Amphiphilic pyridinium salts block TNFα/NFκB signaling and constitutive hypersecretion of interleukin-8 (IL-8) from cystic fibrosis lung epithelial cells

Article abstract of DOI:10.1016/j.bcp.2005.05.002

Cystic fibrosis (CF) is a common, lethal genetic disease, which is due to mutations in the CFTR gene. The CF lung expresses a profoundly proinflammatory phenotype, due to constitutive hypersecretion of IL-8 from epithelial cells lining the airways. In a systematic search for candidate drugs that might be used therapeutically to suppress IL-8 secretion from these cells, we have identified a potent and efficacious series of amphiphilic pyridinium salts. The most potent of these salts is MRS2481, an (R)-1-phenylpropionic acid ester, with an IC₅₀ of ca. 1 μM. We have synthesized 21 analogues of MRS2481, which have proven sufficient to develop a preliminary structure-activity relationship (SAR). For optimal activity, we have found that the ester must be connected to the pyridinium derivative by an eight-carbon chain. An optical isomer of the lead compound, containing an (S)-1-phenylpropionic acid ester, has been found to be a much less active. The mechanism of action of MRS2481 appears to involve inhibition of signaling of the NFκB and AP-1 transcription factors to the IL-8 promoter. MRS2481 is a potent inhibitor of TNFα-induced phosphorylation and proteosomal destruction of IκBα. Inasmuch as IκBα is the principal inhibitor of the NFκB signaling pathway, preservation of intact IκBα would serve to keep the IL-8 promoter silent. We also find that MRS2481 blocks TNFα-activated phosphorylation of JNK, the c-JUN kinase. The IL-8 promoter is also activated by an AP-1 site, which requires a phospho-c-JUN/c-FOS dimer for activity. We therefore interpret these data to suggest that the mechanism of MRS2481 action is to inhibit both NFκB and AP-1 signaling on the IL-8 promoter. Given the medicinally promising properties of water-solubility, potency in the low μM concentration range, and high efficacy, we anticipate that MRS2481, or a further optimized derivative, may find an important place in the armamentarium of pharmaceutical strategies yet to be arrayed against the inflammatory phenotype of the CF lung.

Full text of DOI:10.1016/j.bcp.2005.05.002

Biochemical Pharmacology 70 (2005) 381–393
www.elsevier.com/locate/biochempharm
Amphiphilic pyridinium salts block TNFa/NFkB signaling and
constitutive hypersecretion of interleukin-8 (IL-8) from
cystic fibrosis lung epithelial cells
Susanna Tchilibon a, Jian Zhang b, QingFeng Yang b, Ofer Eidelman b, Haksung Kim a,
Hung Caohuy b, Kenneth A. Jacobson a, Bette S. Pollard c, Harvey B. Pollard b,
*
a Laboratory of Bioorganic Chemistry, NIDDK, National Institutes of Health, Bethesda, MD 20892, USA
b Department of Anatomy, Physiology and Genetics, and Institute of Molecular Medicine, F. Edward Hebert School of Medicine,
Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, MD 20814, USA
c Office of Information Technology, Equal Employment Opportunity Commission Headquarters, Washington, DC 20507, USA
Received 25 February 2005; accepted 2 May 2005
Abstract
Cystic fibrosis (CF) is a common, lethal genetic disease, which is due to mutations in the CFTR gene. The CF lung expresses a
profoundly proinflammatory phenotype, due to constitutive hypersecretion of IL-8 from epithelial cells lining the airways. In a systematic
search for candidate drugs that might be used therapeutically to suppress IL-8 secretion from these cells, we have identified a potent and
efficacious series of amphiphilic pyridinium salts. The most potent of these salts is MRS2481, an (R)-1-phenylpropionic acid ester, with an
IC50 of ca. 1 mM. We have synthesized 21 analogues of MRS2481, which have proven sufficient to develop a preliminary structure–
activity relationship (SAR). For optimal activity, we have found that the ester must be connected to the pyridinium derivative by an eight-
carbon chain. An optical isomer of the lead compound, containing an (S)-1-phenylpropionic acid ester, has been found to be a much less
active. The mechanism of action of MRS2481 appears to involve inhibition of signaling of the NFkB and AP-1 transcription factors to the
IL-8 promoter. MRS2481 is a potent inhibitor of TNFa-induced phosphorylation and proteosomal destruction of IkBa. Inasmuch as IkBa
is the principal inhibitor of the NFkB signaling pathway, preservation of intact IkBa would serve to keep the IL-8 promoter silent. We also
find that MRS2481 blocks TNFa-activated phosphorylation of JNK, the c-JUN kinase. The IL-8 promoter is also activated by an AP-1
site, which requires a phospho-c-JUN/c-FOS dimer for activity. We therefore interpret these data to suggest that the mechanism of
MRS2481 action is to inhibit both NFkB and AP-1 signaling on the IL-8 promoter. Given the medicinally promising properties of water-
solubility, potency in the low mM concentration range, and high efficacy, we anticipate that MRS2481, or a further optimized derivative,
may find an important place in the armamentarium of pharmaceutical strategies yet to be arrayed against the inflammatory phenotype of
the CF lung.
Published by Elsevier Inc.
Keywords: Cystic fibrosis; Inflammation; Lung; Interleukin 8; Epithelium; NFkB; AP-1
1. Introduction
epithelial cells lining the CF airways [3–10]. IL-8 is the
most potent known chemoattractor for leucocytes into
inflammatory foci [11]. The IL-8 gene is principally driven
by the NFkB signaling pathway, and to a lesser extent by
the AP-1 (c-FOS/c-JUN) pathway, and certain other tran-
scription factors [12,13]. The most common CF mutation,
[DF508]CFTR, generates a mutant CFTR protein that fails
to traffick properly to apical plasma membranes in epithe-
lial cells of the CF airway [14]. The cAMP-activated
chloride channel activity of the CFTR protein is thereby
lost. However, while this mutant trafficking defect appears
Cystic fibrosis (CF) is the most common, lethal auto-
somal recessive disease in the United States and Europe
[1,2]. The disease is due to mutations in the CFTR gene,
which causes a profoundly proinflammatory phenotype in
the CF lung. The proinflammatory mechanism is based on
constitutive hypersecretion of interleukin-8 (IL-8) from the
* Corresponding author. Tel.: +1 301 295 3661; fax: +1 301 295 2822.
E-mail address: hpollard@usuhs.mil (H.B. Pollard).
0006-2952/$ – see front matter. Published by Elsevier Inc.
doi:10.1016/j.bcp.2005.05.002

382
S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
to be correlated with the pro-inflammatory disease pheno-
type, it is possible to reduce IL-8 hypersecretion pharma-
ceutically from CF lung cells with little or no correction of
the trafficking defect.
observed substantial and informative variation in activ-
ities. We anticipate that MRS2481 may not only be useful
as a tool to investigate the mechanism of the CF pro-
inflammatory phenotype, but may also prove to be a
prototype for a novel CF pharmaceutical.
The candidate CF drug CPX illustrates how a close
relationship can exist between correction of the trafficking
defect, activation of the cAMP-dependent chloride channel
activity, and suppression of IL-8 hypersecretion [15–22].
CPX is a xanthine with a ka for trafficking rescue of ca. 1–
5 mM, which also suppresses IL-8 hyperexpression. The
mechanism is by inhibition of the TNFa/NFkB signaling
pathway [22]. The same mechanism is responsible for
inhibition of IL-8 secretion from CF lung epithelial cells
by gene therapy with [wildtype]CFTR [22].
2. Materials and methods
2.1. Chemical synthesis
2.1.1. Chemical reagents and instrumentation
Reagents and solvents were purchased from Sigma-
1
Aldrich (St. Louis, MO). H NMR spectra were obtained
The candidate CF drug digitoxin is an example of how
IL-8 hypersecretion can be suppressed in CF cells with
only modest rescue of mutant CFTR [23]. Digitoxin
potently suppresses hyperexpression of IL-8 (ka = ca.
1 nM; Ref. [23]). Nonetheless, digitoxin also returns to
control levels 62% of the genes that are affected in
expression by the mutant CFTR [23]. The mechanism of
digitoxin action on the expression of IL-8 is also to block
the TNFa/NFkB signaling pathway. Other compounds
have been studied with respect to their ability to correct
trafficking and/or channel activity of mutant CFTR, but
have not yet been characterized in terms of their capacity to
suppress IL-8 hypersecretion from CF cells [24–29].
The fundamental importance of the inflammatory arm
of CF for lung disease is particularly demonstrated by
clinical experience with ibuprofen. For those CF patients
able to tolerate the required massively high doses, ibu-
profen stops the progressive loss of lung function typical
of the disease [30]. The mechanism of action of ibuprofen
is to reduce the number of inflammatory leucocytes in the
circulation that are available to be recruited into sites of
IL-8 expression in the CF airway. We have therefore
interpreted these data to suggest that successful CF drug
discovery can be achieved by specifically targeting con-
stitutive hypersecretion of IL-8 by CF lung epithelial
cells. Accurso et al. [31] have emphasized that the most
important property of a clinically useful CF drug is the
ability to specifically suppress the pro-inflammatory CF
phenotype.
with a Varian Gemini-300 spectrometer (300 MHz) with
D2O, CDCl3, CD3OD, and DMSO-d6 as a solvent. The
chemical shifts are expressed as ppm downfield from
tetramethylsilane. Purity of compounds was checked with
a Hewlett-Packard 1090 HPLC equipped with a Luna 5 m
RP-C18(2) analytical column (250 mm 4.6 mm; Phe-
nomenex, Torrance, CA). System A: linear gradient sol-
vent system: H2O/CH3CN from 95/5 to 20/80 in 20 min;
the flow rate was 1 ml/min. System B: linear gradient
solvent system: 5 mM tetrabutylammonium phosphate/
CH3CN from 80/20 to 20/80 in 20 min, then isocratic
for 2 min; the flow rate was 1 ml/min. Peaks were detected
by UV absorption with a diode array detector. All deriva-
tives tested for biological activity showed >96% purity
in the HPLC systems. TLC analysis was carried out
on aluminum sheets precoated with silica gel F254
(0.2 mm) from Aldrich. Low-resolution FAB mass spectro-
metry was performed with a JEOL SX102 spectrometer
with 6 kV Xe atoms following desorption from a glycerol
matrix.
2.1.1.1. Method A. General procedure for the synthesis of
the compounds 1–14 and 16–18.
R-8-Bromo-n-octyl-a-methyl-2-phenylacetate (25). R-
(À)-2-Phenylpropionic acid (210 mg, 1.4 mmol) and
1,8-dibromo-octane (0.26 ml, 1.4 mmol) were put in a
5 ml round bottom flask with a methanolic solution of
benzyltrimethylammonium hydroxide (0.65 ml, 40%).
Tetrabutylammonium iodide (14 mg, 0.038 mmol) was
added and the mixture was stirred for 3 days. The mixture
was poured in water (30 ml), and the aqueous solution was
extracted with ethyl acetate (10 ml 3Â). The organic phase
was dried over Na2SO4, filtered and concentrated. The
residue was purified using preparative thin layer chroma-
tography (silica gel, eluting with petroleum ether:ethyl
acetate, 20:1) obtaining 181 mg of pure 25 (yield 38%).
1H NMR (CDCl3): d 7.34–7.24 (m, 5H), 4.05
(t, J = 6.3 Hz, 2H), 3.70 (q, J = 6.9 Hz), 3.41 (t, J =
6.3 Hz, 2H), 1.90–1.80 (m, 1H), 1.56–1.37 (m, 7H), 1.35–
1.20 (m, 6H). MS (m/e) (positive FAB) 341.1 (M + H)+.
R-8-Iodo-n-octyl-a-methyl-2-phenylacetate (26). 25
(200 mg, 0.58 mmol) was dissolved in acetone (5 ml)
In this paper we describe the discovery of MRS2481, a
new amphiphilic pyridinium salt that potently and effica-
ciously suppresses IL-8 secretion. Although somewhat
less potent than digitoxin, MRS2481 is analogous to
digitoxin in that its mechanism involves the blockade of
the TNFa/NFkB signaling pathway. Furthermore, like
digitoxin, MRS2481 has only modest correction activity
on the mutant CFTR trafficking defect. The structural
requirements responsible for MRS2481 potency and effi-
cacy have been investigated in order to establish a struc-
ture–activity relationship (SAR). To this end we
synthesized 22 different MRS2481 homologues by the
methods of classical organic chemical synthesis. We

S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
383
and NaI (mg 90, 10.6 mmol) was added. The mixture was
stirred at r.t. overnight. The solvent was concentrated and
water (20 ml) was added, and the aqueous phase was
extracted with ethyl acetate (10 ml Â3). The organic phase
was dried over Na2SO4, filtered and concentrated. The
product was sufficiently pure to be used without further
purification.
1.24–1.17 (m, 8H). MS (m/e) (positive FAB) 356.2
(M À I)+.
1
8-Pyridinium-n-octyl-propionate iodide (7). H NMR
(D2O) d 8.88 (d, J = 6.6 Hz, 2H), 8.66 (t, J = 9 Hz, 1H),
8.14 (t, J = 6.9 Hz, 2H), 7.47–7.38 (m, 5H), 4.55
(t, J = 7.2 Hz, 2H), 4.25 (t, J = 4.2 Hz, 2H), 3.85 (s, 2H),
2.12–1.98 (m, 2H), 1.56–1.37 (m, 4H), 1.30–1.19 (m, 6H).
MS (m/e) (positive FAB) 326.2 (M À I)+.
1H NMR (CDCl3): d 7.32–7.24 (m, 5H), 4.05
(t, J = 6.3 Hz, 2H), 3.70 (q, J = 6.9 Hz), 3.18 (t, J = 6.3 Hz,
2H), 1.84–1.79 (m, 2H), 1.56–1.37 (m, 7H), 1.35–1.23
(m, 6H). MS (m/e) (positive FAB) 389.1 (M + H)+.
R-8-Pyridinium-n-octyl-a-methyl-2-phenylacetate
iodide (3). 26 (100 mg, 0.26 mmol) was dissolved in
acetone (10 ml) and pyridine (0.3 ml) was added. The
solution was stirred at 50 8C for three days. The solvent
was evaporated and the residue dissolved in water (20 ml).
The aqueous phase was washed with ether (10 ml Â3) and
lyophilized to give 54 mg of pure 3 (yield 45%).
1H NMR (D2O): d 8.80 (d, J = 6.6 Hz, 2H), 8.57
(t, J = 9 Hz, 1H), 8.09 (t, J = 6.9 Hz, 2H), 7.45–7.31
(m, 5H), 4.53 (t, J = 7.2 Hz, 2H), 4.19–4.05 (m, 2H), 3.88
(q, J = 6.9 Hz, 1H), 1.94–1.88 (m, 2H), 1.61–1.56 (m, 2H),
1.49–1.42 (m, 7H), 1.27–1.23 (m, 4H). MS (m/e) (positive
FAB) 340.1 (M À I)+.
R-8-Pyridinium-n-octyl-a-ethyl-2-phenylacetate iodide
(8). 1H NMR (D2O) d 9.31 (d, J = 6.6 Hz, 2H), 8.49
(t, J = 9 Hz, 1H), 8.10 (t, J = 6.9 Hz, 2H), 7.31–7.22
(m, 5H), 4.93 (t, J = 7.2 Hz, 2H), 4.06–4.01 (m, 2H),
3.48–3.41 (m, 1H), 2.11–1.90 (m, 3H), 1.84–1.74
(m, 1H), 1.61–1.46 (m, 4H), 1.32–1.20 (m, 6H), 0.89
(t, J = 7.2 Hz, 2H). MS (m/e) (positive FAB)355.2(M À I)+.
S-8-Pyridinium-n-octyl-a-methoxy-2-phenylacetate
1
iodide (9). H NMR (D2O) d 8.82 (d, J = 6.6 Hz, 2H),
8.57 (t, J = 9 Hz, 1H), 8.09 (t, J = 6.9 Hz, 2H), 7.48–7.42
(m, 5H), 4.59 (t, J = 7.2 Hz, 2H), 4.29–4.10 (m, 3H), 3.41
(s, 3H), 2.02–1.88 (m, 2H), 1.61–1.52 (m, 2H), 1.29–1.12
(m, 8H). MS (m/e) (positive FAB) 356.2 (M À I)+.
S-8-Pyridinium-n-octyl-2-benzyloxy-propionate iodide
1
(10). H NMR (D2O) d 8.81 (d, J = 6.6 Hz, 2H), 8.57
(t, J = 9 Hz, 1H), 8.07 (t, J = 6.9 Hz, 2H), 7.45–7.40
(m, 5H), 4.62–4.58 (m, 4H), 4.25 (q, J = 6.9 Hz, 1H),
4.18 (t, J = 7.2 Hz, 2H), 2.02–1.94 (m, 2H), 1.72–1.60
(m, 2H), 1.40 (d, J = 7.2 Hz, 3H), 1.38–1.26 (m, 8H). MS
(m/e) (positive FAB) 370.3 (M À I)+.
R-8-Pyridinium-n-butyl-a-methyl-2-phenylacetate
1
iodide (1). H NMR (D2O): d 8.67 (d, J = 6.6 Hz, 2H),
8.57 (t, J = 9 Hz, 1H), 8.07 (t, J = 6.9 Hz, 2H), 7.43–7.31
(m, 5H), 4.45 (t, J = 7.2 Hz, 2H), 4.29–4.09 (m, 2H), 3.90
(q, J = 6.9 Hz, 1H), 1.92–1.82 (m, 2H), 1.72–1.65 (m, 2H),
1.49 (d, J = 7.2 Hz, 3H). MS (m/e) (positive FAB) 284.1
(M À I)+.
1
8-Pyridinium-n-octyl-benzoate iodide (11). H NMR
(D2O) d 8.81 (d, J = 6.6 Hz, 2H), 8.51 (t, J = 9 Hz, 1H),
8.03 (m, 4H), 7.65–7.55 (m, 1H), 7.56–7.51 (m, 2H),
4.59–4.55 (m, 2H), 4.34 (t, J = 7.2 Hz, 2H), 4.05
(q, J = 6.9 Hz, 1H), 2.02–1.94 (m, 2H), 1.78–1.74
(m, 2H), 1.62–1.26 (m, 8H). MS (m/e) (positive FAB)
312.3 (M À I)+.
R-8-Pyridinium-n-hexyl-a-methyl-2-phenylacetate
1
iodide (2). H NMR (D2O) d 8.81 (d, J = 6.6 Hz, 2H),
8.57 (t, J = 9 Hz, 1H), 8.08 (t, J = 6.9 Hz, 2H), 7.45–7.30
(m, 5H), 4.3 (t, J = 7.2 Hz, 2H), 4.21–4.12 (m, 2H), 3.89
(q, J = 6.9 Hz, 1H), 1.94–1.86 (m, 2H), 1.62–1.55 (m, 2H),
1.48 (d, J = 7.2 Hz, 3H), 1.27–1.22 (m, 4H). MS (m/e)
(positive FAB) 312.1 (M À I)+.
R-8-Pyridinium-n-octyl-2-hydroxy-3-phenylpropionate
iodide (12) and S-8-pyridinium-n-octyl-2-hydroxy-3-phe-
nylpropionate iodide (13). 1H NMR (D2O) d 8.80
(d, J = 6.6 Hz, 2H), 8.57 (t, J = 9 Hz, 1H), 8.09 (t, J =
6.9 Hz, 2H), 7.38–7.22 (m, 5H), 4.65 (t, J = 7.2 Hz, 2H),
4.44 (q, J = 6.8 Hz, 1H), 4.14 (t, J = 7.2 Hz, 2H), 3.18–3.02
(m, 2H), 1.95–1.82 (m, 2H), 1.69–1.62 (m, 2H), 1.45–1.15
(m, 8H). MS (m/e) (positive FAB) 356.2 (M À I)+.
8-Pyridinium-n-octyl-a-methyl-2-[4-(2-methylpropyl)-
phenyl]acetate iodide (14). 1H NMR (CD3Cl3) d 9.35 (br s,
2H), 8.56–8.54 (m, 1H), 8.14 (br s, 1H), 7.18 (d, J =
8.1 Hz, 2H), 7.08 (d, J = 8.1 Hz, 2H), 4.98–4.81
(m, 2H), 4.05–3.98 (m, 2H), 3.94–3.62 (m, 1H), 2.43
(d, J = 7.2 Hz, 1H), 2.15 (br s, 2H), 1.88–1.78 (m, 1H),
1.59–1.18 (m, 10H), 0.88 (d, J = 6.6 Hz, 6H). MS (m/e)
(positive FAB) 396.3 (M À I)+.
R-8-Pyridinium-n-decyl-a-methyl-2-phenylacetate
iodide (4). 1H NMR (D2O) d 8.98–8.87 (m, 2H), 8.61–8.57
(m, 1H), 8.11–8.07 (m, 2H), 7.33–7.14 (m, 5H), 4.69–4.55
(m, 2H), 4.09–3.92 (m, 2H), 3.66 (q, J = 6.9 Hz, 1H), 2.02–
1.92 (m, 2H), 1.49–1.08 (m, 17H). MS (m/e) (positive
FAB) 368.3 (M À I)+.
S-8-Pyridinium-n-octyl-a-methyl-2-phenylacetate
1
iodide (5). H NMR (D2O) d 8.80 (d, J = 6.6 Hz, 2H),
8.57 (t, J = 9 Hz, 1H), 8.09 (t, J = 6.9 Hz, 2H), 7.45–7.31
(m, 5H), 4.53 (t, J = 7.2 Hz, 2H), 4.19–4.05 (m, 2H), 3.88
(q, J = 6.9 Hz, 1H), 1.94–1.88 (m, 2H), 1.61–1.56 (m, 2H),
1.49–1.42 (m, 7H), 1.27–1.23 (m, 4H). MS (m/e) (positive
FAB) 340.2 (M À I)+.
R-8-Pyridinium-n-octyl-a-hydroxy-methyl-2-phenyla-
8-Pyridinium-n-octyl-a-methyl-2-[4-(N-Boc amino)-
phenyl]acetate iodide (16). 1H NMR (D2O) d 8.82
(d, J = 6.6 Hz, 2H), 8.57 (t, J = 9 Hz, 1H), 8.08 (t, J =
6.9 Hz, 2H), 7.38–7.31 (m, 4H), 4.63 (t, J = 7.2 Hz, 2H),
4.29–4.20 (m, 1H), 4.18–4.02 (m, 1H), 3.88 (q, J = 6.9 Hz,
1
cetate iodide (6). H NMR (D2O) d 8.84 (d, J = 6.6 Hz,
2H), 8.56 (t, J = 9 Hz, 1H), 8.08 (t, J = 6.9 Hz, 2H), 7.57–
7.39 (m, 5H), 4.60 (t, J = 7.2 Hz, 2H), 4.15 (t, J = 4.2 Hz,
2H), 2.02–1.96 (m, 2H), 1.83 (s, 3H), 1.59–1.55 (m, 2H),

384
S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
1H), 1.98–1.88 (m, 2H), 1.61–1.56 (m, 13H), 1.27–1.23
(m, 9H). MS (m/e) (positive FAB) 455.3(M À I)+.
S-8-(4-n-Propylpyridinium)-n-octyl-a-methyl-2-(4-[2-
methylpropyl]phenyl)acetate bromide (19). S-8-Bromo-n-
octyl-a-methyl-2-(4-[2-methylpropyl]phenyl)acetate
(35 mg, 0.088 mmol) and 4-n-propylpyridine (0.2 ml)
were dissolved in acetone (3 ml). Tetrabutyl ammonium
iodide (5 mg) was added, and the mixture was stirred for 2
days at 50 8C. Acetone was removed in vacuo. The product
(19, 16 mg, 0.028 mmol, 35% yield) was isolated using
preparative thin layer chromatography (silica gel, eluting
with chloroform:methanol, 5:1).
8-Pyridinium-n-octyl-a-methyl-2-[4-ammonium
tri-
fluoroacetate-phenyl]acetate iodide (17). 1H NMR
(D2O) d 8.85 (d, J = 6.6 Hz, 2H), 8.57 (t, J = 9 Hz, 1H),
8.08 (t, J = 6.9 Hz, 2H), 7.49–7.35 (m, 4H), 4.62
(t, J = 7.2 Hz, 2H), 4.14 (t, J = 7.2 Hz, 2H), 3.95 (q, J =
6.9 Hz, 1H), 1.98–1.88 (m, 2H), 1.61–1.49 (m, 5H),
1.31–1.25 (m, 8H). MS (m/e) (positive FAB) 355.3
(M À I-CF3COO)+.
8-(3-Carboxyamido-pyridinium)-n-octyl-a-methyl-2-
1H NMR (CD3OD) d 9.32 (d, J = 6.6 Hz, 2H), 7.84
(d, J = 6.3 Hz, 2H), 7.19 (d, J = 7.8 Hz, 2H), 7.09 (d, J =
7.8 Hz, 2H), 4.91 (t, J = 7.2 Hz, 2H), 4.02 (t, J = 7.2 Hz,
2H) 3.70 (q, J = 6.9 Hz, 1H), 3.39–3.34 (m, 2H), 2.86
(t, J = 7.2 Hz, 2H), 2.45 (d, J = 7.2 Hz, 2H), 2.03–1.98
(m, 2H), 1.87–1.64 (m, 3H), 1.56–1.30 (m, 5H), 1.33–
1.25 (m, 8H), 1.01 (t, J = 7.2 Hz, 2H), 0.90 (d, J =
6.6 Hz, 6H). MS (m/e) (positive FAB) 438.3 (M À Br)+.
S-8-[4-(2-hydroxy-ethyl)pyridinium)-n-octyl-a-methyl-
2-(4-[2-methylpropyl]phenyl)acetate bromide (20). 1H
NMR (CDCl3) d 8.86 (d, J = 6.6 Hz, 2H), 7.98 (d, J =
6.3 Hz, 2H), 7.19 (d, J = 7.8 Hz, 2H), 7.09 (d, J = 7.8 Hz,
2H), 4.69 (t, J = 7.2 Hz, 2H), 4.05–3.92 (m, 3H), 4.62–4.59
(m, 1H), 3.19–3.10 (m, 2H), 2.86 (t, J = 7.2 Hz, 1H),
2.45 (d, J = 7.2 Hz, 2H), 2.03–1.98 (m, 2H), 1.49–1.25
(m, 14H), 0.90 (d, J = 6.6 Hz, 6H). MS (m/e) (positive
FAB) 440.3 (M À Br)+.
1
[4-(2-methylpropyl)phenyl]-acetate iodide (18). H NMR
(CDCl3) d 10.21 (s, 1H), 9.16 (d, J = 9 Hz, 1H), 8.98
(d, J = 6.6 Hz, 2H), 8.58 (s, 1H), 8.19 (t, J = 6.9 Hz,
2H), 7.21–7.08 (m, 4H), 6.53 (s, 1H), 4.88 (t, J =
7.2 Hz, 2H), 4.07–4.03 (m, 2H), 3.70 (q, J = 6.8 Hz,
1H), 2.44 (d, J = 9 Hz, 2H), 2.18–2.02 (m, 2H), 1.85–
1.72 (m, 5H), 1.57–1.26 (m, 9H), 0.89 (d, J = 6.6 Hz, 6H).
MS (m/e) (positive FAB) 439.3 (M À I)+.
2.1.1.2. Method B. General procedure for the synthesis of
the compounds 15 and 19–22.
S,S-1,8-Di(a-methyl-2-(4-[2-methylpropyl]phenyl)ace-
toxy)-n-octane (22). S-a-Methyl-2-(4-[2-methylpropyl]-
benzene)acetic acid (206 mg, 1 mmol), 1,8-dibrom-
ooctane (0.184 ml) were combined. A methanolic solution
of benzyltriethylammonium methoxide (453 mg, 40%) and
tetrabutyl ammonium iodide (10 mg) was added. The mix-
ture was stirred at room temperature for 3 days. The product
(34, 167 mg, 0.42 mmol, 42% yield) was isolated using
preparative thin layer chromatography (silica gel, eluting
with hexanes:ethyl acetate, 20:1). A minor product, a
dimeric molecule 22 was also isolated (34 mg, 7%).
1H NMR (CDCl3) d 7.20 (d, J = 7.8 Hz, 2H), 7.09
(d, J = 7.8 Hz, 2H), 4.05 (t, J = 6.3 Hz, 2H), 3.70 (q, J =
6.9 Hz), 2.45 (d, J = 7.2 Hz, 2H), 1.89–1.80 (m, 1H), 1.57–
1.48 (m, 5H), 1.26–1.20 (m, 4H), 0.90 (d, J = 6.6 Hz, 6H).
MS (m/e) (positive FAB) 523.4 (M + H)+.
S-8-[N-Methylmorpholinium)-n-octyl-a-methyl-2-(4-
1
[2-methylpropyl]phenyl)acetate bromide (21). H NMR
(CDCl3) d 7.19 (d, J = 7.8 Hz, 2H), 7.09 (d, J = 7.8 Hz,
2H), 4.13–3.98 (m, 6H), 3.89–3.78 (m, 4H), 3.71–3.55
(m, 6H), 3.40–3.24 (m, 2H), 2.45 (d, J = 7.2 Hz, 2H), 1.86–
1.26 (m, 13H), 1.02 (t, J = 7.2 Hz, 2H), 0.90 (d, J = 6.6 Hz,
6H). MS (m/e) (positive FAB) 418.3 (M À Br)+.
2.2. Cells and culture methods
The CF lung epithelial cells IB-3 and AAV-[wildty-
pe]CFTR-repaired IB-3/S9 have been previously described
[23,32,33]. Both IB-3 and IB-3/S9 cells were grown in
serum-free LHC-8 medium (Biofluids, Bethesda, MD,
USA), formulated without gentimycin. HeLa cells were
obtained from the ATCC, and cultured in Dulbecco’s
modified Eagle’s medium, supplemented with 10% fetal
bovine serum, 2 mM glutamine, penicillin (100 U/ml), and
streptomycin (100 mg/ml).
S-8-Pyridinium-n-octyl-a-methyl-2-(4-[2-methylpro-
pyl]benzene)acetate bromide (15). S-8-Bromo-n-octyl-a-
methyl-2-(4-[2-methylpropyl]benzene)acetate
(35 mg,
0.088 mmol) and pyridine (0.2 ml, 2.5 mmol) were dis-
solved in acetone (3 ml). Tetrabutyl ammonium iodide
(10 mg) was added, and the mixture was stirred for 2 days
at 50 8C. Acetone was removed in vacuo. The product (15,
16 mg, 0.035 mmol, 40% yield) was isolated using pre-
parative thin layer chromatography (silica gel, eluting with
chloroform:methanol, 5:1).
2.3. Default conditions and controls for screening
paradigm
1H NMR (CD3OD) * d 9.02 (d, J = 6.6 Hz, 2H), 8.59
1
(t, J = 9 Hz, H), 8.09 (t, J = 6.9 Hz, 2H), 7.19 (d, J =
7.8 Hz, 2H), 7.09 (d, J = 7.8 Hz, 2H), 4.62 (t, J = 7.2 Hz,
2H), 4.10–4.01 (m, 2H), 3.70 (q, J = 6.9 Hz, 1H), 2.45
(d, J = 7.2 Hz, 2H), 2.03–1.98 (m, 2H), 1.87–1.78 (m, 1H),
1.59–1.51 (m, 2H), 1.43 (d, J = 7.2 Hz. 3H), 1.33–1.25
(m, 8H), 0.90 (d, J = 6.6 Hz, 6H). MS (m/e) (positive FAB)
396.3 (M À Br)+.
Initial assays for drug effects in the screening paradigm
were performed in duplicate at 10, 3, 1, and 0.3 mM
concentrations on IB-3 cells grown in 96-well microtiter
plates. Each individual plate contained IB-3 and IB-3/S9
cells to set the boundary conditions of the assay. ‘‘Hits’’ in
the screen were chosen on the basis of at least a 50%

S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
385
reduction in constitutive IL-8 secretion. Following a posi-
tive retest, candidate compounds were subjected to more
detailed analysis under the same assay conditions. Many of
the compounds we have screened require organic solvents
such as DMSO or ethanol for initial dissolution. In most
circumstances we use a maximum final solvent concentra-
tion of 0.1% DMSO or 0.1% EtOH. Higher concentrations
of DMSO substantially suppress IL-8 secretion. The sol-
vent control, which at the concentration selected minimally
deviates from medium alone, is then used as a basis for
100% activity.
lysate from each sample were fractionated by SDS-PAGE
and immunoblotted. The blots were visualized with
chemiluminescent substrate (Pierce).
2.6. Data analysis
IC50 values obtained in IL-8 radioimmunoassays were
calculated using the GraphPad software package (Graph-
Pad, San Diego, CA). All concentration-effect curves were
repeated in at least three separate experiments, carried out
in duplicate or triplicate, and the data normalized to
propidium iodide binding by DNA in the adherent cells
to correct for cell recovery.
2.4. Detection of IL-8
The details of this assay have been described in a
previous publication [22]. Briefly, IB-3 cells were grown
in 96-well microtiter plates to 80% confluency. Drugs,
diluted in LHC-8 medium at the given concentrations,
were added, and the cells incubated for an additional 24 h.
To initiate the experiment, cells were washed with fresh
LHC-8 medium, and then incubated for an additional 16 h
in the same medium supplemented with drug. At the end of
the time period, the supernatant solutions were collected
and assayed for IL-8 by an ELISA assay. Alternatively, the
samples were immediately frozen at À80 8C. No quanti-
tative differences have been noted when both paradigms
were applied to the same sample handled in either manner.
The IL-8 ELISA assay was assembled from bulk materials
purchased from R&D (Boston, MA, USA), and performed
exactly according to the manufacturer’s specifications. To
detect possible drug toxicity on cells, as well as normalize
to total cells per well, we measured the double stranded
DNA content of cells attached to the plate after removal of
supernatant solutions for IL-8 assay. Cells remaining on
the plate were then fixed with 10% formalin for 30 min at
room temperature before being incubated with propidium
iodide (kit from Boehringer Mannheim). Propidium
iodide content in each well was then measured with an
automated fluorescence plate reader (FLUOstar Optima,
MBG LabTechnologies, Durham, NC, USA). Final IL-8
data were calculated as a ratio of secreted IL-8 to DNA
for each well. On occasion, cell proliferation assays were
also carried out using the MTT assay (Roche, Basel,
Switzerland).
3. Results
3.1. Amphiphilic pyridinium salts are inhibitors of CF
IL-8 production
Employing a systematic paradigm designed for library
screening, we identified an amphiphilic pyridinium salt
that suppressed constitutive hypersecretion of IL-8 from
CF lung epithelial IB3-1 cells. The strategy behind assem-
bling this library was to construct binary species with
molecules having history of being both biocompatible
and pharmacologically active. The two species were then
connected by a hydrophobic carbon chain, in an approach
similar to that we have previously used to enhance affinity
in muscarinic acetylcholine receptor antagonists [34].
Based on an initial ‘‘hit’’, we synthesized 21 structural
derivatives, hoping to increase the potency. The structures
for the pyridinium derivatives 1–22 are shown in Table 1,
along side their potencies in the IL-8 suppression assay.
The derivatives consisted of a hydrophobic ester moiety
and a hydrophilic pyridinium moiety linked through an
alkyl chain of 4–10 carbons (see compounds 1–4). The
chemical strategies for this set of syntheses are summar-
ized in Fig. 1, and are labeled as schemes A–C. The
structure with the currently optimum activity is compound
3, MRS2481, with an IC50 of ca. 1.8 mM.
Modification of the ester group of 9, leading to 10, failed
to restore the activity. Removal of two carbons at the a-
position of 3, to result in 11, eliminated activity of the
molecule. The stereoselectivity of the effects of two enan-
tiomeric 1-hydroxy-2-phenylethyl derivatives, i.e. 12 and
13, could not be assessed, since no activity was evident.
Substitution at the p-position of the phenyl ring with a
branched alkyl group resulted in the pure S-enantiomer 15,
which was roughly equipotent to MRS2481. The corre-
sponding racemic derivative 14 was toxic to cells in the
relevant concentration range. Thus its effects on IL-8 could
not be assessed. It is likely that the toxicity originates in the
R-enantiomer, which was not prepared. The substitution of
the p-position of the phenyl ring in two racemic derivatives
indicated that the introduction of a urethane group in 16
2.5. Western blot analysis
HeLa cells, IB3-1 and IB3-1/S9 cells were pretreated
with MRS compounds, as described above, and the experi-
ments initiated by incubation with TNFa (R&D Systems)
for 15 min. Cells were washed with PBS, and then col-
lected and lysed in M2 buffer (20 mM pH 7.0 Tris, 0.5%
NP-40, 250 mM NaCl, 3 mM EDTA, 3 mM EGTA, 2 mM
dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride,
20 mM b-glycerol phosphate, 1 mM sodium vanadate,
1 mg/ml of leupeptin). Twenty micrograms of the cell

386
S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
Table 1
Structures of pyridinium compounds prepared for testing in the IL-8 assay
a
R1
Compound
n
IC50
Unsubstituted pyridinium salts (R2 = H)
1
2
4
6
8
>30
>30
3, MRS 2481
4
1.81 Æ 0.58
10
2.52 Æ 0.39
5, MRS 2485
8
>25
>30
6
8
7
8
8
8
12 Æ 0.8
3.16 Æ 0.52
9
8
8
>30
>30
10
11
12
8
8
>30
>30
13
14
8
8
>30
Toxic at 1 mM
15
16
8
8
2.2 Æ 0.8
4.6 Æ 0.9

S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
387
Table 1 (Continued )
R1
Compound
n
IC50
17
8
>30
R1
Compound
R2
IC50
Substituted pyridinium salts and other derivatives, n = 8
18
3-CONH2
5.56 Æ 0.98
19
20
21
p-(CH2)2CH3
p-(CH2)2-OH
3.3 Æ 0.5
18 Æ 0.9
24 Æ 1.0
22
>30
a
The IC50 determinations were performed in duplicate at concentrations ranging from 0.3 to 10 mM on IB-3 cells grown in 96-well microtiter plates (n = 3, or
more). The solvent control containing EtOH or DMSO for initial dissolution of the compounds, which at the concentration selected minimally deviates from
medium alone, is then used as a basis for 100% activity.
was possible with retention of the activity, while the small
hydrophilic p-amino group in 17 eliminated activity.
Finally, addition of a p-substituent of the pyridinium
moiety of 15 led to either retention of potency (n-propyl
19) or an eight-fold loss of potency (2-hydroxyethyl 20)
relative to 15. The pyridinium moiety could be replaced
with an N-methylmorpholino moiety in 21, with only an
11-fold loss of potency in comparison to 15, while the
replacement with an uncharged moiety identical to the
ester side of the molecule to give a dimeric structure in 22
eliminated the activity. Thus, the activity is associated
with the presence of a positively charged ammonium
group.
3.3. The (R)-enantiomer of MRS2481 is much more
active than the (S)-enantiomer
We noted that the optimal species 3 (MRS2481) had an
asymmetric carbon in the R1 moiety (i.e., an (R)-1-phe-
nylpropionic acid ester). To test whether this asymmetric
carbon might have importance for the function of the
compound we changed the group to an (S)-1-phenylpro-
pionic acid ester (see 5, MRS2485). A titration of both
MRS2481 and MRS2485 is shown in Fig. 3. As summar-
ized in Table 1, this change leads to a profound reduction in
the apparent IC50, from ca. 1 mM to >25 mM. We also
replaced the methyl group with an –H, thereby eliminating
the chiral center. This step, as indicated by molecule 7, led
to a six-fold loss of potency. We therefore conclude that
this chiral group is critical for interacting with a site
somewhere in the CF cell that regulates the expression
of IL-8.
3.2. The optimum chain length between pyridinium
and hydrophobic moieties is eight carbons
The original decision to link the pyridinium and hydro-
phobic moieties by an eight-carbon chain was based on
previous experience. However, we decided to test the
contribution of this chain to potency by varying the chain
length. As shown in Fig. 2, we tested molecules with four
different chain lengths. These included 1 (n = 4); 2 (n = 6);
3 (n = 8); and 4 (n = 10). The data show that the eight-
carbon species (MRS2481) is slightly more potent than the
10-carbon species 4. However, the six-carbon species 2 is
much less active, and the four-carbon species 1 is even less
active than the six-carbon species. We summarize these
data as follows: [C8 > C10 ꢀ C6 > C4]. Thus the choice
of an n-octyl linker is optimal.
3.4. Potency is affected by the size and polarity of the
substituent on the asymmetric carbon
To further probe the significance of the structures around
the asymmetric carbon, we questioned whether the size or
polarity of the carbon at this site had any consequences for
potency or efficacy. The optimum structure, 3, MRS2481,
has a methyl group on an asymmetric carbon in the
hydrophobic domain, and proceeded to substitute this
methyl group with either an –OCH3 (viz, 9) or an –
C2H5 (viz, 8). As shown in Fig. 4, increasing the size of

388
S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
Fig. 1. Chemical strategies for syntheses of MRS2481 and analogs. (A) Preparation of analogues of variable spacer chain lengths and variable substitution of the
pyridinium moiety. For the structure of R1, refer to Table 1. Reagents: (a) benzyltrimethylammonium hydroxide, tetrabutylammonium iodide at r.t. for 2 days;
(b) sodium iodide in acetone; (c) pyridine derivative in acetone at 50 8C for 3 days. (B) Preparation of compound 17, which contains a chemically reactive aryl
amino group. Reagents: (a) tin, acetic acid, HCl at 100 8C for 1.5 h; (b) di-t-butyl-dicarbonate in methanol at 45 8C for 1 h; (c) benzyltrimethylammonium
hydroxide, tetrabutylammonium iodide at r.t. for 2 days; (d) sodium iodide in acetone; (e) pyridine in acetone at 50 8C for 3 days; (f) trifuoroacetic acid. (C)
Preparation of 4-substituted pyridinium salts 19 and 20 and non-pyridinium derivatives, including the symmetric diester 22. Reagents: (a) 1,8-dibromooctane,
benzyltrimethylammonium hydroxide, tetrabutylammonium iodide at r.t. for 3 days; (b) N-methylmorpholine, tetrabutylammonium iodide in acetone at 50 8C
for 2 days; (c) 4-propyl-pyridine, tetrabutylammonium iodide in acetone at 50 8C for 3 days; (d) 4-(2-hydroxyethyl)-pyridine, tetrabutylammonium iodide in
acetone at 50 8C for 3 days.

S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
389
Fig. 2. Optimum chain length between pyridinium and hydrophobic moieties of MRS2481. The carbon chain length was varied systematically between 4 and
10 carbons. Based on IC50 values, an eight-carbon chain length is optimal.
the substituent from one carbon to two carbons increases
the IC50 by ca. two-fold. By contrast substituting the
methyl with –OCH3 profoundly reduces potency. We
conclude that both the orientation and the polarity of the
organic substituent on the asymmetric carbon significantly
affects the potency of the compound.
MRS2481 had such influence on the potency of the com-
pound. Such a fact often indicates that a compound with
these properties might have a specific site in the cell upon
which to bind and act. We have hypothesized in previous
reports that constitutive hyperexpression of IL-8 in CF lung
epithelial cells was due to a dysfunctional TNFa/NFkB
signaling pathway [22]. We therefore decided to test
whether MRS2481 might inhibit NFkB signaling in CF
and control cells. As shown in Fig. 5a, HeLa cells express
substantial amounts of IkBa. This means that NFkB is not
activated under control conditions. As expected, when
these cells are exposed to TNFa, the IkBa is vastly
reduced, and NFkB is activated. However, when the cells
are preincubated with 5 mM MRS2481 for up to 6 h, there
is a modest protection when TNFa is added (see Fig. 5b).
3.5. MRS2481 inhibits TNFa-dependent NFkB
signaling in CF and control cells
We were much impressed by that fact that the size,
polarity and orientation around the asymmetric carbon in
Fig. 4. Effect of size and polarity of the substituent at the asymmetric
carbon on the activity of MRS2481 and its analogues. The methyl sub-
stituent on the optically active carbon of MRS2481 was replaced with a
more hydrophobic ethyl group (MRS2591) or a more polar O-Me
(MRS2506). The latter substitution abolished activity.
Fig. 3. Activity comparison of (R)- and (S)-enantiomers of MRS2481. The
(R)-enantiomer, MRS2481, and the (S)-enantiomer, MRS2485, were
synthesized and the IL-8 suppression activity titrated.

390
S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
Fig. 5. Influence of MRS2481 on TNFa-induced IkBa degradation in HeLa
cells by MRS2481. (a) HeLa cells were pretreated with MRS2481 (5 mM)
for 0, 1, 3, or 6 h, and then treated with 2 ng/ml TNFa for 15 min. Reaction
mixtures were then subjected to polyacrylamide gel electrophoresis, blotted
onto nitrocellulose, and probed with anti-IkBa antibody. (b) HeLa cells were
pretreated with the less potent analogue MRS2485 (50 mM) for 0, 1, 3, or
6 h, and then treated with 2 ng/ml TNFa for 15 min. Reaction mixtures were
then subjected to polyacrylamide gel electrophoresis, blotted onto nitro-
cellulose, and probed with anti-IkBa antibody.
Fig. 6. Influence of MRS2481 on TNFa-induced IkBa degradation and
JNK phosphorylation in CF IB3-1 cells. (a) IB3-1 cells were pretreated with
either 5 or 10 mM MRS2481, or the less potent analogue MRS2485 for 6 h.
The cells were then treated with 10 ng/ml TNFa for 15 min. Reaction
mixtures were then subjected to polyacrylamide gel electrophoresis, blotted
onto nitrocellulose, and probed with anti-IkBa antibody. The left-most ‘‘0’’
is an untreated condition. The inner ‘‘0’’ is upon the addition of TNFa only.
(b) IB3-1 cells were pretreated with either 5 or 10 mM MRS2481, or the less
potent analogue MRS2485 for 6 h. The cells were then treated with 10 ng/
ml TNFa for 15 min. Reaction mixtures were then subjected to polyacry-
lamide gel electrophoresis, blotted onto nitrocellulose, and probed with anti-
phospho-JNK or anti-JNK1 antibodies. The left-most ‘‘0’’ is an untreated
condition. The inner ‘‘0’’ is upon the addition of TNFa only.
With the less active enantiomeric species, MRS2485, one
can get some protection as well, except that 50 mM rather
than 5 mM is needed to obtain a comparable protective
effect. These data are consistent with the differential
potency of these two species previously noted for suppres-
sion of IL-8 from CF cells.
A qualitatively similar differential effect of these two
compounds is also seen when IB3-1 or IB3-1/S9 cells are
tested in the same way. As shown in Fig. 6a, preincubation
of IB3-1 cells with 5 mM MRS2481 is more effective than
an equivalent amount of MRS2485 at protecting IkBa
from TNFa-induced destruction. However, the distinction
is lost when cells are preincubated with 10 mM of each
compound. As shown in Fig. 7a, an equivalent effect is also
seen in CFTR-repaired IB3-1/S9 cells when the cells are
pretreated with 5 mM or 10 mM of each compound. The
data with the CF and repaired cells are particularly con-
sistent with the differential potency noted in these same
cells by the two compounds.
3.6. MRS2481 inhibits TNFa-dependent activation of
AP-1 signaling
Fig. 7. Influence of MRS2481 on TNFa-induced IkBa degradation and
JNK phosphorylation in repaired IB3-1/S9 cells. (a) IB3-1/S9 cells were
pretreated with either 5 or 10 mM MRS2481, or the less potent analogue
MRS2485 for 6 h. The cells were then treated with 10 ng/ml TNFa for
15 min. Reaction mixtures were then subjected to polyacrylamide gel
electrophoresis, blotted onto nitrocellulose, and probed with anti-IkBa
antibody. The left-most ‘‘0’’ is an untreated condition. The inner ‘‘0’’ is
upon the addition of TNFa only. (b) IB3-1/S9 cells were pretreated with
either 5 or 10 mM MRS2481, or the less potent analogue MRS2485 for 6 h.
The cells were then treated with 10 ng/ml TNFa for 15 min. Reaction
mixtures were then subjected to polyacrylamide gel electrophoresis, blotted
onto nitrocellulose, and probed with anti-phospho-JNK or anti-JNK1 anti-
bodies. The left-most ‘‘0’’ is an untreated condition. The inner ‘‘0’’ is upon
the addition of TNFa only.
The IL-8 promoter is driven by both NFkB and AP-1
signaling, and we therefore decided to test the AP-1
signaling system for sensitivity to MRS2481. The proximal
step in AP-1 signaling is based on phosphorylation of Jun
kinase (JNK1) and subsequent phosphorylation of c-Jun by
the activated phospho-JNK. As shown in Fig. 6b, IB3-1
cells under control conditions show very little evidence of
JNK phosphorylation, indicating that AP-1 signaling is not
activated. However, when TNFa is added phospho-JNK is

S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
391
made from JNK, and the AP-1 site is activated. The data
show that MRS2481 and MRS2485, both at 5 and 10 mM,
suppress this reaction equivalently. Equivalent data are
shown in Fig. 7b for IB3-1/S9 cells treated in the same
manner. Thus the AP-1 signaling pathway appears to show
less discrimination between the two enantiomeric species.
lie upstream of or near the NFkB/AP-1 signaling bifurca-
tion. The exact site of MRS2481 action in this pathway is
still not known, but is under active investigation in our
laboratory.
4.2. Classification of IL-8 suppression candidate CF
drugs
4. Discussion
There now appear to be at least two classes of candidate
CF drugs that suppress the constitutive hypersecretion of
IL-8 from CF lung epithelial cells. Class C (for CPX)
compounds are typified by the xanthine CPX [22]. CPX,
and the more potent homologue DAX, binds to the first
nucleotide binding-domain of CFTR (viz, NBD1/NBF1)
near the mutant DF508 site [19]. The folding defect is
apparently rescued, thereby permitting efficient trafficking
of the mutant protein to the plasma membrane. The rescued
mutant CFTR now functions to both transport chloride and
suppress IL-8 expression. We have hypothesized that both
rescued mutant CFTR and [wildtype]CFTR suppress IL-8
expression by similar mechanisms [22]. In fact, genomic
analysis has shown that both CPX and [wildtype]CFTR act
by suppressing the TNFa/NFkB pathway [22]. However, it
is not yet clear whether CPX only suppresses IL-8 produc-
tion by virtue of its CFTR trafficking rescue property. In
addition, it is not yet known to what extent other trafficking
rescue/corrector compounds may also suppress IL-8
expression. These other compounds include the less active
xanthines such as IBMX (ka = ca. 1 mM: Ref. [24]); fla-
vones such as genistein (ka = ca. 50–75 mM: Ref. [25,26]);
benzo[c]quinolizium (ka = ca. 50–75 mM; Ref.27); phenyl-
butyrate (ka = ca. 1 mM: Ref. [28]); and the benzothiazines
(ka = ca. 0.5 mM: Ref. [29]).
We report here the discovery of a water-soluble, potent
and efficacious compound, MRS2481, which has the prop-
erty of suppressing the constitutitive hypersecretion of IL-8
from cystic fibrosis lung epithelial cells. MRS2481 is an
amphiphilic pyridinium salt, connected by an eight-carbon
chain to an (R)-1-phenylpropionic acid ester. The indivi-
dual components of this compound are inactive in the assay
at concentrations as high as 25–30 mM, and the eight-
carbon chain connecting the two principal components is
absolutely required for optimal potency. We have pio-
neered the concept of increasing drug potency by the
functionalized congener approach [34], and pyridinium
salts have historically been used to create water-soluble
pro-drugs of otherwise less soluble compounds [35]. How-
ever, we find that the positive charge on the pyridinium
moeity is absolutely required for activity. In addition, we
also find that an enantiomer of the lead compound,
MRS2485, containing an (S)-1-phenyl propionic acid ester,
exhibits greatly reduces IL-8 suppression activity in CF
cells. These results indicate that the more potent enantio-
mer, MRS2481, appears to require all aspects of the
structure for its primary action, and that it very likely
binds to a specific site within the cell.
Class D (for digitoxin) compounds are typified by
digitoxin [23]. These compounds have the most modest
effects, if any, on rescue of mutant CFTR trafficking. Yet
digitoxin profoundly suppresses IL-8 hypersecretion from
CF lung epithelial cells. Into the latter class we can now
place the new candidate CF drug MRS2481. Both digitoxin
and MRS2481 act by suppressing TNFa/NFkB signaling.
However ongoing comparative experiments indicate that
they appear to act at different sites on this complex path-
way. The fact that the sites of action may be different
suggests that there are multiple approaches available for
cytokine suppression in CF. This is important because
some of these compounds may function as site-specific
mimics for CFTR suppression of inflammation without
actually affecting rescue of the mutant CFTR trafficking
defect per se. Details of the exact locations of action of
these two compounds are presently being studied.
4.1. Mechanism of action of MRS2481
Consistently, the mechanism of action of MRS2481
involves the inhibition of at least two signaling pathways,
NFkB and AP-1. Both of these pathways are known to be
potent drivers of the IL-8 promoter [12,13]. Our data
indicate that in both HeLa cells and CF IB3-1 cells,
MRS2481 blocks the TNFa-activated destruction of IkBa.
When IkBa is phosphorylated by the IKKa,b,g complex,
the inactive cytosolic localization of the NFkB p65/NFkB
p50 is lost, and the dimer translocates to the nucleus. Once
in the nucleus the NFkB dimer can bind to the kB site on
the IL-8 promoter, and can become a powerful driver of IL-
8 expression.
Our data also indicate that MRS2481 blocks the TNFa-
activated phosphorylation of JNK. Phosphorylation of JNK
activates the kinase activity to phosphorylate c-JUN. Phos-
pho-cJUN forms a dimer with phospho-c-FOS, which
binds to the AP-1 site on the IL-8 promoter, thereby also
driving IL-8 expression. TNFa is known to drive both
NFkB and AP-1 activation. Since MRS2481 blocks both
pathways, we suggest that the site of MRS2481 action must
4.3. Conclusion
IL-8 is the most powerful known physiological attractant
for inflammatory cells, and the importance of this discov-
ery lies in the fact that the lung pathophysiology of cystic

392
S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
[11] Cruse JM, Lewis RE. Illustrated dictionary of immunology. Boca
Raton: CRC Press, 1995, Appendix 3.
fibrosis is based substantially on an intrinsically inflam-
matory phenotype in the CF airway. Thus, the ability of
MRS2481, and other candidate drugs with this property, to
potently and efficaciously suppress constitutive hyperse-
cretion of IL-8 secretion from the CF airway epithelial
cells carries the promise of a strategy for the development
of a potential CF therapeutic. Preliminary exploration of
the mechanism of action of these amphiphilic pyridinium
salts implicates both NFkB and AP-1 pathways, which are
known to control IL-8 gene transcription. The possible
usefulness of this class of compounds for other disorders of
baseline IL-8 expression and resultant inflammation
remain to be investigated.
[12] Roebuck KA. Regulation of interleukin-8 gene expression. J Inter-
feron Cytokine Res 1999;19:429–38.
[13] Hoffmann E, Dittrich-Breiholz O, Holtmann H, Kracht M. Multiple
control of interleukin-8 gene expression. J Leukoc Biol 2002;72:
847–55.
[14] Cheng SH, Gregory RJ, Marshall J, Souza DW, White GA, O’Riordan
CR, et al. Defective intracellular transport and processing of CFTR is
the molecular basis of most cystic fibrosis. Cell 1990;63:827–34.
[15] Eidelman O, Guay-Broder C, van Galen PJM. Jacobson KA, Turner
RJ, Cabantchik ZI, et al. A1-adenosine antagonists activate chloride
efflux from cystic fibrosis cells. Proc Nat Acad Sci USA
1992;89:5562–6.
[16] Guay-Broder C, Jacobson KA, BarNoy S, Cabantchik ZI, Guggino
WB, Zeitlin PL, et al. Al-Receptor antagonist 8-cyclopentyl-1,3-
dipropyl-xanthine (CPX) selectively activates chloride efflux from
human epithelial and mouse fibroblast cell lines expressing the CFTR
(DF508) mutation, but not the wild type CFTR. Biochemistry
1995;34:9079–87.
Acknowledgements
[17] Jacobson KA, Guay-Broder C, van Galen PJM. Gallo-Rodriguez C,
Melman N, Jacobson MA, et al. Stimulation of alkylxanthines of
chloride efflux from CFPAC-1 cells does not involve A1-adenosine
receptors. Biochemistry 1995;34:9088–94.
Drs. Tchilibon, Zhang and Yang should be considered
joint first authors on this paper. This work was supported
by a grant from the Cystic Fibrosis Foundation, Bethesda,
MD (to HBP), and by the NIH (RO1-DK53051, to HBP;
NO1-HV-28187, to HBP). Mass spectral measurements
were carried out by Dr. Victor Livengood, and NMR
measurements by Wesley White (NIDDK). We thank Heng
Duong (NIDDK) for proofreading the manuscript.
[18] Haws CM, Nepomuceno I, Krause ME, Wakelen H, Law T, Xia Y, et
al. DF508-CFTR channels: kinetics, activation by forskolin, and
potentiation by xanthines. Am J Physiol 1996;270:C1544–55.
[19] Cohen BE, Lee G, Jacobson KA, Kim Y-C, Huang Z, Sorscher EJ, et
al. CPX (1,3-dipropyl-8-cyclopentyl xanthine) and other alkyl-
xanthines differentially bind to the wild type and DF508 mutant first
nucleotide binding fold (NBF-1) domains of the cystic fibrosis trans-
membrane conductance regulator (CFTR). Biochemistry 1997;36:
6455–61.
References
[20] Arispe N, Ma J, Jacobson KA, Pollard HB. Direct activation of cystic
fibrosis transmembrane conductance regulator by 8-cyclopentyl-1,3-
dipropylxanthine and 1,3-diallyl-8-cyclohexylxanthine. J Biol Chem
1998;273:5724–34.
[1] Pilewski JM, Frizzell RA. Role of CFTR in airway disease. Physiol
Rev 1999;79:S215–55.
[2] Welsh MJ, Ramsey BW, Accurso F, Cutting GR. Cystic fibrosis. In:
The metabolic and molecular bases of inherited diseases (Scriver CL,
Beaudet AL, Valle D, Sly WS, eds; Childs B, Kinzler KW, Vogelstein
B, Assoc Eds) 2001; eighth ed. 5121–5188, McGraw-Hill, New York.
[3] Dean TP, Dai Y, Shute JK, Church MK, Warner JO. Interleukin-8
concentrations are elevated in bronchoalveolar lavage, sputum, and
sera of children with cystic fibrosis. Pediatr Res 1993;34:159–61.
[4] Richman-Eisenstat JB, Jorens PG, Hebert CA, Ueki I, Nadel JA.
Interleukin-8: an important chemoattractant in sputum of patients
with chronic inflammatory airway diseases. Am J Physiol 1993;
264:L413–8.
[21] McCarty NA, Standaert TA, Teresi M, Tuthill C, Launspach J, Kelley
TJ, et al. A phase I randomized, multicenter trial of CPX in adult
subjects with mild cystic fibrosis. Pediatr Pulmonol 2002;33:90–8.
[22] Eidelman O, Srivastava M, Zhang J, Murthy J, Heldman E, Jacobson
KA, et al. Genes from the TNFaR/NFkB pathway control the pro-
inflammatory state in cystic fibrosis epithelial cells. Mol Med
2001;7:523–34.
[23] Srivastava M, Eidelman O, Zhang J, Paweletz C, Caohuy H, Yang Q-F,
et al. Digitoxin mimics CFTR-gene therapy, and suppresses hyperse-
cretion of proinflammatory interleukin-8 (IL-8) from cystic fibrosis
lung epithelial cells. Proc Nat Acad Sci USA 2004;101:7693–8.
[24] Drumm ML, Wilkinson DJ, Smit LS, Worrell RT, Strong TV, Frizzell
RA, et al. Chloride conductance expressed by DF508 and other mutant
CFTR’s in Xenopus oocytes. Science 1991;254:1797–9.
[25] Hwang T-C, Wang F, Yang I, Reenstra WW. Genistein potentiates
wild-type and DF508-CFTR channel activity. Am J Physiol 1997;
273:C988–98.
[5] Armstrong DS, Grimwood K, Carlin JB, Carzino R, Gutierrez JP, Hull
J, et al. Lower airway inflammation in infants and young children with
cystic fibrosis. Am J Respir Crit Care Med 1997;156:1197–204.
[6] Khan TZ, Wagener JS, Bost T, Martinez J, Accurso FJ, Riches DW.
Early pulmonary inflammation in infants with cystic fibrosis. Am J
Respir Crit Care Med 1995;151:1075–82.
[7] DiMango E, Ratner AJ, Bryan R, Tabibi S, Prince A. Activation of NF-
kappaB by adherent Pseudomonas aeruginosa in normal and cystic
fibrosis respiratory epithelial cells. J Clin Invest 1998;101:2598–605.
[8] Bonfield TL, Panuska JR, Konstan MW, Hilliard KA, Hilliard JB,
Ghnaim H, et al. Inflammatory cytokines in cystic fibrosis lungs. Am J
Respir Care Med 1995;152:2111–8.
[26] Andersson C, Servetnyk Z, Roomans GM. Activation of CFTR by
genistein in human airway epithelial cell lines. Biochem Biophys Res
Commun 2003;308:518–22.
[27] Becq F, Mettey Y, Gray MA, Galietta LJ, Dormer RL, Merten M, et al.
Development of substituted Benzo[c]quinolizinium compounds as
novel activators of the cystic fibrosis chloride channel. J Biol Chem
1999;274:27415–25.
[9] Bonfield TL, Konstan MW, Burfeind P, Panuska JR, Hilliard JB,
Berger M. Normal bronchial epithelial cells constitutively produce
the anti-inflammatory cytokine interleukin-10, which is downregu-
lated in cystic fibrosis. Am J Respir Cell Mol Biol 1995;13:257–61.
[10] Tirouvanziam R, de Bentzmann S, Hubeau C, Hinnrasky J, Jacquot J,
Peault B, et al. Inflammation and infection in naive human cystic
fibrosis airway grafts. Am J Respir Cell Mol Biol 2000;23:121–7.
[28] Rubenstein RC, Zeitlin PL. Sodium 4-phenylbutyrate downregulates
Hsc70: implications for intracellular trafficking of DF508-CFTR. Am
J Physiol Cell Physiol 2000;278:C259–67.
[29] Yang H, Shelat AA, Guy RK, Gopinath VS, Ma T, Du K, et al.
Nanomolar affinity small molecule correctors of defective Delta

S. Tchilibon et al. / Biochemical Pharmacology 70 (2005) 381–393
393
F508-CFTR chloride channel gating. J Biol Chem 2003;278:35079–
85.
[33] Flotte TR, Afione SA, Solow R, Drumm ML, Markakis D, Guggino
WB, et al. Expression of the cystic fibrosis transmembrane conduc-
tance regulator from a novel adeno-associated virus promoter. J Biol
Chem 1993;268:3781–90.
[30] Konstan MW, Byard PJ, Hoppel C-L, Davis PB. Effect of high dose
ibuprofen in patients with cystic fibrosis. New Eng J Med 1995;
332:848–54.
[34] Karton Y, Bradbury BJ, Baumgold J, Paek R, Jacobson KA. Functio-
nalized congener approach to muscarinic antagonists: analogues of
pirenzepine. J Med Chem 1991;34:2133–45.
[31] Accurso FJ, Sagel SD, Sontag MK, Wegener JS. Anti-inflammatory
endpoints in cystic fibrosis (CF): lessons from clinical trials. Pediatr
Pulmonol 2003;S25:104–5.
[35] Davidsen SK, Summers JB, Albert DH, Holms JH, Heyman HR,
Magoc TJ, et al. N-(Acyloxyalkyl)pyridinium salts as soluble prodrugs
of a potent platelet activating factor antagonist. J Med Chem 1994;
37:4423–9.
[32] Zeitlin PL, Lu L, Hwang T-C, Rhim J, Cutting GR, Keiffer KA, et al. A
cystic fibrosis bronchial epithelial cell line: immortalization by
adeno12-SV40 infection. Am J Respir Cell Mol Biol 1991;4:313–9.