Evaluation of 1,2,3-Triazoles as Amide Bioisosteres In Cystic Fibrosis Transmembrane Conductance Regulator Modulators VX-770 and VX-809

Article abstract of DOI:10.1002/chem.201805919

The 1,2,3-triazole has been successfully utilized as an amide bioisostere in multiple therapeutic contexts. Based on this precedent, triazole analogues derived from VX-809 and VX-770, prominent amide-containing modulators of the cystic fibrosis transmembrane conductance regulator (CFTR), were synthesized and evaluated for CFTR modulation. Triazole 11, derived from VX-809, displayed markedly reduced efficacy in F508del-CFTR correction in cellular TECC assays in comparison to VX-809. Surprisingly, triazole analogues derived from potentiator VX-770 displayed no potentiation of F508del, G551D, or WT-CFTR in cellular Ussing chamber assays. However, patch clamp analysis revealed that triazole 60 potentiates WT-CFTR similarly to VX-770. The efficacy of 60 in the cell-free patch clamp experiment suggests that the loss of activity in the cellular assay could be due to the inability of VX-770 triazole derivatives to reach the CFTR binding site. Moreover, in addition to the negative impact on biological activity, triazoles in both structural classes displayed decreased metabolic stability in human microsomes relative to the analogous amides. In contrast to the many studies that demonstrate the advantages of using the 1,2,3-triazole, these findings highlight the negative impacts that can arise from replacement of the amide with the triazole and suggest that caution is warranted when considering use of the 1,2,3-triazole as an amide bioisostere.

Full text of DOI:10.1002/chem.201805919

A Journal of
Accepted Article
Title: Evaluation of 1,2,3-Triazoles as Amide Bioisosteres In Cystic
Fibrosis Transmembrane Conductance Regulator Modulators
VX-770 and VX-809
Authors: Jake E. Doiron, Christina A. Le, Britton K. Ody, Jonathon
B. Brace, Savannah J. Post, Nathan L. Thacker, Harrison
M. Hill, Gary W. Breton, Matthew J. Mulder, Sichen Chang,
Thomas M. Bridges, Liping Tang, Wei Wang, Steven M.
Rowe, Stephen G. Aller, and Mark Turlington
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201805919
Link to VoR: http://dx.doi.org/10.1002/chem.201805919
Supported by

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
Evaluation of 1,2,3-Triazoles as Amide Bioisosteres In Cystic
Fibrosis Transmembrane Conductance Regulator Modulators VX-
770 and VX-809
Jake E. Doiron,[a] Christina A. Le,[b] Britton K. Ody,[a] Jonathon B. Brace,[a] Savannah J. Post,[a] Nathan
L. Thacker,[a] Harrison M. Hill,[a] Gary W. Breton,[a] Matthew J. Mulder,[c] Sichen Chang,[c] Thomas M.
Bridges,[c] Liping Tang,[d,e] Wei Wang,[e,f] Steven M. Rowe,[d,e,f] Stephen G. Aller,*[b] and Mark
Turlington*[a]
suggest that caution is warranted when considering use of the
1,2,3-triazole as an amide bioisostere.
Abstract: The 1,2,3-triazole has been successfully utilized as
an amide bioisostere in multiple therapeutic contexts. Based on
this precedent, triazole analogs derived from VX-809 and VX-
770, prominent amide-containing modulators of the cystic
fibrosis transmembrane conductance regulator (CFTR), were Introduction
synthesized and evaluated for CFTR modulation. Triazole 11,
derived from VX-809, displayed markedly reduced efficacy in
F508del-CFTR correction in cellular TECC assays in comparison
Since the development of efficient synthetic methodologies
to prepare 1,2,3-triazoles in the early 2000’s,[1] the use of the
1,2,3-triazole as an amide bioisostere has proven to be a widely
successful strategy in medicinal chemistry.[2–5] As shown in
Figure 1, the basis of the effective bioisosteric relationship is due
to the similar topological and electronic characteristics of the
1,4-disubstituted 1,2,3-triazole and the trans-amide. Substituents
located at the 1- and 4-positions of the triazole are separated by
~5.0 to 5.1 Å, which is close to the ~3.8 to 3.9 Å distance
between the amide substituents.[3] The triazole also exhibits
similar H-bonding capabilities to the amide, with the triazole N3
serving as an H-bond acceptor in an analogous spatial location
to the amide carbonyl.[6] Furthermore, due to the substantial
dipole moment of the triazole ring calculated to be ~5 Debye,[5]
the C5 hydrogen atom of the triazole has been demonstrated to
be capable of functioning as an H-bond donor that occupies a
comparable position to the amide hydrogen atom.[7]
to VX-809.
Surprisingly, triazole analogs derived from
potentiator VX-770 displayed no potentiation of F508del, G551D,
or WT-CFTR in cellular Ussing chamber assays. However, patch
clamp analysis revealed that triazole 60 potentiates WT-CFTR
similarly to VX-770. The efficacy of 60 in the cell-free patch
clamp experiment suggests that the loss of activity in the cellular
assay could be due to the inability of VX-770 triazole derivatives
to reach the CFTR binding site. Moreover, in addition to the
negative impact on biological activity, triazoles in both structural
classes displayed decreased metabolic stability in human
microsomes relative to the analogous amides. In contrast to the
many studies that demonstrate the advantages of using the
1,2,3-triazole, these findings highlight the negative impacts that
can arise from replacement of the amide with the triazole and
[a]
[b]
J. E. Doiron, B. K. Ody, J. B. Brace, S. J. Post, N. L. Thacker, H. M.
Hill, Prof. Dr. G. W. Breton, Prof. Dr. M. Turlington
Department of Chemistry and Biochemistry
Berry College
Mount Berry, Georgia 30165 (USA)
E-mail: mturlington@berry.edu
C. A. Le, Prof. Dr. S. G. Aller
A
B
Department of Pharmacology and Toxicology
University of Alabama at Birmingham
Birmingham, Alabama 35205 (USA)
E-mail: sgaller@uab.edu
M. J. Mulder, S. Chang, Dr. T. M. Bridges
Vanderbilt Center for Neuroscience Drug Discovery
Nashville, Tennessee 37232 (USA)
[c]
[d]
L. Tang, Prof. Dr. S. M. Rowe
Departments of Medicine and Pediatrics
University of Alabama at Birmingham
Birmingham, Alabama 35205 (USA)
L. Tang, Dr. W. Wang, Prof. Dr. S. M. Rowe
Gregory Fleming James Cystic Fibrosis Research Center
University of Alabama at Birmingham
Birmingham, Alabama 35205 (USA)
Dr. W. Wang, Prof. Dr. S. M. Rowe
Figure 1. Bioisosteric relationship of 1,4-disubstituted 1,2,3-triazoles and
trans-amides. A) Similar R1-R2 distance and arrangement of H-bond donor
and acceptor groups in the trans-amide and 1,2,3-triazole. Adapted from
references [3] and [4]. B) Overlay of trans-N-methylacetamide and 1,4-
dimethyl-1H-1,2,3-triazole in which the amide carbonyl carbon and nitrogen
atom are aligned with C4 and C5 of the 1,2,3-triazole.
[e]
[f]
Department of Cell, Development, and Integrative Biology
University of Alabama at Birmingham
Birmingham, Alabama 35205 (USA)
Impressively, numerous successful applications of the
1,2,3-triazole as an amide bioisostere have been reported as
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
Figure 2. Use of the 1,2,3-triazole as an amide bioisostere allows use of CuAAC or isCC to optimize and/or discover new chemical scaffolds.
recently reviewed by Passarella and coauthors.[2]
These
tract, and sweat glands.[17] Mutations to the CFTR ion channel
are the causative factor for the devastating disease Cystic
Fibrosis (CF).[18,19]
examples span diverse therapeutic contexts including antiviral
agents,[8] chemotherapeutics,[9–11] antibiofilm agents,[12] anti-
nociceptives,[13] and antipsychotics.[14] Advantageously, use of
the 1,2,3-triazole in place of the amide often affords increased
biological activity, as demonstrated for viral infectivity factor (Vif)
inhibitors 1 and 2 (Figure 2).[8] In this example, the potency of
amide 1 (IC50 = 6 M) was improved substantially in triazole 2
(IC50 = 1.2 M). Furthermore, the discovery that the triazole
could be substituted for the amide enabled use of the Cu-
catalyzed azide-alkyne cycloaddition (CuAAC) reaction to rapidly
discover highly potent triazole-containing inhibitor 3 (IC50 = 10
nM).[8] In addition to CuAAC, the triazole linkage also opens the
door to the use of target-guided in situ click chemistry (isCC)[15]
to probe protein binding sites and discover new chemical
scaffolds. For example, the discovery that the amide in imatinib
(4) could be substituted with the 1,2,3-triazole[9] led to further
studies demonstrating that the Abl protein could assemble
triazole inhibitor 5 from click fragments 6 and 7 in situ, making
possible the use of the isCC strategy for the discovery of new
classes of Abl inhibitors.[10] Finally, in addition to enabling rapid
analog exploration using CuAAC or isCC, substitution with the
1,2,3-triazole has also been demonstrated to improve
physiochemical properties such as aqueous solubility[13] and
metabolic stability.[14]
While ~2,000 CFTR mutations have been identified,[20]
deletion of phenylalanine 508 (F508del) from NBD1 is the most
prominent CFTR mutation, with ~90% of CF patients carrying at
least one F508del-CFTR allele.[21] The F508del mutation has
especially severe consequences for protein stability and function.
First, deletion of phenylalanine 508 destabilizes the NBD1
domain and disrupts key domain-domain interactions between
NBD1 and intracellular loop 4 (ICL4) of MSD2.[22]
This
combination of structural defects impairs F508del-CFTR folding,
leading to significant ubiquitin-proteasomal degradation and low
levels of F508del-CFTR expression in the apical membrane of
epithelial cells.[23] Exacerbating the problem, F508del-CFTR
also exhibits a decreased half-life at the plasma membrane[24]
and a gating defect in which Cl transport is impaired due to a
decreased open probability (Po) of the channel.[24–26] These
combined deficiencies result in CFTR mediated Cl secretion in
human bronchial epithelial (HBE) cells derived from patients
homozygous for F508del-CFTR that is only ~3-4% of that
observed in non-CF epithelia.[24,27] Diminished Cl transport in
CF patients results in dehydration of the airway surfaces, which
concurrently leads to decreases in ciliary beat frequency and
mucociliary clearance causing the buildup of thick mucus within
the lungs.[28] This mucus harbors harmful bacterial pathogens,
leading to chronic lung infection and hyperinflammation that
severely damage the lungs and ultimately causes lung failure,
the primary cause of death in CF patients.
Groundbreaking work by Vertex Pharmaceuticals and
others has identified small molecule modulation of mutant forms
of CFTR as a revolutionary strategy to combat CF.[16,18,19,29]
CFTR modulators are grouped into two categories—
potentiators and correctors. Potentiators increase Cl ion
transport by counteracting the gating defect in the mutant
protein while correctors counteract folding deficiencies to
Due to the widespread success of the 1,2,3-triazole as an
amide bioisostere, we sought to evaluate the utility of this
bioisosteric replacement within amide-containing small molecule
modulators of the cystic fibrosis transmembrane conductance
regulator (CFTR) protein. CFTR is a member of the ATP-
binding cassette (ABC) transporter superfamily and is comprised
of two membrane spanning domains (MSD1, MSD2), two
nucleotide binding domains (NBD1, NBD2), and a regulatory
domain that controls protein function via PKA-dependent
phosphorylation.[16] CFTR functions as a Cl and HCO3 ion
channel and plays an essential role in fluid homeostasis across
epithelial cells in the lungs, pancreas, intestines, reproductive

This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
increase expression of the protein. Modulators in both classes
have gained FDA approval.[19, 29] Potentiator VX-770 (Ivacaftor,
Figure 3) has been approved for a number of gating mutations,
and the combination of VX-770 with VX-809 (Lumacaftor,
Figure 3) is approved to treat the F508del mutation that disrupts
protein folding and channel gating. Recently, the combination of
Ivacaftor with the second-generation corrector VX-661
(Tezacaftor, Figure 3), which exhibits an improved safety profile
over VX-809, was also approved to treat patients with the
F508del mutation.
develop
a
new class of correctors, if the 2,2-
difluorobenzo[d][1,3]dioxole cyclopropane 1,2,3-triazole can
serve as a suitable pharmacophore for F508del-CFTR correction.
To investigate this question, we targeted the synthesis of 1,2,3-
triazole containing corrector 11 shown in Figure 4B, which is a
close structural analog of amide-containing VX-809.
Figure 4. A) Structure of corrector 8[30] and C18 (VRT-534)[31,32] bearing the
benzo[d][1,3]dioxole cyclopropane carboxamide pharmacophore. B) Targeted
Figure 3. Structures of potentiator VX-770 and correctors VX-809 and VX-
661.
alkyne
9
and azide 10 to explore the viability of the 2,2-
difluorobenzo[d][1,3]dioxole cyclopropane 1,2,3-triazole pharmacophore. C)
2-azidopyridine and tetrazole tautomers. D) Targeted alkyne 12 to investigate
the importance of the 2,2-difluorobenzo[d][1,3]dioxole moiety for corrector
efficacy.
Despite the recent breakthroughs in CF therapy, the
search for improved CFTR modulators remains a priority.[19]
Given the presence of the amide linkage in both VX-809 and
VX-770, the detailed pharmacological characterization of these
modulators reported in the primary literature,[24,27] and the
promise of the triazole as an amide bioisostere, we investigated
the suitability of the 1,2,3-triazole bioisosteric replacement
within the VX-809 and VX-770 structural classes. Importantly,
demonstration that the triazole substitution is tolerated in these
molecular scaffolds would provide a starting point for future
studies aimed at generating improved classes of modulators
using CuAAC or isCC strategies. Herein, we report the
synthesis and biological evaluation of 1,2,3-triazole analogs of
CFTR modulators within the VX-809 and VX-770 structural
classes.
In addition to the 1,2,3-triazole substitution, 11 differs from
VX-809 in that it contains a central phenyl ring in contrast to the
pyridine found in VX-809. This modification was made to
prevent challenges associated with the use of 2-azidopyridines
in azide-alkyne cycloaddition reactions due to the unfavorable
equilibrium between 2-azidopyridines and the corresponding
tetrazole tautomer (Figure 4C).[33] In addition to alkyne 9 we
also targeted the synthesis of alkyne 12 containing the 4-
methoxyphenyl group in order to gain insight into the importance
of the 2,2-difluorobenzo[d][1,3]dioxole moiety for corrector
efficacy. The 4-methoxyphenyl moiety was selected from the
VX-809 patent literature as it appeared to display efficacy in
F508del-CFTR correction[34] and was readily accessible
synthetically. We hypothesized that 12 might also be a useful
building block for CuAAC or isCC medicinal chemistry efforts.
Results and Discussion
Synthesis of 1,2,3-Triazole Analogs in the VX-809
Structural Class
Design of 1,2,3-Triazole Analogs in the VX-809 Structural
Class
Alkynes 9 and 12 were prepared from the corresponding nitriles
as shown in Scheme 1. Treatment of nitrile 13 with DIBALH at
-78 C followed by quenching with 3 M HCl produced aldehyde
14 in 89% yield. Corey-Fuchs alkynylation using CBr4 and Ph3P,
followed by treatment of the dibromoolefin intermediate with
nBuLi produced targeted alkyne 9 in 60% yield over two steps.
Preparation of alkyne 12 from nitrile 15 was accomplished in the
same manner, with the Corey-Fuchs reaction proving more
efficient for this substrate as 12 was produced in 81% yield over
two steps.
Synthesis of the azide 10 was also accomplished in a
straightforward manner as shown in Scheme 2. Suzuki coupling
of 3-bromo-4-methylaniline with boronic acid 18 produced biaryl
19 in 98% yield. TFA-mediated deprotection of tert-butyl ester
19, followed by diazotization with sodium nitrite and substitution
Among F508del-CFTR correctors, the benzo[d][1,3]dioxole
cyclopropane carboxamide moiety has been maintained as an
essential structural feature across otherwise diverse chemical
scaffolds, suggesting that this moiety is a key pharmacophore
for corrector efficacy. As seen in the comparison of VX-809, VX-
661, corrector 8 reported by AbbVie Pharmaceuticals (Figure
4A),[30] and the widely employed tool compound C18 (VRT-534,
Figure 4A),[31,32] considerable structurally variability is tolerated
within the amide nitrogen substituent provided that the
benzo[d][1,3]dioxole cyclopropane carboxamide moiety is
present. On the basis of this observation, we hypothesized that
CuAAC or isCC methods using 2,2-difluorobenzo[d][1,3]dioxole
cyclopropane alkyne 9 (Figure 4B) have strong potential to
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
3-methylpyridine with boronic acid 18 first produced 28 in high
yield which was then quantitatively converted to N-oxide 29 via
treatment with mCPBA. Amination via reaction with Ms2O and
pyridine in MeCN at 70 C for 1.5 h, followed by treatment with
ethanolamine yielded 2-aminopyridine 30 in 53% yield.
Preparation of acid chloride 31 from carboxylic acid 24 using
SOCl2, and reaction with amine 30 produced amide 32 which
contained the pyridyl core. Finally, hydrolysis of the tert-butyl
ester using TFA afforded 33, the 4-methoxyphenyl analog of VX-
809, in 77% yield.
Scheme 1. Synthesis of Alkynes 9 and 12. Reagents and conditions: a) (i)
DIBALH, toluene, -78 °C, 3 h; (ii) HCl, 0 °C, 15 min, 14: 89%, 16: 88%; b) CBr4,
PPh3, CH2Cl2, 0 °C, 1 h; c) nBuLi (2.2 equiv) THF, -78 °C, 1.5 h, 9: 60% over 2
steps, 12: 81% over 2 steps.
of the aryl diazonium intermediate with sodium azide, produced
azide 10 in 92% yield over the two-step sequence. With the
targeted alkyne and azide intermediates in hand we then
performed the Cu-mediated azide-alkyne cycloaddition using
standard reaction conditions to produce triazoles 11 and 20 in
quantitative yield.
Scheme 4. Synthesis of Corrector 33. Reagents and conditions: (a) 18,
Pd(dppf)Cl2•CH2Cl2, K2CO3, toluene/H2O, reflux, 18 h, 92%; (b) mCPBA,
CH2Cl2, rt, 24 h, 99%; (c) (i) Ms2O, pyridine, MeCN, 70 °C, 1.5 h; (ii)
i
ethanolamine, rt, 2 h, 53% (d) SOCl2, toluene, 60 °C, 2 h. (e) Pr2EtN, DMAP,
CH2Cl2, rt, 18 h, 56%; (f) TFA, CH2Cl2, 0 °C to rt, 2 h, 77%.
Scheme 2. Synthesis of Azide 10 and 1,2,3-Triazole-Containing Correctors
11 and 20. Reagents and conditions: (a) Pd(dppf)Cl2•CH2Cl2, K2CO3,
toluene/H2O (3:1), reflux, 20 h, 98%; (b) TFA, CH2Cl2, 0 °C to rt, 2 h; (c) (i)
EtOAc, HCl, 0 °C, 5 min; (ii) NaNO2, H2O, 0 °C, 1 h; (iii) NaN3, H2O, 0 °C to rt,
In vitro Evaluation of 1,2,3-Triazole Analogs in the VX-809
Structural Class
3
h; 92% over 2 steps; (d) 9 or 12, CuSO4•5H2O, sodium ascorbate,
1,2,3-triazole-containing correctors 11 and 20, amide-containing
correctors 23 and 26 containing the central phenyl ring, and the
amide-containing corrector 33 bearing the central pyridine were
evaluated in transepithelial chloride conductance (TECC)
assays using F508del CFBE41o monolayers. The F508del
CFBE41o cell line is an immortalized cell line derived from
bronchial epithelial cells of a CF patient homozygous for the
F508del CFTR mutation with complementary stable F508del-
CFTR transgene expression,[36] and has been shown to be
clinically relevant in characterizing small molecule CFTR
CH2Cl2/tBuOH/H2O (1:2:2), rt, 24 h, 11: 100%, 20: 100%.
To enable the direct comparison of triazoles 11 and 20 to
their analogous amides, amides 23 and 26 were prepared as
shown in Scheme 3. HATU-mediated coupling of carboxylic
acids 21 and 24 with aniline 19 produced the corresponding
amides in excellent yield. Removal of the tert-butyl ester with
TFA quantitatively produced amide analogs 23 and 26.
modulators.[37,38]
Prior to TECC experiments, F508del
CFBE41o monolayers were incubated (24 h) with varying
concentrations of corrector compound to increase functional
F508del-CFTR present at the apical membrane as a result of
corrector-assisted protein folding. Increased CFTR-mediated
Cl ion passage (due to increased F508del-CFTR protein) was
characterized via measurement of changes in equivalent
current (Ieq). Since CFTR-mediated Cl ion passage and
therefore Ieq is slightly variable between cell cultures, a 3 M
dose of VX-809 was used as a positive control in all
experiments. Data are presented as normalized area under the
curve values following stimulation with the cAMP agonist
forskolin (10 M) and VX-770 (1 M), followed by inhibition with
the CFTR inhibitor CFTRInh-172 (20 M).
Scheme 3. Synthesis of Amide-Containing Correctors 23 and 26. Reagents
and conditions: (a) 19, HATU, Et3N, DMF, rt, 24 h, 22: 99%, 25: 96%; (b) TFA,
CH2Cl2, 0 °C to rt, 2 h, 23: 100%, 26: 99%.
Finally, in order to fully understand the effect of the
replacement of the pyridyl ring in the VX-809 scaffold with the
phenyl ring in amides 23/26 and 1,2,3-triazoles 11/20, we
prepared pyridyl-containing corrector 33 as shown in Scheme 4.
Following the synthetic procedure reported for this class of
correctors in the patent literature,[35] Suzuki coupling of 2-bromo-
As shown in Table 1 and Figure 5, several trends were
observed
regarding
the
importance
of
the
2,2-
difluorobenzo[d][1,3]dioxole moiety for corrector efficacy, the
impact of substituting the central pyridyl ring with the phenyl
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
correction, albeit with significantly reduced efficacy in
comparison to the analogous amide 23. However, it is possible
that optimization of the triazole N1 substituent via the CuAAC
reaction of alkyne 9 with structurally diverse azides could result
in correctors with improved efficacy.
Table 1. In Vitro Potency and Efficacy Data of Amide and Triazole-
Containing Correctors in the VX-809 Chemical Series in F508del
CFBE41o- Cells.
compd
EC50 (μM)[a]
Maximal Response[b]
In contrast, corrector 20 containing the 4-methoxyphenyl
cyclopropane 1,2,3-triazole pharmacophore was inactive. As
the directly analogous 4-methoxyphenyl amide-containing
corrector 26 was significantly less efficacious than 23, the
inactivity of triazole 20 is consistent with the loss in efficacy
observed upon substitution of the amide with the triazole in this
chemical series. In total, these findings demonstrate that while
the triazole can be used to replace the amide in the VX-809
structural class when the 2,2-difluorobenzo[d][1,3]dioxole
moiety is retained, use of the triazole results in lowered efficacy
and is not a suitable replacement for less efficacious amide-
containing correctors.
11
20
23
26
33
0.660 ± 0.287
inactive
48.1%
--
1.960 ± 0.754
2.971 ± 0.900
0.427 ± 0.514
120.1%
54.1%
49.9%
[a] Mean and standard deviation values from at least three determinations
are reported. [b] Maximal response is reported as % activity of the 3 M
VX-809 response.
ring, and the suitability of the 1,2,3-triazole as an amide
bioisostere in the VX-809 chemical series. First, the 4-
methoxyphenyl moiety in amide 33 displayed greatly reduced
corrector efficacy, highlighting the importance of the 2,2-
difluorobenzo[d][1,3]dioxole functionality for F508del-CFTR
correction. Secondly, substitution of the pyridyl ring in VX-809
with the phenyl ring in corrector 23 resulted in minimal loss of
potency and efficacy, as 23 displayed similar dose response
Design of 1,2,3-Triazole Analogs in the VX-770 Structural
Class
We next investigated use of the 1,2,3-triazole as an amide
bioisostere in the VX-770 potentiator series. As shown in Figure
6, tautomer 34 of VX-770, in which the amide carbonyl oxygen
forms an intramolecular H-bond with the quinolinol hydroxyl
group, has been proposed to be the bioactive structural
isomer.[39] Supporting this hypothesis, attempts to substitute the
amide in the VX-770 series with ester, sulfonamide, or reverse
amide linkers essentially abolished potentiator efficacy.[39] We
hypothesized that triazole containing tautomer 35 could form a
similar intramolecular H-bonding interaction to stabilize the
bioactive tautomer and would be a suitable linker replacement.
Additionally, on the basis of the considerable structure activity
relationship (SAR) data published for the amide nitrogen
substituent in the VX-770 series, we also targeted triazole
analogs of potentiators 36 and 37 reported by Hadida and
coworkers during the discovery of VX-770.[39] Amide 36 was
selected as it was reported to be the most potent analog
characterized in F508del-CFTR HBE cells, and 37 was chosen
to explore the indole moiety that was also determined to be
highly effective for F508del-CFTR potentiation.[39]
activity to VX-809.
The minimal impact of the phenyl
substitution on corrector efficacy is also seen in comparison of
the 4-methoxyphenyl amide-containing analogs (pyridyl-
containing 33 and phenyl-containing 26).
Importantly, as shown in Table 1 and Figure 5, triazole-
containing corrector 11 displayed statistically significant
correction of F508del-CFTR above the DMSO response, with
Key
****
****
****
***
150
125
100
75
11
20
23
26
150
100
50
****
****
**
**
33
**
****
50
25
DMSO
0
0
0.1
1
3
10
Concentration (mM)
Figure 5. Dose response effects of triazole and amide-containing correctors
on Cl conductance in F508del CFBE41o monolayers using the TECC assay.
Current was measured by calculating area under the curve (AUC) after
stimulation with forskolin (10 M) + VX-770 (1 M), followed by inhibition with
CFTRInh-172 (20 M). A 3 M dose of VX-809 was used as a positive control
in all experiments. Relative corrector efficacy was determined by calculating
the percentage of the analog’s response against the 3 M VX-809 response.
Asterisks indicate significant (p<0.05, Two-way ANOVA, Tukey’s post-hoc
test) increases in AUC against the vehicle (DMSO) response, n = 3 – 5.
Figure 6. A) Proposed bioactive tautomer of VX-770[39] and triazole
containing analog. B) Structures of highly potent compounds within the VX-
770 structural class.[39]
Synthesis of 1,2,3-Triazole Analogs in the VX-770
Structural Class
an EC50 = 0.660 ± 0.287 M and a maximal efficacy of 48%
compared to the 3 M VX-809 response. This demonstrates
that the 2,2-difluorobenzo[d][1,3]dioxole cyclopropane 1,2,3-
triazole can serve as a pharmacophore for F508del-CFTR
In contrast to the straightforward synthesis of 1,2,3-triazole
analogs within the VX-809 structural class, the synthesis of
triazole derivatives in the VX-770 chemical series proved more
challenging.
We first focused on the synthesis of 3-
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
ethynylquinolin-4(1H)-one 43 that would serve as a key CuAAC
precursor as shown in Scheme 5. Iodination of quinolin-4(1H)-
one with I2 and Na2CO3 allowed facile synthesis of 3-
iodoquinolin-4(1H)-one 39[40] which was subjected to
Sonogoshira reaction conditions as reported by Corelli and
coworkers.[41] Disappointingly while Corelli and coworkers were
able to use microwave irradiation for similar substrates to
achieve Sonogoshira coupling without conversion to furo[3,2-
c]quinoline derivatives such as 41,[42] reaction of 39 with thermal
heating (reflux) under Corelli’s reaction conditions afforded
undesired furo[3,2-c]quinoline 41, which was unambiguously
evolution was observed after the addition of NaNO2 to 51.
Fortunately, use of Zhang and Moses’ method for azide
formation using tBuONO and TMSN3 in MeCN[46] proved
effective, although significant formation of trimethylsilyl protected
phenol 52 was produced (73%) in addition to isolation of the
desired azide 53 (20%). This did not prove to be problematic as
the TMS group could be readily removed by treatment with
TBAF to produce 53 in quantitative yield. Notably aniline 51 is
not a bench stable solid, and must be transformed to the azide
immediately after preparation to prevent decomposition to
quinone 54.[47]
identified by conversion to known compound 42.[43]
The
cyclization likely proceeds through the mechanism shown in
Scheme 5 involving coordination of the acetylene by copper and
deprotonation of the acidic proton on the quinolinone nitrogen
atom.[42] This suggests that use of 43 would lead to a similar
cyclization reaction under the copper-catalyzed azide-alkyne
cycloaddition reaction conditions (see Scheme 5), and this was
found to be the case (see Supporting Information).
Scheme 6. Synthesis of Benzyl Protected 3-ethynylquinolin-4(1H)-one 45 and
Compatibility with CuAAC. Reagents and conditions: (a) (i) NaH, THF, 0 °C, 30
min; (ii) benzyl bromide, 0 °C to rt, 19 h, 77%; (b) trimethylsilylacetylene,
Pd(PPh3)2Cl2, CuI, DIPEA, THF, rt, 18 h, 86%; (c) K2CO3, MeOH, rt, 24 h, 86%.
(d)
4-tert-butylphenyl
azide,
CuSO4•5H2O,
sodium
ascorbate,
CH2Cl2/tBuOH/H2O (1:2:2), rt, 24 h, 66%.
Scheme 5. Unsuccessful Synthesis of 43 and Incompatiblity of 43 with
CuAAC Due to Cyclization of 3-Ethynylquinolin-4(1H)-ones to Furo[3,2-
c]quinolines. Reagents and conditions: (a) I2, Na2CO3, THF, rt, 24 h, 82%; (b)
trimethylsilylacetylene, Pd(PPh3)2Cl2, CuI, DIPEA, THF, rt, 18 h, 64%; (c)
K2CO3, MeOH, rt, 24 h, 67%.
Scheme 7. Synthesis of Azides 48, 50, and 53. Reagents and conditions: (a)
(i) EtOAc, HCl, 0 °C, 5 min; (ii) NaNO2, H2O, 0 °C, 30 min; (iii) NaN3, H2O, 0 °C
to rt, 2 h, 48: 81%; 50: 75%; 53: 0%; (b) TMSN3, tBuONO, MeCN, rt, 18 h, 52:
73%, 53: 20%; (c) TBAF, THF, 0 °C to rt, 2 h, 100%.
With azides 48, 50, and 53 and alkyne 45 in hand we
completed the synthesis of the 1,2,3-triazole containing
potentiators as shown in Scheme 8. While CuAAC could be
achieved with azides 48 and 50 at rt, no product formation was
As a methyl substituent on the quinolin-4(1H)-one nitrogen
prevents the cyclization reaction during the Sonogoshira
coupling,[42] we began to investigate suitable nitrogen protecting
groups. As shown in Supplementary Schemes 1 and 2, use of
the methyl and tert-butyl carbamate protecting groups proved
ineffective (see Supporting Information). In contrast, use of the
benzyl protecting group, which has been successfully used in
the synthesis of VX-770,[44] was effective. Benzyl protection of
39 via deprotonation with NaH and reaction with benzyl bromide
produced 44.[45] Sonogoshira coupling of 44 with trimethylsilyl
acetylene, and TMS deprotection using K2CO3 in MeOH
produced terminal alkyne 45 in 57% total yield over the three
step sequence as shown in Scheme 6. Pleasingly, use of 45 in
the CuAAC reaction with test substrate 4-tert-butylphenyl azide
produced the desired click product 46 in moderate yield (66%),
and the benzyl group could be removed by refluxing 46 with
ammonium formate and Pearlman’s catalyst in DMF. Following
these test reactions, the synthesis of 45 was readily scaled up to
produce gram quantities of 45.
In parallel with the synthesis of alkyne 45, we also pursued
the synthesis of azides 48, 50, and 53 as shown in Scheme 7.
Precursor amines 47, 49, and 51 were prepared according to the
synthetic procedures reported by Hadida and coworkers (see
Supporting Information, Supplementary Scheme 3).[39] The
synthesis of azides 48 and 50 was readily accomplished in good
yield by treatment of the precursor amine with NaNO2 followed
by NaN3. Surprisingly, use of these reaction conditions with
aniline 51 could not produce the desired azide, as vigorous gas
Scheme 8. Synthesis of 1,2,3-Triazole Containing Potentatiors 56, 58, and 60.
Reagents and conditions: (a) CuSO4•5H2O, sodium ascorbate,
CH2Cl2/tBuOH/H2O (1:2:2), rt, 24 h, 55: 69%, 57: 71%, 59: 0% (b) Pd(OH)2,
ammonium formate, DMF, 80 °C, 56: 10 h, 74%, 58: 8 h, 88%, 60: 8.5 h, 87%;
(c) CuSO4•5H2O, sodium ascorbate, CH2Cl2/tBuOH/H2O (1:2:2), 60 °C, 24 h,
45%.
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
reasoned the higher Cl ion conductance might allow the
detection of weak potentiators and differentiate the response
between potentiators to a greater extent than the CFBE41o cell
line.
observed for reaction of alkyne 45 with sterically hindered azide
53 at rt. Fortunately, heating to 60 C for 24 h produced the
desired product 60 in modest yield (45%), but attempts to
improve yield by extending the reaction time (up to 3 days) were
unsuccessful. Finally, removal of the benzyl protection group
using ammonium formate and Pearlman’s catalyst in refluxing
DMF afforded the desired 1,2,3-triazole analogs in good yield.
Additionally, in order to directly compare the potentiator ability of
the triazole derivatives to their parent amides 36 and 37, these
compounds were also prepared. As shown in Scheme 9, HATU-
mediated amide bond formation afforded the desired
compounds in improved yields over the published reaction
procedures.[39]
The effect of the potentiator compounds was quantified
via measurement of changes in short circuit current (∆Isc) due to
increased CFTR-mediated Cl ion secretion. The response
was measured as the cumulative Isc observed with serial
addition of each test agent after stimulation with forskolin, either
100 nM (the EC50) in WT monolayers or 10 M in mutant
(F508del, G551D) monolayers.
As Cl ion transport and
increases in ∆Isc can be variable between cell cultures, VX-770
was used as a positive control in all experiments and relative
potentiator efficacy was determined by calculating the
percentage of each analog’s response against the VX-770
response as shown in Table 2 and Figure 7. Potency was
quantified by determining EC50 values and is shown in Table 2.
Disappointingly, as shown in Table 2 and Figure 7, none
of the 1,2,3-triazoles (56, 58, 60) achieved statistically
significant increases in ∆Isc above the DMSO response in any of
Scheme 9. Synthesis of Amide Containing Potentiators 36 and 37. Reagents
and conditions: (a) 47 or 49, HATU, Et3N, CH2Cl2, 70 °C, 16 h, 36: 87%; 37:
64%.
the CFTR variants or cell types tested.
These findings
demonstrate that use of the 1,2,3-triazole moiety significantly
alters the VX-770 chemical scaffold such that potentiation is
abolished in the in vitro cellular assay. Among the active amide-
containing potentiators, VX-770 consistently displayed the
highest efficacy by 1.5 to 3-fold (Table 2), a finding that to our
knowledge has not previously been reported in the primary
In vitro Evaluation of 1,2,3-Triazole Analogs in the VX-770
Structural Class
1,2,3-triazole containing potentiators 56, 58, and 60 were
evaluated for potentiator activity and compared to amide
potentiators 36, 37, and VX-770 in Ussing Chamber
experiments. VX-770 is known to potentiate WT-CFTR (that
does not exhibit a gating defect) and F508del-CFTR and
G551D-CFTR (that do display the gating defect) with varying
levels of potency and efficacy.[27,48] Accordingly, we evaluated
our amide and triazole-containing potentiators in CFBE41o
monolayers with complementary WT CFTR expression (WT
literature as Hadida and coworkers did not report
a
measurement of efficacy for 36 and 37.[39] Interestingly, 36
(which varies structurally from VX-770 only in the deletion of
one tert-butyl group) did not produce statistically significant
potentiation above the DMSO response except in the G551D
CFBE41o cell line. In contrast, indole 37 displayed more
robust efficacy across all cell lines, although it was less
efficacious than VX-770. However, 37 did display improved
potency over VX-770 in our experiments. In total, these results
demonstrate the essential nature of the amide linker in the VX-
770 series as substitution with the triazole resulted in inactivity
in the cellular assay. These studies also highlight the important
impact of the amide substituent on in vitro efficacy.
CFBE41o), F508del CFBE41o monolayers after 24
h
treatment with VX-809 (3 M) to rescue mutant F508del-CFTR
to the cell surface, and CFBE41o monolayers with
complementary
G551D-CFTR
expression
(G551D
CFBE41o).[49] In addition, we also utilized Fischer Rat Thyroid
(FRT) cells expressing G551D as this cell line is commonly
used to characterize potentiator compounds and exhibits
elevated Cl ion conductance compared to CFBE cells.[50] We
Table 2. In Vitro Potency and Efficacy Data of Amide and Triazole-Containing Potentiators in the VX-770 Chemical Series in CFBE41o and FRT Cells.[a]
EC50 (nM)[b]
Maximal Response[c]
compd
WT CFBE
F508del CFBE
G551D CFBE
77.7 ± 27.2
G551D FRT
WT CFBE
100.0%
F508del CFBE
G551D CFBE
G551D FRT
100.0%
VX-770
36.5 ± 21.3
1889 ± 1388
428.5 ± 167.3
100.0%
100.0%
36
37
--
--
26.7± 79.4
18.5 ± 23.4
--
Inactived
47.5%
Inactived
36.3%
63.5%
72.0%
Inactived
36.7%
21.8 ± 41.2
37.1 ± 48.2
81.45 ± 102.8
56
58
60
Inactive across all cell and CFTR typesd
Inactive across all cell and CFTR typesd
Inactive across all cell and CFTR typesd
aCFBE41o abbreviated as CFBE. bMean and standard deviation values from at least three determinations are reported. cMaximal response is reported as %
activity of the VX-770 response. dNo statistically significant response above DMSO vehicle.
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
WT CFBE41o–
F508del CFBE41o–
Key
60
DMSO
15
10
5
****
****
*
****
125
**
**
****
100
75
50
25
0
VX-770
****
**
100
75
50
25
0
****
*
36
40
37
56
*
58
20
60
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
1
0
0
0
1
0
0
1
0
0
3
1
0
0
1
1
3
-25
-25
Concentration (nM)
G551D CFBE41o–
Concentration (nM)
G551D FRT
-5
-20
4
3
80
60
40
20
0
125
100
75
****
**
125
****
**
*
****
***
****
***
*
****
**
100
75
50
25
0
*
*
****
***
2
50
***
1
25
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
1
0
0
0
1
0
0
0
1
0
0
3
1
1
0
1
3
-25
-25
-20
Concentration (nM)
-1
Concentration (nM)
Figure 7. Dose response effects of triazole and amide-containing potentiators on Cl conductance in WT CFBE41o, F508del CFBE41o following 24 h exposure
to VX-809 (3 M), G551D CFBE41o, and G551D FRT cell lines. Current was measured after stimulation with forskolin (100 nM in WT and 10 M in mutants
F508del and G551D). Inhibition with CFTRInh-172 (10 M) demonstrated that increased current was due to a CFTR-mediated mechanism. Relative potentiator
efficacy was determined by calculating the percentage of the analog’s response against the VX-770 response. Data are normalized according to vehicle (DMSO)
response. Asterisks indicate significant (p<0.05, Two-way ANOVA, Tukey’s post-hoc test) increases in Isc against the vehicle (DMSO) response, n = 3 – 5.
X-Ray Crystal Structures and Natural Bond Orbital
Calculations for VX-770 and 1,2,3-Triazole Analog 60
length and/or the slight change in orientation angle could be
responsible for an unfavorable steric interaction between
potentiator 60 and the binding pocket.
Following our disappointing findings that 1,2,3-triazole
containing potentiators in the VX-770 structural class were
inactive in the cellular Ussing Chamber assays, we verified that
we correctly synthesized the targeted 1,4-substituted 1,2,3-
triazoles by obtaining the crystal structure of triazole 60 as
shown in Figure 8A. With the X-ray structure providing
conclusive proof that we synthesized the targeted triazole
analog, we also obtained the crystal structure of VX-770 as
shown in Figure 8B. We then compared the two structures in
order to gain insight into the structural differences in the solid-
state (an overlay of the structures is shown in Figure 8C). In
line with previous studies comparing the R1-R2 distance in
amide and 1,2,3-triazole analogs,[3] the C8-C12 distance across
the triazole in 60 was determined to be 4.99 Å. In comparison
the C7-C11 distance (i.e. the analogous R1-R2 distance)
across the amide in VX-770 was 3.76 Å, representing a 1.23 Å
lengthening of the linker between the quinolin-4(1H)-one core
and the phenol substituent upon replacement of the amide with
the triazole. Additionally, the orientation angle between the
quinolin-4(1H)-one and phenol substituents as defined in Figure
8D was increased upon substitution of the amide (143.4°) with
the triazole (161.5°). We wondered if the increased linker
More relevant than comparison of the solid-state structures
a structural comparison of the proposed bioactive
was
tautomeric isomer and quantification of whether the triazole
nitrogen atom could participate in H-bonding with the quinolinol
hydroxyl to stabilize the tautomer as we had hypothesized (see
Figure 6). To this end, we first performed a computational
geometry optimization on the solid-state crystal structures of
VX-770 and 60 at the M06-2X/6-31 + G(d) level of theory using
the experimentally-derived X-ray structures as the input
geometries. We then modeled the quinolinol tautomer for both
molecules by tautomerizing the CO bond and rotating the amide
or triazole moiety so that it could participate in an intramolecular
H-bonding interaction (as shown in Figure 9).
These
geometries were then optimized computationally. Additionally,
Natural Bond Orbital (NBO) analysis was performed on both of
the resulting optimized geometries to quantify the degree of
intramolecular H-bonding that stabilizes the tautomeric form.
Stabilization afforded to the molecule by hydrogen bonding of
the quinolinol O-H bond with the triazole nitrogen atom’s lone
pair was estimated to be 27.02 kcal/mol (as determined from
2nd order perturbation theory within the NBO analysis). This
confirmed that
a
substantial intramolecular H-bonding
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
triangle. The C7-C10-C11 angle was measured to be 143.4°. For 60 lines
were generated from C8-C10, C10-C12, and C8-C12 and the C8-C10-C12
angle was measured to be 161.5°.
interaction can exist with the triazole ring as a donor as shown
in Figure 9B which displays the relevant orbital interactions of
the H-bond as provided by NBO analysis. In comparison,
stabilization afforded by hydrogen bonding of the quinolinol O-H
bond with the amide group’s carbonyl oxygen lone pairs was
estimated to be 38.30 kcal/mol (see Figure 9A and Supporting
Information). This analysis confirmed that the 1,2,3-triazole ring
in 60 can participate in a H-bonding interaction with the
quinolinol hydroxyl group to stabilize the bioactive tautomer
similarly to the amide carbonyl within VX-770.
A
B
A
Figure 9. Optimized geometries for tautomeric isomers 34 and 35 at the
M06-2X/6-31 + G(d) level of theory. The resulting hydrogen bonds between
the tautomeric OH bond and the lone-pair donor atom in each molecule are
illustrated by the relevant orbital interactions as provided by NBO analysis. A
second, weaker, lone pair donation from the CO bond of 34 to the OH bond
(see Supporting Information) is not shown for the sake of clarity.
B
We also quantified the difference in energies (including
zero-point energy corrections) between the optimized non-
tautomeric and tautomeric conformations. This calculation
allows comparison of the energetic cost of assuming the
bioactive conformation.
Interestingly, assuming bioactive
conformation 34 from the optimized geometry of VX-770
requires 8.4 kcal/mol in energy, while the energetic cost to
access 35 from 60 was significantly less at 3.8 kcal/mol. This
analysis suggests that 60 is not energetically hindered from
achieving the bioactive tautomeric form, as evidenced by the
lower energetic cost required to assume the quinolinol
tautomeric form in comparison to VX-770.
C
The lower energetic barrier in the conversion of 60 to 35 is
likely due to the fact that an intramolecular H-bond already
exists between the amide and quinolinone carbonyl in VX-770
(red dotted line in Figure 8B). Converting to tautomeric
conformation 34, therefore, involves trading one hydrogen bond
in VX-770 (20.46 kcal/mol) for another, stronger hydrogen bond
in the tautomeric form (38.30 kcal/mol). The net stabilization of
17.84 kcal/mol from the new hydrogen bonding interaction
partially offsets the energetic cost of assuming the tautomeric
form. In contrast, no H-bonding interaction is present in the
non-tautomeric form of triazole analog 60 (Figure 8A) and so
the energy required to populate tautomeric form 35 from 60 is
offset to a greater degree by the formation of the new H-bond
interaction in 35 for a net stabilization of 27.02 kcal/mol.
D
Finally, we quantified the R1-R2 distance between the
amide and triazole substituents and angle of orientation
between the R1 and R2 substituents in the energy optimized
tautomeric forms as we did for the solid state crystal structures.
Interestingly, the C8-C12 distance across the triazole in 35 was
determined to be 4.99 Å and the C7-C11 distance across the
amide in 34 was 3.78 Å, both of which were very similar to the
solid state measurements. The orientation angles were also
comparable between the tautomeric forms and the solid state
as the orientation angle (as defined in Figure 8D) of the amide
Figure 8. A) Crystal structure of 60 showing representation of the thermal
ellipsoids. B) Crystal structure of VX-770 showing representation of the
thermal ellipsoids. H-bonding interaction is shown in red. C) Overlay of
crystal structures of 60 and VX-770. D) Calculation of orientation angle
between quinolin-4(1H)-one and phenol substituents using Olex2.[51] For VX-
770 lines were generated from C7-C10, C10-C11, and C7-C11 to define a
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
was found to be 145.6° and the orientation angle of the triazole
was determined to be 163.7°.
Comparison of the Metabolic Stability of 1,2,3-Triazole and
Amide-Containing CFTR Modulators
In summary, the NBO calculations suggest that triazole 60
should be able to assume the bioactive conformation 35.
Comparison of the energy optimized structures and X-ray
crystal structures highlight the lengthened R1-R2 distance and
increased angle of orientation that would position the triazole
substituents in a slightly different position in the CFTR binding
pocket relative to the amide substituents.
Having uncovered some drawbacks in the use of the 1,2,3-
triazole as an amide bioisostere in the VX-809 and VX-770
structural classes, we investigated whether the incorporation of
the triazole would afford beneficial pharmacokinetic properties
as has been demonstrated in some instances[14] and is widely
perceived to be the case.[2-5] To this end, the metabolic stability
of five triazoles and the analogous amides was evaluated in
human hepatic microsomes.
Patch Clamp Analysis of VX-770 and 1,2,3-Triazole Analog
60
Table 3. Intrinsic Clearance (CLINT) in Human Microsomes and Predicted
Hepatic Clearance (CLHEP) of Analogous Triazole and Amide-Containing
CFTR Modulators.
Given the computational evidence that the 1,2,3-triazole
isomers should be able to assume the bioactive tautomeric
conformation, we investigated the potentiator ability of 60 using
patch clamp analysis in order to study the interaction of our
triazole potentiators directly with the CFTR protein. We
hypothesized that this experiment would eliminate potential
cellular factors that may be responsible for the inactivity of 60 in
vitro and provide insight into whether structural changes arising
from incorporation of the 1,2,3-triazole prevent interaction of the
potentiator with the binding site of CFTR. Macroscopic currents
for WT-CFTR were recorded in the excised inside-out patch as
shown in Figure 10. Interestingly, we observed that triazole-
containing potentiator 60 robustly stimulated WT-CFTR currents
in a dose dependent manner. The potency and efficacy of
triazole 60 was similar to VX-770, as 60 exhibited an EC50 of
5.5 ±0.9 nM in comparison to an EC50 of 1.3 ±0.3 nM for VX-
770, and 60 increased the percent of the PKI control to a
1,2,3-triazole
amide
[a]
[b]
[a]
[b]
CLINT
CLHEP
(mL min-1
kg -1
CLINT
(mL min-1
kg -1
CLHEP
(mL min-1
kg -1
(mL min-1
cmpd
cmpd
kg -1
)
)
)
)
11
20
56
58
60
11.8
23.2
16.3
19.3
56.2
7.55
11.0
9.19
10.1
15.3
23
26
2.49
36.1
12.5
9.35
45.6
2.23
13.3
7.84
6.47
14.4
36
37
VX-770
[a] Intrinsic clearance in human hepatic microsomes. [b] Predicted
hepatic clearance, human.
slightly greater extent than VX-770.
These findings
demonstrate that under the cell-free condition, the 1,2,3-triazole
substitution within the VX-770 chemical series is tolerated and
that triazole-containing analogs can directly interact with CFTR
and potentiate the channel activity. Moreover, they raise the
question as to the cause of inactivity in the cellular assay.
Intrinsic clearance (CLINT) and calculated human hepatic
clearance (CLHEP) are shown in Table 3 (triazole and amide
pairs are listed side-by-side). Surprisingly, in all but one
instance (20 vs. 26) the triazole derivatives showed reduced
microsomal stability (i.e., increased CLINT) in comparison to the
analogous amide. In particular, triazole corrector 11 displayed
substantially increased CLINT in comparison to amide 23, and all
triazole potentiators (56, 58, 60) showed a modest increase in
microsomal clearance. In contrast to the common perception
that the triazole affords increased metabolic stability (due to
elimination of amide hydrolysis), the decreased stability of the
triazole analogs studied here demonstrates that the triazole
may also have a negative effect on hepatic clearance in certain
applications. This impact on hepatic clearance must also be
taken into account when considering the overall impact of the
triazole substitution on metabolic stability.
Conclusions
In conclusion, despite the widespread success of the 1,2,3-
triazole as an amide bioisostere, our results demonstrate that
use of the triazole is not effective in all medicinal chemistry
contexts and highlight several negative impacts that can arise
from this substitution. First, in the context of the VX-809
chemical series, while the 2,2-difluorobenzo[d][1,3]-dioxole
Figure 10. Dose response analysis for VX-770 and 60 by patch clamp
analysis. A/B) VX-770 and 60 titrations for representative patches. Protein
kinase A inhibitor (PKI) was applied to prevent further phosphorylation. C)
Mean titration data fit to Y = Bmax*X/(K1 + X), EC50 =1.3 ±0.3 (n=5) for VX-
770 and 5.5 ±0.9 (n=6) for 60. Data were normalized to the peak current at
100-200 nM. The results are reported as means ±S.E. The lowest dose tested
was 0.2 nM, which increased the current by ~12% for VX-770 and ~7% for 60.
D) Mean data % of maximal stimulation (efficacy) by VX-770 and 60.
cyclopropane 1,2,3-triazole was found to be
a suitable
pharmacophore for F508del-CFTR correction, the triazole
derivative displayed significantly reduced corrector efficacy as
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
1
procedures, compound characterization data, copies of H and 13C NMR
11 increased Cl ion transport in F508del CFBE41o cells to only
48% of the 3 μM VX-809 response. Additionally, triazole
corrector 20, containing the 4-methoxyphenyl group, was
inactive in contrast to the analogous amide 26 which displayed
modest F508del-CFTR correction (54% of the 3 μM VX-809
response). These findings highlight that the triazole, while
tolerated in the VX-809 chemical series when the 2,2-
difluorobenzo[d][1,3]-dioxole cyclopropane moiety is maintained,
is not widely applicable across less efficacious correctors in this
series.
spectra, X-ray crystallography data, and details on computational studies
are provided. Biology: Experimental procedures for cell culture, TECC
assay, Ussing Chamber recordings, patch clamp analysis, and intrinsic
clearance assays are provided.
CCDC data 1879006 (VX-770) and 1879007 (60) contain the
supplementary crystallographic data for this paper. These data are
provided free of charge by The Cambridge Crystallographic Data Centre.
Secondly, in the VX-770 chemical series, the use of the
triazole substitution completely abolished potentiator activity in
the cellular assays. These findings point to a significant
structural change to the VX-770 scaffold resulting from the
triazole substitution. On the basis of Natural Bond Orbital
calculations performed on triazole 60, this structural change is
not likely to be disruption of the proposed bioactive tautomeric
form as the triazole was found to be able to participate in the
necessary intramolecular H-bonding interaction that stabilizes
the tautomer. Furthermore, quantification of the difference in
energy between the optimized non-tautomeric and tautomeric
conformations provided evidence that 60 is not energetically
Acknowledgements
The authors gratefully acknowledge the following sources of
funding: J.E.D. thanks the Richards Scholar Program for
financial support. S.G.A. acknowledges funding by the Cystic
Fibrosis Foundation (ALLER16P0 and ALLER16G0). S.M.R.
acknowledges funding from the NIH (P30 DK072482 and R35
HL135816) and the Cystic Fibrosis Foundation (R464-CR11).
The authors gratefully thank Prof. Craig Lindsley of Vanderbilt
University for access to microsomal stability analysis, compound
purity analysis, and HRMS analysis. The authors thank Dr. John
Bacsa at the X-ray Crystallography Center at Emory University
for X-ray data collection.
hindered from achieving the bioactive conformation.
The
conclusions drawn from these calculations were supported by
the similar potency and efficacy of 60 in comparison to VX-770
in the cell-free patch clamp experiment. The patch clamp results
suggest that the inactivity of triazole potentiators 56, 58, and 60
in the cellular assay could be due to the inability of these
compounds to access the CFTR binding site under cellular
conditions, either due to poor membrane permeability or active
Conflict of interest
The authors declare no conflict of interest.
Keywords: triazole • bioisostere • CFTR • corrector • potentiator
transport out of the cell.
These findings highlight the
considerable impact the triazole substitution can have on crucial
physiochemical properties that impact target engagement.
Thirdly, comparison of the solid-state structures of 60 and
VX-770 and comparison of the energy optimized tautomeric
forms 34 and 35 highlight the structural differences resulting
from the triazole substitution. The distance across the amide
substituents increases by ~1.2 Å in the triazole derivative, and
the angle of orientation between the amide substituents is
increased by ~18.1° in the triazole. While these structural
changes can be tolerated in many instances without negative
consequences on biological activity and/or physiochemical
properties,2-5 this did not prove to be the case in the VX-809 and
VX-770 chemical series.
[1]
[2]
J. E. Hein, V. V. Fokin, Chem. Soc. Rev. 2010, 39, 1302–1315.
E. Bonandi, M. S. Christodoulou, G. Fumagalli, D. Perdicchia, G.
Rastelli, D. Passarella, Drug Discov. Today 2017, 22, 1572–1581.
J. Hou, X. Liu, J. Shen, G. Zhao, P. G. Wang, Expert Opin. Drug
Discov. 2012, 7, 489–501.
[3]
[4]
G. C. Tron, T. Pirali, R. A. Billington, P. L. Canonico, G. Sorba, A. A.
Genazzani, Med. Res. Rev. 2008, 28, 278–308.
[5]
[6]
H. C. Kolb, K. B. Sharpless, Drug Discov. Today 2003, 8, 1128–1137.
E. D. Chrysina, É. Bokor, K.-M. Alexacou, M.-D. Charavgi, G. N.
Oikonomakos, S. E. Zographos, D. D. Leonidas, N. G. Oikonomakos,
L. Somsák, Tetrahedron Asymmetry 2009, 20, 733–740.
W. S. Horne, M. K. Yadav, C. D. Stout, M. R. Ghadiri, J. Am. Chem.
Soc. 2004, 126, 15366–15367.
I. Mohammed, I. R. Kummetha, G. Singh, N. Sharova, G. Lichinchi, J.
Dang, M. Stevenson, T. M. Rana, J. Med. Chem. 2016, 59, 7677–
7682.
[7]
[8]
Finally, despite the widespread perception that use of the
1,2,3-triazole can provide increased metabolic stability, in two
distinct chemical series 1,2,3-triazole derivatives largely
displayed similar or increased metabolic clearance in human
[9]
F. Arioli, S. Borrelli, F. Colombo, F. Falchi, I. Filippi, E. Crespan, A.
Naldini, G. Scalia, A. Silvani, G. Maga, F. Carraro, M. Botta, D.
Passarella, ChemMedChem 2011, 6, 2009–2018.
[10]
[11]
C. Peruzzotti, S. Borrelli, M. Ventura, R. Pantano, G. Fumagalli, M. S.
Christodoulou, D. Monticelli, M. Luzzani, A. L. Fallacara, C. Tintori, M.
Botta, D. Passarella, ACS Med. Chem. Lett. 2013, 4, 274–277.
M. S. Christodoulou, M. Mori, R. Pantano, R. Alfonsi, P. Infante, M.
Botta, G. Damia, F. Ricci, P. A. Sotiropoulou, S. Liekens, B. Botta, D.
Passarella, ChemPlusChem 2015, 80, 938–943.
hepatic microsomal assays. In the case of triazole 11 (CLINT
=
11.8) the increase in intrinsic clearance was substantial (~5x) in
comparison to the analogous amide 23 (CLINT = 2.49). In total,
these findings represent an important contrast to the many
successful uses of the 1,2,3-triazole as an amide bioisostere in
medicinal chemistry.
[12]
[13]
T. E. Ballard, J. J. Richards, A. L. Wolfe, C. Melander, Chem. – Eur. J.
2008, 14, 10745–10761.
C. Mugnaini, S. Nocerino, V. Pedani, S. Pasquini, A. Tafi, M. De
Chiaro, L. Bellucci, M. Valoti, F. Guida, L. Luongo, S. Dragoni,
A.Ligresti, A. Rosenberg, D. Bolognini, M. G. Cascio, R. G. Pertwee,
R. Moaddel, S. Maione, V. Di Marzo, F. Corelli, ChemMedChem 2012,
7, 920–934.
Experimental Section
[14]
[15]
T. M. Keck, A. K. Banala, R. D. Slack, C. Burzynski, A. Bonifazi, O. M.
Okunola-Bakare, M. Moore, J. R. Deschamps, R. Rais, B. S. Slusher,
A. H. Newman, Bioorg. Med. Chem. 2015, 23, 4000–4012.
S. K. Mamidyala, M. G. Finn, Chem. Soc. Rev. 2010, 39, 1252.
Complete experimental procedures and Supplementary Schemes and
Figures are provided in the Supporting Information. Chemistry:
Supplementary Schemes and Figures, synthetic experimental
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
[16]
T.-C. Hwang, J.-T. Yeh, J. Zhang, Y.-C. Yu, H.-I. Yeh, S. Destefano, J.
Gen. Physiol. 2018, jgp.201711946.
D. N. Sheppard, M. J. Welsh, Physiol. Rev. 1999, 79, S23–S45.
P. A. Sloane, S. M. Rowe, Curr. Opin. Pulm. Med. 2010, 16, 591–597.
M. T. Jennings, P. A. Flume, Ann. Am. Thorac. Soc. 2018, 15, 897–
902.
P. R. Sosnay, K. S. Raraigh, R. L. Gibson, Pediatr. Clin. North Am.
2016, 63, 585–598.
J. L. Bobadilla, M. Macek Jr., J. P. Fine, P. M. Farrell, Hum. Mutat.
2002, 19, 575–606.
A. W. R. Serohijos, T. Hegedus, A. A. Aleksandrov, L. He, L. Cui, N.
V. Dokholyan, J. R. Riordan, Proc. Natl. Acad. Sci. 2008, 105, 3256–
3261.
D. Penque, F. Mendes, S. Beck, C. Farinha, P. Pacheco, P. Nogueira,
J. Lavinha, R. Malhó, M. D. Amaral, Lab. Invest. 2000, 80, 857–868.
F. Van Goor, S. Hadida, P. D. J. Grootenhuis, B. Burton, J. H. Stack,
K. S. Straley, C. J. Decker, M. Miller, J. McCartney, E. R. Olson, J. J.
Wine, R. A. Frizzell, M. Ashlock, P. A. Negulescu, Proc. Natl. Acad.
Sci. 2011, 108, 18843–18848.
[34]
R. Hadida, P. D. J. Grootenhuis, J. Zhou, B. Bear, M. Miller, J.
McCartney, Pyridine Compounds as Modulators of CFTR and Their
Preparation and Use in the Treatment of Diseases. WO 2008/141119,
November 20, 2008.
D. Siesel, Processes for Producing Cycloalkylcarboxamido-Pyridine
Benzoic Acids. US 8,124,781 B2, February 28, 2012.
K. K. Goncz, K. Kunzelmann, Z. Xu, D. C. Gruenert, Hum. Mol. Genet.
1998, 7, 1913–1919.
[17]
[18]
[19]
[35]
[36]
[37]
[38]
[39]
[20]
[21]
[22]
C. Ehrhardt, E.-M. Collnot, C. Baldes, U. Becker, M. Laue, K.-J. Kim,
C.-M. Lehr, Cell Tissue Res. 2006, 323, 405–415.
P.-W. Phuan, G. Veit, J. Tan, A. Roldan, W. E. Finkbeiner, G. L.
Lukacs, A. S. Verkman, Mol. Pharmacol. 2014, 86, 42–51.
S. Hadida, F. Van Goor, J. Zhou, V. Arumugam, J. McCartney, A.
Hazlewood, C. Decker, P. Negulescu, P. D. J. Grootenhuis, J. Med.
Chem. 2014, 57, 9776–9795.
A. Almeida, A. Silva, J. Cavaleiro, Synlett 2010, 2010, 462–466.
C. Mugnaini, C. Falciani, M. De Rosa, A. Brizzi, S. Pasquini, F. Corelli,
Tetrahedron 2011, 67, 5776–5783.
[23]
[24]
[40]
[41]
[42]
S. Venkataraman, D. K. Barange, M. Pal, Tetrahedron Lett. 2006, 47,
[25]
[26]
[27]
W. Wang, G. Li, J. P. Clancy, K. L. Kirk, J. Biol. Chem. 2005, 280,
23622–23630.
H. Miki, Z. Zhou, M. Li, T.-C. Hwang, S. G. Bompadre, J. Biol. Chem.
2010, 285, 19967–19975.
F. V. Goor, S. Hadida, P. D. J. Grootenhuis, B. Burton, D. Cao, T.
Neuberger, A. Turnbull, A. Singh, J. Joubran, A. Hazlewood, J. Zhou,
J. McCartney, V. Arumugam, C. Decker, J. Yang, C. Young, E. R.
Olson, J. J. Wine, R. A. Frizzell, M. Ashlock, P. Negulescu, Proc. Natl.
Acad. Sci. 2009, 106, 18825-18830.
7317–7322.
[43]
[44]
K. Beydoun, H. Doucet, Eur. J. Org. Chem. 2012, 2012, 6745–6751.
S. Thatipally, V. K. Reddy, V. L. N. R. Dammalapati, S. Chava,
Processes for the Preparation of Ivacaftor. WO 2017/037672 A1,
March 9, 2017.
D. Audisio, S. Messaoudi, J.-F. Peyrat, J.-D. Brion, M. Alami, J. Org.
Chem. 2011, 76, 4995–5005.
F. Zhang, J. E. Moses, Org. Lett. 2009, 11, 1587–1590.
S. Thatipally, V. K. Reddy, V. L. N. R. Dammalapati, S. R. A. Gorantla,
S. Chava, An Improved Process for the Preparation of 5-Amino-2,4,-
Di-Tert-Butylphenol or an Acid Additional Salt Thereof. WO
2016/075703 A2, May 19, 2016.
K.-Y. Jih, T.-C. Hwang, Proc. Natl. Acad. Sci. 2013, 110, 4404–4409.
L. B. Gottschalk, B. Vecchio-Pagan, N. Sharma, S. T. Han, A. Franca,
E. S. Wohler, D. A. S. Batista, L. A. Goff, G. R. Cutting, J. Cyst. Fibros.
2016, 15, 285–294.
[45]
[46]
[47]
[28]
[29]
[30]
H. Matsui, B. R. Grubb, R. Tarran, S. H. Randell, J. T. Gatzy, C. W.
Davis, R. C. Boucher, Cell 1998, 95, 1005–1015.
P. R. Kym, X. Wang, M. Pizzonero, S. E. Van der Plas, in Prog. Med.
Chem. (Eds.: D.R. Witty, B. Cox), Elsevier, 2018, pp. 235–276.
X. Wang, B. Liu, X. Searle, C. Yeung, A. Bogdan, S. Greszler, A.
Singh, Y. Fan, A. M. Swensen, T. Vortherms, C. Balut, Y. Jia, K.
Desino, W. Gao, H. Yong, C. Tse, P. Kym, J. Med. Chem. 2018, 61,
1436–1449.
T. Okiyoneda, G. Veit, J. F. Dekkers, M. Bagdany, N. Soya, H. Xu, A.
Roldan, A. S. Verkman, M. Kurth, A. Simon, T. Hegedus, J. M.
Beekman, G. L. Lukacs, Nat. Chem. Biol. 2013, 9, 444–454.
P. D. Eckford, M. Ramjeesingh, S. Molinski, S. Pasyk, J. F. Dekkers,
C. Li, S. Ahmadi, W. Ip, T. E. Chung, K. Du, H. Yeger, J. Beekman, T.
Gonska, C. E. Bear, Chem. Biol. 2014, 21, 666–678.
A. Bolje, D. Urankar, J. Košmrlj, Eur. J. Org. Chem. 2014, 8167–8181.
[48]
[49]
[50]
[51]
H. Yu, B. Burton, C.-J. Huang, J. Worley, D. Cao, J. P. Johnson, A.
Urrutia, J. Joubran, S. Seepersaud, K. Sussky, B. J. Hoffman, F. Van
Goor, J. Cyst. Fibros. 2012, 11, 237–245.
O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard, H.
Puschmann, J. Appl. Crystallogr. 2009, 42, 339–341.
[31]
[32]
[33]
This article is protected by copyright. All rights reserved.

10.1002/chem.201805919
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Text for Table of Contents
Author(s), Corresponding Author(s)*
Page No. – Page No.
Title
((Insert TOC Graphic here: max.
width: 5.5 cm; max. height: 5.0 cm))
Layout 2:
FULL PAPER
Jake E. Doiron, Christina A. Le, Britton K.
Ody, Jonathon B. Brace, Savannah J.
Post, Nathan L. Thacker, Harrison M. Hill,
Prof. Dr. Gary W. Breton, Matthew J.
Mulder, Sichen Chang, Dr. Thomas M.
Bridges, Liping Tang, Dr. Wei Wang, Prof.
Dr. Steven M. Rowe, Prof. Dr. Stephen G.
Aller,* and Prof. Dr. Mark Turlington*
Page No. – Page No.
Trying out the triazole as an amide bioisostere in CFTR modulators VX-809 and
VX-770 revealed that triazole derivatives of VX-809 and VX-770 displayed
significantly decreased efficacy or inactivity in cellular assays. In addition, triazole
derivatives displayed decreased metabolic stability in hepatic microsomal assays.
These findings represent an important counterpoint to the many successful uses of
the triazole as an amide bioisostere.
Evaluation of 1,2,3-Triazoles as
Amide Bioisosteres In Cystic Fibrosis
Transmembrane Conductance
Regulator Modulators VX-770 and VX-
809
This article is protected by copyright. All rights reserved.