Structure-activity relationships for enhancement of paracellular permeability by 2-alkoxy-3-alkylamidopropylphosphocholines across Caco-2 cell monolayers

Article abstract of DOI:10.1021/js9900957

The oral route is the preferred route of delivery for a large number of drug molecules. However, the intestinal epithelium presents a formidable barrier for delivery of drugs into systemic circulation. Phospholipids are among compounds that enhance the ab

Full text of DOI:10.1021/js9900957

Structure−Activity Relationships for Enhancement of Paracellular
Permeability by 2-Alkoxy-3-alkylamidopropylphosphocholines across
Caco-2 Cell Monolayers
DONG-ZHOU LIU,†,‡ SUSAN L. MORRIS-NATSCHKE,§ LOUIS S. KUCERA,| KHALID S. ISHAQ,§ AND
,†
DHIREN R. THAKKER*
Contribution from Division of Drug Delivery and Disposition, and Division of Medicinal Chemistry and Natural Products, School
of Pharmacy, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-7360, and Department of
Microbiology and Immunology, Wake Forest University, School of Medicine, Winston-Salem, North Carolina 27103.
Received March 27, 1999. Accepted for publication August 26, 1999.
is particularly formidable because of the presence of the
tight junctions that severely restrict the free passage of
Abstract 0 The oral route is the preferred route of delivery for a
solutes through the paracellular space.1,2 Hence, develop-
ment of absorption enhancers that can improve the absorp-
tion of hydrophilic molecules via the paracellular route is
an active area of research in many laboratories. To date, a
large number of absorption enhancers with diverse chemi-
cal structures have been identified. These include bile salts,
medium chain fatty acids, phospholipids, acyl carnitines,
Ca2+ chelators, surfactants, and detergents.3-23
large number of drug molecules. However, the intestinal epithelium
presents a formidable barrier for delivery of drugs into systemic
circulation. Phospholipids are among compounds that enhance the
absorption of drugs across the intestinal epithelium. In this paper, we
describe structure−activity relationships for phospholipid derivatives
as enhancers of paracellular permeability across Caco-2 cell mono-
layers. In a series of 2-alkoxy-3-alkylamidopropylphosphocholine
derivatives, compounds with a long chain at C-3 (R ) and short chain
3
at C-2 (R ) were potent in causing a decrease in transepithelial
Among many compounds that can enhance the intestinal
absorption of hydrophilic drug molecules, phospholipid-like
agents with medium length fatty acid chains are of special
interest, because they are ubiquitous in living organisms
and are found in the membranes and membranous or-
ganelles of all living matter.7 In two separate studies,
formulations containing dodecanoylphoshatidylcholine
achieved ∼8-13% bioavailability (with respect to the iv
bolus dose) of insulin after intranasal administration to
healthy volunteers.8,9 Lysophosphatidylcholine, containing
a mixture of palmitoyl and stearoyl side chains, signifi-
cantly increased the nasal absorption of human growth
hormone in rats, rabbits, and sheep.10 Palmitoyllysophos-
phatidylcholine11 as well as its analogues12 increased the
permeability of poorly permeable vasopressin analogues
across Caco-2 cell monolayers by almost 2 orders of
magnitude. The increase in the permeability was achieved
at concentrations of the enhancer that did not perturb the
membrane significantly (trypan blue exclusion).12 Lyso-
phoshatidic acid caused a decrease in transepithelial
electrical resistance (TEER) and an increase in sucrose flux
across cultured brain capillary endothelial cells.13 Our
recent studies have shown that dodecylphosphocholine
(DPC) can cause an increase in the permeability of hydro-
philic compounds and a decrease in TEER across Caco-2
cell monolayers by modulation of the tight junctions.14
Very little information is available regarding the mech-
anism by which the phospholipid derivatives are able to
enhance the absorption of drug molecules or the structural
requirements for their activity. In vitro studies11-14 suggest
that these agents increase the paracellular permeability
of compounds by modulation of the tight junctions. Studies
with the brain endothelial capillary cells13 suggest that
such agents may alter the permeability of the tight junc-
tions by intracellular mechanisms such as tyrosine phos-
phorylation of proteins at focal contact, recruitment of focal
adhesion components, and activation of yet undefined
signaling pathways. Our studies have shown that DPC
displaces some of the tight junction (zonula occludens)-
associated protein ZO-1 at a concentration that causes a
decrease in TEER and an increase in paracellular perme-
2
electrical resistance (TEER) and an increase in mannitol transport,
but also showed significant cytotoxicity. Compounds with 9−11 carbons
at C-3 and 6−10 carbons at C-2 provided good separation (up to
2.7-fold) between activity and cytotoxicity. Notably, a good correlation
(r2 ) 0.93) was observed between EC (TEER) [concentration that
50
caused a drop in TEER to 50% of its control (untreated) value] and
EC10× (mannitol) [concentration that caused 10-fold increase in mannitol
transport over the control (untreated) value], confirming that a decrease
in TEER is associated with enhanced permeability of the hydrophilic
compounds across Caco-2 cell monolayers. Compounds with medium
to long carbon chains at C-2 and C-3, and the total carbons in the
alkyl chains > 20, showed poor activity and no cytotoxicity.
Introduction
The oral route is the preferred route of delivery for a
large number of drug molecules. However, the intestinal
epithelium presents a formidable barrier for delivery of
many drugs into systemic circulation. Lipophilic drug
molecules can cross this barrier by diffusion through the
cell membranes, whereas small hydrophilic compounds
diffuse through the intercellular space via the paracellular
route. Some drugs can act as substrates for the carrier
proteins present in the intestinal epithelium (e.g., amino
acid, di/tripeptide, glucose, etc.), and are absorbed by a
carrier-mediated active transport process. For those hy-
drophilic drugs that are likely to traverse the intestinal
epithelium via the paracellular route, the intestinal barrier
* Corresponding author. Tel: (919) 962-0092. Fax: (919) 966-0197.
E-mail: dhiren_thakker@unc.edu.
† Division of Drug Delivery and Disposition, School of Pharmacy,
The University of North Carolina.
‡ Current address: Wyeth-Ayerst Research, 401 North Middletown
Rd., Pearl River, NY 10965.
§ Division of Medicinal Chemistry and Natural Products, School of
Pharmacy, The University of North Carolina.
| Department of Microbiology and Immunology, Wake Forest Uni-
versity.
© 1999, American Chemical Society and
American Pharmaceutical Association
10.1021/js9900957 CCC: $18.00
Published on Web 10/15/1999
Journal of Pharmaceutical Sciences / 1169
Vol. 88, No. 11, November 1999

ability across Caco-2 cells.14 Our studies with DPC14 have
further suggested that the presence of a medium (or long)
alkyl chain and a zwitterionic headgroup are essential for
the activity of phospholipid-like agents as enhancers of
paracellular permeability across Caco-2 cell monolayers.
Similar requirements (i.e., alkyl chain and zwitterionic
headgroup) for enhancement of paracellular permeability
by structurally unrelated acyl carnitines14,15 suggest the
possibility of an emerging structure-activity relationship
for enhancement of paracellular permeability across intes-
tinal epithelium by modulation of tight junctions.
In this paper, we describe a more definitive and detailed
study to delineate the structure-activity relationship for
phospholipid-like agents as enhancers of paracellular
permeability across Caco-2 cell monolayers. An excellent
group of compounds for such a structure-activity relation-
ship study is a series of 2-alkoxy-3-amidopropylphospho-
choline derivatives22 with varying lengths of alkyl chains
at C-2 and C-3 positions. As proposed in our previous
study,14,23 we have applied the concept of potency index in
developing the structure-activity relationship. Embedded
in the potency index are two parameterssone reflecting
the potency of the compounds as modulators of the tight
junctions, i.e., transepithelial electrical resistance (TEER),
and the other reflecting their potency in causing cytotox-
icity by indiscriminate damage to the cell membrane (MTT
test). We report here one of the first attempts to delineate
the relationship between phospholipid structure and en-
hancement of paracellular permeability in relation to the
cytotoxicity of this class of compounds.
Table 1sStructures of 2-Alkoxy-3-alkylamidopropylphosphocholines
group
I
enhancers
R
R
Y
3
2
1
2
C H
CH
N(CH )
3 3
17 35
3
C H
16 33
C H
2
N(CH )
5
3 3
3
4
C H
C H
19 39
C H
N(CH )
3 3
17 35
2
5
C H
N(CH )
2
5
3 3
II
5
6
C H
C H
11 23
C H
N(CH )
3 3
9
19
6
13
C H
N(CH )
8
17
3 3
7
8
C H
C H
C H
10 21
CH N(CH )
11 23
8
17
2
3 3
C H
N(CH )
9
19
3 3
III
9
C H
C H
N(CH )
3 3
17 35
8
17
10
11
12
C H
C H
N(CH )
11 23
10 21
3 3
C H
C H
N(CH )
3 3
9
19
12 25
C H
11 23
C H
12 25
N(CH )
3 3
methyl chloride in pyridine. Reaction of the secondary hydroxyl
with sodium hydride and various alkyl halides gave the C2 alkyl
ether. Deprotection of the primary hydroxyl group with p-tolu-
enesulfonic acid in CH2Cl2/MeOH gave the 3-alkylamido-2-alkoxy-
1-propanol. The phosphocholine was formed in two steps: (1)
reaction with 2-bromoethyl dichlorophosphate and (2) amination
with aqueous trimethylamine in CHCl3/i-PrOH/DMF. Confirma-
tion of chemical structure and determination of purity of the
compounds have been reported.26,27
Materials and Methods
The stock solutions of phospholipids (80 mM) were obtained by
dissolving them in HBSS/ethanol (20:80, v/v) and were stored at
-20 °C. J ust before the experiments, the phospholipid solutions
were thawed and diluted to the appropriate concentrations with
HTB (final ethanol concentrated up to 2% v/v) and sonicated for 5
min in an ice bath.
Ma ter ia lssCaco-2 cells were purchased from American Type
Culture Collection (Rockville, MD). Eagle’s Minimum Essential
Medium (EMEM), 0.25% trypsin/0.02% ethylenediaminetetraacetic
acid-tetrasodium 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, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltet-
razolium bromide (MTT), palmitoyl-DL-carnitine chloride, sodium
dodecyl sulfate (SDS), DPC, 1-myristoyl-sn-glycero-3-phosphate
sodium (MGP), 1-lauroyl-sn-glycero-3-phosphocholine (LGPC),
3-amino-1,2-propanediol, acyl chlorides, triphenylmethyl chloride,
60% sodium hydride (in oil), alkyl halides, p-toluenesulfonic acid
monohydrate, and aqueous (40%) trimethylamine were purchased
from either Sigma (St. Louis, MO) or Aldrich (Milwaukee, WI).
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 (Cambridge, MA). [3H]Mannitol was obtained from DuPont
NEN (Boston, MA).
Ca co-2 Cell Cu ltu r esCaco-2 cells were grown in EMEM
containing 10% FBS, 1% l-glutamine, 1% NEAA and antibiotics
(100 U/mL of penicillin, 100 µg/mL of streptomycin, and 0.25 µg/
mL of amphotericin B) in 75 cm2 culture flasks. The cultures were
kept at 37 °C in an atmosphere 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 of passage number 45-55 were used
in all the studies. Media were changed every 2 days after seeding
until late confluence (20-22 days). J ust before the experiments,
the culture medium was replaced with HBSS transport buffer
(HTB) that contained 1x 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.
Cytotoxicity Assa ysThe cell viability was measured by the
MTT assay.26,27 Approximately 3.0 × 103 Caco-2 cells (in 100 µL
of cell culture medium) were seeded into each of the wells in a
96-well tissue culture plate (Corning-Costar, Cambridge, MA). The
cells were then cultured under the same condition (see Caco-2 Cell
Culture) for 96 h, and 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 compound solution in HTB was
added to each well. After 20 min, 20 µL of a 5 mg/mL MTT solution
was added to each well, and the cells were incubated for another
90 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 compound 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 dehy-
drogenase of the viable cells) using a multiwell scanning spectro-
photometer (Bio-Rad, Hercules, CA). IC50 represented the concen-
tration of the phospholipid derivatives that caused 50% cell death
as measured by the mitochondrial dehydrogenase activity.
EC50 (TEER) Deter m in a tion sTo measure the effect of the
phospholipids on the TEER values of Caco-2 cell monolayers, the
test compounds, dissolved in 0.5 mL of HTB at various concentra-
tions, were applied to the apical side. The monolayers were treated
for 20 min at room temperature and the TEER measured using
an Epithelial Tissue Voltohmmeter (EVOM, World Precision
Instruments, Sarasota, FL) and calculated as Ω‚cm2 by multiplying
it with the surface area of the monolayer (1.0 cm2). The resistance
due to the cell monolayers was determined after subtracting the
contribution of the blank filter and HTB. From the data, EC50 for
TEER, the concentration at which the phospholipid decreases the
TEER of cell monolayer by 50% of the control (untreated) value,
was calculated.
Gen er a l Ch em ica l Meth od ssCompounds 1-12 (Table 1)
were synthesized as described previously.24,25 Briefly, 3-amino-
1,2-propanediol was reacted with various acyl chlorides to give
the corresponding (alkylamido)propanediols. The primary hydroxyl
group was selectively protected as the trityl ether with triphenyl-
Tr a n sp or t Exp er im en tssAll transport studies were per-
formed at room temperature on filter-grown Caco-2 monolayers.
The transport experiments were carried out under sink conditions
such that the concentration of the transported compound in the
1170 / Journal of Pharmaceutical Sciences
Vol. 88, No. 11, November 1999

receiver compartment was less than 10% of its initial concentration
in the donor compartment. Permeability coefficients (Papp) were
calculated using the following equation:
Papp ) (dQ/dt) (1/A) (1/Co)
where dQ/dt (mol transported/s) is the flux of the marker com-
pound across the Caco-2 cell monolayer, A (cm2) represents the
diffusional area of the inserts, and Co (M) denotes the initial
concentration of marker compound in the donor compartment. All
measurements were in triplicate and expressed as mean ( SD
values.
Transport without the Tight J unction BarriersTo evaluate the
maximum transport rate of mannitol without the functional
barrier of epithelial tight junctions, EDTA was used to open up
the tight junction completely.28 A solution of EDTA at concentra-
tion range of 0.001 mM to 15 mM was prepared using HTB without
Ca2+/Mg2+. The solution was applied to the apical and basolateral
sides of cell monolayer for 30 min. Then the solution was removed,
and the cells were washed once with HTB without Ca2+/Mg2+. [3H]-
Mannitol (0.5 mL of 0.1 mM, 0.25 µCi) was added to the apical
compartment, and the transport rate was determined by measur-
ing the radioactivity (PACKARD 1600 TR liquid scintillation
counter, Downers Grove, IL) appearing in the basolateral com-
partment as a function of time.
Absorption EnhancementsTo calculate the compound concen-
tration at which the permeability of mannitol is increased by 10-
fold [EC10× (mannitol)], the compound solution at concentration
range of 0.05 to 5.0 mM in HTB (0.5 mL) was added to the apical
side of the cells. HTB (1.5 mL) was added to the basolateral side.
After treatment for 20 min, the lipid solution was removed and
the monolayer was washed once with fresh HTB. Following the
measurement of the TEER, 0.5 mL of HTB solution containing
0.1 mM [3H]mannitol (0.25 µCi) was applied to the apical side.
Samples from the basolateral side were taken at 10, 20, 40, and
60 min. Transport rates were determined by measuring the
radioactivity in the basolateral side by liquid scintillation counter.
Results
Figure 1sEffect of concentration on TEER and mannitol permeability for
2-alkoxy-3-alkylamidopropylphosphocholines. A, compound 2; B, compound
5; C, compound 9. 9, mannitol permeability; [, TEER. Compounds were
applied to the apical side of the monolayers for 20 min at room temperature.
Previous studies with two structurally distinct com-
pounds, i.e., DPC and PC, have shown that the minimum
structural requirements to cause a decrease in TEER and
a corresponding increase in paracellular permeability
across Caco-2 cells are (i) the presence of a medium to long
alkyl chain and (ii) a zwitterionic group.14,15 The same
structural features may also confer a similar activity to
phospholipid derivatives.7-12,14 Here we report the effect
of varying the length of alkyl chains at C-2 and C-3
positions of phospholipid derivatives on TEER, paracellular
permeability across Caco-2 cell monolayers, and cyto-
toxicity to these cells.
Effect of Syn th etic P h osp h olip id Der iva tives on
TE E R a n d P a r a cellu la r (m a n n it ol) P er m ea b ilit y
a cr oss Ca co-2 CellssThe chemical structures of the
2-alkoxy-3-alkylamidopropylphosphocholines studied are
shown in Table 1. The effect of these compounds on TEER
across Caco-2 cell monolayers was determined by exposing
the cells to various concentrations of these agents on the
apical side for 20 min prior to the measurement of TEER.
The time of exposure was determined based on our studies
with DPC which showed that maximal effect on TEER was
achieved in 20 min.14 Under these conditions, the mannitol
permeability across Caco-2 cell monolayers was also de-
termined. The typical TEER-concentration profiles as well
as mannitol permeability-concentration profiles observed
are shown in Figure 1. Based on the TEER/mannitol
permeability-concentration profiles, the synthetic ether
phospholipids were classified into three groups. Compounds
in group I (1-4, Table 1) caused a precipitous drop in TEER
and a steep rise in mannitol permeability with concentra-
tion (e.g., compound 2, Figure 1A). These compounds are
characterized by a short alkyl chain (1-2 carbons) as the
R2 substituent and a long alkyl chain (16-19 carbons) as
the R3 substituent. In contrast, compounds in group II (5-
8, Table 1) caused a more gradual decrease in TEER and
increase in mannitol permeability with concentration (e.g.,
compound 5, Figure 1B). These compounds contain medium
length alkyl chains as the R2 (6-10 carbons) and R3 (9-
11 carbons) substituents, with the total carbons in both
the alkyl chains < 20. Compounds in group III (9-12, Table
1) had little or no effect on TEER or mannitol permeability
over the concentration range examined (e.g., compound 9,
Figure 1C). These compounds contain a medium length
alkyl chain as the R2 substituent (8-12 carbons) and a
medium to large alkyl chain as the R3 substituent (9-17
carbons) with the total number of carbons in both alkyl
chains > 20.
For comparison, the effect of EDTA, a known modulator
of paracellular permeability via tight junction modulation,28
on TEER and mannitol permeability is shown in Figure 2.
The mannitol permeability reached a plateau at high
concentrations of EDTA; no such plateau was reached with
2-alkoxy-3-alkylamidopropylphosphocholines (Figures 1
and 2). The compounds in groups I and II appear to be more
potent than EDTA in causing a drop in TEER and an
increase in the mannitol permeability across Caco-2 cell
monolayers.
Com p a r ison of th e Effect of Syn th etic P h osp h o-
lip id s on TEER a n d Ma n n itol P er m ea bility a cr oss
Journal of Pharmaceutical Sciences / 1171
Vol. 88, No. 11, November 1999

Figure 3sRelationship between EC (TEER) and EC (mannitol) for
50
10x
Figure 2sEffect of EDTA concentration on TEER and mannitol permeability.
EDTA was applied on both apical and basolateral side of the monolayers for
30 min at room temperature. All measurements were in triplicate and expressed
as mean ± SD.
2-alkoxy-3-alkylamidopropylphosphocholines. EC (TEER) and EC10x (mannitol)
50
have been defined in Materials and Methods.
Table 2. Values of EC , EC10×, IC , and PI of the Synthetic
50
50
Phospholipids
group enhancers EC (TEER)a EC × (mannitol)a IC (MTT)a PIb
50
10
50
I
1
2
3
4
5
6
7
8
9
0.14 (0.02)
0.12 (0.01)
0.16 (0.01)
0.28 (0.02)
0.42 (0.03)
0.25 (0.02)
0.42 (0.03)
0.80 (0.02)
no effect
0.21 (0.02)
0.20 (0.02)
0.29 (0.02)
0.42 (0.01)
0.73 (0.05)
0.70 (0.03)
0.73 (0.05)
1.35 (0.12)
0.16 (0.02) 1.14
0.11 (0.01) 0.83
0.16 (0.02) 1.02
0.25 (0.02) 0.89
1.10 (0.05) 2.62
0.68 (0.02) 2.72
1.10 (0.10) 2.62
1.40 (0.25) 1.75
II
III
c
ND
>2.0
−
10
11
12
1.45 (0.02)
1.6 (0.1)
no effect
ND
ND
ND
1.50 (0.05) 1.03
Figure 4sRelationship between concentration and mitochondrial dehydro-
genase activity (MTT) for 2-alkoxy-3-alkylamidopropylphosphocholines that are
typical of three groups of compounds within this series. All measurements
were in triplicate and expressed as mean ± SD.
2.0 (0.2)
no effect
1.25
−
a Values of EC50 (TEER), EC × (mannitol) and IC50 (MTT) are expressed
10
as mM; the numbers in parentheses indicate SD. b PI: the ratio of IC /EC
50
50
based on the data from Caco-2 cells. c ND: not detectable because of the
insolubility of the lipids at higher concentrations.
compounds were reversible. After removal of the com-
pounds and recovery of the cells in complete cell culture
medium (10% fetal calf serum, 1% NEAA in EMEM) for 2
h, the decreased TEER and the enhanced mannitol trans-
port returned to their initial values (approximately 85-
95% TEER recovery and 90-95% recovery of mannitol
permeability). However, on similar treatment of the cell
monolayers with group I compounds, the TEER and man-
nitol transport values did not recover, indicating that tight
junction integrity cannot be restored after exposure of the
cells to group I compounds.
Cytotoxicity of th e Syn th etic P h osp h olip id ssTo
evaluate cytotoxicity of the synthetic phospholipids toward
Caco-2 cells, the viability of the cells was evaluated using
the MTT method.26,27 MTT is a tetrazolium salt that is
cleaved by mitochondrial dehydrogenase in living cells to
give a dark blue formazan product. Damaged or dead cells
show reduced or no mitochondrial dehydrogenase activity.
Cytotoxicity of compounds 2, 5, and 9, representing the
compounds in groups I, II, and III, respectively, as a
function of their concentration is shown in Figure 4. The
toxicity of these compounds is expressed as IC50, which is
defined as the concentration at which the compounds
caused 50% decrease in mitochondrial dehydrogenase
activity in the MTT test. The IC50 values of all the
compounds tested are given in Table 2. Among the three
compounds shown in Figure 4, compound 2 was most
cytotoxic (IC50 ) 0.11 mM), compound 5 showed moderate
cytotoxicity (IC50 ) 1.1 mM), and compound 9 did not show
any toxicity even at the highest concentration (2.5 mM)
tested. The cytotoxicity of compounds 2, 5, and 9 was
representative of the cytotoxicity exhibited by compounds
in their respective groups.
Ca co-2 CellssTo express the effect of these compounds
on TEER and paracellular permeability quantitatively, we
have defined the terms EC50 (TEER)14 and EC10× (manni-
tol).22 EC50 (TEER) is the concentration of an enhancer at
which TEER drops to 50% of its control value (untreated
cells), and EC10× (mannitol) is the concentration at which
mannitol permeability is increased by 10-fold over the
control value (untreated cells). In case of mannitol perme-
ability, the term EC10× was defined instead of EC50 because
the plateau in the mannitol permeability was not always
easy to obtain (see Figure 1); hence, a percentage of
maximal effect could not always be determined. Further-
more, 10-fold increase in mannitol permeability over the
control value (∼2.5 × 10-7 cm/s) represented 50% of the
maximal mannitol permeability (25 × 10-7 cm/s) obtained
by opening up the tight junctions with EDTA (Figure 2).
The EC50 (TEER) and EC10× (mannitol) for the synthetic
phospholipids are shown in Table 2. By both these param-
eters, the compounds in group I were the most potent
phospholipid derivatives tested, the compounds in group
II showed moderate potency, and those in group III were
weakly active or inactive.
The relationship between the EC50 (TEER) and the EC10x
(mannitol) is shown in Figure 3. Clearly, an excellent
correlation exists between the EC10x (mannitol) values and
the EC50 (TEER) values. Hence, EC50 (TEER) can serve as
a reliable index of the potency of the compounds as
enhancers of paracellular permeability if the data on
mannitol permeability are not available.
Interestingly, at their EC50 concentrations, the changes
in TEER and mannitol transport caused by group II
1172 / Journal of Pharmaceutical Sciences
Vol. 88, No. 11, November 1999

paracellular permeability is very similar to the concentra-
tion at which they are cytotoxic. In contrast, the compounds
in group II, with medium length alkyl chains at C-2 and
C-3 (total carbons < 20), have PI values in the range of
1.75-2.72 and achieve the best separation of activity as
paracellular permeation enhancer and cytotoxicity among
all the compounds tested including those tested in our
previous study.14 This PI range is significantly greater than
that for the extensively investigated enhancer PC (PI =
1). Compounds 6 and 7 exhibit comparable PI values
despite the fact that their cationic and anionic centers are
separated by two and three carbons, respectively. Thus a
small change in the distance between the cationic and the
anionic centers does not lead to any change in activity or
toxicity. Some compounds in group III (medium to long
alkyl chains at C-2 and C-3 and total carbons > 20) have
PI values of approximately 1, whereas others are too
inactive to determine either the EC50 or the IC50 value.
In conclusion, with this limited series of compounds, we
have been able to demonstrate that the structure-activity
(modulation of paracellular permeability) relationship for
synthetic phospholipid derivatives is distinct from the
structure-toxicity relationship. Interestingly, the separa-
tion of activity and toxicity is most pronounced (>2.5-fold)
for moderately active compounds, while the activity and
the toxicity appear to converge among the more potent
compounds. Hence, further assessment of the structure-
activity relationships should focus on phospholipid deriva-
tives with medium-size alkyl substituents at C-2 and C-3.
The relationship between relative position and distance of
the anionic and cationic centers and the activity/toxicity
should be further explored.
Figure 5sRelationship between R2 and R3 substituents in 2-alkoxy-3-
alkylamidopropylphosphocholines and the potency index, PI [IC /EC (TEER)].
50
50
The compounds are in numerical order (1 to 12) from left to right. *The
trimethylammonium group is separated from the phosphate group by a three-
carbon chain instead of a two-carbon chain
Discussion
A decrease in TEER and an increase in mannitol
permeability across Caco-2 cell monolayers caused by
2-alkoxy-3-alkylamidopropylphosphocholines suggest that
this class of compounds has the potential to act as absorp-
tion enhancers. Their effect on TEER and mannitol perme-
ability suggests that they are acting via modulation of the
tight junctions; however, the evidence is entirely circum-
stantial. In a previous study,14 we had presented evidence
that a long alkyl chain and a zwitterionic functionality are
essential for the activity of such compounds as modulators
of paracellular permeability. In the present study, we have
examined a limited structure-activity relationship (2-
alkoxy-3-alkylamidopropylphosphocholines) that provides
more insights on the effect of branching of the alkyl chain
on the activity (as enhancers of paracellular permeability)
and toxicity of phospholipid derivatives. The branching is
not achieved with branched alkyl chains, but rather by
introducing an ether functionality at C-2 and an amido
functionality at C-3 of a propylphosphocholine.
Surprisingly, introduction of a small alkyl group at the
C-2 position (R2) significantly enhanced the potency of
phospholipids in causing a decrease in TEER and an
increase in mannitol permeability across Caco-2 cell mono-
layers. For example, the EC50 (TEER) value of compound
2 (0.12 mM) was 3.5- to 5-fold lower than the previously
reported values for PC (0.42 mM), DPC (0.65 mM), or
LGPC (0.65 mM). Unfortunately, this structural feature
also caused the phosphocholine derivatives to be more
potent cytotoxic agents. Thus the IC50 value for compounds
2 (0.11 mM) is much lower than that for PC (0.45 mM),
DPC (0.92 mM), and LGPC (0.84 mM).14 Increasing the
length of C-2 alkyl chain to 6-8 carbon atoms while
simultaneously decreasing the length of C-3 alkyl chain
caused a decrease in the potency of the compounds as
modulators of the paracellular permeability with a concur-
rent decrease in cytotoxicity. Finally, further increase in
the length of the C-2 or C-3 alkyl chains rendered the
compound inactive.
References and Notes
1. Van Hoogdalem, E. J ., De Boer, A. G., Breimer, D. D.
Intestinal drug absorption enhancement: an overview. Phar-
macol. Ther. 1989, 44, 407-443.
2. Lutz, K. L.; Siahaan, T. J . Molecular structure of the apical
junction complex and its contribution to the paracellular
barrier. J . Pharm. Sci. 1997, 86, 977-984.
3. Lee, V. H. L.; Yamamoto, A.; Kompella, U. B. Mucosal
penetration enhancers for facilitation of peptide and protein
drug absorption. Crit. Rev. Ther. Drug Carrier Syst. 1991, 8
(2), 91-192.
4. LeCluyse, E. L.; Sutton, S. C. In vitro models for selection of
development candidates. Permeability studies to define
mechanisms of absorption enhancement. Adv. Drug Delivery
Rev. 1997, 23, 163-183.
5. Swenson, E. C.; Curatolo, W. J . Intestinal permeability
enhancement for proteins, peptides and other polar drugs:
Mechanisms and potential toxicity. Adv. Drug Delivery Rev.
1992, 8 (1), 39-92.
6. Van Hoogdalem, E. J .; Wackwitz, A. T. E.; De Boer, A. G.;
Breimer, D. D. Topical effects of absorption enhancing agents
on the rectal mucosa of rats in vivo. J . Pharm. Sci. 1990, 79
(10), 866-70.
7. Weltzien, H. U. Cytolytic and membrane-perturbing proper-
ties of lysophosphatidylcholine. Biochim. Biophys. Acta 1979,
559, 259-287.
8. Drejer, K.; Vaag, A.; Bech, K.; Hansen, P.; Sørensen, A. R.;
Mygind, N. Intranasal administration of insulin with phos-
pholipids as absorption enhancer: pharmacokinetics in
normal subjects. Diabetic Med. 1992, 9, 335-340.
9. J acobs, M. A.; Schreuder, R. H.; J ap-A-J oe, K.; Nauta, J . J .;
Andersen, P. M.; Heine, R. J . The pharmacodynamics and
activity of intranasally administered insulin in healthy male
volunteers. Diabetes 1993, 42, 1649-1655.
10. Fisher, A. N.; Farraj, N. F.; O′Hagan, D. T.; J abbal-Gill, I.;
J ohnsen, B. R.; Davis, S. S.; Illum, L. Effect of L-R-lysophos-
phatidylcholine on the nasal absorption of human growth
hormone in three animal species. Int. J . Pharm. 1991, 74,
147-156.
In developing a structure-activity relationship for para-
cellular permeation enhancers, the potency of these agents
must be addressed in relation to their cytotoxicity. Hence,
we have defined the term potency index (PI), which is a
ratio of the IC50 (MTT) value of a compound to its EC50
(TEER) value.14,22 For the synthetic phospholipid deriva-
tives, the PI values are listed in Table 2 and plotted in
Figure 5. Clearly, for all compounds having a short alkyl
chain at C-2 (group I), the PI value is ∼1; i.e., the
concentration at which they can exhibit enhancement in
11. Brondsted, H.; Nielsen, H. M.; Hovgaard, L. Drug delivery
studies in Caco-2 monolayers. III. Intestinal transport of
various vasopressin analogues in the presence of lysophos-
phatidylcholine. Intl. J . Pharm. 1995, 114, 115-157.
Journal of Pharmaceutical Sciences / 1173
Vol. 88, No. 11, November 1999

12. Hovgaard, L.; Bronsted, H.; Nielsen, H. M. Drug delivery
studies in Caco-2 monolayers. II. Absorption enhancer effects
of lysophosphatidylcholine. Int. J . Pharm. 1995, 114, 141-
149.
22. Liu, D. Z.; Morris-Natschke, S. L. Ishaq, K. S.; Thakker, D.
R. Structural requirements for synthetic phospholipids that
modulate tight junction permeability across Caco-2 cells.
PharmSci 1998, 1 (1), S-453.
13. Schulze, C.; Smales, C.; Rubin, L. L.; Staddon, J . M. Lyso-
phosphatidic acid increases tight junction permeability in
cultured brain endothelial cells. J . Neurochem. 1997, 68,
991-1000.
23. Liu, D. Z.; Thakker, D. R. Modulation of Caco-2 cell tight
junctions by dodecanoylphosphorylcholine (DPC). Pharm.
Res. 1997, 14, S-23.
24. Meyer, K. L.; Marasco, C. J .; Morris-Natschke, S. L.; Ishaq,
K, S.; Piantadosi, C.; Kucera, L. S. In vitro evaluation of
phosphocholine and quaternary ammonium containing lipids
as novel anti-HIV agents. J . Med. Chem. 1991, 34 (4), 1377-
83.
25. Surles, J . R.; Morris-Natschke, S.; Marx, M. H.; Piantadosi,
C. Multigram synthesis of 1-alkylamido phospholipids. Lipids
1993, 28 (1), 55-57.
26. Mosmann, T. Rapid colorimetric assay for cellular growth
and survival: application to proliferation and cytotoxicity
assays. J . Immunol. Methods 1983, 65, 55-63.
27. Andenberg, E. K.; Artursson, P. Epithelial transport of drugs
in cell culture: VIII. Effects of the pharmaceutical surfactant
sodium dodecyl sulfate on cell membrane and tight junctional
permeability in human intestinal epithelial (caco-2) cells. J .
Pharm. Sci. 1993, 82, 392-398.
28. Artursson, P.; Magnusson, C. Epithelial transport of drugs
in cell culture. II. Effect of extracellular calcium concentra-
tion on the paracellular transport of drugs of different
lipophilicities across monolayers of intestinal epithelial
(Caco-2) cells. J . Pharm. Sci. 1990, 79, 595-600.
14. Liu, D.-Z.; LeCluyse, E. L.; Thakker, D. R. Dodecylphospho-
choline-mediated enhancement of paracellular permeability
and cytotoxicity in Caco-2 cell monolayers. J . Pharm. Sci.
1999, 88, 1161-1168.
15. Hochman, J . H.; Fix, J .; LeCluyse, E. L. In vitro and in vivo
analysis of the mechanism of absorption enhancement by
palmitoyl carnitine. J . Pharmacol. Exp. Ther. 1994, 269 (2),
813-822.
16. Shimazaki, T.; Tomita, M.; Sadahiro, S.; Hayashi, M.; Awazu,
S. Absorption-enhancing effects of sodium caprate and
palmitoyl carnitine in rat and human colons. Dig. Dis. Sci.
1998, 43, 641-645.
17. Knipp, G. T.; Ho, N. F.; Barsuhn, C. L.; Borchardt, R. T.
Paracellular diffusion in Caco-2 cell monolayers: effect of
perturbation on the transport of hydrophilic compounds that
vary in charge and size. J . Pharm. Sci. 1997, 86 (10), 1105-
10.
18. Anderberg, E. K.; Lindmark, T.; Artursson, P. Sodium
caprate elicits dilatations in human intestinal tight junctions
and enhances drug absorption by paracellular route. Pharm.
Res. 1993, 10, 857-864.
19. Tomita, M.; Hayashi, M.; Awazu, S. Absorption-enhancing
mechanism of EDTA, caprate and decanoylcarnitine in
Caco-2 cells. J . Pharm. Sci. 1996, 85 (6), 608-611.
20. Lindmark, T.; Kimura, Y.; Artursson, P. Absorption enhance-
ment through intracellular regulation of tight junction
permeability by medium chain fatty acids in Caco-2 cells. J .
Pharmacol. Exp. Ther. 1998, 284, 362-369.
21. Gan, L.-S.; Thakker, D. R. Applications of the Caco-2 model
in the design and development of orally active drugs:
Elucidation of biological and physical barriers posed by the
intestinal epithelium. Adv. Drug Delivery Rev. 1997, 23, 80-
100.
Acknowledgments
The authors gratefully acknowledge Mr. Tim Tippin (Glaxo
Wellcome and UNC) for a careful review of the manuscript and
suggestions.
J S9900957
1174 / Journal of Pharmaceutical Sciences
Vol. 88, No. 11, November 1999