Structure-activity relationship for enhancement of paracellular permeability across Caco-2 cell monolayers by 3-alkylamido-2-alkoxypropylphosphocholines

Article abstract of DOI:10.1021/jm020001x

Paracellular permeability enhancers have been used to improve the oral bioavailability of hydrophilic drugs; however, the mechanism of action of many enhancers is poorly understood. In this study, highly potent enhancers of paracellular permeability were

Full text of DOI:10.1021/jm020001x

J . Med. Chem. 2002, 45, 2857-2866
2857
Str u ctu r e-Activity Rela tion sh ip for En h a n cem en t of P a r a cellu la r P er m ea bility
a cr oss Ca co-2 Cell Mon ola yer s by 3-Alk yla m id o-2-a lk oxyp r op ylp h osp h och olin es
Hui Ouyang,†,‡ Susan L. Morris-Natschke,† Khalid S. Ishaq,† Peter Ward,§, Dongzhou Liu,‡,|
Sarah Leonard,§ and Dhiren R. Thakker*,‡
Divisions of Medicinal Chemistry and Natural Products, and Drug Delivery and Disposition, School of Pharmacy, and
Department of Pharmacology, School of Medicine, The University of North Carolina at Chapel Hill,
Chapel Hill, North Carolina 27599
Received J anuary 2, 2002
Paracellular permeability enhancers have been used to improve the oral bioavailability of
hydrophilic drugs; however, the mechanism of action of many enhancers is poorly understood.
In this study, highly potent enhancers of paracellular permeability were identified in the
3-alkylamido-2-alkoxypropylphosphocholine series, and a structure-activity relationship was
developed for enhancement of paracellular permeability across Caco-2 cell monolayers.
Compounds with short (<5 carbons) hydrocarbon chains at both C-2 and C-3 were generally
inactive. The potency exhibited a parabolic relationship with respect to the chain length at
either C-2 or C-3. Linear molecules (i.e., compounds with a short hydrocarbon chain at C-2 or
C-3 and a long hydrocarbon chain on C-3 or C-2, respectively) were more potent than the
corresponding branched molecules with the same carbon load. The efficacy of 3-alkylamido-
2-alkoxypropylphosphocholines as enhancers of paracellular permeability was not dependent
on their existence in micellar form or their ability to alter the fluidity of cell membrane.
Previously, a correlation between the potency of alkylphosphocholines as enhancers of
paracellular permeability and the inhibitors of phospholipase C (PLC) was established in Madine
Darby canine kidney (MDCK) cell monolayers. The potencies of selected 3-alkylamido-2-alkoxypro-
pylphosphocholines as inhibitors of PLC and enhancers of paracellular permeability fit well
into this correlation. Therefore, phosphocholines are likely to increase paracellular permeability
by modulating the signal transduction pathway initiated by a PLC-catalyzed reaction rather
than by physically altering the cell membrane.
In tr od u ction
larger hydrophilic ones.4,5 It also acts as an intramem-
brane diffusion barrier that restricts the intermixing of
apical and basolateral membrane components,6,7 and
thus maintains polarity of enterocytes.8 The restricted
movement of ions across the tight junctions gives rise
to transepithelial electrical resistance (TEER), which is
often used as an index of the integrity of the tight
junctions in an epithelial or endothelial tissue. The
function of the tight junction is complex; this complexity
is also reflected in its multiprotein architecture that
spans extracellular, transmembrane, and intracellular
domains of the cells. The tight junction is composed of
a group of both transmembrane and cytosolic proteins
that interact not only with each other but also with the
membrane and cytoskeleton.9
The intestinal epithelium constitutes a major barrier
to oral drug delivery.1 Molecules cross the intestinal
epithelium into the systemic circulation primarily by
three pathways: passive diffusion across the cell mem-
branes (transcellular pathway), passive diffusion be-
tween adjacent cells (paracellular pathway), or carrier-
mediated transport (carrier-mediated transcellular
pathway).2 Lipophilic molecules easily cross the cell
membrane via transcellular diffusion. On the other
hand, hydrophilic molecules, if not recognized by a
carrier, cannot partition into the hydrophobic mem-
brane, and thus traverse the epithelial barrier via the
paracellular pathway. The transport of hydrophilic
molecules via the paracellular pathway, however, is
severely restricted by the presence of tight junctions.3
In recent years, the morphology and regulation of the
tight junction has been studied in great detail. The tight
junction allows passage of small hydrophilic compoundss
ions, nutrients, and drugssbut acts as a barrier to
Many useful therapeutic agents (e.g., peptides, certain
cephalosporin antibiotics, etc.) are hydrophilic10-14 and
thus cannot be delivered via the oral route, which is the
most favored route of drug delivery. One way to increase
the absorption of hydrophilic drugs across the intestinal
epithelium is to make them more lipophilic by designing
a prodrug, thus increasing transcellular flux of the
drug.11,13 Another approach involves redesigning the
drug so that it is a substrate for a carrier (or linked to
a substrate for a carrier).10,14 The controlled and revers-
ible opening of the tight junction represents an alterna-
tive approach to increase the absorption of hydrophilic
drugs across the intestinal epithelium.12,15,16 This ap-
proach is attractive because it could be applied to many
* To whom correspondence should be addressed. Tel: (919)962-0092.
Fax: (919)966-6919. E-mail: dhiren_thakker@unc.edu.
† Division of Medicinal Chemistry and Natural Products, School of
Pharmacy.
‡ Division of Drug Delivery and Disposition, School of Pharmacy.
§ Department of Pharmacology, School of Medicine.
| Current address: Drug Delivery & Biopharmaceutics, Wyeth-
Ayerst Research, Pearl River, NY 10965.
Current address: Arvna Pharmaceuticals, Inc., San Diego, CA
92121.
10.1021/jm020001x CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/25/2002

2858 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13
Ouyang et al.
amido-1,2-propanediol was synthesized by using acetic
anhydride and sodium carbonate in methanol (method
C).24,25
The primary hydroxyl group of 4 was then protected
as the trityl ether (5).26 Any reaction at the secondary
hydroxyl group was minimized by avoiding long reaction
times (>10 h) and high temperatures (>50 °C). Next,
the secondary hydroxyl group was alkylated with an
alkyl bromide and NaH in tetrahydrofuran (THF). The
resulting compounds (6) were detritylated with p-
toluenesulfonic acid in CH2Cl2/methanol to give the
1-alkylamido-2-alkoxypropanols (7).
The final phosphocholines (2) were prepared by two
methods. Method D was a one pot synthesis of J ia et
al.27 using 2-chloro-1,3,2-dioxaphospholan-2-one and
trimethylamine gas. However, in our hands, acceptable
and more dependable yields were obtained with method
E, a two step approach with 2-bromoethyl dichlorophos-
phate (9) and aqueous trimethylamine.28 All final
products were characterized by nuclear magnetic reso-
nance (NMR) and high-resolution mass spectroscopy
(HRMS).
F igu r e 1. Structures of (1) alkylphosphocholines and (2)
3-alkylamido-2-alkoxypropylphosphocholines.
different hydrophilic drugs. In addition, this approach
would avoid degradation of the active agent by intra-
cellular enzymes, another barrier for absorption of
drugs. In recent years, identification of compounds that
selectively open the tight junction has been actively
pursued.4 Coadministration of these compounds with
hydrophilic drugs improves the oral bioavailability of
these drugs in vivo. For example, palmitoyl carnitine
increased the oral bioavailability of cefoxitin from less
than 5% to as much as 70% in rats.17 Similarly, a
medium chain glyceride increased the absorption of
cefmetazole sodium from 5.6 to 64.8% across dog duode-
num.18
Several phospholipids have been used to increase
paracellular permeability. Lysolethicin, a lumenal di-
gestive product of dietary phospholipids and a common
component of intestinal lumen, has been explored for
use in enhancing oral drug delivery. Talbot et al.19
reported that relatively low concentration of lysolethicin
(0.05-0.20%) increased horseradish peroxidase penetra-
tion in guinea pig proximal jejunum without morpho-
logical damage. Similarly, alkylphosphocholines (Figure
1) have been found to effectively increase paracellular
permeability of hydrophilic compounds across Caco-2
cell monolayers.20 Furthermore, a limited structure-
activity relationship (SAR) study with a series of 3-alkyl-
amido-2-alkoxypropylphosphocholines (Figure 1) indi-
cated that it was possible to vary potency of these
compounds as paracellular permeability enhancers by
varying the alkyl chain length at the C-2 (alkoxy) or
C-3 (alkylamido) positions.20 In the present study, we
report a comprehensive SAR for the 3-alkylamido-2-
alkoxyphosphocholines as enhancers of paracellular
permeability across Caco-2 cell monolayers, identify
potent permeability enhancers, and establish that phos-
phocholines containing linear alkyl chains are much
more potent than the corresponding derivatives with
branched chains. In addition, this study also provides
evidence that these compounds increase paracellular
permeability via specific modulation of the tight junc-
tions, and not by nonspecific perturbation of the cell
membrane.
Biology
Effect of 3-Alk yla m id o-2-a lk oxyp r op ylp h osp h o-
ch olin es on P a r a cellu la r P er m ea bility a cr oss Ca -
co-2 Cell Mon ola yer s. In this study, the target com-
pounds have been evaluated as enhancers of paracellular
permeability across intestinal epithelium. Caco-2 cell
monolayers grown in a transwell, a well-accepted model
for intestinal epithelium, were used to evaluate the
potency and selectivity of the test compounds as para-
cellular permeability enhancers. This model was previ-
ously used to assess the potency and mechanism of
paracellular permeability enhancers of similar com-
pounds.20,29 In addition, the effect of these compounds
on paracellular permeability was also assessed in Ma-
dine Darby canine kidney (MDCK) cell monolayers,
particularly in experiments relating the enhancer activ-
ity to inhibition of phospholipase C (PLC) (cf. section
on Determination of Adenosine 3′,5′-Triphosphate (ATP)-
Stimulated PLC Activity in MDCK Cells). MDCK cell
monolayers, like Caco-2 cell monolayers, have also been
used as an in vitro model for assessment of transport
rates and transport mechanisms of compounds across
epithelial tissues, including intestinal epithelium.30,31
The effect of target compounds on paracellular perme-
ability was determined by measurement of the decrease
in TEER and/or the increase in mannitol (a paracellular
marker that traverses the epithelial tissue predomi-
nantly via the paracellular route) flux from apical to
basolateral (absorptive direction) across Caco-2 cell
monolayers.
Ch em istr y
3-Alkylamido-2-alkoxypropylphosphocholines were syn-
thesized in six steps as shown in Figure 221 with
necessary modifications as noted below. 3-Alkylamido-
1,2-propanediols (4) were synthesized by reacting com-
mercially available 3-amino-1,2-propanediol (3) with an
appropriate acid chloride or anhydride. For acid chlo-
rides with more than eight carbons, pyridine/dimeth-
ylformamide (DMF) at room temperature gave accept-
able yields (method A, 47-70%). With acid chlorides
containing fewer than eight carbons, these conditions
failed to produce the desired product; however, sodium
carbonate in water/chloroform (method B) gave com-
pounds in good yields.22,23 The water soluble 3-ethan-
Effect on TEER. The effect of 3-alkylamido-2-alkoxy-
propylphosphocholines on TEER was evaluated by a 20
min treatment of the Caco-2 cell monolayers on the
apical side with the compounds.20 The treatment time
was determined based on our previous studies20,32 with
structurally similar compounds, which showed that the
drop in TEER was maximal after a 20 min treatment.
These studies have also shown that the effect of these
compounds is reversed over several hours in a static cell
culture system. The active compounds decreased TEER
in a concentration-dependent manner from a control

Permeability across Caco-2 Cell Monolayers
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13 2859
F igu r e 2. Synthesis of 3-alkylamido-2-alkoxypropylphosphocholines. Reagents: (a) CH3(CH2)nCOCl/pyridine/DMF; (b) Ph3CCl/
pyridine; (c) CH3(CH2)mBr/NaH/THF; (d) p-toluenesulfonic acid/CH3OH/CH2Cl2; (e) ethyl ether/THF/pyridine; (f) triethylamine
(40% aqueous solution)/DMF/CHCl3/isopropanol; (g) triethylamine/acetonitrile.
value of 600-800 Ω‚cm2 down to 20-100 Ω‚cm2. The
drop in TEER occurred over a 3-4-fold concentration
range for the most potent compounds and over a broader
concentration range for less potent compounds. The
potency of 3-alkylamido-2-alkoxypropylphosphocholines,
expressed as the concentration that decreased TEER by
50% of control value (EC50), varied markedly (Table 1).
Compounds with a short alkyl ether chain at C-2 (4-6
carbons) and a short alkylamido chain at C-3 (4-8
carbons) were inactive. Similarly, compounds with
moderate to long hydrocarbon chains on both C-2 and
C-3, such that the total carbons exceed 22, were also
inactive. In the 2-ethoxy series (compounds 10-15), the
compound with the C-3 alkylamido group containing 10
carbons was practically inactive with the next higher
homologue showing modest activity. However, the po-
tency jumped dramatically when the C-3 alkylamido
chain length increased from 12 to 14 carbons; no further
increase in potency was observed as the length of the
alkylamido chain increased from 14 to 18 carbons, and
the potency dropped when the alkylamido chain length
increased further to 20 carbons. A similar trend was
observed when the 3-alkylamido functionality was kept
short (two carbons) and the length of the 2-alkoxy
functionality was varied (compounds 32, 37, and 41).
For a series in which the alkylamido functionality was
fixed at 10 carbons with increasing number of carbons
in the 2-alkoxy functionality (carbons 2-16, compounds
10, 18, 22, 25, 29, 33, 36, and 39), the potency exhibited
a bell-shaped relationship with the maximum potency
observed for compounds with 6-8 carbons in the
2-alkoxy functionality. It appears that compounds with
a total of 16-20 carbons in the C-2 and C-3 substituents
show maximum potency. However, in this subset, the
compounds that are branched (i.e., approximately equal
numbers of carbons in C-2 and C-3 substituents) have
a much lower potency than those that are more linear
(i.e., either C-2 or C-3 is very short, i.e., a two carbon
chain length).
E ffect on Ma n n it ol F lu x. As expected, 3-alkyl-
amido-2-alkoxypropylphosphocholines increased man-
nitol flux in a concentration-dependent manner. The
potency of these compounds to increase tight junction
permeability was determined by the concentration that
increased the mannitol flux by 10-fold (EC10x) of the
control value.20,32 The potency for causing the increased
mannitol flux varied with changes in the length of the
alkoxy and alkylamido substituents in a manner similar
to that observed when TEER was used as an indicator
of tight junction permeability, except that generally
greater concentrations were required to cause a 10-fold
increase in mannitol flux than those required to cause
50% decrease in TEER (Table 1).
Rela tion sh ip betw een Cr itica l Micelle Con cen -
tr a tion (CMC) a n d En h a n cem en t of P a r a cellu la r
P er m eability by 3-Alkylam ido-2-alkoxypr opylph os-
p h och olin es. Because of the amphiphilic nature of
3-alkylamido-2-alkoxypropylphosphocholines, they could
interact with the cell membrane; thus, it is conceivable
that the enhancement of paracellular permeability by
this class of compounds is due to compromised integrity
of the tight junctions caused by perturbation of the cell
membrane. If this were the case, it would be reasonable
to speculate that efficacy of these compounds as para-
cellular permeability enhancers may be related to their
ability to form micelles. Hence, the CMC of a selected

2860 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13
Ouyang et al.
Ta ble 1. EC50, EC10x, and Potency Index of
3-Alkylamido-2-alkoxypropylphosphocholines in Caco-2 Cell
Monolayersa
chain length
at C2, C3b
EC50
mM
EC10x
mM
potency index
IC50/EC50
compd
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
02, 10
02, 12
02, 14
02, 16
02, 18
02, 20
04, 04
04, 08
04, 10
04, 12
04, 16
06, 08
06, 10
06, 12
06, 16
08, 10
08, 12
08, 16
08, 18
10, 10
10, 12
10, 16
12, 02
12, 10
12, 12
12, 16
14, 10
16, 02
16, 06
16, 10
16, 12
18, 02
18, 16
>5.00
>5.00
NAc
1.22
2.00
2.20
4.00
0.89
NAc
NAc
1.24
1.55
1.11
NAc
2.38
0.84
0.60
1.27
1.24
NAc
NAc
2.12
1.18
NAc
1.72
1.92
NAc
NAc
NAc
1.43
0.93
NAc
NAc
1.11
insoluble
1.79 (0.45) >5.00
0.08 (0.00) 0.89 (0.27)
0.05 (0.03) 0.30 (0.12)
0.04 (0.04) 0.26 (0.03)
0.28 (0.01) 0.42 (0.01)
>5.00
>5.00
3.19 (0.04) >5.00
0.31 (0.06) 1.87 (0.50)
0.09 (0.02) 0.48 (0.08)
>5.00
0.42 (0.03) 0.73 (0.05)
0.19 (0.11) 1.01 (0.24)
0.52 (0.40) 1.61 (0.68)
0.30 (0.08) 1.21 (0.20)
0.55 (0.40) 1.19 (0.30)
>5.00
>5.00
>5.00
>5.00
>5.00
>5.00
>5.00
0.66 (0.14) 1.17 (0.13)
1.27 (0.35) >3.00d
F igu r e 3. Concentration-dependent effects of selected com-
pounds on the steady state fluorescence anisotropy measured
by (a) DPH and (b) TMA-DPH in Caco-2 cells (ambient
temperatures). Data represent the means of three measure-
ments with different cell preparations. The standard devia-
tions of data points were 1-4% of means. Symbols that
represent the compounds are as follows: (, cholesterol; 9, SDS;
2, HPC; b, compound 14; *, compound 25.
>2.00d
>2.00d
0.78 (0.25) 4.27 (0.67)
1.04 (0.11) >3.00d
>2.00d
>1.00d
>2.00d
>2.00d
>1.00d
>2.00d
0.07 (0.02) 0.35 (0.23)
0.30 (0.21) 1.35 (0.04)
>2.00d
>2.00d
propylphosphocholines. However, compounds exhibited
activity at concentrations above or below CMC. For
example, the EC50 values of compounds 14 and 25 were
4- and 6-fold higher than their respective CMC values.
On the other hand, the EC50 value of compound 12 was
approximately half that of its CMC value (Table 2).
These results suggested that the efficacy of the phos-
phocholines as paracellular permeability enhancers is
not dependent on their ability to form micelles.
>0.50d
>0.50d
0.09 (0.05) 0.34 (0.10)
insoluble insoluble
a
Results are reported as mean ( SD (in parentheses) in
b
triplicate for three separate experiments. See Figure 1, general
structure 2. c Not applicable. The compound is insoluble above
d
this concentration.
Ta ble 2. CMC Values of
3-Alkylamido-2-alkoxypropylphosphocholines
Cyt ot oxicit y of 3-Alk yla m id o-2-a lk oxyp r op yl-
p h osp h och olin es to Ca co-2 Cells. The toxicity of
3-alkylamido-2-alkoxypropylphosphocholines toward Ca-
co-2 cells was evaluated by the 3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay,
which measures the mitochondrial dehydrogenase ac-
tivity as a surrogate for cell viability by a colorimetric
assay.45-47 IC50 was determined as the concentration
that decreases the activity of mitochondrial dehydro-
genase by 50% as compared to the control (untreated)
value. The potency index (Table 1) was defined as IC50/
EC50. For most 3-alkylamido-2-alkoxypropylphospho-
cholines, the potency index was found to be between 0.5
and 2.0. The only exception was 3-octadecanamido-2-
ethoxypropylphosphocholine (compound 14), with IC50/
EC50 ) 4.0 (IC50 ) 0.16 mM, EC50 ) 0.04 mM).
Rela tion sh ip betw een th e Ability of 3-Alk yla -
m id o-2-a lk oxyp r op ylp h osp h och olin es t o Affect
Mem br a n e F lu id ity a n d Ca u se P a r a cellu la r P er -
m ea bility En h a n cem en t. The effect of 3-alkylamido-
2-alkoxypropylphosphocholines on cell membrane flu-
idity was determined in Caco-2 cells by fluorescence
anisotropy (r) measurement with DPH or trimethylam-
monium-1,6-diphenyl-1,3,5-hexatriene tosylate (TMA-
DPH) as probes (Figure 3).35,36 DPH incorporates into
the hydrophobic region of the lipid bilayer; therefore,
its fluorescence anisotropy provides a measure of the
compd
chain length at C2, C3
CMC, µMa,b
10
18
25
11
12
13
14
02, 10
04, 10
08, 10
02, 12
02, 14
02, 16
02, 18
4700 (110)
1170 (338)
55.3 (26.7)
1030 (222)
148 (81.5)
34.7 (2.03)
10.5 (1.70)
a
CMC values were determined by fluorescence assay using
DPH as a probe. Results are reported as mean ( SD determined
in triplicate for three separate experiments.
b
group of 3-alkylamido-2-alkoxypropylphosphocholines
was determined (Table 2) with 1,6-diphenyl-1,3,5-
hexatriene (DPH) as a probe,33 whose fluorescence yield
was significantly greater in the lipid environment than
in the aqueous environment. The CMC values of 3-alkyl-
amido-2-alkoxypropylphosphocholines decreased 3-5-
fold with the addition of every two methylene units on
the hydrocarbon chain at either the C-2 or the C-3
position, indicating that the more lipophilic phospho-
lipids form micelles at a lower concentration. These
results are consistent with the report by Stafford et al.34
that the CMC values of 1-acyl-sn-glycero-3-phosphocho-
lines decreased 10-fold with the addition of every two
methylene units on the acyl chain.
In general, the EC50 values varied in the same
direction as the CMC values of 3-alkylamido-2-alkoxy-

Permeability across Caco-2 Cell Monolayers
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13 2861
average molecular packing order of the hydrophobic
regions of the cell membranes.37 By contrast, TMA-DPH
is localized at the lipid-water interface and its fluores-
cence anisotropy provides a measure of the lipid order
at the surface of the cells (i.e., hydrophilic region).38
Cholesterol, which is known for increasing the pack-
ing order of membranes,37 significantly decreased the
membrane fluidity of both hydrophilic and hydrophobic
regions of Caco-2 cell membranes, as indicated by the
increased anisotropy observed upon treatment with both
DPH and TMA-DPH. In contrast, sodium dodecyl sul-
fate (SDS), an anionic detergent that increases cellular
permeability by damaging cell membranes,39 signifi-
cantly decreased the anisotropy measured by TMA-DPH
without affecting the fluorescence anisotropy measured
by DPH, indicating that SDS selectively increased the
fluidity of the hydrophilic region of the cell membrane.
These results confirmed that fluorescence anisotropy
measured with DPH and TMA-DPH was a good indica-
tor of increased or decreased fluidity of cell membranes
with respect to control. One of the potent 3-alkylamido-
2-alkoxypropylphosphocholines (compound 14, linear
chain) had little effect on membrane fluidity at its EC50
(i.e., 40 µM). In contrast, one of the less potent 3-alkyl-
amido-2-alkoxypropylphosphocholines (compound 25,
branched chain) significantly increased the membrane
fluidity in both regions at a concentration near its EC50
(i.e., 300 µM). These results suggest that the paracelluar
permeability enhancement by 3-alkylamido-2-alkoxy-
propylphosphocholines is not the direct result of the
alteration of the cell membrane fluidity. The results also
suggest that the potent 3-alkylamido-2-alkoxypropyl-
phosphocholines may increase paracellular permeability
by a mechanism other than grossly altering the struc-
ture of cell membranes, while the less potent ones may
achieve their effects, at least in part, by interacting with
cell membranes. Our results with a related potent
enhancer of paracellular permeability, hexadecylphos-
phocholine (HPC),40 confirmed this conclusion in that
HPC caused only a slight increase in membrane fluidity
in the hydrophilic region, and this effect was indepen-
dent of concentration.
F igu r e 4. Correlation between the IC50(PLC) and the EC50
values for U73122 (a potent PLC inhibitor), five alkylphos-
phocholines (decylphosphocholine, dodecylphoaphocholine, tet-
radecylphosphocholine, hexadecylphosphocholine, and octa-
decylphosphocholine), and two 3-alkylamido-2-alkoxypropyl-
phosphocholines (compounds 12 and 14). The data points for
the title compounds are highlighted with rectangles. Data
points represent mean ( SD in triplicate from three separate
experiments.
Caco-2 cells.40,43,44 On the other hand, both Caco-2 cells
and MDCK cells have been used for assessment of
transport properties of compounds across epithelial
tissue, including intestinal epithelium.30,31 These com-
pounds, similar to alkylphosphocholines, inhibited PLC
and increased paracellular permeability in a concentra-
tion-dependent manner. The IC50(PLC), which was de-
fined as the concentration that inhibited ATP-stimu-
lated inositol triphosphate formation by 50%, was 30 (
17 and 10 ( 4 µM, respectively, for compounds 12 and
14. The EC50 in MDCK cell monolayers for these
compounds was 51 ( 9 and 13 ( 4 µM, respectively,
and the EC10x was 121 ( 16 and 27 ( 5 µM, respectively.
The potencies of these compounds to increase paracel-
lular permeability and inhibit PLC maintained the same
rank order, suggesting the existence of an association
between inhibition of PLC and increase in tight junction
permeability caused by these compounds. In addition,
the potencies to cause inhibition of PLC and increase
in paracellular permeability of 3-alkylamido-2-alkoxy-
propylphosphocholines fit very well to the relationship
between the potencies of alkylphosphocholines to cause
inhibition of PLC and enhancement of paracellular
permeability in MDCK cell monolayers (Figure 4),
suggesting that inhibition of PLC might be the bio-
chemical mechanism underlying the ability of 3-alkyl-
amido-2-alkoxypropylphosphocholines to increase para-
cellular permeability.
Rela tion sh ip betw een In h ibition of ATP -Stim -
u la ted P LC a n d En h a n cem en t of P a r a cellu la r
P er m ea bility a cr oss MDCK Cell Mon ola yer s. We
have demonstrated that alkylphosphocholines increase
tight junction permeability across MDCK cell monolay-
ers through inhibition of PLC.40 In addition, PLC has
been implicated in the modulation of paracellular
permeability by fatty acids and acyl carnitines.41,42
Alkylphosphocholines are structurally similar to 3-alkyl-
amido-2-alkoxypropylphosphocholines containing a short
branch. To determine if this same biochemical mecha-
nism is behind the ability of these compounds to
increase tight junction permeability, the effect of two
3-alkylamido-2-alkoxypropylphosphocholines containing
a short branch [e.g., 3-tetradecanamido-2-ethoxypropyl-
phosphocholine (compound 12) and 3-octadecanamido-
2-ethoxypropylphosphocholine (compound 14)] on para-
cellular permeability and ATP-stimulated PLC activity
was determined in MDCK cell monolayers. MDCK cells
were used because the relationship between PLC activ-
ity and paracellular permeability enhancement has been
more extensively investigated in this model than in
Discu ssion
There are numerous reports indicating that phospho-
lipids and related compounds enhance permeability of
tight junctions in epithelial tissue.4 Hence, many phos-
pholipid derivatives have been investigated as enhanc-
ers of oral or nasal delivery of hydrophilic drugs with
poor absorption. The choice of these agents as absorption
enhancers is frequently empirical, and the mechanism
of their action is poorly understood. In a previous paper,
we showed that a long alkyl chain (C-12) and a zwitte-
rionic headgroup are sufficient structural features for
causing enhancement in tight junction permeability.20,32
In the accompanying study, we also showed that glyc-
erophospholipid analogues containing a short alkyl
ether functionality at C-2 (methoxy or ethoxy) and a

2862 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13
Ouyang et al.
long alkylamido moiety at C-3 (amides of 16-19 carbon
acids) were quite potent as enhancers of tight junction
permeability across Caco-2 cell monolayers.20 We fur-
ther showed that derivatives with longer alkyoxy chain
length at C-2 were less potent. The results with a
limited series of 3-alkylamido-2-alkoxyphosphocholines
suggested that branching at C-2 might be detrimental
to the potency of these compounds as enhancers of tight
junction permeability. In the present study, the effect
of the alkyl chain length at C-2 and C-3 on the potency
as enhancers of paracellular permeability is systemati-
cally investigated with 32 phosphocholine derivatives.
This study represents the first attempt at systematically
evaluating the importance of linear vs branched alkyl
chains in defining the potency of phosphocholines as
enhancers of tight junction permeability. The 3-alkyla-
mido-2-alkoxyphsophocholines were chosen for two rea-
sons: (i) The glycerol skeleton allowed access to syn-
thetic schemes for varying alkyl chain lengths at C-2
and C-3; this enabled us to design compounds with
varying extent of branching. (ii) These compounds
represent metabolically stable analogues of naturally
occurring glycerophospholipids.
of P2Y2 receptors by ATP.48 The activity of different
families of PLC is regulated through different receptor-
mediated pathways. For example, the activities of PLC-â
and PLC-γ are regulated through G-protein-coupled
receptors (e.g., purinergic receptor) and receptor ty-
rosine kinases (e.g., epidermal growth factor receptor),
respectively.49 The PLC-dependent pathway has been
implicated in the assembly of the tight junction,50-54
particularly during development, although definitive
evidence for its role in the function of the mature tight
junctions is lacking. HPC, an inhibitor of PLC,55,56 has
structural features (i.e., a long alkyl chain and a
zwitterionic functionality) that have been previously
identified to cause an increase in tight junction perme-
ability.20,32 HPC was found to be a potent enhancer of
tight junction permeability across MDCK cell monolay-
ers and a potent inhibitor of PLC in these cells. Our
recent studies with analogues of HPC, containing alkyl
chains of different lengths, established a clear link
between inhibitory activity of alkylphosphocholines
toward PLC-â and their activity as enhancers of tight
junction permeability across MDCK cell monolayers.40
Additionally, these studies provided evidence that in-
hibition of PLC-â activity by these compounds increased
tight junction permeability via changes in organization
of actin filament network. Because the 3-alkylamido-
2-alkoxypropylphosphocholines share some of the struc-
tural features of alkylphosphocholines, we investigated
the possibility that enhancement of tight junction
permeability caused by the title compounds is also
related to inhibition of PLC-â by these compounds. As
shown in Figure 4, not only did these compounds inhibit
PLC-â, but the relationship between their potency as
PLC-â inhibitors (IC50(PLC)) and their potency as en-
hancers of tight junction permeability (EC50) fit well
with the correlation between these two parameters
obtained with the alkylphosphocholines. While these
results do not prove that the title compounds cause an
increase in tight junction permeability via inhibition of
PLC-â, they provide a very strong lead for a possible
mechanism of action underlying their activity as en-
hancers of tight junction permeability.
The potency of these compounds as enhancers of
paracellular permeability varied significantly as the
length of the alkyl chains at C-2 and C-3 was varied.
With a short alkoxy chain at C-2, the potency exhibited
a parabolic relationship with respect to the chain length
at C-3. A similar behavior was observed for compounds
with a short alkylamido chain at C-3 and varying chain
length at C-2. Interestingly, the compounds with a
linear configuration (i.e., a short chain at C-2 and a long
chain at C-3 or vice versa) were much more potent
enhancers of paracellular permeability than those with
a branched configuration for the same number of total
carbons at the two positions. The potent enhancers
identified in this study are at least an order of magni-
tude more potent than previously known amphiphilic
enhancers such as palmitoyl carnitine.16,17,32
While EC50 and EC10x values of these compounds
appear to be related to their CMC values, the com-
pounds are efficacious as enhancers above and below
their CMC values, suggesting that the enhancement of
paracellular permeability was not simply due to strong
interaction of these compounds with the cell membranes
in micellar form. Furthermore, the concentrations at
which the glycerophospholipid analogues caused changes
in the cell membrane fluidity (as evidenced by the
changes in fluorescence anisotropy of DPH and DPH-
TMA) did not appear to be related to the concentrations
at which more potent compounds caused enhanced
paracellular permeability. For less potent compounds,
changes in membrane fluidity may be contributing to
their effect on paracellular permeability. Thus, it is
reasonable to conclude that the enhancement of para-
cellular permeability caused by 3-alkylamido-2-alkoxy-
propylphsosphocholines is not due to gross perturba-
tion of the cell membrane, but rather, it is due to specific
changes in the structure and/or function of the tight
junction.
The present study provides some preliminary insights
into the structural features of metabolically stable
glycerophospholipid analogues that may contribute to
the potency of these compounds as enhancers of para-
cellular permeability. It is encouraging that the en-
hancement of the paracellular permeability is not due
to gross changes in the cell membrane integrity. At least
one of the mechanisms by which these compounds may
cause enhanced paracellular permeability involves in-
hibition of the regulatory enzyme PLC. Thus, these
compounds will provide leads for identifying mecha-
nism-based approaches to open tight junctions in a
controlled and reversible fashion. Such agents are likely
to be useful in increasing the oral absorption of hydro-
philic drugs, such as peptides, provided that the separa-
tion between efficacy and cytotoxicity is significantly
improved. The best potency index (IC50/EC50) achieved
with this series of compounds was 4, with most of the
compounds showing a potency index in the range of 1-2.
It would be important to assess the in vivo relevance to
this in vitro potency index before significant effort is
invested in improving the potency index.
PLC is an important regulatory enzyme that catalyzes
hydrolysis of phosphatidylinositol-4,5-bisphosphate into
inositol triphosphate and diacylglycerol in response to
the stimulation of a variety of receptors, e.g., stimulation

Permeability across Caco-2 Cell Monolayers
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13 2863
was diluted with 100 mL of water and extracted three times
with 100 mL of chloroform. The combined extracts were
washed with 50 mL each of cold, 5% HCl and saturated NaCl,
dried over sodium sulfate, filtered, and evaporated to dryness.
The crude residue was recrystallized two times from hexanes
giving light yellow products.
r a c-3-H e xa d e ca n a m id o-1-t r ip h e n ylm e t h oxy-2-p r o-
p a n ol. Yield, 53%; Rf ) 0.60 in chloroform. 1H NMR (300 MHz,
CDCl3): δ 0.89 (t, 3H, J ) 6. 9 Hz), 1.27 (m, 24H), 1.57 (m,
2H), 2.09 (m, 2H), 3.10-4.20 (m, 5H), 5.64 (t, 1H, J ) 8.1 Hz),
7.20-7.50 (m, 15H). MS (FAB): m/z 572.4 (M+1).
3-Alk yla m id o-2-a lk oxy-1-t r ip h en ylm et h oxyp r op a n es
(6). A solution of 3-alkylamido-1-triphenylmethoxy-2-propanol
(0.045 mol) in 100 mL of THF was added to a slurry of 85%
NaH (0.05 mol) in 10 mL of THF. After it was stirred for 1 h
at room temperature, alkyl bromide (0.05 mol) was added and
the reaction mixture was heated to 50 °C for 4 h. An additional
0.01 mol of NaH and 0.02 mol of alkyl bromide was added,
and heating was continued overnight. After it was cooled,
water was added slowly to decompose any residual NaH.
Diethyl ether (100 mL) was added, and the layers were
separated. The aqueous layer was re-extracted with ether (2
× 100 mL), and the organic extracts were combined, washed
with brine (2 × 100 mL), and dried over sodium sulfate. The
crude product was dissolved in hot hexanes, and a small
amount of insoluble material was filtered and discarded. The
product was purified by silica gel chromatography using
hexane:ethyl acetate (9:1 to 4:1).
In the intestinal inflammatory diseases such as
irritable bowel syndrome, compromised tight junctions
are implicated in the etiology of the disease. It is unclear
whether the compromised tight junctions are the con-
sequence of the disease or whether they play a role in
the progression of the disease. A series of compounds,
such as the ones reported in this study, with a well-
defined SAR with respect to their ability to cause graded
increase in tight junction permeability, could serve as
an important mechanistic probe for the construction of
experimental models to evaluate the role of the com-
promised tight junctions in the disease progression.
Exp er im en ta l Section
Gen er a l Syn th etic Meth od s. Starting materials were
purchased from Aldrich Chemical Co. (Milwaukee, WI) or
Fluka Chemical Co. (Milwaukee, WI). Unless otherwise indi-
cated, all of the chemicals were used without further purifica-
tion. THF was distilled from sodium and benzophenone. Thin-
layer chromatography (TLC) was conducted on Aldrich TLC
silica gel plates, and fractions were visualized by UV or the
spray of either sulfuric acid/methanol (95: 5 v/v) or phospho-
molybdate reagent followed by heating. Column chromatog-
raphy was performed on silica gel 60 (230-400 mesh) and
eluted with a mixture of ethyl acetate and hexanes unless
otherwise indicated. 1H NMR spectra were recorded on a
Varian 300 MHz spectrometer. Chemical shifts were reported
in parts per million (δ) relative to tetramethylsilane. Mass
spectrum was conducted on a VG Analytical ZAB double-
focusing spectrometer. For each typical reaction, only the
spectral data of one representative compound are provided (the
detailed characterization of the final products is provided in
Supporting Information). The identity of the test compounds
was established by accurate mass measurement and NMR
spectra, and their purity was >98% as evidenced by TLC
analysis.
3-Alk yla m id o-1,2-p r op a n ed iols (4). Meth od A. To a
magnetically stirred solution of 3-amino-1,2-propanediol (0.32
mol) in 100 mL of pyridine and 350 mL of DMF was added a
solution of acid chloride (0.33 mol) in 150 mL of DMF. After 2
h, the total volume of solvent was reduced by three-fourths.
The gelatinous mass was filtered, washed with water, and
dried in the air. The solid was recrystallized successively from
350 mL of ethanol, 350 mL of 2-propanol, and 100 mL of
chloroform to give a white powder.
Meth od B. A solution of 3-amino-1,2-propanediol (0.17 mol)
and K2CO3 (0.18 mol) in 120 mL of water was kept on ice-
bath. Acid chloride (0.18 mol) in 120 mL of chloroform was
added dropwise. After 1 h, the ice-bath was removed and the
reaction mixture was kept at 40 °C overnight. The solvent was
removed under reduced pressure, and the crude product was
filtered to remove the insoluble inorganic salt. The product
was obtained as a colorless liquid in quantitative yield.
Meth od C. A solution of 3-amino-1,2-propanediol (0.10 mol)
in acetic anhydride (25 mL) was stirred overnight at room
temperature. Acetic acid and acetic anhydride were removed
under reduced pressure. Methanol (50 mL) and solid potassium
carbonate (10.0 g) were added, and the suspension was stirred
for an additional 2 days. The solvent was then removed to give
a colorless liquid in quantitative yield. The crude product
mixture was used in the next reaction without further puri-
fication.
r a c-3-H e xa d e ca n a m id o-2-oct a d e coxy-1-t r ip h e n yl-
m eth oxyp r op a n e. Yield, 63%; Rf ) 0.30 in hexane:ethyl
1
acetate ) 7:1. H NMR (300 MHz, CDCl3): δ 0.89 (t, 6H, J )
6.5 Hz), 1.27 (broad m, 54H), 1.45-1.70 (m, 4H), 2.09 (t, 2H,
J ) 7.8 Hz), 3.00-3.70 (m, 7H), 5.77 (t, 1H, J ) 5.4 Hz), 7.20-
7.50 (m, 15H). MS (FAB): m/z 824.2 (M+1).
3-Alk yla m id o-2-a lk oxy-1-p r op a n ols (7). p-Toluenesulfon-
ic acid (0.08 mmol) was added to a solution of 3-alkylamido-
2-alkoxy-1-triphenylmethoxypropane (0.32 mmol) in methyl-
ene chloride (10 mL) and methanol (0.5 mL). The solution was
stirred for 8 h at room temperature. Aqueous solution of
saturated sodium bicarbonate (2 mL) was added and stirred
for 0.5 h. The layers were separated, and the organic fraction
was washed with brine. After it was dried over sodium sulfate,
the solution was concentrated under reduced pressure. Chro-
matography on silica gel with a gradient of methylene chloride:
methanol (100:0 to 95:5) gave a white solid.
r a c-3-Hexadecan am ido-2-octadecoxy-1-pr opan ol. Yield,
84%; Rf ) 0.35 in methylene chloride:methanol ) 100:2. 1H
NMR (300 MHz, CDCl3): δ 0.89 (m, 6H), 1.27 (broad, 54H),
1.45-1.70 (m, 4H), 2.23 (t, 2H), 3.20-3.80 (m, 7H), 5.77 (t,
1H, J ) 5.7 Hz). MS (FAB): m/z 582.2 (M+1).
3-Alk yla m id o-2-a lk oxyp r op ylp h osp h och olin es
(2).
Meth od D. To a dry ice-bath-cooled solution of 3-alkylamido-
2-alkoxy-1-propanol (0.17 mmol) and trimethylamine (5.1
mmol) in 10 mL of anhydrous acetonitrile was added 2-chloro-
2-oxo-1,3,2-dioxophospholane (0.25 mmol). The glass pressure
tube was sealed, and the reaction mixture was stirred at 65
°C for 16 h. Then, the tube was cooled on dry ice and broken.
After it was warmed up to room temperature, methanol was
added to dissolve the precipitate. The solvent was removed,
and the product was purified by silica gel chromatography
using a gradient of methylene chloride:methanol (5:1 to 5:3).
Meth od E. To a solution of 3-alkylamido-2-alkoxy-1-pro-
panol (7.8 mmol) in anhydrous diethyl ether/THF (2:1, 225 mL)
at 0 °C was added dry pyridine (10 mL) followed by 2-bromo-
ethyl dichlorophosphate (3.3 mL, 21 mmol). The reaction
mixture was warmed to room temperature and heated at a
gentle reflux for 4 h. After it was cooled in an ice-bath, 6 mL
of water was added dropwise and stirring was continued for
another hour. The solvent was removed on a rotary evaporator,
the residue was taken up in 100 mL of chloroform/methanol
(2:1, v/v), and the solution was extracted with 20 mL of water
while the pH of the organic phase was kept at 5 by addition of
2 N HCl. The aqueous phase was re-extracted with two
portions of chloroform/methanol (2:1, v/v, 50 mL), and the
r a c-3-Hexa d eca n a m id o-1,2-p r op a n ed iol. Yield, 62%; Rf
) 0.60 in chloroform:methanol ) 4:1. 1H NMR (300 MHz,
CDCl3): δ 0.89 (t, 3H, J ) 6.8 Hz), 1.26 (broad m, 24H), 1.65
(m, 2H), 2.26 (m, 2H), 3.20-4.20 (m, 5H), 6.12 (m, 1H). MS
(FAB): m/z 330.5 (M+1).
3-Alk yla m id o-1-tr ip h en ylm eth oxy-2-p r op a n ols (5). Tri-
tyl chloride (0.11 mol) was added to a stirring solution of
3-alkylamido-1,2-propanediol (0.10 mol) in 250 mL of pyridine.
The reaction mixture was kept at 45-50 °C for 10 h. After
the pyridine was removed under reduced pressure, the residue

2864 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13
Ouyang et al.
combined organic phases were dried over sodium sulfate and
concentrated. The residue was purified on silica gel column
with a gradient of chloroform:methanol (100:0 to 3:1), giving
colorless oil or white solid. In some cases, product was used
in the next reaction without thorough purification and char-
acterization.
To a mixture of 3-alkylamido-2-alkoxy-1-propyl-2′-bromo-
ethyl phosphate (5.7 mmol) in chloroform:2-propanol:DMF (3:
5:5, 195 mL) was added trimethylamine (40% aqueous solu-
tion, 41 mL). The solution was heated to 65 °C for 5 h and
then allowed to be cooled to room temperature before silver
carbonate (196 mg) was added. Heat was reapplied for 1 h.
The mixture was cooled and filtered. The solvent was removed
in vacuo, and the residue was purified on silica gel column
using chloroform:methanol:NH4OH (75:25:5).
buffer. EC50 was calculated as the concentration at which the
phospholipid decreases the TEER of cell monolayer by 50% of
the control (untreated) value. All measurements were in
triplicate and expressed as mean ( SD values.
Mea su r em en t of Ma n n itol Tr a n sp or t Ra te (EC10x De-
ter m in a tion ). All transport studies were performed at 37 °C
under sink condition (i.e., the concentration of the transported
compound in the receiver compartment was less than 10% of
its initial concentration in the donor compartment). Perme-
ability coefficients (Papp) were calculated using the following
equation: Papp ) dQ/(dt × C0 × A), where dQ/dt (mol trans-
ported/sec) is the flux of the marker compound across the
Caco-2 cell monolayer, A (cm2) represents the diffusional area
of the inserts, and C0 (M) denotes the initial concentration of
marker compound in the donor compartment. To calculate the
enhancer concentration at which the permeability of mannitol
is increased by 10-fold (EC10x), the enhancer solution at
concentration range of 0.01-5.0 mM in HBSS (0.5 mL) was
added to the apical side of the cells. HBSS (1.5 mL) was added
to the basolateral side. After treatment for 20 min, the
enhancer solution in the apical compartment was removed and
replaced with fresh HBSS after washing the cell monolayer
once with HBSS. Following the measurement of TEER, the
HBSS on the apical side was replaced with 0.5 mL of HBSS
solution containing 0.1 mM [3H]mannitol (0.25 µCi). Samples
from the basolateral side were taken at 20, 40, 60, and 90 min.
Transport rates were determined by measuring the radioactiv-
ity in the basolateral side with a liquid scintillation counter.
All measurements were in triplicate and expressed as mean
( SD values.
Cytotoxicity Assa y. The cell viability was measured by
the MTT assay.45-47 Approximately 3.0 × 103 Caco-2 cells (in
100 µL of cell culture medium) were seeded into each well in
a 96 well tissue culture plate (Corning-Costar). The cells were
cultured in the same manner as described previously (Caco-2
and MDCK cell cultures) for 96 h. The culture medium was
changed once. J ust prior to the start of each experiment, the
medium was removed from the wells, and 100 µL of the
3-alkylamido-2-alkoxypropylphosphocholine solutions in HBSS
was added to each well. After 20 min, 20 µL of MTT solution
(5 mg/mL) was added to each well, and the cells were incubated
for another 120 min. Then, 100 µL of 10% SDS in 0.02 M HCl/
isobutanol (1:1, v/v) solution was added to stop the reaction.
The cells without the treatment of any phospholipid were
harvested as above and were used as controls. Absorbance was
measured at 590 nm (indicative of the formation of the
formazan product by mitochondrial dehydrogenase of the
viable cells) using a multiwell scanning spectrophotometer
(Bio-Rad, Hercules, CA). IC50 represents the concentration of
the phospholipid that causes 50% cells to be incompetent with
respect to mitochondrial dehydrogenase activity. All measure-
ments were in triplicate and expressed as mean ( SD values.
CMC Deter m in a tion . The aqueous solutions of 3-alkyl-
amido-2-alkoxypropylphosphocholine at appropriate concen-
tration range were prepared. To each 2 mL solution of
phosphocholine at various concentrations was added 1 µL stock
solution of 10 mM DPH in THF. The solutions were mixed
thoroughly and incubated in darkness at room temperature
for 30 min before fluorescence was measured by a Perkin-
Elmer Spectrofluorometer set at excitation wavelength of 365
nm and emission wavelength of 460 nm. The excitation and
emission bandwidths were set at optimal width to achieve
highest sensitivity. Linear regression analysis on the sets of
points below and above the inflection point of fluorescence
intensity formed two straight lines that intersected at the
CMC. Mean CMC was calculated from triplicates measure-
ments for each compound.
Mem br a n e F lu id ity Mea su r em en t. Caco-2 cells cultured
on Transwell membranes were treated with 0.2% EDTA in
PBS solution for 5 min and trypsinized. The cells were
suspended, collected by centrifugation, washed twice with
HBSS, and diluted in HBSS to a concentration of 2.0 ×105
cells/mL. Caco-2 cells were labeled with DPH by adding 2.5
µL of 1 mM freshly prepared stock solution in THF to 2.5 mL
r a c-3-Hexa d eca n a m id o-2-octa d ecoxyp r op ylp h osp h o-
ch olin e (42). Yield, 46%; Rf ) 0.35 in chloroform:methanol:
1
ammonium hydroxide ) 75:25:5. H NMR (300 MHz, DMSO-
d6): δ 0.85 (m, 6H), 1.27 (m, 54H), 1.45-1.70 (m, 4H), 2.30 (t,
2H, J ) 6.5 Hz), 3.33 (s, 9H), 3.30-3.60 (m, 6H), 3.60-4.00
(m, 3H), 4.35 (m, 2H). MS (FAB): m/z 747.6 (M+1). HRMS
(FAB) calcd for C42H88N2O6P, 747.6380; found, 747.6324.
P h ysicoch em ica l a n d Biologica l Eva lu a tion . Caco-2
cells were purchased from American Type Culture Collection
(Rockville, MD). Eagle’s Minimum Essential Medium (EMEM),
0.25% trypsin/0.02% ethylenediaminetetraacetic acid-tetra-
sodium salt (EDTA-4Na), and fetal bovine serum (FBS) were
obtained from Gibco (Grand Island, NY). Nonessential amino
acids (NEAA), Hank’s balanced salt solution (HBSS), antibiotic
antimycotic solution, MTT, SDS, hexadecylphosphocholine,
DPH, and TMA-DPH were purchased from Sigma (St. Louis,
MO). Dodecylphosphocholine was purchased from Avanti Polar
Lipids Inc. (Alabaster, AL). N-(2-Hydroxyethyl) piperazine-N′-
2-ethanesulfonic acid (HEPES) was obtained through Tissue
Culture Facilities (UNC at Chapel Hill, NC). Transwell plates
and inserts (12 wells/plate, 3.0 µm pore and 1.0 cm2 area,
polycarbonate) were purchased from Corning-Costar (Cam-
bridge, MA). [3H]Mannitol was purchased from DuPont NEN
(Boston, MA). The stock solutions of 3-alkylamido-2-alkoxy-
propylphosphocholines (50 mM) were obtained by dissolving
the phospholipids in HBSS/ethanol (50:50, v/v) and were stored
at -20 °C.
Ca co-2 a n d MDCK Cell Cu ltu r es. Caco-2 cells were
grown in EMEM containing 10% FBS, 1% l-glutamine, 1%
NEAA, and antibiotics (100 U/mL of penicillin, 100 mg/mL of
streptomycin, and 0.25 g/mL of amphotericin B) in 75 cm2
culture flasks. The cultures were kept at 37 °C in an atmo-
sphere of 5% CO2, 95% air, and 90% relative humidity. Cells
were passaged after 95% confluency and were seeded with a
density of 1.0 × 105 cells/mL on to porous polycarbonate filter
membranes with a pore size of 3.0 µm and a surface area of
1.0 cm2. Cells from passage number 45-55 were used in all of
the studies. Media were changed every 2 days after seeding
until late confluence (20-22 days). J ust before the experi-
ments, the culture medium was replaced with HBSS buffer
that contained 1 × HBSS, 25 mM HEPES, and 25 mM glucose
at pH 7.4 and incubated for 1 h at 37 °C. The cell monolayers
with TEER values in the range of 600-800 Ω‚cm2 were used
for the experiments. MDCK cells were passaged and seeded
in the same manner as Caco-2 cells, except that MDCK cells
were maintained on filter membrane for 4 days to reach
confluence. MDCK cells from passage number 65-75 with
TEER values in the range of 250-350 Ω‚cm2 were used in all
the studies
TEER Mea su r em en t (EC50 Deter m in a tion ). To measure
the effect of the 3-alkylamido-2-alkoxypropylphosphocholines
on the TEER values of cell monolayers, the phospholipids were
dissolved in 0.5 mL of HBSS at various concentrations and
applied to the apical side of the cell monolayers for 20 min at
room temperature. The TEER value was measured with an
Epithelial Tissue Voltohmmeter (EVOM, World Precision
Instruments, Sarasota, FL) and calculated as Ω‚cm2. The
resistance caused by the cell monolayer was determined after
subtracting the contribution of the blank filter and HBSS

Permeability across Caco-2 Cell Monolayers
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13 2865
of cell suspension in HBSS and incubated at room temperature
for 30 min before fluorescence measurements were made.
Alternatively, Caco-2 cells were labeled with TMA-DPH by
adding 2.5 µL of 0.5 mM freshly prepared stock solution in
DMF to 2.5 mL of cell suspension in HBSS and incubated at
room temperature for 2 min before fluorescence measurements
were made. These labeling conditions have been found to allow
maximal fluorescent probe incorporation and to allow stable
measurements over the experiment period.57
Refer en ces
(1) Granger, D. N.; Kvietys, P. R.; Perry, M. A.; Barrowman, J . A.
The microcirculation and intestinal transport. Physiology of the
Gastrointestinal Tract, 2nd ed.; Raven Press: New York, 1987;
pp 1671-1697.
(2) Gangwar, S.; Pauletti, G. M.; Wang, B.; Siahaan, T. J .; Stella,
V. J .; et al. Prodrug strategies to enhance the intestinal
absorption of peptides. Drug Discovery Today 1997, 2, 148-155.
(3) Lapierre, L. A. The molecular structure of the tight junction.
Adv. Drug Delivery Rev. 2000, 41, 255-264.
(4) Ward, P. D.; Tippin, T. K.; Thakker, D. R. Enhancing paracel-
lular permeability by modulating epithelial tight junctions.
Pharm. Sci. Technol. Today 2000, 3, 346-358.
Fluorescence anisotropy was measured with an SLM-
AMINCO Subnanosecond Lifetime Fluorometer, model 4800c,
equipped with a Zenith 158 computer for data analysis as
described by Audus et al.35,36 In brief, photomultiplier tubes
were placed to the right and left of the dual chamber sample
cell with Glans-Thompson polarizers inserted in emission and
excitation beams. A Schott KV-389 filter was inserted into the
horizontally polarized emission beam, and an SLM MC320
monochromator set at 430 nm was placed in the vertically
polarized emission beam. Fluorescence intensity was first
measured with the sample excited at 360 nm with a horizon-
tally polarized beam of light and then measured a second time
with the sample excited with a vertically polarized beam of
light. Corrected fluorescence anisotropy data were calculated
from the relationships: P ) (A/B - 1)/(A/B + 1), r ) 2P/3 - P,
where P was the fluorescence polarization, A was the ratio of
fluorescence intensities parallel and perpendicular to the plane
of vertically polarized excitation light, B was the ratio of
fluorescence intensities parallel and perpendicular to the plane
of horizontally polarized excitation light, and r was the
fluorescence anisotropy. Under these conditions, the effect of
light scattering will be negligible. Each sample was measured
six times, and the average was calculated. Caco-2 cells were
maintained in suspension by stirring, for both labeling and
fluorescence experiments, with a magnetic stirrer. Two sample
cuvettes containing cell suspension were always used as
follows: one received indicated compound and the other, as a
control, received an equivalent amount of solvent in which the
compound was dissolved.
(5) Simon, D. B.; Lu, Y.; Choate, K. A.; Velazquez, H.; Al-Sabban,
E.; et al. Paracellin-1, a renal tight junction protein required
for paracellular Mg2+ resorption. Science 1999, 285, 103-106.
(6) van Meer, G.; Gumbiner, B.; Simons, K. The tight junction does
not allow lipid molecules to diffuse from one epithelial cell to
the next. Nature 1986, 322, 639-641.
(7) Dragsten, P. R.; Blumenthal, R.; Handler, J . S. Membrane
asymmetry in epithelia: is the tight junction
a barrier to
diffusion in the plasma membrane? Nature 1981, 294, 718-722.
(8) Mandel, L. J .; Bacallao, R.; Zampighi, G. Uncoupling of the
molecular ‘fence’ and paracellular ‘gate’ functions in epithelial
tight junctions. Nature 1993, 361, 552-555.
(9) Anderson, J . M.; Van Itallie, C. M. Tight junctions and the
molecular basis for regulation of paracellular permeability. Am.
J . Physiol. 1995, 269, G467-G475.
(10) Huff, J . R. HIV protease: a novel chemotherapeutic target for
AIDS. J . Med. Chem. 1991, 34, 2305-2314.
(11) Wang, W.; J iang, J .; Ballard, C. E.; Wang, B. Prodrug approaches
to the improved delivery of peptide drugs. Curr. Pharm. Des.
1999, 5, 265-287.
(12) Swenson, E. S.; Curatolo, W. J . Means to enhance penatration
(2) Intestinal permeability enhancement for proteins, peptided
and other polar drugs: mechanisms and potential toxicity. Adv.
Drug Delivery Rev. 1992, 8, 39-92.
(13) Oliyai, R. Prodrugs of peptides and peptidomimetics for improved
formulation and delivery. Adv. Drug Delivery Rev. 1996, 19,
275-286.
(14) Ojima, I.; Chakravarty, S.; Dong, Q. Antithrombotic agents: from
RGD to peptide mimetics. Bioorg. Med. Chem. 1995, 3, 337-
360.
(15) Lee, V. H.; Yamamoto, A.; Kompella, U. B. Mucosal penetration
enhancers for facilitation of peptide and protein drug absorption.
Crit. Rev. Ther. Drug Carrier Syst. 1991, 8, 91-192.
(16) LeCluyse, E. L.; Sutton, S. C. In vitro models for selection of
development candidates. Permeability studies to define mech-
anisms of absorption enhancement. Adv. Drug Delivery Rev.
1997, 23, 163-183.
(17) Sutton, S. C.; LeCluyse, E. L.; Engle, K.; Pipkin, J . D.; Fix, J . A.
Enhanced bioavailability of cefoxitin using palmitoylcarnitine.
II. Use of directly compressed tablet formulations in the rat and
dog. Pharm. Res. 1993, 10, 1516-1520.
(18) Sekine, M.; Terashima, H.; Sasahara, K.; Nishimura, K.; Okada,
R.; et al. Improvement of bioavailability of poorly absorbed drugs.
II. Effect of medium chain glyceride base on the intestinal
absorption of cefmetazole sodium in rats and dogs. J . Pharma-
cobiodyn. 1985, 8, 286-295.
(19) Talbot, R. W.; Foster, J . R.; Hermon-Taylor, J .; Grant, D. A.
Induced mucosal penetration and transfer to portal blood of
luminal horseradish peroxidase after exposure of mucosa of
guinea pig small intestine to ethanol and lysolecithin. Dig. Dis.
Sci. 1984, 29, 1015-1022.
(20) Liu, D. Z.; Morris-Natschke, S. L.; Kucera, L. S.; Ishaq, K. S.;
Thakker, D. R. Structure-activity relationships for enhance-
ment of paracellular permeability by 2-alkoxy-3-alkylamidopro-
pylphosphocholines across Caco-2 cell monolayers. J . Pharm. Sci.
1999, 88, 1169-1174.
(21) Surles, J . R.; Morris-Natschke, S.; Marx, M. H.; Piantadosi, C.
Multigram synthesis of 1-alkylamido phospholipids. Lipids 1993,
28, 55-57.
(22) J anusz, J . M.; Buckwalter, B. L.; Young, P. A.; LaHann, T. R.;
Farmer, R. W.; et al. Vanilloids. 1. Analogues of capsaicin with
antinociceptive and antiinflammatory activity. J . Med. Chem.
1993, 36, 2595-2604.
(23) Yous, S.; Andrieux, J .; Howell, H. E.; Morgan, P. J .; Renard, P.;
et al. Novel naphthalenic ligands with high affinity for the
melatonin receptor. J . Med. Chem. 1992, 35, 1484-1486.
(24) Mbappe, M.; Scisic, S. Enzymatic resolutions in 3-amino-1,2-
propanediol series. Tetrahedron: Asymmetry 1993, 4, 1035-
1040.
Deter m in a tion of ATP -Stim u la ted P LC Activity in
MDCK Cells. The activity of PLC in MDCK cells was
determined by an adaptation of a previously published method.58
MDCK cells were seeded at 400 000 cells/well in a 12 well plate
and subsequently cultured for 4 days. MDCK monolayers were
then labeled with [3H]myo-inositol (1.6 µCi/well in 0.4 mL of
inositol-free media) for 24 h at 37 °C. Assays were initiated
on labeled cells that were removed from the incubator by
immediately supplementing the cells with 100 µL of 250 mM
HEPES (pH 7.3), containing 100 mM LiCl (to inhibit hydrolysis
of inositol monophosphate to inositol) with or without the test
compound. Subsequently, the cells were incubated for 30 min
at 37 °C. Immediately after the termination of this incubation,
ATP (final concentration: 100 µM) was added. The cells were
then incubated at 37 °C for 15 min to allow accumulation of
[3H]inositol phosphates. Incubations were terminated by as-
piration of the media and the addition of 1 mL of boiling 10
mM EDTA (pH 8.0). The supernatant was applied to AG1 × 8
formate columns for chromatographic isolation of [3H]inositol
phosphates.59 The amount of [3H]inositol phosphates was
measured by liquid scintillation counting in a Packard Tri
Carb 4000 Series spectrophotometer. Data from each experi-
ment were normalized to the response observed with 100 µM
ATP (without treatment with test compounds) and were
reported as the mean ( SD of three experiments performed
in triplicate. The IC50(PLC) was determined by the concentration
of compound that causes a 50% decrease in ATP-stimulated
accumulation of [3H]inositol phosphates.
Ack n ow led gm en t. The financial support by Glaxo-
Wellcome, Inc. is gratefully acknowledged.
(25) Crans, D. C.; Whitesides, G. M. Glycerol kinase: substrate
specificity. J . Am. Chem. Soc. 1985, 107, 7008-7018.
(26) Chaudhary, S. K.; Hernandez, O. A simplified procedure for the
preparation of triphenylmethyl ethers. Tetrahedron Lett. 1979,
2, 95-98.
Su p p or tin g In for m a tion Ava ila ble: Detailed character-
ization of the final products is provided. This material is
available free of charge via the Internet at http://pubs.acs.org

2866 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 13
Ouyang et al.
(27) J ia, C.; Haines, A. H. Diamide alalogues of phosphatidyl choline
as potential anti-AIDS agents. J . Chem. Soc., Perkin Trans. 1
1993, 2521-2523.
(45) Mosmann, T. Rapid colorimetric assay for cellular growth and
survival: application to proliferation and cytotoxicity assays. J .
Immunol. Methods 1983, 65, 55-63.
(28) Hansen, W. J .; Murari, R.; Wedmid, Y.; Baumann, W. J . An
improved procedure for the synthesis of choline phospholipids
via 2-bromoethyl dichlorophosphate. Lipids 1982, 17, 453-459.
(29) Gan, L. L.; Thakker, D. R. Applications of the Caco-2 model in
the design and development of orally active drugs: elucidation
of biochemical and physical barriers posed by the intestinal
epithelium. Adv. Drug Delivery Rev. 1997, 23, 77-98.
(30) Hidalgo, I. J . Assessing the absorption of new pharmaceuticals.
Curr. Top. Med. Chem. 2001, 1.
(46) Scudiero, D. A.; Shoemaker, R. H.; Paull, K. D.; Monks, A.;
Tierney, S.; et al. Evaluation of a soluble tetrazolium/formazan
assay for cell growth and drug sensitivity in culture using human
and other tumor cell lines. Cancer Res. 1988, 48, 4827-4833.
(47) Alley, M. C.; Scudiero, D. A.; Monks, A.; Hursey, M. L.;
Czerwinski, M. J .; et al. Feasibility of drug screening with panels
of human tumor cell lines using a microculture tetrazolium
assay. Cancer Res. 1988, 48, 589-601.
(48) Nicholas, R. A.; Watt, W. C.; Lazarowski, E. R.; Li, Q.; Harden,
K. Uridine nucleotide selectivity of three phospholipase C-
activating P2 receptors: identification of a UDP-selective, a
UTP-selective, and an ATP- and UTP-specific receptor. Mol.
Pharmacol. 1996, 50, 224-229.
(49) Katan, M. Families of phosphoinositide-specific phospholipase
C: structure and function. Biochim. Biophys. Acta 1998, 1436,
5-17.
(50) Thackeray, J . R.; Gaines, P. C.; Ebert, P.; Carlson, J . R. small
wing encodes a phospholipase C-(gamma) that acts as a negative
regulator of R7 development in Drosophila. Development 1998,
125, 5033-5042.
(51) Wang, S.; Gebre-Medhin, S.; Betsholtz, C.; Stalberg, P.; Zhou,
Y.; et al. Targeted disruption of the mouse phospholipase C beta3
gene results in early embryonic lethality. FEBS Lett. 1998, 441,
261-265.
(52) Emori, Y.; Sugaya, R.; Akimaru, H.; Higashijima, S.; Shishido,
E.; et al. Drosophila phospholipase C-gamma expressed pre-
dominantly in blastoderm cells at cellularization and in endo-
dermal cells during later embryonic stages. J . Biol. Chem. 1994,
269, 19474-19479.
(53) Cereijido, M.; Gonzalez-Mariscal, L.; Contreras, R. G.; Gallardo,
J . M.; Garcia-Villegas, R.; et al. The making of a tight junction.
J . Cell Sci. Suppl. 1993, 17, 127-132.
(54) Balda, M. S.; Gonzalez-Mariscal, L.; Contreras, R. G.; Macias-
Silva, M.; Torres-Marquez, M. E.; et al. Assembly and sealing
of tight junctions: possible participation of G- proteins, phos-
pholipase C, protein kinase C and calmodulin. J . Membr. Biol.
1991, 122, 193-202.
(31) Bohets, H.; Annaert, P.; Mannens, G.; Van Beijsterveldt, L.;
Anciaux, K.; et al. Strategies for absorption screening in drug
discovery and development. Curr. Top. Med. Chem. 2001, 1,
367-383.
(32) Liu, D. Z.; LeCluyse, E. L.; Thakker, D. R. Dodecylphosphocho-
line-mediated enhancement of paracellular permeability and
cytotoxicity in Caco-2 cell monolayers. J . Pharm. Sci. 1999, 88,
1161-1168.
(33) Chattopadhyay, A.; London, E. Fluorimetric determination of
critical micelle concentration avoiding interference from deter-
gent charge. Anal. Biochem. 1984, 139, 408-412.
(34) Stafford, R. E.; Fanni, T.; Dennis, E. A. Interfacial properties
and critical micelle concentration of lysophospholipids. Biochem-
istry 1989, 28, 5113-5120.
(35) Audus, K. L.; Guillot, F. L.; Braughler, J . M. Evidence for 21-
aminosteroid association with the hydrophobic domains of brain
microvessel endothelial cells. Free Radical Biol. Med. 1991, 11,
361-371.
(36) Audus, K. L.; Gordon, M. A. Effect of tricyclic antidepressant
drugs on lymphocyte membrane structure. J . Immunopharma-
col. 1984, 6, 105-132.
(37) Van Blitterswijk, W. J .; Van Hoeven, R. P.; Van der Meer, B.
W. Lipid structural order parameters (reciprocal of fluidity) in
biomembranes derived from steady-state fluorescence polariza-
tion measurements. Biochim. Biophys. Acta 1981, 644, 323-332.
(38) Prendergast, F. G.; Haugland, R. P.; Callahan, P. J . 1-[4-
(Trimethylamino)phenyl]-6-phenylhexa-1,3,5-triene: synthesis,
fluorescence properties, and use as a fluorescence probe of lipid
bilayers. Biochemistry 1981, 20, 7333-7338.
(39) Kakemi, K.; Sezaki, H.; Muranishi, S.; Tsujimura, Y. Absorption
and excretion of drugs. XL. Enhancement of the rectal absorption
of pharmaceutical amines with lauryl sulfate and saccharinate
anions. Chem. Pharm. Bull. (Tokyo) 1969, 17, 1641-1650.
(40) Ward, P. D.; Ouyang, H.; Thakker, D. R. Unpublished results.
(41) Lindmark, T.; Kimura, Y.; Artursson, P. Absorption enhance-
ment through intracellular regulation of tight junction perme-
ability by medium chain fatty acids in Caco-2 cells. J . Pharmacol.
Exp. Ther. 1998, 284, 362-369.
(55) Berkovic, D.; Haase, D.; Hiddemann, W.; Fleer, E. A. Induction
of resistance in the human leukemia cell line HL60 towards
hexadecylphosphocholine and other ether phospholipid ana-
logues. J . Exp. Ther. Oncol. 1996, 1, 368-375.
(56) Pawelczyk, T.; Lowenstein, J . M. Inhibition of phospholipase C
delta by hexadecylphosphorylcholine and lysophospholipids with
antitumor activity. Biochem. Pharmacol. 1993, 45, 493-497.
(57) Audus, K. L.; Holthaus, S. R.; van Bree, J . M.; Shinogle, J . A.
Aluminum effects on growth of brain microvessel endothelial
cells in primary culture. Res. Commun. Chem. Pathol. Pharma-
col. 1988, 60, 71-85.
(42) Tomita, M.; Hayashi, M.; Awazu, S. Absorption-enhancing
mechanism of EDTA, caprate, and decanoylcarnitine in Caco-2
cells. J . Pharm. Sci. 1996, 85, 608-611.
(43) Peterson, M. W.; Walter, M. E. Calcium-activated phosphati-
dylcholine-specific phospholipase C and D in MDCK epithelial
cells. Am. J . Physiol. 1992, 263, C1216-C1224.
(58) Schachter, J . B.; Sromek, S. M.; Nicholas, R. A.; Harden, T. K.
HEK293 human embryonic kidney cells endogenously express
the P2Y1 and P2Y2 receptors. Neuropharmacology 1997, 36,
1181-1187.
(44) Huang, C. F.; Cabot, M. C. Phorbol diesters stimulate the
accumulation of phosphatidate, phosphatidylethanol, and dia-
cylglycerol in three cell types. Evidence for the indirect formation
of phosphatidylcholine-derived diacylglycerol by a phospholipase
D pathway and direct formation of diacylglycerol by a phospho-
lipase C pathway. J . Biol. Chem. 1990, 265, 14858-14863.
(59) Berridge, M. J .; Dawson, R. M.; Downes, C. P.; Heslop, J . P.;
Irvine, R. F. Changes in the levels of inositol phosphates after
agonist-dependent hydrolysis of membrane phosphoinositides.
Biochem. J . 1983, 212, 473-482.
J M020001X