Accepted Manuscript

Self-Assembling Bubble Carriers for Oral Protein Delivery

Er-Yuan Chuang, Kun-Ju Lin, Po-Yen Lin, Hsin-Lung Chen, Shiaw-Pyng Wey,
Fwu- Long Mi, Hsu-Chan Hsiao, Dr. Chiung-Tong Chen, Dr. Hsing-Wen
Sung, PhD
PII:
DOI:
Reference:

S0142-9612(15)00549-9
10.1016/j.biomaterials.2015.06.035
JBMT 16918

To appear in:

Biomaterials

Received Date: 14 April 2015

Revised Date: 11 June 2015
Accepted Date: 18 June 2015

Please cite this article as: Chuang E-Y, Lin K-J, Lin P-Y, Chen H-L, Wey S-P, Mi F-L, Hsiao H-C,
Chen
C-T, Sung H-W, Self-Assembling Bubble Carriers for Oral Protein Delivery, Biomaterials (2015),
doi:
10.1016/j.biomaterials.2015.06.035.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a
service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.

ACCEPTED MANUSCRIPT
Self-Assembling Bubble Carriers for Oral Protein Delivery

a,

Er-Yuan Chuang,

b,c,

Kun-Ju Lin,

a,

Po-Yen Lin,
a

Hsin-Lung Chen, Shiaw-Pyng Wey,

e,

a,

Fwu-Long Mi, Hsu-Chan Hsiao, Chiung-Tong Chen, * and Hsing-Wen Sung *

a

Department of Chemical Engineering and Institute of Biomedical Engineering, National

Tsing Hua University, Hsinchu, Taiwan (ROC)

Department of Medical Imaging and Radiological Sciences, College of Medicine, Chang

Gung University, Taoyuan, Taiwan (ROC)

Department of Nuclear Medicine and Molecular Imaging Center, Chang Gung Memorial
Hospital, Taoyuan, Taiwan (ROC)

Department of Biochemistry and Molecular Cell Biology, College of Medicine, Taipei

Medical University, Taipei, Taiwan (ROC)

Institute of Biotechnology and Pharmaceutical Research, National Health Research

Institutes, Zhunan, Miaoli, Taiwan (ROC)

*Correspondence to:
Hsing-Wen Sung, PhD
Department of Chemical Engineering
National Tsing Hua University
Hsinchu, Taiwan 30013
Phone: +886-3-574-2504
E-mail: [email protected]

The first three authors contributed equally.

*To whom correspondence should be addressed: [email protected] (Dr. H.W. Sung)

and [email protected] (Dr. C.T. Chen).

ABSTRACT
Successful oral delivery of therapeutic proteins such as insulin can greatly improve the
quality of life of patients. This study develops a bubble carrier system by loading diethylene
triamine pentaacetic acid (DTPA) dianhydride, a foaming agent (sodium bicarbonate; SBC), a
surfactant (sodium dodecyl sulfate; SDS), and a protein drug (insulin) in an enteric-coated
gelatin capsule. Following oral administration to diabetic rats, the intestinal fluid that has
passed through the gelatin capsule saturates the mixture; concomitantly, DTPA dianhydride
produces an acidic environment, while SBC decomposes to form CO2 bubbles at acidic pH.
The gas bubbles grow among the surfactant molecules (SDS) owing to the expansion of the
generated CO2. The walls of the CO2 bubbles consist of a self-assembled film of water that is
in nanoscale and may serve as a colloidal carrier to transport insulin and DTPA. The grown
gas bubbles continue to expand until they bump into the wall and burst, releasing their
transported insulin, DTPA, and SDS into the mucosal layer. The released DTPA and SDS
function as protease inhibitors to protect the insulin molecules as well as absorption enhancers
to augment their epithelial permeability and eventual absorption into systemic circulation,
exerting their hypoglycemic effects.

Keywords: foaming agent; colloidal carrier; oral protein delivery; surfactant; biodistribution

ACCEPTED MANUSCRIPT
1. Introduction
Oral delivery of therapeutic proteins such as insulin can avoid the pain and discomfort
associated with injections and greatly improve the quality of life of patients. Nevertheless,
oral absorption of insulin is limited by hostile gastric and intestinal environments and by the
poor epithelial permeability [1]. Experiments have shown that enteric-coated gelatin capsules
containing insulin with protease inhibitors and absorption enhancers orally administered to
rats can achieve increased plasma insulin levels and corresponding reductions in blood
glucose [2]. The enteric coating is a pH-responsive polymer barrier, so the capsule remains
intact in the highly acidic pH environment of the stomach; conversely, as the body fluid from
the intestinal tract (neutral or slightly basic pH) comes into contact with the enteric-coated
capsule, the water soluble polymer and gelatin dissolve and the drug load diffuses through the
resulting pores.
However, the enteric polymer coated on the gelatin capsule does not dissolve instantly or
completely in the small intestine because of its partial contact with the body fluid. Thus, a
certain amount of the encapsulated protein drugs within the capsule might remain stuck within
the partially dissolved capsule (Fig. 1). Another problem is that protein molecules may
aggregate during their exposure to body fluid, which can limit their interaction with and
entrance into epithelial cells and reduce their oral bioavailability. Therefore, increasing the
bioavailability of oral protein drugs to a therapeutically acceptable level is still challenging.
To provide quick and effective delivery of a drug load from enteric-coated gelatin
capsules, this study develops a novel carrier system that involves a foaming agent (sodium
bicarbonate; SBC) that can generate CO2 bubbles upon exposure to an acidic aqueous
environment. The walls of these CO2 bubbles consist of a nanoscale self-assembled film of
water that can be sandwiched between two layers of surfactant molecules to form a colloidal
carrier for transporting protein drugs. Absorption can be increased by formulating the drug as
a solubilizate within a colloidal dispersion [3]. This bubble carrier system is prepared by

ACCEPTED MANUSCRIPT
mixing diethylene triamine pentaacetic acid (DTPA) dianhydride, SBC, a surfactant (sodium
dodecyl sulfate; SDS), and a protein drug (insulin). The DTPA dianhydride is an acidic oxide
that dissolves in water to form an acid (DTPA) solution [4]. When reacting with the acid,
SBC produces a bubbly CO2 gas [5].
Figure 1 shows a schematic of this bubble carrier system and how it works. When the
water that passes through the gelatin capsule saturates the mixture, nanosized CO2 bubbles are
generated. As the gas pressure opens the capsule, the load is quickly released. The gas
bubbles grow among the surfactant molecules, expanded by the generated CO2. The reaction
also proceeds outside the formed bubbles, which are therefore effectively suspended in a gasrich environment. The SDS is an anionic surfactant consisting of a hydrophilic (water-loving)
head and a hydrophobic (water-hating) tail. As the gas bubbles expand, the hydrophobic tail
of each surfactant molecule is attracted to the gas-rich phase while their hydrophilic head is
attracted to the water phase. Therefore, a bubble carrier system comprising a water film
containing insulin and DTPA is self-assembled between double layers of surfactant molecules
(SDS). The gas bubbles continue to expand until they come in contact with the mucosal layer
covering the intestinal wall and eventually burst and release their loads of insulin, DTPA, and
SDS. In the mucosal layer, DTPA serves a paracellular enhancer for delivering insulin
molecules through the epithelial cells, and the amphiphilic surfactant (SDS) enhances their
paracellular and transcellular permeability. Both DTPA and SDS also protect insulin
molecules by functioning as protease inhibitors [2,6].
We hypothesize that the proposed bubble carrier system can mediate temporary
alterations in the morphology of epithelial cell membranes and transient opening of their
apical junctional complexes (AJCs), which facilitates the encapsulated insulin molecules in
crossing the epithelial barrier and eventually being absorbed into systemic circulation, where
they exert hypoglycemic effects.

2. Materials and Methods

2.1. Materials
The SBC, DTPA dianhydride, SDS, and insulin (from bovine pancreas, 27.4 IU/mg) were
obtained from Sigma-Aldrich (St. Louis, MO, USA). The fluorescein isothiocyanate (FITC)insulin (bovine) and cyanine 3 (Cy3) were purchased from Invitrogen Corp. (Carlsbad, CA,
USA) and Lumiprobe Corp. (Broward, FL, USA), respectively, and

123

I was acquired from

Institute of Nuclear Energy Research (Taoyuan, Taiwan). All other chemicals and reagents
used were of analytical grade.
2.2. Preparation of test capsules
Hard gelatin capsules (size 9; Torpac Inc., Fairfield, NJ, USA) were manually filled with
a mixture of SBC (8.0 mg), DTPA dianhydride (8.0 mg), SDS (4.0 mg), and insulin (0.38 mg)
according to manufacturer instructions. The capsules without SBC were used as a control. The

prepared capsules were immersed in a methanol solution of Eudragit L100-55 (15% w/v,
Evonik Industries, Parsippany, NJ, USA) and then dried at room temperature using an airblower; this procedure was repeated three times.
2.3. Characterization of test capsules
The in vitro dissolution of test capsules was performed in test tubes containing a
phosphate buffered saline (PBS) solution at pH 2.0, pH 6.6, or pH 7.4 (adjusted by HCl); the
test tubes were immersed in a water-filled tank at 37 C. An ultrasound imaging system with a
7 MHz linear-array transducer (Z-one, Zonare, Mountain View, CA, USA) was used to
visualize the generated CO2 bubbles; consecutive ultrasonic B-mode images were recorded
with a computer.
2.4. Structural analysis of bubble carriers
The formation of bubble carriers and their structural changes as the gas bubbles expanded
were examined by fluorescence microscopy (Axio Observer; Carl Zeiss, Jena, Germany),

transmission electron microscopy (TEM; JEOL 2010F, Tokyo, Japan), and small angle X-ray
scattering (SAXS). For the TEM observation, a drop of sample solution immediately after
reaction was placed on the copper grid, and blot-dried using a filter paper, followed by
dropping 1% osmium tetroxide (OsO4), a preferential staining agent for the double bonds of
unsaturated compounds (e.g., SDS) [7], solution onto the grid and blot-dried again.
For the SAXS experiments, test samples were injected into sample cells composed of two
Kapton windows [8]. The measurements were performed using the BL23A1 beamline at the
National Synchrotron Radiation Research Center (NSRRC), Hsinchu, Taiwan. The energy of
the beam source was 15.0 keV, and the sample-to-detector distance was 3060.3 mm. The
scattering intensity profile was obtained after corrections for sample transmission, empty cell
transmission, empty cell scattering, and the sensitivity of the detector. Scattering intensity
I(Q) was plotted as a function of scattering vector, Q = (4/)sin(/2) ( = scattering angle).
The resolvable time scale for this instrument is under 10 sec. A two-dimensional Pilatus 1M
detector with 981 1043 pixel resolution was used to record the SAXS profiles.
2.5. Investigation of the mechanism of epithelial permeability
The Caco-2 cell monolayers were grown on a tissue-culture-treated polycarbonate filter
in a Costar Transwell with twelve wells/plate (Corning Costar Corp., NY, USA) and then
treated with test samples (DTPA or SDS, 2.0 mg/mL, 1.0 mL) containing FITC-insulin for 2 h;
translocation of their AJC proteins was then analyzed by immunofluorescence staining. After
washing the cells three times with pre-warmed PBS, the cells were fixed and permeabilized in
100% acetone for 10 min at 20 C. Non-specific binding was blocked in 5% normal goat
serum (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) in PBS for 60 min at
37 C. The cells were then incubated overnight in primary antibodies [mouse anti-claudin-4
(Invitrogen) and donkey anti-E-cadherin (Abcam, Cambridge, MA, USA)] at 4 C. Results
were visualized by applying appropriate Alexa-Fluor-conjugated secondary antibodies
(Invitrogen); the cells were also counterstained with DAPI (Invitrogen) for examination by

ACCEPTED MANUSCRIPT
inverted confocal laser scanning microscopy (CLSM; Zeiss LSM780, Carl Zeiss, Jena GmbH,
Germany).
Extracellular calcium levels were detected by staining the treated cells with 40.0 mM
Alizarin Red S (ARS; Sigma-Aldrich), pH 4.0, for 10 min followed by washing with 70%
ethanol and with PBS to remove non-specific staining. Samples were then visualized by an
inverted phase-contrast light microscope (Nikon Eclipse-Ti, Nikon Instruments Inc., Melville,
NY, USA).
2.6. Evaluation of inhibition of proteolytic degradation
In vitro inhibition of intestinal proteases by DTPA and SDS was separately studied in the
proximal portion of the small intestine freshly isolated from rats (Wistar). The intestinal
tissues were incubated with a KrebsRinger buffer containing insulin (1.0 mg/mL, 1.0 mL)
and DTPA or SDS (2.0 mg/mL, 1.0 mL) at 37 C. The remaining amount of intact insulin was
quantified by high-performance liquid chromatography (HPLC) analysis of test solutions
(50.0 L) withdrawn at different time intervals, based on a protocol described in the literature
[9].
2.7. Animal study
All animal studies were performed according to the Guide for the Care and Use of
Laboratory Animals, which was prepared by Institute of Laboratory Animal Resources,
National Research Council, and published by the National Academy Press in 1996. All studies
were also approved by the Institutional Animal Care and Use Committee of National Tsing
Hua University (protocol number 10046).
2.8. In vivo dissolution of test capsules and biodistribution of insulin
Insulin was radiolabeled with Iodine-123 (

123

I) using an iodogen-tube (Pierce Iodination

Tubes, Thermo Fisher Scientific, Rockford, IL, USA) according to a method reported in the
literature [10]; previous radiochemical stability tests in human serum showed that 91% of the
isotope-labeled compound remained intact after incubation for 24 h. After purification, the
7

ACCEPTED MANUSCRIPT
obtained

123

I-insulin was lyophilized for further studies. The dynamic plain X-ray and

scintigraphy images of rats (Wistar, ca. 250 g, n = 3) were acquired at predetermined time
intervals after oral administration of test capsules containing

123

I-insulin and barium sulfate, a

radiocontrast agent.
The biodistribution of insulin in the intestinal tract was investigated by in vivo imaging
system (IVIS). At predetermined time points after oral ingestion of test capsules containing
FITC-insulin, the rats (n = 3) were sacrificed. The gastrointestinal (GI) tracts were then
harvested, irradiated (FITC excitation wavelength, 495 nm), and imaged with an appropriate
emission filter (519 nm) to obtain FITC images. The retrieved tissues were also washed three
times with PBS, fixed in 3.7% paraformaldehyde, processed for cryosectioning, and then
examined under CLSM.
The biodistribution of insulin absorbed into systemic circulation was examined by a
single-photon emission computed tomography (SPECT)/computed tomography (CT) scanner
(n = 3 rats). After the last SPECT scan, additional whole-body CT images of the internal
organs were acquired after intravenous administration of the Conray contrast agent
(Mallinckrodt, MO, USA). The amount of contrast fluid stasis in the vasculature was
maximized by slowly injecting the contrast agent via the tail vein (2.0 mL over 40 s). After
receiving half of the Conray dose, each animal was sacrificed by co-injection of potassium
chloride under deep anesthesia. The image scanning was performed using the protocol
described in a previous work by this research group [11].
2.9. In vivo toxicity
The toxicity of the test capsule was evaluated in rats, which were randomly divided into
two groups. For 14 days, the experimental group orally received daily a dose of the capsules
that contained DTPA dianhydride, SDS and SBC (100 mg/kg, n = 6 for each gender), while
the other untreated group was a control. Animals were observed carefully for the onset of any
signs of toxicity and monitored for changes in body weight. After the animals had been
8

ACCEPTED MANUSCRIPT
sacrificed, internal organs of each animal were harvested and observed grossly. For sectional
histology, intestine, liver and kidney specimens were fixed in 10% phosphate buffered
formalin, embedded in paraffin, cross-sectioned at 5 m and then stained with hematoxylin
and eosin (H&E).
2.10. Examination of the pharmacokinetic (PK) and pharmacodynamic (PD) profiles
Diabetes was induced in rats via intraperitoneal injection of streptozotocin as described
previously [12]. The rats were then fasted overnight and for the remainder of the experiment
but with free access to the water. The following formulations were administered individually
to the diabetic rats: an enteric-coated capsule containing free-form insulin powder alone (30.0
IU/kg, oral); an enteric-coated capsule filled with DTPA dianhydride, SDS, insulin (30.0
IU/kg), together with or without SBC (oral); and a subcutaneous (SC) injection of free-form
insulin solution (5.0 IU/kg) (n = 6 for each studied group). Blood samples were taken from
the tail veins before administering the drug and at varying time intervals after dosing. The
blood glucose levels were then measured with a glucose meter (LifeScan, Milpitas, CA, USA).
Plasma insulin levels were determined by centrifuging blood samples (10,000 rpm, 4 C, for
15 min) with subsequent quantification using ELISA (bovine insulin kit, Mercodia AB,
Uppsala, Sweden). The formula used to calculate the relative bioavailability (BAR) of insulin
after the treatments was
BAR =

AUC
AUC

Dose
oral
SC Dose SC 100%
oral

where AUC represents the total area under the plasma insulin concentration vs. time curve.
2.11. Statistical analysis
Average data were reported as means SD. Statistical analyses were performed with
SPSS (Chicago, IL, USA). One-tailed Student t test was used for statistical comparisons of
the two groups. A P value less than 0.05 was considered statistically significant.

3. Results and Discussion

This study evaluated the feasibility of using the proposed bubble carrier system for oral
insulin delivery. The DTPA dianhydride, SBC (foaming agent), SDS, and insulin were
thoroughly mixed and loaded in hard gelatin capsules. The capsules were then coated with an

enteric polymer (Eudragit L100-55), which is dissoluble above pH 5.5 for protecting the
loaded contents from gastric fluid [13,14]. The control was enteric-coated capsules without
SBC.
3.1. In vitro dissolution of test capsules
The dissolution of the enteric-coated capsules with or without SBC was evaluated in vitro
in conditions simulating the pH environments in the GI tract. In this experiment, trypan blue
was included in the test capsules as a dye indicator. The generation of CO2 bubbles was
investigated using an ultrasound machine. According to Fig. 2a, both test capsules remained
intact at pH 2.0 (mimicking the acidic milieu in the stomach) [15], confirming the

gastroresistant feature of Eudragit L100-55. In the pH environments of the small intestine

(pH 6.6 and pH 7.4) [16], the load diffused through the pore that developed as the enteric
polymer coated on the control capsule (containing no SBC) slowly dissolved. The absence of
ultrasound signals in this case suggests that no bubbles were generated.
Conversely, following partial dissolution of the enteric polymer and penetration of water
into the gelatin capsule containing DTPA dianhydride and SBC, gas bubbles were
immediately produced (Fig. 2b). When DTPA dianhydride is exposed to water, an acidic
environment is produced [17], in which SBC decomposes to form CO2 bubbles that are
hyperechogenic and could act as a contrast enhancer for ultrasound imaging [18]. The
resulting gas pressure forced the capsule to open and quickly released its loaded contents.

10
10

3.2. Formation of bubble carriers

To identify their locations within the generated gas bubbles, insulin was labeled with a
green fluorescence dye (FITC) while DTPA was marked with a red fluorescence dye (Cy3);
the test samples were then viewed under a fluorescence microscope. Structural changes
caused by expansion of the gas bubbles were observed by TEM and by SAXS. Figure 3a
shows that the insulin molecules released from the capsule without SBC tended to aggregate
upon exposure to water. A well-known effect of water sorption by protein molecules is the
production of structural transitions that cause aggregation and other deleterious phenomena
[19].
Conversely, as the contents of the capsule with the foaming agent were released and
exposed to water, CO2 bubbles in nanoscale were self-assembled (Fig. 3b). The obtained
TEM and fluorescence micrographs demonstrate that the gas bubbles grew from nanoscale
(Fig. 3b) to microscale (Figs. 3c and 3d) among the surfactant molecules (SDS) owing to the
expansion of the generated CO2. The gas bubbles were surrounded by a layer (thickness,
approximately 150 nm) of green (Fig. 3c) and red (Fig. 3d) fluorescence indicating insulin
and DTPA molecules, respectively. As the gas bubbles expanded, the hydrophilic heads of
SDS remained solvated in the water phase, and the hydrophobic tails were extended into the
gas-rich environment which is generally considered hydrophobic [20], as schematically
illustrated in Fig. 3e. Thus, a self-assembled film of water sandwiched between the two layers
of surfactant molecules stabilized the generated gas bubbles; the self-assembled water film
could also serve as a colloidal carrier to encapsulate the hydrophilic insulin and DTPA
molecules. The locations of insulin and DTPA after bubble formation are important for their
subsequent delivery.
Figure 3f shows the results of time-resolved SAXS performed to reveal the structural
changes in gas bubbles during their formation. When the SDS in powder form came in contact
with water (t = 0 sec, first row in Figs. 3c3e), the SAXS profile showed two sharp peaks

ACCEPTED MANUSCRIPT
corresponding to (001) and (002) diffractions of a multilamellar structure, signifying that the
surfactant molecules in the powder had self-assembled into a multilamellar structure with an
interlamellar distance of ca. 4.0 nm. At 30 sec after reaction, the original primary peak
transformed into a broad halo and the second-order peak vanished. This result indicates that
the multilamellar structure was disrupted by the generated CO2 bubbles to yield oligolamellar
carriers composed of gas bubbles surrounded by water films (second row in Figs. 3c3e). The
progressive diminishment of the primary peak in the SAXS profiles suggested that the number
of layers of surfactant molecules decreased as the reaction proceeded. After 240 sec, the
SAXS peaks were no longer visible, which indicated the formation of unilamellar carriers
(third row, Figs. 3c3e).
3.3. Enhanced epithelial permeability
Drug permeation may occur across the intestinal epithelium and through either
transcellular or paracellular routes [21]. To elucidate the precise mechanisms of interactions
between the bubble carrier system and epithelial cells and its subsequent transport of the drug,
the effects of DTPA and SDS on epithelial permeability were individually studied using
Caco-2 cell monolayers. Monolayers of Caco-2 cells derived from human colorectal
adenocarcinoma are widely used in in vitro models for predicting intestinal drug permeability
[22].
The AJC, which is comprised of adherens junctions (AJs) and tight junctions (TJs), is an
important regulator of the intestinal cell structure and the epithelial barrier function. The AJs
cause plasma membranes to combine, and they enhance TJ assembly; alteration of AJs
therefore modulates the TJ structure and epithelial permeability [12]. Claudin-4 (CLDN4) is a
2+

TJ protein whereas E-cadherin [23], a calcium (Ca )-dependent adhesion molecule, is the
major transmembrane protein of AJs. Accordingly, extracellular Ca

2+

is essential for

ACCEPTED MANUSCRIPT
formation of AJC and maintenance of cellcell junctions [24]. Both AJs and TJs can rapidly
disassemble and reorganize in response to various extracellular stimuli [25].
Figure 4 shows the extracellular Ca

2+

2+

levels (as indicated by the ARS-labeled Ca ) and

the immunofluorescence staining of Caco-2 cell monolayers before and after exposure to
FITC-insulin combined with either DTPA or SDS. Following treatment with test samples, the
ultrastructural changes in Caco-2 cell monolayers were directly visualized by CLSM. The
DTPA treatment significantly reduced extracellular calcium levels; additionally, both CLDN4
and E-cadherin clearly revealed intracellular internalization. Since DTPA is a well-known
2+

complexing agent that chelates divalent metal ions such as Ca [2], it disrupted the epithelial
AJCs and thereby enhanced the permeability of insulin via the paracellular pathway.
Conversely, SDS treatment caused insulin molecules to permeate Caco-2 cells via both
the paracellular and transcellular routes. Previous studies reported that SDS enhances protein
drug absorption across the intestinal epithelium [26]. The increased paracellular permeability
caused by SDS might result from transient depletion of cellular ATP whereas enhanced
transcellular absorption is associated with an altered cell morphology [27]; however, this
alteration is reversible [28].
3.4. Inhibition of proteolytic degradation
Inhibiting the proteolytic degradation of insulin in the intestinal fluid and the mucosal
layer is critical for increasing its oral bioavailability. The inhibiting effects of DTPA and SDS
on intestinal proteases were separately studied in in vitro studies of the proximal portion of
the small intestine (i.e., the duodenum) freshly isolated from rats. The various digestive
enzymes secreted from the pancreas into the duodenum are well-recognized [29]. The HPLC
was used to quantify the remaining amount of intact insulin. According to Fig. 5, free-form
insulin molecules were rapidly degraded by the proteolytic enzymes that were present in the
mucosal layer that covered the intestinal wall; however, the addition of DTPA or SDS in

ACCEPTED MANUSCRIPT
insulin produced a strong protective effect against the intestinal proteases (P < 0.05). Insulin
2+

is highly sensitive to trypsin and chymotrypsin, which are Ca -dependent enzymes in the
intestinal fluid and in the mucosal layer [30]. One strategy for inhibiting intestinal protease
2+

activities is using complexing agents such as DTPA to remove essential metal ions (e.g., Ca )
from the enzyme structure. A previous study indicated that SDS has a binding affinity for
both free enzymes and the enzyme-substrate binary complex. Proposed molecular
mechanisms of proteolytic inhibition by SDS include local conformational changes or altered
charge distributions of the proteases induced by their specific binding of SDS [31].
3.5. In vivo dissolution of the test capsule and its subsequent release of insulin
In the in vivo dissolution study, the capsule with SBC was orally administered to rats
using a gavage needle; its counterpart containing no SBC served as a control. In this
experiment, insulin was radiolabeled with

123

I; additionally, a radiocontrast agent, barium

sulfate, was incorporated in the test capsules. A SPECT/CT scanner was used for imaging.
The dynamic plain X-ray and scintigraphy images of the rats were acquired sequentially at 2
min intervals for 1 h after oral administration of the test capsules. Figure 6a shows that, once
the capsule left the stomach, the CO2 generated in the pH environment of the small intestine
immediately disintegrated the SBC capsule; the

123

I-labeled insulin was then rapidly released

(white arrow). During this time, the capsule without the foaming agent remained intact.
3.6. Intestinal biodistribution and absorption of the released insulin
Next, FITC-labeled insulin was used to visualize the biodistribution of the insulin
released from test capsules in the intestinal tract and its subsequent absorption into the villi.
At predetermined time points after oral ingestion of the test capsules, the rats were sacrificed,
and their GI organs were retrieved and examined with an IVIS. Cryosections of the retrieved
intestinal villi were also inspected with a CLSM. Figure 6b shows that the FITC-insulin was
released into the intestinal tract more rapidly and at higher concentrations in the group that

ACCEPTED MANUSCRIPT
received the capsule with SBC compared to the group that received the capsule without SBC.
In the former case, the difference most likely resulted from formation of gas bubbles which
generated forces that increased dispersion of the released insulin molecules and avoided their
aggregation. After oral ingestion, dispersion of the drug molecules in the intestinal milieu is a
prerequisite for their absorption [32].
The bubble carriers formed with a water film containing insulin and DTPA sandwiched
between two layers of SDS (Fig. 3). They then drifted in the intestinal lumen until they
eventually bumped into the intestinal wall and burst, releasing their loaded insulin, DTPA,
and SDS into the mucosal layer (Fig. 1). The released insulin, DTPA, and SDS then infiltrated
the mucosal layer and accessed the epithelium surface. The presence of green fluorescence
(FITC-insulin) on the lacteal side of the villi demonstrate that DTPA and SDS protect insulin
molecules from proteolytic degradation and enhance their epithelial permeability and
absorption into systemic circulation (Fig. 6c). Notably, the green fluorescence intensity was
higher in the group treated with the SBC capsule compared to the group treated with the
control capsule containing no SBC, indicating that the developed bubble carrier system is a
promising vehicle for orally administered insulin.
3.7. Systemic biodistribution of the absorbed insulin
The systemic biodistribution of the absorbed

123

I-insulin was investigated by SPECT/CT

over a 4-h period after ingestion of the test capsules. The radioactivity of

123

I-insulin was

presented in rainbow pseudo-color scale and superimposed on the CT images (gray color) to
distinguish anatomic localizations. In both groups, the accumulation of radioactivity in the
heart, kidney, and urinary bladder indicated systemic absorption of

123

I-insulin (Fig. 6d). The

rapid release of the drug load as the test capsule was disintegrated by the bubble pressure
produced a high local concentration of the drug. Since transport to the blood is a function of
the local drug concentration, rapid release of a high concentration results in rapid transport of

ACCEPTED MANUSCRIPT
the drug. Therefore, the radioactivity of

123

I-insulin in systemic circulation was earlier and

stronger in the group that received the SBC test capsule compared to the control group that
received the capsule without SBC.
For a further illustration of the enhanced uptake of insulin in systemic circulation
conferred by the bubble carriers in the group that orally received the capsule containing SBC,
the Supplementary Information provides a time-lapse 3D video; the group orally treated with
the capsule without SBC was used as a control. The video shows that, after ingestion, the test
capsule with SBC disintegrated faster and more completely compared to the control capsule
without the foaming agent, producing a higher local concentration of the released insulin.
Thus, absorption of insulin into systemic circulation in the case incorporating the foaming
agent was much quicker and more pronounced than in the case containing no foaming agent.
3.8. In vivo toxicity
In the in vivo toxicity investigation, for 14 days, rats were treated daily with a dose of the
capsules that contained SBC; untreated rats formed a control group. According to Fig. S1a,
the body weights of the rats in the experimental group (male and female) increased slightly
over time in a manner similar to those of the untreated control group (P > 0.05). The
reduction in body-weight is commonly regarded a simple and sensitive index of in vivo
toxicity [11]. No apparent tissue damage or toxic effects in organs (small intestine, liver and
kidney) were observed (Fig. S1b). Generally, a specific tissue-level toxicological study is
concerned with inflammation (intestine), hepatotoxicity (liver), and nephrotoxicity (kidney)
responses [11]. The analytical results herein suggest no significant in vivo toxicity following
treatment with the SBC capsule.
3.9. In vivo PK and PD profiles
The PK and PD profiles of the diabetic rats that received the capsules with or without
SBC were also analyzed. The control groups received free-form insulin solution via SC
16
16

ACCEPTED MANUSCRIPT
injection or the capsule containing the free-form insulin powder alone via oral ingestion.
Figure 7 shows that SC injection of free-form insulin solution resulted in a maximum plasma
concentration at 1 h post administration, which caused a sharp decrease in the blood glucose
level within 13 h before returning to its basal level. As expected, oral administration of the
capsule filled with the free-form insulin powder produced no hypoglycemic effects, indicating
the poor oral absorption of insulin in the absence of a suitable delivery system.
Conversely, measurable concentrations of insulin were obtained within an hour of
administration of the capsules containing protease inhibitors/absorption enhancers (DTPA and
SDS), producing a steady plasma concentration of drug for over 10 h with maximum insulin
concentrations at 45 h. This fact caused a slow but prolonged reduction in blood glucose
levels, which indicated the continuing pharmacological activity of the absorbed insulin.
Compared with the group given the capsule without SBC, the group given the capsule with
SBC revealed a faster and greater absorption of insulin, resulting in a more substantial
hypoglycemic effect. For the cases without and with SBC, Table 1 shows that the maximum
plasma concentrations were 34.0 6.4 and 68.4 6.8 IU/mL, the AUC (010 h) values
were 123.8 28.4 and 249.2 34.1 IU h/mL, and the corresponding relative BAR values
were 10.8 3.1 and 21.7 1.7% (P < 0.05), respectively.
Conversely, administration of the capsules containing protease inhibitors/absorption
enhancers (DTPA and SDS) produced a slow but prolonged reduction in blood glucose levels,
which indicated the continuing pharmacological activity of the absorbed insulin. Compared
with the group given the capsule without SBC, the group given the capsule with SBC revealed
a faster and greater absorption of insulin, resulting in a more substantial hypoglycemic effect.
For the cases without and with SBC, Table 1 shows that the AUC (010 h) values were 123.8
28.4 and 249.2 34.1 IU h/mL, and the corresponding relative BAR values were 10.8
3.1 and 21.7 1.7% (P < 0.05), respectively. These results suggest that oral administration of

ACCEPTED MANUSCRIPT
insulin using our bubble carriers may reduce the risk of hyperinsulinemia which has been
commonly observed in patients receiving insulin subcutaneously

ACCEPTED MANUSCRIPT
4. Conclusion
The above results reveal that the bubble carrier system developed in the study
considerably enhances the dispersion of insulin molecules released from an enteric-coated
capsule transiting the intestinal tract. Since the transported DTPA and SDS serve as both
protease inhibitors and absorption enhancers, they improve the oral bioavailability of insulin
and induce hypoglycemic effects. In addition to oral administration of insulin, the proposed
bubble carrier system has potential applications as a platform for oral administration of other
therapeutic proteins.

Acknowledgements
The authors would like to thank the Ministry of Science and Technology, Taiwan, for
financially supporting this research under Contract No. NSC 103-2120-M-007-009-CC1. The
molecular-imaging

study

was

partially

supported

by

grants

CMRP391513

and

CMRPG3A0512 from Chang Gung Memorial Hospital, Linkou, Taiwan. The NSRRC is
gratefully acknowledged for providing technical support for the synchrotron X-ray scattering
experiment.

ACCEPTED MANUSCRIPT
REFERENCES
[1]

des Rieux A, Fievez V, Garinot M, Schneider YJ, Prat V. Nanoparticles as potential

oral delivery systems of proteins and vaccines: a mechanistic approach. J Control
Release
2006;116:127.
[2] Su FY, Lin KJ, Sonaje K, Wey SP, Yen TC, Ho YC, et al. Protease inhibition and
absorption enhancement by functional nanoparticles for effective oral insulin delivery.
Biomaterials 2012;33:280111.
[3] Pouton CW. Lipid formulations for oral administration of drugs: non-emulsifying, selfemulsifying and self-microemulsifying drug delivery systems. Eur J Pharm Sci
2000;11:S938.
[4] Harper JF, Brooker G. Femtomole sensitive radioimmunoassay for cyclic AMP and
cyclic GMP after 2'0 acetylation by acetic anhydride in aqueous solution. J Cyclic
Nucleotide Res 1975;1:20718.
[5] Ke CJ, Su TY, Chen HL, Liu HL, Chiang WL, Chu PC, et al. Smart multifunctional
hollow microspheres for the quick release of drugs in intracellular lysosomal
compartments. Angew Chem Int Edit 2011;123:82369.
[6] Deane SM, Robb FT, Woods DR. Production and activation of an SDS-resistant alkaline
serine exoprotease of vibrio alginolyticus. J Gen Microbiol 1987;133:3918.
[7] Riemersma JC. Osmium tetroxide fixation of lipids: nature of the reaction products. J
Histochem Cytochem 1963;11:43642.
[8] Peng SF, Yang MJ, Su CJ, Chen HL, Lee PW, Wei MC, et al. Effects of incorporation
of poly(-glutamic acid) in chitosan/DNA complex nanoparticles on cellular uptake and
transfection efficiency. Biomaterials 2009;30:1797808.
[9] Marschutz MK, Caliceti P and Bernkop-Schnurch A. Design and in vivo evaluation of an
oral delivery system for insulin. Pharm Res 2000;17:146874.
[10] Jang B, Park S, Kang SH, Kim JK, Kim SK, Kim IH, et al. Gold nanorods for target
selective SPECT/CT imaging and photothermal therapy in vivo. Quant Imaging Med
Surg 2012;2:111.
[11] Chuang EY, Lin KJ, Su FY, Mi FL, Maiti B, Chen CT, et al. Noninvasive imaging oral
absorption of insulin delivered by nanoparticles and its stimulated glucose utilization in
controlling postprandial hyperglycemia during OGTT in diabetic rats. J Control Release
2013;172:51322.
[12] Chuang EY, Lin KJ, Su FY, Chen HL, Maiti B, Ho YC, et al. Calcium
depletion- mediated protease inhibition and apical-junctional-complex disassembly via
an EGTA- conjugated carrier for oral insulin delivery. J Control Release 2013;169:296
305.
[13] Khan MZI, Prebeg , Kurjakovi N. A pH-dependent colon targeted oral drug delivery
system using methacrylic acid copolymers: I. manipulation of drug release using

Eudragit L100-55 and Eudragit S100 combinations. J Control Release 1999;58:215

22.
[14] De Jaeghere F, Allmann E, Kubel F, Galli B, Cozens R, Doelker E, et al. Oral
bioavailability of a poorly water soluble HIV-1 protease inhibitor incorporated into pHsensitive particles: effect of the particle size and nutritional state. J Control Release
2000;68:2918.
[15] Evans DF, Pye G, Bramley R, Clark AG, Dyson TJ, Hardcastle JD. Measurement of
gastrointestinal pH profiles in normal ambulant human subjects. Gut 1988;29:103541.
[16] Sonaje K, Chen YJ, Chen HL, Wey SP, Juang JH, Nguyen HN, et al. Enteric-coated
capsules filled with freeze-dried chitosan/poly(-glutamic acid) nanoparticles for oral
insulin delivery. Biomaterials 2010;31:338494.
20
20

ACCEPTED MANUSCRIPT

[17] Diop CIK, Li HL, Xie BJ, Shi J. Effects of acetic acid/acetic anhydride ratios on the
properties of corn starch acetates. Food Chem 2011;126:16629.

21
21

[18] Chen KJ, Chuang EY, Wey SP, Lin KJ, Cheng F, Lin CC, et al. Hyperthermia-mediated
local drug delivery by a bubble-generating liposomal system for tumor-specific
chemotherapy. ACS Nano 2014;8:510515.
[19] Flores-Fernndez GM, Sol RJ, Griebenow K. The relation between moisture-induced
aggregation and structural changes in lyophilized insulin. J Pharm Pharmacol
2009;61:155561.
[20] Knebel D, Sieber M, Reichelt R, Galla HJ, Amrein M. Scanning force microscopy at the
air-water interface of an air bubble coated with pulmonary surfactant. Biophys J
2002;82:47480.
[21] Arajo F, Fonte P, Santos HA, Sarmento B. Oral delivery of glucagon-like peptide-1 and
analogs: alternatives for diabetes control? J Diabetes Sci Technol 2012;6:148697.
[22] Roger E, Lagarce F, Garcion E, Benoit JP. Lipid nanocarriers improve paclitaxel
transport throughout human intestinal epithelial cells by using vesicle-mediated
transcytosis. J Control Release 2009;140:17481.
[23] Hsu LW, Ho YC, Chuang EY, Chen CT, Juang JH, Su FY, et al. Effects of pH on
molecular mechanisms of chitosanintegrin interactions and resulting tight-junction
disruptions. Biomaterials. 2013;34:78493.
[24] Brown RC, Davis TP. Calcium modulation of adherens and tight junction function: a
potential mechanism for blood-brain barrier disruption after stroke. Stroke
2002;33:170611.
[25] Ivanov AI, Nusrat A, Parkos CA. Endocytosis of epithelial apical junctional proteins by a
clathrin-mediated pathway into a unique storage compartment. Mol Biol Cell
2004;15:17688.
[26] Thanou M, Verhoef JC, Junginger HE. Oral drug absorption enhancement by chitosan
and its derivatives. Adv Drug Deliv Rev 2001;52:11726.
[27] Ward PD, Tippin TK, Thakker DR. Enhancing paracellular permeability by modulating
epithelial tight junctions. Pharm Sci Technolo Today 2000;3:34658.
[28] Trier S, Linderoth L, Bjerregaard S, Andresen TL, Rahbek UL. Acylation of glucagonlike peptide2: interaction with lipid membranes and in vitro intestinal permeability.
PLoS ONE 2014;9:e109939.
[29] Ihse I, Lilja P, Lundquist I. Trypsin as a regulator of pancreatic secretion in the rat. Scand
J Gastroenterol 1979;14:87380.
[30] Yamagata T, Morishita M, Kavimandan NJ, Nakamura K, Fukuoka Y, Takayama K, et
al.
Characterization of insulin protection properties of complexation hydrogels in gastric
and intestinal enzyme fluids. J Control Release 2006;112:3439.
[31] Mu H, Zhou SM, Xia Y, Zou H, Meng F, Yan YB. Inactivation and unfolding of the
hyperthermophilic inorganic pyrophosphatase from thermus thermophilus by sodium
dodecyl sulfate. Int J Mol Sci 2009;10:284959.
[32] Porter CJH, Trevaskis NL, Charman WN. Lipids and lipid-based formulations:
optimizing the oral delivery of lipophilic drugs. Nat Rev Drug Discov 2007;6:23148.

Table
Table 1. The PK parameters of insulin in diabetic rats given varying insulin formulations (n =
6).

Dose [IU/kg]
c)

Cmax [IU/mL]
d)

Tmax [h]
e)

AUC(010 h) [IU h /mL]

f)

BAR [%]
a)

Free-Form
Insulin (SC)a)

Free-Form
Insulin (Oral)b)

w/o Foaming
Agent (Oral)b)

w/ Foaming
Agent (Oral)b)

5.0

30.0

30.0

30.0

119.6 3.8

1.8 0.7

34.0 6.4

68.4 6.8

1.0

5.0

5.0

191.3 30.7

6.5 2.7
0.5 0.3

123.8 28.4
10.8 3.1

100
b)

IP

4.0

249.2 34.1
21.7 1.7

c)

Subcutaneous injection; Oral administration of test capsules; Maximum plasma

d)
e)
concentration; Time at which Cmax is attained; Area under the plasma concentration-time
f)
curve; Relative bioavailability of insulin.

ACCEPTED MANUSCRIPT
Figures

U
N
A
M

R
C
S

T
IP

Figure 1. Schematic illustrations showing the formation and expansion of bubble carriers (see
subfigures a, b, and c) developed from the orally administered gelatin capsule and their
mechanism for delivering insulin across the epithelial barrier. TJ: tight junction; AJ: adherens
junction.

C
S
U
N
A
M

I
PT
R

Figure 2. (a) Results of the dissolution and disintegration of test capsules with or without
SBC evaluated in vitro under varying pH to simulate the pH environments in the GI tract; (b)
ultrasound images showing the generation of CO2 bubbles.

E
C
C

E
T
P

U
N
A
M
D

I
PT
R
C
S

Figure 3. (a) Fluorescence images showing the dispersion of the FITC-labeled insulin
released from the capsule with or without the foaming agent (SBC) upon exposure to water;
(b) TEM micrographs showing the evolution and expansion of nanobubbles generated by the
capsule containing SBC. Fluorescence images revealing encapsulation of both (c) FITCinsulin and (d) Cy3-DTPA within the water film of the bubble carriers; (e) corresponding
schematic illustrations displaying the bubble carriers produced upon exposure to water; (f)
SAXS profiles showing structural changes in the formation and expansion of bubble carriers.

C
S
U
N
A
M

T
IP
R

2+

Figure 4. Photomicrographs showing the extracellular calcium levels (red ARS-labeled Ca )

and immunofluorescence staining results for Caco-2 cell monolayers treated with DTPA or
SDS in the presence of FITC-insulin.

U
N
A

T
IP
R
C
S

Figure 5. Results of insulin degradation in the presence or absence of DTPA or SDS in

proximal intestinal segments freshly isolated from rats. N.S.: not significant; *: statistically
significant (P < 0.05).

T
P
E

U
N
A
M
D
E

R
C
S

T
IP

Figure 6. (a) Sequential plain X-ray and X-ray/scintigraphy images of test rats orally treated
123
with the capsules containing barium sulfate and I-isulin (white arrow) with or without
foaming agent. (b) The IVIS fluorescent images of FITC-insulin released from capsules with
or without foaming agent during their transit in the GI tract; (c) CLSM images of frozen
intestinal sections with schematic diagram of insulin absorption into the villi. (d)
123
Biodistribution of
I-insulin illustrated by contrast-enhanced tomographic images. The
123
intensities of I-insulin in different organs are shown in rainbow pseudo-color scale.

T
P

Figure 7. Plasma insulin level vs. time profiles and blood glucose change vs. time profiles of
diabetic rats treated with varying insulin formulations. Oral: oral administration of test
capsules and SC: subcutaneous injection.

ACCEPTED MANUSCRIPT

Supporting Information
Self-Assembling Bubble Carriers for Oral Protein Delivery

Er-Yuan Chuang , Kun-Ju Lin , Po-Yen Lin , Hsin-Lung Chen, Shiaw-Pyng Wey,
*

Fwu-Long Mi, Hsu-Chan Hsiao, Chiung-Tong Chen , and Hsing-Wen Sung

A
M
D
E
T
P

R
C
S
U
N

Figure S1. (a) Changes in body weight with time and (b) representative
photomicrographs of the intestine, liver and kidney sections (H&E staining) of rats
before and after treated with test capsules in a dose of 100 mg/kg per day for 14 days.