ARTICLE IN PRESS

Il Farmaco "" (2005) """-"""

http://france.elsevier.com/direct/FARMAC/

Original article

Absorption enhancement, mechanistic and toxicity studies

of medium chain fatty acids, cyclodextrins and bile salts
as peroral absorption enhancers
Pradeep Sharma, Manthena V.S. Varma, Harmander P.S. Chawla, Ramesh Panchagnula *
Department of Pharmaceutics, National Institute of Pharmaceutical Education and Research (NIPER), Sector No. 67, SAS Nagar 160 062, Punjab, India
Received 19 July 2005; revised and accepted 19 August 2005

Dedicated to Professor A.T. Florence, Dean, The School of Pharmacy, London University, on the eve of his 65th birthday and retirement

Abstract

The objective of the present investigation was to evaluate an oral ‘drug delivery’ approach, which involves co-administration of absorption
enhancers (AEs). The representative low permeable hydrophilic (biopharmaceutic classification system (BCS) Class III) drugs used in the
study comprised of cefotaxime sodium and ceftazidime pentahydrate, whereas low permeable lipophilic (BCS Class IV) drugs include
cyclosporin A and lovastatin. AEs from three different chemical classes, namely, medium chain fatty acids (sodium caprylate and caprate),
cyclodextrins (b-cyclodextrin, hydroxypropyl b-cyclodextrin) and bile salts (sodium cholate and deoxycholate) were evaluated for absorption
enhancement efficacy, mechanism of action and toxicity using in vitro everted intestinal sac model. These AEs were found to enhance intes-
tinal permeability of drugs from 2- to 27-fold. Light microscopy studies of intestinal sac incubated with AEs for 120 min revealed morpho-
logical changes in absorptive mucosa and rank order of toxicity were cyclodextrins > bile salts ≅ medium chain fatty acids. Fluorescence
polarization studies indicated that brush bordered membrane vesicles labeled with lipophilic (DPH, 12AS) and hydrophilic dyes (ANS), when
treated with AEs exhibited concentration and time dependent decrease in fluorescence polarization. Total protein released in presence of AEs
was more than control but considerably less than EDTA (0.58% w/v), which is known to cause toxic release of proteins from cell. Overall, AEs
were found to significantly enhance drug permeability by decreasing lipid membrane fluidity and/or interacting with hydrophilic domains of
membrane, and has the potential to improve oral delivery.
© 2005 Elsevier SAS. All rights reserved.

Keywords: Peroral absorption enhancers; BCS Class III and IV drugs; Fluorescence polarization

1. Introduction newly emerging diverse compounds for regulatory authori-

Complex molecules like oligonucleotides, vaccines, chi-
meric proteins and simple peptides exhibit different absorp-
tion pattern than conventional synthetic small molecular
weight organic molecules. However, in all cases it is possible
to correlate solubility as well as permeability of drug mol-
ecule with its absorption profile. To rationalize science of drug
delivery and simplify complications in drug registration of
* Corresponding author. Present address: School of Biomedical Scien-
ces, University of Ulster, Cromore Road, Coleraine BT52 1SA, UK. Tel.:
+44 28 7032 4128; fax: +44 28 7032 4965.
E-mail address: panchagnula@yahoo.com (R. Panchagnula).
0014-827X/$ - see front matter © 2005 Elsevier SAS. All rights reserved.
doi:10.1016/j.farmac.2005.08.008
ties, all drugs are now classified according to biopharmaceu-
tic classification system (BCS) [1,2] into four categories on
the basis of solubility and permeability. Among the various
classes of BCS, peroral delivery of Class III and IV drugs is
partially or completely hindered due to their poor intestinal
permeability. Many drug molecules show poor permeability
because of their unfavorable physicochemical and chemical
features, which are difficult to change, thus an external excipi-
ent may be added to increase permeation transiently [3–5].
The compounds that reversibly increase the permeation of
drugs across biomembranes are known as absorption enhanc-
ers (AEs). Various surfactants, fatty acids and alkaloids have
been reported to be effective permeation enhancers (AE) for
FARMAC-2440
 ARTICLE IN PRESS
2 P. Sharma et al. / Il Farmaco "" (2005) """-"""
these classes of drugs [6]. Therefore, development of safer
and effective permeation enhancer is an active area of research
and will provide impetus to peroral pharmaceutical technol-
ogy.
In present study, permeation enhancement of representa-
tive low permeable hydrophilic model drugs of BCS Class III
(cefotaxime sodium, CX and ceftazidime pentahydrate, CZ)
and low permeable lipophilic model drugs of BCS Class IV
(cyclosporin A (CSA) and lovastatin (LV)) was investigated
in vitro in everted intestinal sac model by the use of AEs. We
attempt to propose mechanisms of action of selected AEs by
employing time and concentration dependent fluorescence
polarization studies of brush bordered membrane vesicles
(BBMVs) prepared from rat intestine. Local acute toxicity of
AEs on intestinal epithelia was also evaluated on the basis of
total protein release as intracellular biochemical marker and
by light microscopy of intestinal mucosa exposed to AEs.
2. Materials and methods
2.1. Chemicals
Lupin Laboratories Ltd. (India) generously supplied cefo-
taxime sodium (CX) and ceftazidime pentahydrate (CZ) as
gift sample. CSA and LV were provided by Dabur Research
Foundation (India) and Dr. Reddy’s Research Laboratories
Ltd. (India), respectively. Sodium caprylate (sodium oc-
tanoate), sodium caprate (sodium decanoate), b-cyclodextrin
(b-CD), were purchased from Sigma Chemicals (USA). Fluo-
rescent probes 12-AS (12-(9anthroxyloxy) stearic acid), 1,6–
diphenyl-1,3,5-hexatriene (DPH) and 8-anilino-1-napthalene
sulfonate (ANS) were purchased from Sigma Chemicals
(USA), Aldrich Chem Co. (Germany) and Fluka Biochemika
(Germany), respectively. Sodium cholate and sodium deoxy-
cholate were purchased from Acros Organics (USA). Hydrox-
ypropyl b-cyclodextrin (HP-b-CD) was kindly supplied by
Lupin Laboratories Ltd. (India) and Wacker Cyclodextrins
(Germany), respectively. All other chemicals were of analyti-
cal grade purchased from Hi-Media (India) and Ranbaxy Labs
Ltd (India). HPLC grade solvents from J.T. Baker (USA) were
used for drug analysis.
2.2. In vitro everted rat intestinal sac absorption
All animal studies were done according to the guidelines
of the Institutional Animal Ethical Care Committee of
National Institute of Pharmaceutical Education and Research
(NIPER), Punjab. In vitro permeation studies were per-
formed using everted rat jejunal segments [7] prepared from
male Sprague–Dawley rats (250–350 g). They were fasted
for 16 h and water was allowed ad libitum. After sacrificing
animals by excessive ether inhalation, intestine was immedi-
ately removed from the rat and placed in beaker with ice-cold
KRB solution, which was continuously aerated. The jejunal
segments (approximately 6 cm in length) were selected and
everted with help of glass rod. The distal end of everted seg-
ment was tied and attached to a 2 g weight and the proximal
end was attached to the cannula. It was suspended in 40 ml of
the drug containing KRB solution so as to immerse the sac
completely in KRB and 3 ml of KRB was placed into the
serosal compartment. Samples (1.6 ml) were collected from
serosal compartment at the end of each time interval and equal
volume of KRB was replaced and was analyzed by RP-HPLC
methods [8,9]. Cumulative amount of drug permeated through
the sac per unit area (µg cm–2) was plotted against time (min).
Slope of linear portion in the graph was taken as permeation
flux (Jw, µg cm–2 min–1) and apparent permeability coeffi-
cient (Papp) was calculated using Eq. (1).
Jw −1
Papp = cm min (1)
IDC
Where, IDC = Initial drug concentration in mucosal solu-
tion (µg ml–1).
2.3. Light microscopy
Rat jejunal segments were incubated with AEs and intes-
tinal mucosa was exposed to them in vitro in rat intestinal sac
for 120 min. Treated Segments were fixed in formal saline
(10% v/v formaldehyde in physiological saline) for 12 h. They
were serially dehydrated by increasing concentrations of etha-
nol, embedded in paraffin, cut and stained with hematoxylin
and eosin. All slides were observed under 400 and 1000× in
Leica Microscope (Leica MPS 48 with Leica MPS 52 Cam-
era). Quantitative parameters (Nucleo-apical distance, NPA
and Villi index, Vi) as well as semi-quantitative parameters
(morphological scoring) were determined as described by [10]
and employed to interpret acute local toxic profile of AEs on
intestinal tissue.
2.4. Preparation of BBMVs
Mucosal scrapings of the proximal half of the small intes-
tine of 6–12 rats were pooled, homogenized, and treated with
10 mM of CaCl2. After homogenization at high speed, the
microvillus membranes were isolated by differential centrifu-
gation and suspended in 13 mM Tris buffer of pH 7.4 [11].
BBMVs were characterized by determination of their total
protein content and enrichment of specific enzyme (alkaline
phosphatase) [12]. The protein concentration of vesicles was
determined by colorimetric method considering bovine serum
albumin as standard [13]. For further florescence polariza-
tion mechanistic studies, the quantity of BBMVs was ex-
pressed in terms of amount equivalent to total protein. BBMVs
are considered to be of high purity, suitable for biochemical
work and mechanistic studies, if specific activity of alkaline
phosphatase is about 10–15 times more in BBMVs when com-
pared to mucosal homogenates [12]. In present work, enzy-
matic conversion of p-nitrophenyl phosphate into
p-nitrophenol by alkaline phosphatase was used for its spec-
 ARTICLE IN PRESS
P. Sharma et al. / Il Farmaco "" (2005) """-""" 3

trophotometric determination, as described by using UV/VIS 2.6. Total protein release

Spectrophotometer (Perkin–Elmer Lambda 20) [14]. The
enrichment factor of alkaline phosphatase was 14.5 in pre-
pared BBMVs.
DPH and 12-AS were dissolved in tetrahydrofuran at con-
centrations of 3.2 and 1 mM, respectively. ANS was dis-
solved in isotonic phosphate buffer (pH 7.4) consisting of
120 mM NaCl, 3 mM K2HPO4, 1 mM MgCl2 and 10 mM
Tris (hydroxymethyl) aminomethane to obtain concentration
of 1 mM [11]. An aliquot of this probe solution (200 µl) was
added to 50 ml of BBMV solution equivalent to 80 mg of
total protein content. The suspension was then incubated for
30 min at 37 °C. Labeled vesicles were separated from unla-
beled vesicles by washing the preparation with buffer and
then centrifuging the suspension at 27,000 × g for 30 min.
This procedure was repeated for three times at 4 °C and resi-
due was suspended in 10 ml of phosphate buffer solution (pH
7.4). To 1 ml of this solution, 3.5-ml phosphate buffer (pH
7.4) and 0.5 ml of AE solution were added. An aliquot of
200 µl of this solution was diluted to 3.0 ml and readings
were taken in luminescence spectrometer (Perkin–Elmer
LS50B). The excitation light for DPH, 12-AS and ANS were
380, 390 and 380 nm, while emission light were 455, 460 and
480 nm, respectively. Furthermore, a cut off filter of 430 nm
was used for emission light in case of all three fluorescent
probes.
Rat intestinal mucosa was exposed to AEs in everted sac
absorption model and extent of total protein released was
determined by withdrawing aliquots from mucosal solution
and analyzing by Lowry’s method [13] considering bovine
serum albumin as standard. The cumulative amount of total
protein released in presence of AE divided by surface area of
sac was taken as total protein release index of that AE in mg
cm–2. EDTA (0.58% w/v) is known to cause high release of
proteins and is taken as positive control to compare total pro-
tein release indices of other AEs [15].
2.7. Statistical analysis
All experiments were done in triplicate unless specified
and the values are expressed as mean ± S.D. Data were sub-
jected to statistical analysis by Student’s t-test, One way
ANOVA test or Mann–Whitney rank sum test (non-
parametric), wherever applicable, at a significance level of
P < 0.05 using Sigmastat® (Jandel Scientific, USA).
3. Results
3.1. In vitro everted rat intestinal sac absorption studies

As shown in Table 1, intestinal permeability of CX, CZ,

2.5. Concentration and time dependent fluorescence
polarization studies
BBMVs were treated with three different concentrations
of AE, denoted as lower, medium and higher concentrations
and they were (in % w/v); sodium caprylate (0.063, 0.125,
0.25), sodium caprate (0.063, 0.125, 0.25), b-CD (0.45, 0.9,
1.8), HP b-CD (0.45, 0.9, 1.8), sodium deoxycholate (0.25,
0.5, 1.0) and sodium cholate (0.25, 0.50, 1.0). Probe labeled
BBMVs were treated with higher concentrations of AEs and
fluorescence polarization measurements were taken at 37 °C
after 1, 15, 30, 45 and 60 min to observe time dependent effect
of AEs.
CSA and LVs were enhanced to varied extent by different
categories of AEs. Furthermore, intestinal permeability of
hydrophilic drugs was significantly (P < 0.05) enhanced by
AEs from 2- to 15-fold and permeation of lipophilic drugs
was increased (P < 0.05) from 2- to 27-fold.
3.2. Light microscopy
Acute local toxicity of AEs on intestinal mucosa was evalu-
ated with the help NPA, Vi and morphological scoring. NPA
is the distance between the nucleus and the apical membrane
of enterocytes determined using an eyepiece micrometer [10].
It was always calculated as mean of five measurements in

Table 1
Increase in intestinal permeability of Class III and IV drugs by AEs in vitro during everted intestinal sac studies
Treatmenta CX CZ CSA LV
Papp × 107 (cm min–1) AERb Papp × 107 (cm min–1) AER Papp × 107 (cm min–1) AER Papp × 107 (cm min–1) AER
Control 1.69 (0.05) ---- 1.46 (0.10) ---- 9.62 (0.04) ---- 69.11 (0.35) ----
AE1 7.43 (0.41) 4.40 4.37 (0.35) 2.99 86.30 (1.28) 8.97 270.15 (2.60) 3.91
AE2 7.94 (0.22) 4.70 2.60 (0.04) 1.78 105.78 (0.85) 11.004 510.45 (2.02) 7.39
AE3 7.15 (0.76) 4.23 3.39 (0.57) 2.32 254.77 (1.97) 26.48 1463.61 (3.21) 21.18
AE4 3.23 (0.13) 1.91 3.04 (0.21) 2.08 262.47 (3.03) 27.28 1726.10 (4.71) 24.98
AE5 25.70 (1.26) 15.21 14.10 (0.42) 9.66 22.87 (0.18) 2.38 255.02 (0.59) 3.70
AE6 4.93 (0.21) 2.92 2.87 (0.21) 1.97 20.36 (0.06) 2.12 118.71 (0.55) 1.72
1. N = 3, S.D. is indicated in parentheses. 2. Drug solutions used were in concentration, CX: 500 µM, CZ: 500 µM, CSA: 4.99 µM, LV: 0.87 µM.
a
Concentrations of various AEs (% w/v); AE1: sodium caprylate [0.25], AE2: sodium caprate [0.25], AE3: b-CD [1.8], AE4: hydroxypropyl b-CD [1.8],
AE5: sodium deoxycholate [1.0], AE6: sodium cholate [1.0].
b
AER = absorption enhancement ratio was defined as ratio of Papp of drug in presence of AE to that of Papp in absence of AE.
 ARTICLE IN PRESS
4 P. Sharma et al. / Il Farmaco "" (2005) """-"""
Fig. 1. Morphological changes in intestinal mucosa of rat when incubated
with AEs in vitro. The relative damage to intestinal mucosa was assessed by
(A) nucleo-apical distance (B) Villi index and (C) morphological scoring
(N = 3, values are shown as mean ± S.D.). AE1: sodium caprylate [0.25%
w/v], AE2: sodium caprate [0.25% w/v], AE3: b-CD [1.8% w/v], AE4:
hydroxypropyl b-CD [1.8% w/v], AE5: sodium deoxycholate [1.0% w/v],
AE6: sodium cholate [1.0% w/v].
each villus and three villi from different intestinal sections
were measured. There was decrease in NPA of enterocytes
on treatment with AEs (Fig. 1A) and bile salts exhibited maxi-
mum decrease in NPA. Furthermore, Vi was determined by
measurement of height and width (at one half of height of
villus) of enterocytes. Further, measurement of five villi was
done in each intestinal section and three such sections were
counted. Vi was then calculated using following formula:
Height
Vi = (2)
Width
Vi of intestinal tissue was found to decrease in presence of
AEs (Fig. 1B). Alterations in architecture of intestinal mucosa
produced by AEs were also investigated by relative scoring
method [16,17]. To facilitate understanding of such morpho-
Table 2
Morphological scoring to assess intestinal tissue damage on exposure to AEs
Grade of Description of the damageb
damagea Epithelium Villus
0 No damage No damage
1 Single cells extruded Slight oedema
2 Groups of cells extruded Villus height decreased
and/or thin epithelium and/or several villi show
oedema
3 Sheets of cells extruded Villus height greatly redu-
and/or thin epithelium ced and/or severe oedema
4 Total loss of epithelium Villus structure flattened
and/or severe oedema
a
The grading system followed was as described by Fix et al. [16] and
Henrikson et al. [17].
b
Although, grade of ‘0’ is included in scale, control samples were also
found to show some level of damage higher than ‘0’ due to mechanical han-
dling of tissue.
logical changes at different hierarchical levels of tissue orga-
nization, observations were made systematically in follow-
ing successive steps:
• Damage to epithelial monolayer only.
• Gross damage to entire villus.
A scale from ‘0–4’ was used, where 0 indicates normal
undamaged segment and ‘4’ a severely damaged segment
(Table 2). In general, rank order of morphological changes
induced by AEs on jejunal mucosa (Fig. 1C) displayed fol-
lowing trend:
cyclodextrins>bile salts ≅ medium chain fatty acids
Photomicrographs of intestinal mucosa exposed to differ-
ent AEs are shown in Fig. 2. Since the observations were based
on scoring or ranking non-parametric test – Mann–Whitney
rank order test was applied to test the significance of differ-
ence between control and treatment. There was no statisti-
cally significant difference (P > 0.05) when compared with
morphological score of control as verified by Mann–Whit-
ney rank order test.
3.3. Concentration and time dependent fluorescence
polarization studies
The decrease in fluorescence polarization of BBMVs due
to treatment of AEs, represented as DP was calculated as dif-
ference between fluorescence polarization of control and AE
treated BBMVs. Exposure of BBMVs labeled with DPH,
12-AS and ANS to AEs resulted in increase of DP (Figs. 3
and 4). The extent of decrease (DP) and its reversal was depen-
dent upon the nature of AE, concentration and time.
3.4. Total protein release
Total protein released by AEs from intestinal mucosa was
found to be higher than control, but significantly (P > 0.05)
lower than positive control, i.e. EDTA (Fig. 5).
4. Discussion
CX and CZ are water-soluble and their peroral absorption
is mainly limited due to low intestinal permeability. How-
 ARTICLE IN PRESS
P. Sharma et al. / Il Farmaco "" (2005) """-""" 5

Fig. 2. Photomicrographs (×400) of jejeunal segment of rat from in vitro everted intestinal sac model after incubation for 120 min in Kreb’s ringer bicarbonate
solution. (A) Control (B) 0.25% w/v sodium caprylate (C) 0.25% w/v sodium caprate (D) 1.8% w/v b-CD (E) 1.8% w/v hydroxypropyl b-CD (F) 1.0% w/v
sodium deoxycholate (G) 1.0% w/v sodium cholate. Black arrowhead indicates cell sloughing at villi tips and white arrowhead show oedema of villi.

ever, in case of CSA and LV poor absorption is not only due
to low intestinal permeability but also because of poor solu-
bility. Since, present investigation was aimed specifically
towards permeability enhancement of Class III and IV drugs
without solubility considerations, aqueous solutions of Class
IV model drugs (below their saturation solubility) were used
in the permeation experiments. Hence, absorption across
everted sac was no more restricted by dissolution rate and
permeability becomes rate-limiting barrier to absorption.
Results of in vitro permeation studies, summarized in
Table 1, indicate that the intestinal permeability of selected
model drugs lie in range of 1.46–69.11 × 10–7 cm min–1.
Medium chain fatty acids significantly (P < 0.05) enhanced
permeation of hydrophilic (CZ and CX) and lipophilic (LV
and CSA) drugs up to 5- and 11-fold, respectively. Further-
more, sodium caprate was more effective than sodium capry-
late to enhance permeability of drugs except in case of CZ
and such a difference in efficacy may be due to the difference
in mechanism of action of two AEs. In case of CDs, perme-
ability of hydrophilic and lipophilic drugs was significantly
(P < 0.05) enhanced up to 4- and 27-fold, respectively. Pre-
sumably, such a preference for permeation enhancement of
lipophilic drugs compared to hydrophilic drug was due to pre-
dominant action of CDs on transcellular mechanisms of
absorption through which lipophilic drugs are absorbed. At
the same time, comparison of permeation enhancement effi-
cacies of b-CD and HP-b-CD within drug classes show that
b-CD enhanced permeability of hydrophilic drugs more effec-
tively than HP-b-CD, whereas, HP-b-CD was more effective
than b-CD for lipophilic drugs. Although, on the basis of avail-
 ARTICLE IN PRESS
6 P. Sharma et al. / Il Farmaco "" (2005) """-"""

Fig. 4. Time dependent effect of AEs on fluorescence polarization of BBMVs

labeled with (A) DPH (B) 12-AS and (C) ANS (N = 3, mean ± S.D.). AE1:
sodium caprylate [0.25% w/v], AE2: sodium caprate [0.25% w/v], AE3: b-CD
[1.8% w/v], AE4: hydroxypropyl b-CD [1.8% w/v], AE5: sodium deoxy-
Fig. 3. Concentration dependent effect of AEs on fluorescence polarization cholate [1.0% w/v], AE6: sodium cholate [1.0% w/v].
of BBMVs labeled with (A) DPH (B) 12-AS and (C) ANS (N = 3,
mean ± S.D.). LC: lower concentration, MC: medium concentration and HC:
higher concentration. The actual concentrations of AEs are shown in paren-
theses, AE1: sodium caprylate [0.063, 0.125, 0.25% w/v], AE2: sodium
caprate [0.063, 0.125, 0.25% w/v], AE3: b-CD [0.45, 0.9, 1.8% w/v], AE4:
hydroxypropyl b-CD [0.45, 0.9, 1.8% w/v], AE5: sodium deoxycholate [0.25,
0.5, 1.0% w/v], AE6: sodium cholate [0.25, 0.5, 1.0% w/v].

able results, it is difficult to explain this trend, such type of

drug specific permeation enhancement of drugs by AEs had
been observed by different workers with respect to other cat-
egories of AEs also and reported in literature [18]. Interest-
ingly, bile salts exhibited significant (P < 0.05) permeation
enhancement of hydrophilic and lipophilic drugs up to 15-
and fourfold, respectively. Similar to CDs, sodium deoxycho-
late specifically increased permeability of CX to a very large
extent (15-fold) compared to other drugs. Moreover, sodium
cholate (trihydroxy bile salt) was less effective than sodium
deoxycholate (dihydroxy bile salt) in permeation enhance-
ment of drugs. In contrast to trihydroxy bile salts, dihydroxy
Fig. 5. Total protein released from intestinal mucosa when incubated with
AEs in vitro, expressed as percentage of protein released by EDTA [0.58%
w/v], which was taken as positive control (N = 3, mean ± S.D.). AE1: sodium
caprate [0.25% w/v], AE2: sodium caprylate [0.25% w/v], AE3: b-CD [1.8%
w/v], AE4: hydroxypropyl b-CD [1.8% w/v], AE5: sodium deoxycholate
[1.0% w/v], AE6: sodium cholate [1.0% w/v].
bile salts are more hydrophobic, rapidly absorbed and more
effective as AEs.
Morphological evaluation of rat jejunal segments was done
by light microscopy before and after incubation with differ-
 ARTICLE IN PRESS
P. Sharma et al. / Il Farmaco "" (2005) """-"""
ent AEs for 120 min in rat intestinal sac absorption model to
get quantitative (NPA and Vi) and semi-quantitative (morpho-
logical scoring) parameters which provided an idea of changes
in mucosa. The effect of various AEs on these morphometric
parameters of intestinal mucosa is shown in Fig. 1.
NPA is an important morphological parameter of cells,
which is defined as the distance between nucleus and apical
membrane in enterocytes, and it was determined using an eye-
piece micrometer [10]. NPA was decreased to variable extent
on treatment with different categories of AEs. However, the
extent of decrease in NPA by AEs was not statistically sig-
nificant when compared to control (P > 0.05). Moreover,
repair of superficial damage to gastrointestinal mucosa occurs
by a process called restitution and rapid epithelial restitution
is considered as one of the primary defense mechanisms of
the stomach and small intestine. In general, during restitution
phenomenon epithelial cells tend to flatten out, after exten-
sive epithelial exfoliation or epitheliolysis, to re-establish cell-
to-cell contact by pseudopod formation (amoeboid move-
ment) along the basement lamina. Such a mechanism is
responsible for restoration of epithelial continuity in spite of
extensive cell loss. Apparently, flattening of cells is reflected
as decrease in NPA of enterocytes on exposure to AEs
(Fig. 1A). In conclusion, lack of significant decrease in NPA
by AEs compared to control reflects that they have no toxic
effect on intestinal mucosa.
Second crucial morphometric parameter used in present
investigation was Vi, which is generally used to monitor
changes in the shape and surface area of villi. Vi of intestinal
tissue was found to decrease in presence of AEs but no sta-
tistically significant difference was observed (P > 0.05,
Fig. 1B). Moreover, Vi that is indicative of surface area of
absorptive intestinal mucosa exposed to external milieu in
gastrointestinal tract tends to decrease in presence of toxi-
cants. Hence, in the present study, absence of any significant
decrease in Vi of intestinal mucosa exposed to AEs con-
firmed that they are non-toxic to enterocytes.
Although, control was not exposed to AEs, it still dis-
played morphological score of higher than ‘0’. Such an obser-
vation may be consequence of mechanical handling during
processing of tissue for preparation of slides wherein some
extent of damage incurred and which was visible in form of
single cell extrusions in villi of control specimen (Fig. 2A).
However, there was no oedema of intestinal tissue and villi
height was unaffected. Further, light microscopic sections of
intestinal segments treated with sodium caprylate and sodium
caprate (Fig. 2B, C), demonstrated extrusion of single cell
and group of cells from villi when compared to morphology
of untreated segments. In addition, slight oedema and decrease
in villi length was observed in intestinal tissue treated with
medium chain fatty acids. However, the changes induced by
sodium caprylate and sodium caprate were mild and may not
be very significant in vivo.
At the same time CD treated mucosa showed structurally
altered “ballooned” absorptive cells and exfoliation of epi-
thelial cells (Fig. 2D, E). The “patchy” defects exposing the
7
lamina densa of basement membrane due to denudation of
epithelial lineage were also observed in sections. In addition
to this, CDs also caused villous atrophy, oedema and reduc-
tion of villous height in intestinal mucosa. The sections clearly
demonstrated increased number of goblet cells (visible as
large ‘blank’ circular cells) in epithelial cell lining. It was
suggested that goblet cells might play an important role in
restitution process due to CD induced epitheliolysis. The
higher secretion of mucin by densely populated goblet cells
form thick pellicle over mucosa and protect it from solubiliz-
ing action of CDs and this was further substantiated by the
fact that the increased mucin on enterocyte surfaces inhibit
penetration of CDs into intestinal mucosa [19].
Sodium deoxycholate caused extensive cell sloughing,
shortening of villi height and oedema (Fig. 2F, G). On the
other hand, sodium cholate induced morphological changes
were less prominent than sodium deoxycholate and included
cell extrusions and slight oedema. It is noteworthy that unlike
CDs, no overgrowth of goblet cells was visible in bile salt
treated intestinal mucosa. Since bile salts have higher pen-
etration through mucus secreted by goblet cells, it may not
offer barrier to their membrane solubilization potential as
observed in case of CDs.
Despite of observed morphological alterations induced by
AEs there was no statistically significant difference (P > 0.05)
when compared with morphological score of control as veri-
fied by Mann–Whitney rank sum test (non-parametric), and
hence they appear as safe additives in peroral preparations
with respect to their acute local effects on epithelial integrity.
Fluorescence polarization is extensively used to study lipid
composition and molecular dynamics in membranes [20]. In
these techniques fluorescent probes are embedded in the dif-
ferent regions of biomembrane and excited with polarized
monochromatic light, which results in excitation of probes to
higher energy levels and to dissipate energy in form of emit-
ted light. The intensity of emitted light was determined
through analyzer, oriented parallel (III) and perpendicular
(I⊥) to the direction of polarization of the excitation light.
The steady state fluorescence polarization was determined
from Eq. (3).
III − I⊥
FP = (3)
III + I⊥
In principle, when incorporated in membrane and excited
by polarized light, the fluorescent probe reorients during the
time elapsing between excitation and emission. The plane of
polarization of emitted light depends on the degree of molecu-
lar reorientation of probe in the membrane and steady state
fluorescence polarization of probe represents the degree of
structural order in the “domain” (homogenous phase e.g.
hydrophobic or lipid domain) in which it resides. Thus
decrease in polarization signifies an increase in the probe rota-
tional cone angle and consequently, a reduction in the pack-
ing order of that domain. In present investigation, changes in
membrane fluidity, detected by fluorescence polarization stud-
 ARTICLE IN PRESS
8 P. Sharma et al. / Il Farmaco "" (2005) """-"""
ies, were evaluated to understand molecular basis of augmen-
tation in permeability by AEs.
With aim to comprehensively understand changes in vari-
ous domains of membrane, different fluorescent probes that
reside in these domains were employed. Of these, DPH and
12-AS are lipophilic probes which selectively partitions into
hydrophobic core in the interior of biomembranes [21]. On
the other hand, ANS is hydrophilic probe, which localizes on
the exterior surface of lipid bilayer and at lipid–protein inter-
faces. Further, it serves as a marker of molecular interactions
at interface of membrane and aqueous bulk phase [22].
BBMVs were labeled with DPH, 12-AS and ANS sepa-
rately and treated with different concentrations of AEs at 37 °C
in 13 mM Tris/chloride buffer (pH 7.4) and fluorescence polar-
ization was measured to observe concentration dependent
effect (dose dependent studies). Furthermore, BBMVs were
incubated with AEs for different time intervals to observe
effect of exposure time on fluorescence polarization (time
dependent studies). For purpose of interpretation of results,
the decrease in fluorescence polarization of BBMVs, repre-
sented as DP was calculated as difference between fluores-
cence polarization of control and AE treated BBMVs.
Although, insertion of probe into lipid bilayer per se may be
visualized to bring minor changes in membrane disordered-
ness, such probe induced local perturbation effects were mini-
mized in present investigation by using very low concentra-
tions of probes (< 3 mM). In addition to this, any artifact in
data due to probe related changes was removed by inclusion
of control in studies and measurement of decrease in fluores-
cence polarization of vesicles due to exposure of AEs with
respect to it.
Sodium caprylate and sodium caprate resulted in signifi-
cant increase (P < 0.05) in DP of DPH, 12-AS and ANS
labeled vesicles indicating that medium chain fatty acids act
both on hydrophobic and hydrophilic domains of membrane
(Fig. 3A–C). Whereas sodium caprylate induced significant
increase (P < 0.05) in DP of 12-AS vesicles, sodium caprate
augmented DP of DPH vesicles with increase in dose. This
reflects that sodium caprylate preferentially bring membrane
disorderedness in lipid domain proximal to ‘head’ or polar
groups where 12-AS resides, while sodium caprate predomi-
nantly act in interior of hydrophobic core of the membrane
[23]. These findings indicate that medium chain fatty acids
partition into lipid bilayer and disrupt intermolecular forces
among membrane phospholipids. Subsequently ‘sol’state pre-
vails in quasi-static mosaic membrane, and this fluidized state
offers less impedance to transcellular passage of drugs
[21,23,24].
Both of medium chain fatty acids caused progressive
increase in DP of DPH and 12-AS labeled vesicles with
increasing time of exposure up to 1 h (P < 0.05) when com-
pared to control (Fig. 4A, B). In spite of increase in DP of
ANS labeled vesicles by medium chain fatty acids, neither
dose dependent nor time progressive changes were observed.
It is speculated on the basis of their interactions with globu-
lar proteins in bulk solution as reported earlier [25], that they
may be interacting with extrinsic membrane proteins to bring
conformational change in their tertiary structure. Conse-
quently, such a change may influence the orientation of phos-
pholipid molecules in membrane proximity, resulting in
increased DP of ANS, which resides in lipid–protein inter-
face of the membrane.
Similar to medium chain fatty acids, both CDs (b-CD and
HP-b-CD) increased DP of DPH, 12-AS and ANS labeled
vesicles significantly (P < 0.05). As shown in Fig. 3A, B,
b-CD increased DP in case of DPH and 12-AS probes to
higher extent when compared to HP-b-CD. Nevertheless, it
was concluded on the basis of these observations that both
CDs act on hydrophobic and hydrophilic domains of biomem-
brane. Since, CDs have affinity to form inclusion complexes
with lipoidal compounds [24], increase in DP was implica-
tion of this behavior. Furthermore, both CDs resulted in sig-
nificant increase (P < 0.05) of DP of DPH and 12-AS vesicles
with increasing increments of dose. Moreover, they also sig-
nificantly enhanced DP of DPH and 12-AS with increasing
time of exposure (P < 0.05), implying increase in membrane
disorderedness by CDs with higher exposure times.
Bile salts significantly (P < 0.05) enhanced DP of DPH,
12-AS and ANS vesicles (Fig. 3A–C), confirming their action
on both hydrophilic and hydrophobic domains of membrane.
Furthermore, they exhibited dose dependent increase in DP
of DPH and 12-AS vesicles significantly (P < 0.05). This con-
sistent decrease in fluidity was probably due to sequence of
events involving embedding of AE in lipid bilayer, formation
of micelle and extraction of micelle out of membrane remov-
ing lipid components and hence prominent decrease in fluid-
ity of lipid bilayer. Moreover, bile salts exhibited very high
increase in DP, when concentration was increased from sub-
micellar (i.e. below CMC) to above CMC value.
The CMC values of sodium deoxycholate and sodium cho-
late are 10 and 13 mM, respectively. These fluorescence stud-
ies were done at 0.25% w/v (6.06 mM), 0.5% w/v (12.06 mM)
and 1.0% w/v (24.12 mM) for sodium deoxycholate and
0.25% w/v (5.8 mM), 0.5% w/v (11.61 mM) and 1.0% w/v
(23.23 mM) for sodium cholate. At lower concentrations it
may be only embedding into lipid bilayer resulting in break-
age of bonds among tightly packed complex lipids of bilayer.
However, at concentrations above CMC, bile salts might be
effective to bring micellar solubilization and removal of mem-
brane components. Subsequent breakdown of membrane
structure due to perturbation of lipid components and high
mobility of phospholipids was perceived as sudden abrupt
increase in DP. Furthermore, sodium deoxycholate and sodium
cholate produced time dependent increase in DP of 12-AS
and ANS vesicles (P < 0.05). The distinct hyperbolic trend in
DP of ANS was observed in case of sodium cholate (Fig. 4C),
which may be due to complex sodium deoxycholate–mem-
brane protein interactions.
Proteins are major constituents of sub-cellular structures
and cytosolic ground matrix of cells. Therefore, perturbation
of membrane by AEs will result into release of proteins from
enterocytes into the lumen. In principle, extent of protein
 ARTICLE IN PRESS
P. Sharma et al. / Il Farmaco "" (2005) """-"""
released is considered as direct measure of cytolytic poten-
tial of compounds [25]. Intestinal mucosa was exposed to AEs
in vitro and amount of protein released was determined by
Lowry’s colorimetric method. Furthermore, EDTA (0.58%
w/v) was taken as positive control in present investigation to
compare protein release by AEs from intestinal mucosa, as
reported previously [15].
The extent of protein released from intestinal mucosa on
exposure to different categories of AEs, as shown in Fig. 5
was expressed as percentage of protein released by positive
control, i.e. EDTA. Results obtained from this study demon-
strated that sodium caprylate and sodium caprate resulted in
higher protein release compared to control (P < 0.05), which
may be due to mild surfactant activity associated with them.
However, extent of protein release was significantly less than
positive control (P < 0.05) and they resulted in lowest pro-
tein release among all the AEs. Based on these observations,
it was proposed that medium chain fatty acids, which consti-
tute 2–3% of total fatty acid content of dairy products, are
non-toxic [26].
b-CD as well as HP-b-CD released higher levels of pro-
tein from the intestinal mucosa, which was not statistically
different from other categories of AEs (P > 0.05). This was
due to extraction of lipoidal constituents by phenomenon of
inclusion complex formation leading to ultrapores in mem-
brane, which not only resulted in increased permeability of
drugs in vitro but also leached intracellular contents out of
cells. However, CDs (which resulted in significantly less bio-
chemical damage than positive control, P < 0.05) may be con-
sidered safe adjuvants due to the presence of homeostatic
mechanisms that are operative in vivo to reverse back super-
ficial damage and rejuvenate intestinal mucosa rapidly.
Sodium deoxycholate produced higher release than sodium
cholate, which may be due to more lipophilicity of sodium
deoxycholate (dihydroxy bile salt) than sodium cholate (tri-
hydroxy bile salt). It is also noteworthy that sodium deoxy-
cholate has shown higher AER of drugs than sodium cholate
in vitro (Table 1). As in vivo bile secretions are mixture of
different bile salts mixed with intestinal contents, they are
less toxic than single component bile salt acting in intestinal
sac. Based on these findings it was concluded that, although
AEs resulted in release of protein from enterocytes, the extent
of release was significantly lower than positive control
(P < 0.05) and hence they do not cause local toxicity to intes-
tinal mucosa.
In conclusion, intestinal permeability of hydrophilic and
lipophilic drugs was significantly enhanced by AEs in vitro
during everted sac permeation studies. CDs showed large
increase in the permeability for lipophilic drugs, LV and CSA,
while sodium deoxycholate showed a profound effect on the
permeability of hydrophilic drugs, CX and CZ. Total protein
release index and light microscopy proved that AEs are non-
toxic to enterocytes and do not induce any significant changes
in morphology of intestinal mucosa of rat. Fluorescence polar-
ization studies using BBMVs of rat jejunum had demon-
strated that AEs result in decrease in membrane fluidity, which
is responsible for increased passive permeability of drugs.
9
Acknowledgements
One of the authors P. Sharma is grateful to Council of Sci-
entific and Industrial Research (CSIR), New Delhi for finan-
cial assistance during project work.
References
[1] G.L. Amidon, H. Lennernas, V.P. Shah, J.R. Crison, A theoretical
basis for a biopharmaceutical drug classification: the correlation of in
vitro drug product dissolution and in vivo bioavailability, Pharm. Res.
12 (1995) 413–420.
[2] M.V.S. Varma, S. Khandavilli, Y. Ashokraj, A. Jain, A. Dhanikula,
A. Sood, N.S. Thomas, O.P. Pillai, P. Sharma, R. Gandhi, S. Agrawal,
V. Nair, R. Panchagnula, Biopharmaceutic classification system: a
scientific framework for pharmacokinetic optimization in drug
research, Curr. Drug Metab. 5 (2004) 375–388.
[3] B.J. Aungst, Novel formulation strategies for improving oral bioavail-
ability of drugs with poor membrane permeation or presystemic
metabolism, J. Pharm. Sci. 82 (1993) 979–987.
[4] B.J. Aungst, H. Saitoh, D.L. Burcham, S.M. Huang, S.A. Mousa,
M.A. Hussain, Enhancement of the intestinal absorption of peptides
and non-peptides, J. Control. Release 41 (1996) 19–31.
[5] A. Leone-Bay, D.R. Paton, J.J. Weidner, The development of delivery
agents that facilitate the oral absorption of macromolecular drugs,
Med. Res. Rev. 20 (2000) 169–186.
[6] E.J. van Hoogdalem, M.A. Hardens, A.G. de Boer, D.D. Breimer,
Absorption enhancement of rectally infused cefoxitin sodium by
medium-chain fatty acids in conscious rats: concentration–effect rela-
tionship, Pharm. Res. 5 (1988) 453–456.
[7] K.M. Hillgren, A. Kato, R.T. Borchardt, In vitro systems for studying
intestinal drug absorption, Med. Res. Rev. 15 (1995) 83–109.
[8] P. Sharma, H.P.S. Chawla, R. Panchagnula, LC determination of
cephalosporins in vitro in rat intestinal sac absorption model, J.
Pharm. Biomed. Anal. 27 (2002) 39–50.
[9] P. Sharma, H.P.S. Chawla, R. Panchagnula, Analytical method for
monitoring concentrations of cyclosporin and lovastatin in vitro in
everted intestinal sac absorption model, J. Chromatogr. B. 768 (2002)
349–359.
[10] B.L. Polentarutti, A.L. Peterson, A.K. Sjoberg, E.K.I. Anderberg,
L.M. Utter, Evaluation of viability of excised rat intestinal segments in
the using chamber: investigation of morphology, electrical param-
eters, and permeability characteristics, Pharm. Res. 16 (1999) 446–
454.
[11] H. Kajii, T. Horie, M. Hatashi, S. Awazu, Effects of salicylic acid on
the permeability of the plasma membrane of the small intestine of the
rat: a fluorescence spectroscopic approach to elucidate the mechanism
of promoted drug absorption, J. Pharm. Sci. 75 (1986) 475–478.
[12] M. Tomita, M. Hayashi, T. Horie, T. Ishizawa, A. Awazu, Enhance-
ment of colonic drug absorption by the transcellular permeation route,
Pharm. Res. 5 (1998) 786–789.
[13] O. Lowry, N. Rosebrough, A. Farr, R. Randall, Protein measurement
with the Folin phenol reagent, J. Biol. Chem. 193 (1951) 265–275.
[14] C.L. Venable, E.U. Frevert, Y. Kim, B.M. Fischer, S. Kamatkar,
B.G. Neel, B.B. Kahn, Overexpression of protein–tyrosine
phosphatase-1B in adipocytes inhibits insulin-stimulated phosphoi-
nositide 3-kinase activity without altering glucosa transport or
Akt/protein kinase B activation, J. Biol. Chem. 275 (2000) 18318–
18326.
[15] A. Yamamoto, T. Okagawa, A. Kotani, T. Uchiyama, T. Shimura,
S. Tabata, S. Muranishi, Effect of absorption enhancers on the pre-
meation of ebiratide, an ACTH analogue, across intestinal mem-
branes, J. Pharm. Pharmacol. 49 (1997) 1057–1061.
 ARTICLE IN PRESS
10 P. Sharma et al. / Il Farmaco "" (2005) """-"""
[16] J.A. Fix, K. Engle, P.A. Porter, P.S. Leppert, S.J. Selk, C.R. Gardner,
J. Alexander, Acylcarnitines: drug absorption enhancing agents in
gastrointestinal tract, Am. J. Physiol. 251 (1986) G332–G340.
[17] C.K. Henrikson, R.A. Argenzio, J.A. Liacos, J. Khosla, Morphologic
and functional effects of bile salt on ten porcine colon during injury
and repair, Lab. Invest. 60 (1989) 72–87.
[18] A.F. Kotze, H.L. Lueben, B.J. de Leeuw, A.G. de Boer, J.C. Verhoef,
H.E. Junginger, N-trimethyl chitosan chloride as a potential absorp-
tion enhancer across mucosal surfaces: in vitro evaluation in intestinal
epithelial cells (Caco-2), Pharm. Res. 14 (1997) 1197–1202.
[19] K. Nakanishi, T. Nadai, M. Masada, K. Miyajima, Effect of cyclodex-
trins on biological membranes. II. Mechanism of absorption enhance-
ment of the intestinal absorption of non-absorbable drug by cyclodex-
trins, Chem. Pharm. Bull. (Tokyo) 40 (1992) 1252–1256.
[20] W.J. Van Blitterswijk, R.P. Van Hoeven, B.W. Van der Meer, Lipid
structural order parameters (reciprocal of fluidity) in biomembranes
derived from steady state fluorescence polarization measurements,
Biochim. Biophys. Acta 644 (1981) 323–332.
[21] H.J. Lee, S.V. Balasubramanian, H. Murer, J. Biber, M.E. Morris,
Modulation of sulfate renal transport by alterations in cell membrane
fluidity, J. Pharm. Sci. 88 (1999) 976–980.
[22] T.M. Turunen, A. Urtti, P. Paronen, K.L. Audus, J.H. Rytting, Effect of
some penetration enhancers on epithelial membrane lipid domains:
evidence from fluorescence spectroscopy studies, Pharm. Res. 11
(1994) 288–294.
[23] T. Lindmark, T. Nikkila, P. Artursson, Mechanisms of absorption
enhancement by medium chain fatty acids in intestinal epithelial
Caco-2 cell monolayers, J. Pharmacol. Exp. Ther. 275 (1995) 958–
964.
[24] T. Lindmark, N. Schipper, L. Lazorova, A.G. de Boer, P. Artursson,
Absorption enhancement in intestinal epithelial Caco-2 monolayers
by sodium caprate: assessment of molecular weight dependence and
demonstration of transport routes, J. Drug Target. 5 (1998) 215–223.
[25] T. Irie, K. Uekama, Cyclodextrins in peptide and protein delivery,
Adv. Drug Del. Rev. 36 (1999) 101–123.
[26] M. Sakai, T. Imai, H. Ohtake, M. Otagiri, Cytotoxicity of absorption
enhancers in Caco-2 cell monolayers, J. Pharm. Pharmacol. 50 (1996)
1101–1108.