ± 0.0337
0.0670 ± 0.0089
23.9698 ± 4.8783
1
3
5
≥
2.5b
6
0.6721 ± 0.1696
1.0223 ± 0.1403
1.3123 ± 0.2440
0.3512 ± 0.0426
0.1272 ± 0.0104
7
8
9
10
11
12
13
14
15
≥
2.5b
0.3320 ± 0.0482
2.5b
0.2515 ± 0.0558
≥
[a] Inhibition of CYP3A4 was quantified in an assay with recombinant
CYP3A4. Experiments were performed in duplicate at 37 °C. [b] The
highest tested concentration due to solubility limitations under the assay
conditions.
These results suggest that, in general, imidazole-based antifungal azoles are more effective
inhibitors of CYP3A4 and are, therefore, less CYP51-selective than the corresponding triazole-
based antifungal azoles. These results also suggest that the residual antifungal activity of the
15

ACCEPTED MANUSCRIPT
coumarin-based azoles bearing an imidazole-ring in the strain lacking CYP51 may stem from
non-selective inhibition of other heme-containing monooxygenases in fungal cells.
Imidazole ring is associated with toxicity to mammalian cells
We next quantified the effects of the antifungal azoles on the viability mammalian HepG2
cells, an immortalized human liver hepatocellular carcinoma cell line, and HEK-293T cells, an
immortalized human embryonic kidney cell line. We studied coumarin-based azoles 10 and 14,
which bear 1,2,4-triazole rings, and the corresponding imidazole analogues 11 and 15,
respectively, as well as 1, 5, 6, and 7. The cells were treated with each of the azoles at
concentrations ranging from 5 to 50 µM for 24 h, after which cell viability was evaluated using
the 3-(4,5- dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The maximal
tested concentration of 50 µM was determined based on solubility limitations of some of the
coumarin-based azoles and control drugs. Notably, at the maximal concentration tested, none of
azole antifungals reduced viability of the mammalian cells tested by more than 40% (Figure 3).
Of the clinically used azoles, 1 and 7 were the most toxic, and 5 had the smallest effect on
cell viability (Figure 3). The clinically used azoles 5, 6, and 7 share the same pharmacophore
composed of a 1,2,4-triazole and difluorophenyl ring. The additional segments attached to the
pharmacophore in these clinically used antifungals are distinctly different (Figure 1). The fact
that 7 was more toxic to the tested mammalian cells than was 5 indicates that although the
imidazole ring has a major effect on mammalian cell toxicity the moieties attached to the
pharmacophore moderate cell toxicity. Of coumarin-based azoles tested, 7-
diethylaminocoumarin-based imidazole 11 was the most toxic to both tested cells lines, and the
corresponding triazole 10 was the least toxic; the differences in the toxicity of the 7-(1-
ethylpropoxy)-coumarin-based azoles 14 and 15 were less pronounced (Figure 3).
16

ACCEPTED MANUSCRIPT
Figure 3. Viability of HEK-293T and HepG2 cells in the presence of azole antifungals. Cells
were allowed to grow in the presence of indicated concentration of azole derivative for 24 h at
37ꢀ°C. Viability is plotted as percent of vehicle-treated control cells. Experiments were
performed in triplicate, and the results were obtained from two independent experiments.
Subcellular distribution in live Candida cells is not affected by the type of azole ring
We previously demonstrated the use of inherently fluorescent antifungal azoles for
determination of the subcellular distribution of the molecule in live Candida cells during the first
few hours after exposure to the azole.[29,30,39] Notably, the subcellular distributions of the free
17

ACCEPTED MANUSCRIPT
fluorescent dyes differed from those of the corresponding fluorescent azole antifungals
suggesting that the staining pattern is not determined by the dye segment alone.
Since there are distinctive differences in biological activities of antifungal azoles depending
on the type of the azole ring, we asked if the azole ring affects subcellular distribution by
analyzing the subcellular localization of azoles 8 and 9, which differ in their azole ring (1,2,4-
triazole and imidazole, respectively) contain the 7-diethylaminocoumarin fluorophore that
proved suitable for live cell imaging experiments in Candida yeast cells. Subcellular localization
was evaluated in Candida albicans SN152 and in Candida glabrata 2001 yeast cells. A distinct
and typical circular ER structure around the nuclear envelope with peripheral cortical extensions
[39] was clearly visible in both C. glabrata and C. albicans yeast cells stained with the azoles
(Figure 4). Previously, we showed that azole antifungals with different fluorophores including
dansyl and cyanine 5 can localize to different organelles.[29] However, staining patterns were
similar for azoles 8 and 9. Although the activities of these two azoles varied significantly in most
of the assays in this study, the results of the fluorescent imaging experiments clearly indicated
that the type of azole ring did not affect the subcellular distribution.
18

ACCEPTED MANUSCRIPT
Figure 4. Fluorescent imaging of azoles 8 and 9 in live Candida cells. A-B) C. albicans
SN152 cells incubated for 60 min with azole 8 (cyan, 10 µM), C-D) C. albicans SN152 cells
incubated for 60 min with azole 9 (yellow, 10 µM), E-F) C. glabrata 2001 cells incubated for 60
min with azole 8 (cyan, 10 µM), and G-H) C. glabrata 2001 cells incubated for 60 min with
azole 9 (yellow, 10 µM). For each azole and yeast cell strain, the bright field image is shown on
the left and the fluorescent image is on the right. The bandpass filters used to image the
fluorescent azoles were ex: 427 nm and em: 510 nm. Scale bars, 5 µm.
CONCLUSIONS
A series of 11 coumarin-based and three 2-quinolone-based azoles were synthesized by
systematically altering four structural motifs. From analyses of biological activities of these
compounds, we were able to generate high-resolution structure activity relationship insights. The
imidazole-containing antifungal azoles had higher efficacy against a diverse panel of Candida
pathogens than their corresponding 1,2,4-triazole analogues. Unlike the MIC values of the
19

ACCEPTED MANUSCRIPT
triazoles, the MIC values of the imidazole-bearing compounds were more similar at 24 and 48
hours, indicating that the yeast strains are less tolerant to the imidazole-based azoles. Since drug
tolerance is associated with persistence and reoccurrence of infection, this feature is of particular
importance. The antifungal activity of the triazole-based azoles was dependent on expression of
CYP51, the target of the azole antifungals. In contrast, imidazole-based compounds displayed
antifungal activity against a mutant lacking CYP51 indicating that imidazole-based azole
antifungals have additional modes of action. The imidazole-based compounds were more toxic to
mammalian cells and more potently inhibited CYP3A4 than the triazole-based antifungals.
CYP3A4 is one of the most abundant human cytochrome P450s, a family of enzymes that are
important for the metabolism of endogenous and exogenous compounds.
By taking advantage of the suitability of the 7-diethylaminocoumarin fluorescence for live-
cell imaging, we showed that although the type of the azole ring affects the antifungal activity,
selectivity, and mode of action, it does not influence the subcellular distribution of the azole
compounds in yeast cells. The two derivatives evaluated were identical except that one had a
triazole and the other an imidazole ring; both localized to the ER. This is of particular
importance since we recently reported that by targeting the drug to the ER, which harbors
CYP51, the antifungal activity of an azole antifungal can be enhanced. To conclude, this study
suggests that coumarin residues are promising scaffolds for development of novel azole-based
antifungals that effectively localize to the endoplasmic reticulum.
20

ACCEPTED MANUSCRIPT
EXPERIMENTAL SECTION
Chemistry
1
General methods and instrumentation: H-NMR spectra were recorded on BrukerAvance™
400 or 500 MHz spectrometers, and chemical shifts (reported in ppm) were calibrated to CDCl3
(d = 7.26) when CDCl3 was the solvent, to CD3OD (d = 3.31) when CD3OD was the solvent and
to DMSO (d = 2.50) when DMSO was the solvent. 13C-NMR spectra were recorded on
BrukerAvance™ 400 or 500 MHz spectrometers at 100 or 125 MHz. 19F-NMR spectra were
recorded on BrukerAvance™ 400 or 500 MHz spectrometers at 375 or 470 MHz. Multiplicities
are reported using the following abbreviations: s = singlet, d = doublet, t = triplet, q = quartet, dd
= doublet of doublets, ddd = doublet of doublet of doublets, td = triplet of doublets, quin =
quintet, m = multiplet. Coupling constants (J) are given in Hertz. High-resolution electrospray
ionization (HRESI) mass spectra were measured on a Waters Synapt instrument. Low-resolution
electrospray ionization mass spectra (ESI-MS) were measured on a Waters 3100 mass detector.
Chemical reactions were monitored by TLC (Merck, Silica gel 60 F254). Visualization was
achieved using a cerium molybdate stain (5 g (NH4)2Ce(NO3)6, 120 g (NH4)6Mo7O24·4H2O, 80
mL H2SO4, 720 mL H2O) or with UV lamp. All chemicals, unless otherwise stated, were
obtained from commercial sources. Compounds were purified using Geduran® Si 60
chromatography (Merck). The preparative reverse-phase high-pressure liquid chromatography
(RP-HPLC) system used was an ECOM system equipped with a 5-µm, C-18 Phenomenex Luna
Axia column (250 mm x 21.2 mm). The flow rate was 15 mL/min. Solvent A was H2O, solvent
B was acetonitrile. Analytical RP-HPLC was performed on a VWR Hitachi instrument equipped
with a diode array detector and an Alltech Apollo C18 reversed-phase column (5 m, 4.6 x 250
21

ACCEPTED MANUSCRIPT
mm). The flow rate was 1 mL/min. Solvent A was 0.1% TFA in H2O, solvent B was acetonitrile.
The SpectraMax i3x Platform spectrophotometer from Molecular Devices was used for the
measurement of the excitation and absorbance spectra of the fluorescent probes.
Compound 9. 7-Diethylaminocoumarin-3-carboxylic acid[40] (61 mg, 0.23 mmol) was
dissolved in dry DMF (5 mL) under argon, treated with HATU (149 mg, 0.39 mmol) and
triethylamine (0.11 mL, 0.78 mmol), then stirred for 10 min at 0 °C. To the reaction mixture, 5c
(50 mg, 0.19 mmol) was added, and the solution was stirred at room temperature for 24 h. The
reaction was monitored by TLC (MeOH/DCM, 1:9). Upon completion, the product was
extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated to give the
crude product. The product was isolated by column chromatography on SiO2 using a gradient of
MeOH/DCM as eluent to afford compound 9 (70 mg, 71%). HRESI-MS m/z calculated for
1
C26H27F2N4O4, 497.2000; found for [M+H]+, 497.1996. H NMR (400 MHz, CD3OD) δ 8.55 –
8.51 (m, H-11, 1H), 7.53 – 7.43 (m, H-1, H-4, H-12, 3H), 6.99 – 6.91 (m, H-6, 1H), 6.90 (s, H-2,
1H), 6.84 (td, J = 8.4, 2.4 Hz, H-5, 1H), 6.80 – 6.73 (m, H-3, H-13, 2H), 6.51 (d, J = 3.2 Hz, H-
14, 1H), 4.53 (d, J = 14.5 Hz, H-7, 1H), 4.38 (d, J = 14.4 Hz, H-7, 1H), 3.99 (d, J = 14.0 Hz, H-
8, 1H), 3.89 (d, J = 14.0 Hz, H-8, 1H), 3.50 (q, J = 7.1 Hz, H-15, 4H), 1.21 (t, J = 7.0 Hz, H-16,
1
3
6H). 13C NMR (100 MHz, CD3OD) δ 164.84, 162.97 (dd, JC-F=234.1 Hz, JC-F=12.5 Hz),
1
3
162.54, 159.09 (dd, JC-F=244.9 Hz, JC-F=12.0 Hz), 157.78, 153.32, 148.08, 137.95, 131.29,
130.42, 126.59, 124.18, 120.70, 110.84, 110.30, 108.07, 103.52, 95.85, 75.28, 53.43, 44.58,
11.26. 19F NMR (375 MHz, CD3OD) δ -110.14 (m, Fpara), -113.29 (m, Fortho).
Compound 10. 7-Diethylaminocoumarin-3-carboxylic acid[40] (38 mg, 0.15 mmol) was
dissolved in dry DMF (4 mL) under argon, treated with HATU (93 mg, 0.25 mmol) and
triethylamine (0.07 mL, 0.49 mmol), then stirred for 10 min at 0 °C. To the reaction mixture, 5b
22

ACCEPTED MANUSCRIPT
(35 mg, 0.12 mmol) was added and the solution was stirred at room temperature for 24 h. The
reaction was monitored by TLC (petroleum ether/ethyl acetate, 2:8). Upon completion, the
product was extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated
to give the crude product. The product was isolated by column chromatography on SiO2 using a
gradient of petroleum ether/ethyl acetate as eluent to afford compound 10 (65 mg, 100%).
1
HRESI-MS m/z calculated for C25H25Cl2N5O4Na, 552.1181; found for [M+Na]+, 552.1174. H
NMR (500 MHz, CDCl3) δ 9.17 (t, J = 6.0 Hz, H-9, 1H), 8.55 (s, H-10, 1H), 8.19 (s, H-2, 1H),
7.81 (s, H-1, 1H), 7.76 (d, J = 8.6 Hz, H-3, 1H), 7.40 – 7.33 (m, H-5, H-11, 2H), 7.15 (dd, J =
8.6, 2.1 Hz, H-4, 1H), 6.69 – 6.57 (m, H-8, H-12, 2H), 6.45 (d, J = 2.3 Hz, H-13, 1H), 4.79 (s, H-
6, 2H), 4.23 (dd, J = 14.6, 6.3 Hz, H-7, 1H), 3.97 (dd, J = 14.6, 5.9 Hz, H-7, 1H), 3.44 (q, J = 7.1
13
Hz, H-14, 4H), 1.22 (t, J = 7.1 Hz, H-15, 6H). C NMR (125 MHz, CDCl3) δ 166.41, 162.35,
157.88, 153.02, 151.25, 148.49, 144.74, 137.08, 134.65, 131.17, 130.84, 130.57, 129.95, 129.60,
127.53, 127.28, 110.19, 108.26, 96.58, 77.93, 55.02, 47.27, 45.16, 12.40.
Compound 11. 7-Diethylaminocoumarin-3-carboxylic acid[40] (54 mg, 0.21 mmol) was
dissolved in dry DMF (5 mL) under argon, treated with HATU (132 mg, 0.35 mmol) and
triethylamine (0.1 mL, 0.71 mmol), then stirred for 10 min at 0 °C. To the reaction mixture, 5d
(50 mg, 0.18 mmol) was added, and the solution was stirred at room temperature for 24 h. The
reaction was monitored by TLC (MeOH/DCM, 1:9). Upon completion, the product was
extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated to give the
crude product. The product was isolated by column chromatography on SiO2 using a gradient of
MeOH/DCM as eluent to afford compound 11 (88 mg, 95%). HRESI-MS m/z calculated for
C26H27Cl2N4O4, 529.1409; found for [M+H]+, 529.1413. 1H NMR (400 MHz, CD3OD) δ 8.53 (s,
H-11, 1H), 7.64 (d, J = 8.7 Hz, H-12, 1H), 7.52 – 7.45 (m, H-1, H-4, H-6, 3H), 7.19 (dd, J = 8.7,
23

ACCEPTED MANUSCRIPT
2.2 Hz, H-5, 1H), 6.90 (s, H-2, 1H), 6.81 – 6.71 (m, H-3, H-13, 2H), 6.51 (d, J = 2.1 Hz, H-14,
1H), 4.91 (d, J = 14.5 Hz, H-7, 1H), 4.59 (s, H-9, 1H), 4.42 (d, J = 14.5 Hz, H-7, 1H), 4.17 (q, J
= 14.2 Hz, H-8, 2H), 3.49 (q, J = 7.1 Hz, , H-15, 4H), 1.20 (t, J = 7.1 Hz, H-16, 6H). 13C NMR
(100 MHz, CD3OD) δ 165.17, 162.72, 157.94, 153.50, 148.26, 138.03, 137.13, 134.43, 131.47,
131.08, 130.15, 129.50, 127.08, 126.63, 120.74, 110.47, 108.14, 96.01, 76.96, 52.21, 45.71,
44.74, 11.42.
Compound 12. 7-(1-Ethylpropoxy)-coumarin-3-carboxylic acid (29 mg, 0.10 mmol) was
dissolved in dry DMF (1 mL) under argon, treated with HATU (74 mg, 0.19 mmol) and N,N-
diisopropylethylamine (0.06 mL, 0.34 mmol), then stirred for 10 min at 0 °C. To the reaction
mixture, 5a[41] (22 mg, 0.09 mmol) was added, and the solution was stirred at room temperature
for 24 h. The reaction was monitored by TLC (MeOH/DCM, 1:9). Upon completion, the product
was extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated to give
the crude product. The product was isolated by column chromatography on SiO2 using a gradient
of MeOH/DCM as eluent to afford compound 12 (33 mg, 72%). HRESI-MS m/z calculated for
C26H27F2N4O5, 513.1950; found for [M+H]+, 513.1951. 1H NMR (500 MHz, CD3OD) δ 8.71 (s,
H-10, 1H), 8.35 (s, H-2, 1H), 7.78 (s, H-1, 1H), 7.69 (d, J = 8.8 Hz, H-11, 1H), 7.51 (td, J = 9.0,
6.6 Hz, H-3, 1H), 7.01 – 6.91 (m, H-5, H-12, H-13, 3H), 6.85 (td, J = 8.4, 2.4 Hz, H-4, 1H), 4.78
(d, J = 14.4 Hz, H-6, 1H), 4.68 (d, J = 14.4 Hz, H-6, 1H), 4.41 (quin, J = 5.8 Hz, H-14, 1H), 4.02
(d, J = 13.9 Hz, H-7, 1H), 3.95 (d, J = 14.1 Hz, H-7, 1H), 1.78 – 1.67 (m, H-15, 4H), 0.96 (t, J =
7.4 Hz, H-16, 6H). 13C NMR (125 MHz, CD3OD) δ 164.81, 163.69, 163.00 (dd, 1JC-F=246.6 Hz,
1
3
3JC-F=12.2 Hz), 161.54, 159.43 (dd, JC-F=245.6 Hz, JC-F=11.9 Hz), 156.88, 150.03, 148.37,
144.75, 131.35, 130.00, 123.94, 114.69, 113.22, 111.87, 110.75, 103.65, 101.19, 80.94, 74.70,
55.59, 46.13, 25.65, 8.29. 19F NMR (470 MHz, CD3OD) δ -109.26 (m, Fpara), -112.84 (m, Fortho).
24

ACCEPTED MANUSCRIPT
Compound 13. 7-(1-Ethylpropoxy)-coumarin-3-carboxylic acid (33 mg, 0.12 mmol) was
dissolved in dry DMF (1 mL) under argon, treated with HATU (78 mg, 0.21 mmol) and N,N-
diisopropylethylamine (0.07 mL, 0.40 mmol), then stirred for 10 min at 0 °C. To the reaction
mixture, 5c (29 mg, 0.11 mmol) was added, and the solution was stirred at room temperature for
24 h. The reaction was monitored by TLC (MeOH/DCM, 1:9). Upon completion, the product
was extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated to give
the crude product. The product was isolated by column chromatography on SiO2 using a gradient
of MeOH/DCM as eluent to afford compound 13 (34 mg, 60%). HRESI-MS m/z calculated for
C27H28F2N3O5, 512.1997; found for [M+H]+, 512.1995. 1H NMR (400 MHz, CD3OD) δ 8.76 (s,
H-11, 1H), 7.73 (d, J = 8.8 Hz, H-12, 1H), 7.57 – 7.47 (m, H-1, H-4, 2H), 7.05 – 6.98 (m, H-6,
H-13, 2H), 6.97 (d, J = 2.3 Hz, H-14, 1H), 6.94 (s, H-2, 1H), 6.88 (td, J = 8.4, 2.4 Hz, H-5, 1H),
6.79 (s, H-3, 1H), 4.58 (d, J = 14.6 Hz, H-7, 1H), 4.44 (m, H-7, H-15, 2H), 4.04 (d, J = 14.0 Hz,
H-8, 1H), 3.97 (d, J = 14.1 Hz, H-8, 1H), 1.82 – 1.69 (m, H-16, 4H), 1.00 (t, J = 7.4 Hz, H-17,
13
1
3
6H). C NMR (100 MHz, CD3OD) δ 164.82, 163.70, 163.00 (dd, JC-F=246.7 Hz, JC-F=12.2
1
3
Hz), 159.15 (dd, JC-F=247.1 Hz, JC-F=11.9 Hz), 156.87, 148.35, 137.96, 131.33, 130.48,
126.67, 120.67, 114.68, 113.27, 111.89, 110.85, 103.54, 101.22, 80.97, 75.15, 53.35, 46.36,
25.65, 8.26. 19F NMR (375 MHz, CD3OD) δ -110.11 (m, Fpara), -113.24 (m, Fortho).
Compound 14. 7-(1-Ethylpropoxy)-coumarin-3-carboxylic acid (34 mg, 0.12 mmol) was
dissolved in dry DMF (1 mL) under argon, treated with HATU (83 mg, 0.22 mmol) and N,N-
diisopropylethylamine (0.07 mL, 0.40 mmol), then stirred for 10 min at 0 °C. To the reaction
mixture, 5b (34 mg, 0.12 mmol) was added, and the solution was stirred at room temperature for
24 h. The reaction was monitored by TLC (MeOH/DCM, 1:9). Upon completion, the product
was extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated to give
25

ACCEPTED MANUSCRIPT
the crude product. The product was isolated by column chromatography on SiO2 using a gradient
of MeOH/DCM as eluent to afford compound 14 (48 mg, 73%). HRESI-MS m/z calculated for
C26H27Cl2N4O5, 545.1359; found for [M+H]+, 545.1355. 1H NMR (400 MHz, CD3OD) δ 8.73 (s,
H-10, 1H), 8.39 (s, H-2, 1H), 7.81 (s, H-1, 1H), 7.69 (t, J = 8.5 Hz, H-3, H-11 2H), 7.49 (d, J =
2.2 Hz, H-5, 1H), 7.25 (dd, J = 8.6, 2.2 Hz, H-4 1H), 7.00 (dd, J = 8.7, 2.3 Hz, H-12, 1H), 6.95
(d, J = 2.2 Hz, H-13, 1H), 5.12 (d, J = 14.4 Hz, H-6, 1H), 4.82 (d, J = 14.4 Hz, H-6, 1H), 4.43
(quin, J = 5.8 Hz, H-14, 1H), 4.28 (d, J = 14.2 Hz, H-7, 1H), 4.20 (d, J = 14.2 Hz, H-7, 1H), 1.81
13
– 1.69 (m, H-15, 4H), 0.99 (t, J = 7.4 Hz, H-16, 6H). C NMR (125 MHz, CD3OD) δ 164.83,
163.86, 161.55, 156.90, 150.02, 148.40, 144.83, 136.85, 134.33, 131.40, 130.83, 130.25, 126.89,
114.69, 113.12, 111.87, 101.19, 80.93, 76.17, 54.28, 45.33, 25.64, 8.28.
Compound 15. 7-(1-Ethylpropoxy)-coumarin-3-carboxylic acid (27 mg, 0.10 mmol) was
dissolved in dry DMF (1.3 mL) under argon, treated with HATU (64 mg, 0.17 mmol) and N,N-
diisopropylethylamine (0.06 mL, 0.34 mmol), then stirred for 10 min at 0 °C. To the reaction
mixture, 5d (25 mg, 0.09 mmol) was added, and the solution was stirred at room temperature for
24 h. The reaction was monitored by TLC (MeOH/DCM, 1:9). Upon completion, the product
was extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated to give
the crude product. The product was isolated by column chromatography on SiO2 using a gradient
of MeOH/DCM as eluent to afford compound 15 (30 mg, 63%). HRESI-MS m/z calculated for
C27H28Cl2N3O5, 544.1406; found for [M+H]+, 544.1403. 1H NMR (500 MHz, CD3OD) δ 8.72 (s,
H-11, 1H), 7.69 (d, J = 8.8 Hz, H-12, 1H), 7.65 (d, J = 8.7 Hz, H-4, 1H), 7.47 (m, H-1, H-6, 2H),
7.21 (dd, J = 8.7, 2.2 Hz, H-5, 1H), 6.99 (dd, J = 8.8, 2.4 Hz, H-13, 1H), 6.93 (d, J = 2.2 Hz, H-
14, 1H), 6.91 (s, H-2, 1H), 6.75 (s, H-3, 1H), 4.93 (d, J = 14.5 Hz, H-7, 1H), 4.48 – 4.36 (m, H-
7, H-15, 2H), 4.23 (d, J = 14.2 Hz, H-8, 1H), 4.18 (d, J = 14.2 Hz, H-8, 1H), 1.78 – 1.66 (m, H-
26

ACCEPTED MANUSCRIPT
16, 4H), 0.96 (t, J = 7.4 Hz, H-17, 6H). 13C NMR (125 MHz, CD3OD) δ 164.83, 163.84, 161.58,
156.90, 148.39, 137.88, 136.86, 134.32, 131.31, 130.96, 130.14, 126.99, 126.63, 120.56, 114.69,
113.15, 111.88, 101.20, 80.94, 76.64, 51.96, 45.50, 25.64, 8.28.
Compound 16. Coumarin-3-carboxylic acid[42] (74 mg, 0.39 mmol) was dissolved in dry DMF
(8 mL) under argon, treated with HATU (298 mg, 0.79 mmol) and triethylamine (0.22 mL, 1.57
mmol), then stirred for 10 min at 0 °C. To the reaction mixture, 5a[41] (100 mg, 0.39 mmol) was
added, and the solution was stirred at room temperature for 24 h. The reaction was monitored by
TLC (petroleum ether/ethyl acetate, 1:9). Upon completion, the product was extracted with ethyl
acetate, washed with H2O, dried over MgSO4, and concentrated to give the crude product. The
product was isolated by column chromatography on SiO2 using a gradient of petroleum
ether/ethyl acetate as eluent to afford compound 16 (158 mg, 95%). HRESI-MS m/z calculated
for C21H17F2N4O4, 427.1218; found for [M+H]+, 427.1213. 1H NMR (500 MHz, CDCl3) δ 9.16
(s, H-9, 1H), 8.82 (s, H-10, 1H), 8.10 (s, H-2, 1H), 7.84 (s, H-1, 1H), 7.72 – 7.65 (m, H-11, H-
13, 2H), 7.61(m, H-3, 1H), 7.47 – 7.31 (m, H-12, H-14, 2H), 6.89 – 6.75 (m, H-4, H-5, 2H), 5.98
(s, H-8, 1H), 4.63 (q, J = 14.1 Hz, H-6, 2H), 3.95 (d, J = 6.1 Hz, H-7, 2H). 13C NMR (125 MHz,
1
3
1
CDCl3) δ 164.14, 162.95 (dd, JC-F=246.0 Hz, JC-F=12.3 Hz), 161.03, 158.73(dd, JC-F=247.4
3
Hz, JC-F=11.5 Hz), 154.56, 151.63, 149.07, 144.58, 134.63, 130.07, 129.96, 125.48, 123.63,
118.37, 117.23, 116.76, 111.78, 104.23, 56.04, 47.81. 19F NMR (470 MHz, CDCl3) δ -109.52 (m,
Fpara), -109.92 (m, Fortho).
Compound 17. 2-Oxo-1,2-dihydroquinoline-3-carboxylic acid (55 mg, 0.29 mmol) was
dissolved in dry DMF (2 mL) under argon, treated with HATU (167 mg, 0.44 mmol) and N,N-
diisopropylethylamine (0.17 mL, 0.97 mmol), then stirred for 10 min at 0 °C. To the reaction
mixture, 5a[41] (64 mg, 0.25 mmol) was added, and the solution was stirred at room temperature
27

ACCEPTED MANUSCRIPT
for 24 h. The reaction was monitored by TLC (MeOH/DCM, 1:9). Upon completion, the product
was extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated to give
the crude product. The product was isolated by column chromatography on SiO2 using a gradient
of MeOH/DCM as eluent to afford compound 17 (43 mg, 40%). HRESI-MS m/z calculated for
C21H18F2N5O3, 426.1378; found for [M+H]+, 426.1375. 1H NMR (400 MHz, DMSO-d6) δ 12.37
(s, H-15, 1H), 9.94 (t, J = 5.7 Hz, H-9, 1H), 8.75 (s, H-10, 1H), 8.28 (s, H-2, 1H), 7.89 (dd, J =
8.0, 1.0 Hz, H-11, 1H), 7.74 (s, H-1, 1H), 7.62 (ddd, J = 8.5, 7.3, 1.4 Hz, H-13, 1H), 7.44 – 7.34
(m, H-3, H-14, 2H), 7.28 – 7.22 (m, H-12, 1H), 7.15 (ddd, J = 11.9, 9.2, 2.5 Hz, H-5, 1H), 6.92
(td, J = 8.5, 2.5 Hz, H-4, 1H), 6.28 (s, H-8, 1H), 4.66 (d, J = 14.4 Hz, H-6, 1H), 4.53 (d, J = 14.3
13
Hz, H-6, 1H), 3.85 (d, J = 5.8 Hz, H-7, 2H). C NMR (100 MHz, DMSO-d6) δ 163.72, 162.34
(dd, 1JC-F=244.5 Hz, 3JC-F=12.7 Hz), 162.26, 159.51 (dd, 1JC-F=245.9 Hz, 3JC-F=12.1 Hz), 151.07,
145.41, 144.46, 139.86, 133.23, 130.72, 130.22, 125.12, 123.27, 121.59, 118.96, 115.70, 111.30,
19
104.42, 74.77, 55.70, 46.26. F NMR (375 MHz, DMSO-d6) δ -107.97 (m, Fpara), -112.31 (m,
Fortho).
Compound 18. 7-Methoxy-coumarin-3-carboxylic acid[43] (54 mg, 0.25 mmol) was dissolved
in dry DMF (3 mL) under argon, treated with HATU (166 mg, 0.44 mmol) and N,N-
diisopropylethylamine (0.14 mL, 0.80 mmol), then stirred for 10 min at 0 °C. To the reaction
mixture, 5a[41] (53 mg, 0.21 mmol) was added, and the solution was stirred at room temperature
for 24 h. The reaction was monitored by TLC (MeOH/DCM, 1:9). Upon completion, the product
was extracted with ethyl acetate, washed with H2O, dried over MgSO4, and concentrated to give
the crude product. The product was isolated by column chromatography on SiO2 using a gradient
of MeOH/DCM as eluent to afford compound 18 (78 mg, 82%). HRESI-MS m/z calculated for
C22H19F2N4O5, 457.1324; found for [M+H]+, 457.1317. 1H NMR (400 MHz, DMSO-d6) δ 8.88 –
28

ACCEPTED MANUSCRIPT
8.80 (m, H-9, H-10, 2H), 8.32 (s, H-2, 1H), 7.91 (d, J = 8.8 Hz, H-11, 1H), 7.78 (s, H-1, 1H),
7.43 (dd, J = 15.9, 8.9 Hz, H-3, 1H), 7.19 (ddd, J = 11.8, 9.2, 2.4 Hz, H-5, 1H), 7.11 (d, J = 2.2
Hz, H-13, 1H), 7.05 (dd, J = 8.7, 2.3 Hz, H-12, 1H), 6.98 (td, J = 8.6, 2.4 Hz, H-4, 1H), 6.35 (s,
H-8, 1H), 4.68 (d, J = 14.4 Hz, H-6, 1H), 4.58 (d, J = 14.4 Hz, H-6, 1H), 3.95 – 3.79 (m, H-7, H-
1
14, 5H). 13C NMR (125 MHz, CDCl3) δ 165.35, 164.88, 162.89 (dd, JC-F=250.0 Hz, 3JC-F=12.0
1
3
Hz), 161.43, 158.73 (dd, JC-F=246.3 Hz, JC-F=11.8 Hz), 156.90, 151.52, 148.95, 144.59,
131.13, 130.36, 123.74, 114.32, 113.35, 112.16, 111.71, 104.19, 100.35, 56.11, 47.85. 19F NMR
(470 MHz, CDCl3) δ -109.49 (m, Fpara), -110.11 (m, Fortho).
Compound 19. 7-Methoxy-2-oxo-1,2-dihydroquinoline-3-carboxylic acid (38 mg, 0.17 mmol)
was dissolved in dry DMF (2 mL) under argon, treated with HATU (146 mg, 0.39 mmol) and
N,N-diisopropylethylamine (0.12 mL, 0.69 mmol), then stirred for 10 min at 0 °C. To the
reaction mixture, 5a[41] (46 mg, 0.18 mmol) was added, and the solution was stirred at room
temperature for 24 h. The reaction was monitored by TLC (MeOH/DCM, 1:9). Upon
completion, the product was extracted with ethyl acetate, washed with H2O, dried over MgSO4,
and concentrated to give the crude product. The product was isolated by column chromatography
on SiO2 using a gradient of MeOH/DCM as eluent to afford compound 19 (24 mg, 30%).
1
HRESI-MS m/z calculated for C22H20F2N5O4, 456.1483; found for [M+H]+, 456.1480. H NMR
(400 MHz, DMSO-d6) δ 9.98 (t, J = 5.3 Hz, H-9, 1H), 8.67 (s, H-10, 1H), 8.33 (s, H-2, 1H), 7.80
(d, J = 8.7 Hz, H-11, 1H), 7.75 (s, H-1, 1H), 7.42 (dd, J = 15.9, 8.9 Hz, H-3, 1H), 7.16 (ddd, J =
11.7, 9.2, 2.4 Hz, H-5, 1H), 7.00 – 6.82 (m, H-4, H-12, H-13, 3H), 6.46 (s, H-8, 1H), 4.68 (d, J =
13
14.3 Hz, H-6, 1H), 4.56 (d, J = 14.3 Hz, H-6, 1H), 3.94 – 3.76 (m, H-7, H-15, 5H). C NMR
1
3
(100 MHz, DMSO-d6) δ 164.33, 163.37, 162.84, 162.30 (dd, JC-F=244.2 Hz, JC-F=12.4 Hz),
1
3
159.50 (dd, JC-F=246.0 Hz, JC-F=12.2 Hz), 151.04, 145.43, 143.99, 142.49, 131.80, 130.75,
29