Journal of Colloid and Interface Science xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

Disposition and crystallization of saturated fatty acid in mixed micelles

of relevance to lipid digestion
Stephanie Phan a, Stefan Salentinig a,⇑, Elliot Gilbert b, Tamim A. Darwish c, Adrian Hawley d,
Reece Nixon-Luke e, Gary Bryant e, Ben J. Boyd a,f,⇑
a
Monash Institute of Pharmaceutical Sciences, Monash University, 381 Royal Parade, Parkville, VIC 3052, Australia
b
Bragg Institute, Australian Nuclear Science and Technology Organisation, Locked Bag 2001, Kirrawee DC, NSW 2232, Australia
c
National Deuteration Facility, Australian Nuclear Science and Technology Organisation, Locked Bag 2001, Kirrawee DC, NSW 2232, Australia
d
SAXS/WAXS beamline, Australian Synchrotron, 800 Blackburn Rd, Clayton, VIC 3168, Australia
e
Centre for Molecular and Nanoscale Physics, School of Applied Sciences, RMIT University, GPO Box 2476, Melbourne, Victoria 3001, Australia
f
ARC Centre of Excellence in Convergent Bio-Nano Science and Technology, Monash Institute of Pharmaceutical Sciences, Monash University (Parkville Campus), 381 Royal
Parade, Parkville, VIC 3052, Australia

a r t i c l e i n f o a b s t r a c t

Article history: During lipolysis of triglyceride by lipase, monoglyceride and fatty acids are produced which combine
Received 9 October 2014 with gastrointestinal fluids to form self-assembled structures. These solubilize hydrophobic food
Accepted 10 November 2014 components to promote their absorption. The aim of this study was a detailed understanding of structure
Available online xxxx
formation from triglyceride digestion products with saturated short-, medium- and long chain fatty acids.
Complementary characterization methods have been applied comprising small angle X-ray and neutron
Keywords: scattering – the latter involving the contrast matching technique using fully deuterated fatty acids –
Lipolysis
polarized and depolarized dynamic light scattering and cryogenic-transmission electron microscopy.
Lipid digestion
Lipid-based drug delivery
Shape, size and solubilization capacity of the self-assembled structures was dependent on composition
Synchrotron small angle Xray scattering and lipid chain length. Crystallization of fatty acid was observed when the solubility limit in the mixed
Small angle neutron scattering bile salt micelles was exceeded; however, increasing pH and temperature increased the fatty acid solu-
Cryogenic transmission electron microscopy bility. The results provide insight into structure formation and crystallization of incorporated lipolysis
products; this is important for a detailed understanding of food structure and nutrition, as well as the
rational design of lipid based drug delivery systems.
Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction for metabolic diseases such as atherosclerosis, type 2 diabetes

and obesity [3]. Fatty acids can also affect the development and
Triglyceride lipids are responsible for 30% of the daily calorie progression of cancer and cardiovascular disease [4–6].
intake in the Western diet [1,2]. The digestion of triglycerides by In the small intestine, fatty acid and monoglyceride combine
lipases in the gastro-intestinal tract leads to generation of fatty with endogenous amphiphilic molecules secreted from the gall-
acids and monoglycerides. These digestion products are required bladder, such as bile salt, cholesterol and phospholipid. These mix-
for membrane synthesis, elements of cells and tissues, production tures spontaneously self-assemble into thermodynamically stable
of signaling compounds and a source of energy [1]. The disposition structures such as mixed micelles of varying size and shape, vesi-
of fatty acids may impact on lipid metabolism with implications cles and more complex liquid crystalline phases [7–9]. This struc-
ture formation has been confirmed in vivo in killifish [10] and
human duodenal aspirates following a lipid-rich meal [9,11–13].
Abbreviations: Cryo-TEM, cryogenic-transmission electron microscopy; DLS,
dynamic light scattering; DDLS, depolarized dynamic light scattering; SANS, small The phase behavior has been studied in systems containing combi-
angle neutron scattering; SAXS, small angle X-ray scattering; BS, bile salt; MG, nations of bile salt, monoglyceride and cholesterol, where it has
monoglyceride; FA, fatty acid. been shown that colloidal structure and size depends on composi-
⇑ Corresponding authors at: Monash Institute of Pharmaceutical Sciences,
tion [14–17]. Bile salt micelles have been shown to elongate with
Monash University (Parkville Campus), 381 Royal Parade, Parkville, VIC 3052,
Australia. Fax: +61 3 99039583.
increasing proportions of phospholipid, finally resulting in vesicles
E-mail addresses: [email protected] (S. Salentinig), ben.boyd@ [18–20]. Bile salt micelles have also been found to increase in size
monash.edu (B.J. Boyd). with addition of monoglyceride and fatty acid [7].

http://dx.doi.org/10.1016/j.jcis.2014.11.026
0021-9797/Ó 2014 Elsevier Inc. All rights reserved.

Please cite this article in press as: S. Phan et al., J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.11.026
2 S. Phan et al. / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

Such structures facilitate transport of hydrophobic molecules in

the gastrointestinal tract and uptake through the enterocyte mem-
brane. They have high solubilizing capacity for hydrophobic mole-
cules including cholesterol, dietary lipids [21,22] and fat-soluble
vitamins A, D, E and K [23,24], and thus act as their carrier to pre-
vent precipitation and enhance bioavailability. They also have a
high solubilization capacity for hydrophobic drugs, which makes
them potential vehicles for drug delivery [7,25]. Co-administration
of drugs with lipids has been proposed to reduce precipitation of
the former during lipid digestion and increase bioavailability, ren-
dering them of interest in the drug delivery field [26,27].
Many saturated triglycerides are present in the Western diet
[28], and the fatty acids produced on their digestion in the gut have
melting points higher than physiological temperature e.g. C12, C14,
C16 and C18 [29]. They are protonated at physiological pH, which Fig. 1. Variation in pH in the gastrointestinal tract. Image adapted with permission
can result in crystallization of lipid and excretion in the faeces, from [37].
rather than absorption [30,31]. The melting points of C8 and C10
fatty acids are lower than physiological temperature enabling them
to be better solubilized and absorbed than long chain fatty acids. 2.2. Deuteration method
However there is a lack of understanding of the influence of struc-
ture formation on crystallization of saturated fatty acids and mono- The deuteration of the three fatty acids were achieved following
glycerides in biologically relevant multi-component systems a procedure published elsewhere [36]. In summary, a mixture of
containing bile salt, phospholipid, monoglyceride and fatty acid the appropriate acid, Pt/activated carbon and NaOD in D2O was
under physiological pH conditions. Solubilization of lipid digestion subjected to two hydrothermal H/D exchange cycles in a Parr pres-
products such as fatty acids is required for uptake and use by the sure reactor at 220 °C (23 bar) for 3 days each. Thin layer chroma-
body; thus crystallization is undesirable. In this study, the shape tography was used (referenced with the protonated compound) to
and size of the resulting colloidal structures and the crystallization develop separation protocols. Purification of the deuterated
of saturated fatty acids in biologically relevant mixtures have been compounds was performed on silica gel columns, eluted with
studied using scattering methods and electron microscopy. the appropriate solvents. 1H NMR (400 MHz), and 2H NMR
The pH in the gastrointestinal tract has been shown to vary (61.4 MHz) spectra were recorded on a Bruker 400 MHz spectrom-
between 2 and 8 [32] (Fig. 1), which will influence its partition eter at 298 K. More details on the characterization are presented in
coefficient between the aqueous bulk phase and the lipid areas of the Supporting information. The 1H NMR and 2H NMR spectra, as
the colloidal structure, and potentially induce crystallization of well as the mass spectrometry data are shown in Figs. SI-1–SI-6.
the fatty acid. In addition, pH has an impact on the protonation
state of the fatty acid which can modify the packing geometry 2.3. Sample preparation
based on the effective charge per molecule at the interface
[33–36], thus influencing the size and shape of structures. Fed simulated intestinal fluid was prepared using bile salt (BS,
To the best of our knowledge, this is the first study investigating sodium taurodeoxycholate) and phospholipid (PL, DOPC) concen-
the disposition and crystallization of lipolysis products in mixed trations at a ratio of 20 mM:5 mM to represent concentrations rel-
micellar structures in biologically relevant conditions of the evant after food intake, in digestion buffer (50 mM Tris maleate,
human intestine. An understanding of how intermolecular assem- 150 mM NaCl, 6 mM NaN3 as an antibacterial agent) [38–41]. The
bly of digestion products is influenced by composition, lipid chain PL was weighed into a round bottom flask and dissolved in chloro-
length, pH and temperature will provide insight into the digestion form which was subsequently evaporated to produce a thin film of
process, and further the understanding for the delivery of hydro- PL. BS and digestion buffer was added, and the PL and BS were dis-
phobic molecules, including drugs, to the body. persed in a sonicator bath for 30 min. The equilibrium systems
reflecting the endpoint of the lipid digestion process were prepared
by adding 10 mM monoglycerides (C8, C12 or C14) and 20 mM deu-
2. Experimental methods terated fatty acids (C8, C12 or C14) [7,42–44] to high levels of bile
salt/phospholipid (BS/PL) micelles expected in the fed state simu-
2.1. Materials lating the gastrointestinal state at the nominal end point of diges-
tion of triglycerides (C8, C12 or C14). The 1:2 ratio of monoglyceride
Deuterated caprylic acid (C8 fatty acid, C8HD15O2), lauric acid to fatty acid represents the nominal end point of digestion of the
(C12 fatty acid, C12HD23O2), and myristic acid (C14 fatty acid corresponding triglyceride, and thus is most commonly used. How-
C14HD27O2), were prepared as described below. Monomyristin ever there is evidence that digestion can proceed past this, thus it
(C14 monoglyceride > 97%) was obtained from TCI Co., Ltd. (Kawag- may not necessarily be entirely reflective of the end point of diges-
uchi, Saitama, Japan). Bile salt (sodium taurodeoxycholate > 95%), tion but some transient composition towards the end of digestion
monocaprylin (C8 monoglyceride 99%), monolaurin (C12 monoglyc- [9,43,45,46].
eride 99%), tris maleate (reagent grade), NaOH (p.a. grade) and HCl Sonication with a tip sonicator for 30 s at 100 W was used to
(p.a. grade) were purchased from Sigma Aldrich (St. Louis, MO, solubilize the components. Samples were equilibrated for 24 h
USA). Phospholipid (1,2-dioleoyl-sn-glycero-3-phosphocholine, and pH was adjusted with concentrated NaOH and HCl.
DOPC) was from Trapeze Association Pty. Ltd., (Clayton, Victoria,
Australia). Sodium chloride was purchased from Chem Supply 2.4. Small angle neutron scattering (SANS)
(Gillman, SA, Australia). Sodium azide was purchased from Merck
Schuchardt OHG (Eduard-Buchner-Straße, Hohenbrunn, Germany). SANS experiments were performed on the Quokka instrument
Ultra-pure water (resistivity > 18 MX cm) was used for the prepa- at OPAL [47] at a wavelength of 5.0 Å and 10% wavelength
ration of all samples. resolution. Two instrument configurations were used with a

Please cite this article in press as: S. Phan et al., J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.11.026
 S. Phan et al. / Journal of Colloid and Interface Science xxx (2014) xxx–xxx 3

source-to-sample distance of 8 m and sample-to-detector dis- of 1 min duration were performed, and the average of the data sets
tances of 8 m and 2 m, the latter with a 300 mm detector offset was taken for each angle. DDLS takes advantage of the fact that the
to increase maximum q. These configurations provide a continuous light scattered from optically homogeneous spherical particles
q range of 0.01–0.5 Å1 where q is the length of the scattering vec- maintains the same polarization as the incident light. However,
tor, defined by q = 4p/k sin (h/2), k is the wavelength and h the scat- optically anisotropic particles such as cylindrical micelles lead to
tering angle. All samples were enclosed in Hellma cuvettes with a depolarized scattering. The correlation function measured in DDLS
1 mm path length. Samples were studied at several temperatures is a combination of translational diffusion and rotational diffusion,
with control provided with a Julabo thermostatted bath. All data and the decay rate of the depolarized field correlation function, C,
were corrected for blocked beam measurements, normalized and is given by:
radially averaged using a package of macros in Igor software
(Wavemetrics, Lake Oswego, Oregon, USA) and modified to accept C ¼ q2 DT þ 6DR
HDF5 data files from Quokka. Scattering curves are plotted as a
function of absolute (SANS) intensity, I, versus q. C was determined from the correlation function by cumulant
A Guinier plot of the SANS data where ln (I) is plotted against q2 expansion [50].
was produced to determine the radius of gyration (Rg). This
provides information about the particle size and was obtained 2.7. Cryogenic-transmission electron microscopy (cryo-TEM)
from the slope of the Guinier plot, where the q range used was
qRg < 1.3 for a reliable fit [48]. Cryo-TEM was used to view the samples. A laboratory-built
humidity-controlled vitrification system was used to prepare the
2.5. Small angle X-ray scattering (SAXS) samples. Humidity was kept close to 80% for all experiments, and
ambient temperature was 22 °C.
SAXS measurements were performed at the SAXS/WAXS beam- Copper grids (200-mesh) coated with perforated carbon film
line [49] at the Australian Synchrotron. An X-ray beam with a (Lacey carbon film: ProSciTech, Qld, Australia) were glow dis-
wavelength of 1.1271 Å (11 keV) was used. A sample to detector charged in nitrogen to render them hydrophilic. Aliquots (4 lL)
distance of 1651 mm covered the q-range 0.007 < q < 0.5 Å1. A of the sample were pipetted onto each grid prior to plunging. After
thermostatted metal heating block controlled by a Peltier system 30 s adsorption time the grid was blotted manually using What-
accurate to ±0.1 °C was used for measurements taken at 37 °C. man 541 filter paper, for approximately 2 s. Blotting time was opti-
The 2D SAXS patterns were acquired over 1 s using a Pilatus 1 M mized for each sample. The grid was then plunged into liquid
detector with active area 169  179 mm2 and with a pixel size of ethane cooled by liquid nitrogen. Frozen grids were stored in liquid
172 lm. 2D scattering patterns were integrated into the 1D scat- nitrogen until required.
tering function I(q) using the in-house developed software package The samples were examined using a Gatan 626 cryoholder
scatterBrain. Scattering curves are plotted as a function of relative (Gatan, Pleasanton, CA, USA) and Tecnai 12 Transmission Electron
X-ray intensity, I, versus q. Microscope (FEI, Eindhoven, The Netherlands) at an operating volt-
age of 120 kV. At all times low dose procedures were followed,
2.6. Dynamic light scattering (DLS) using an electron dose of 8–10 electrons/Å2 for all imaging. Images
were recorded using FEI Eagle 4 k  4 k CCD camera at magnifica-
DLS was used to determine the mean hydrodynamic radius (RH) tions ranging from 15,000 to 50,000. Images were recorded
and the size distribution width of the particles. DLS measurements using a Megaview III CCD camera and AnalySIS camera control
were performed on a Malvern Zetasizer NanoZS to determine the software (Olympus.).
hydrodynamic radius (RH) and particle size distribution. Samples Negative stain TEM was used to determine crystallinity. Carbon
were centrifuged for 5 min at 3500 rpm prior to the measurement coated 300-mesh copper grids were glow-discharged in nitrogen to
to remove any trace dust particles. A laser power of 4 mW was render the carbon film hydrophilic. A 4 lL aliquot of the sample
used at a back scattering angle of 173° at room temperature. The was pipetted onto each grid. After 30 s adsorption time the excess
average diffusion coefficient, D, was obtained by the cumulant was drawn off using Whatman 541 filter paper, followed by stain-
analysis of the correlation functions [50]. The mean hydrodynamic ing with 2% aqueous potassium phosphotungstate at pH 7.2, for
radius RH was deduced from the diffusion coefficient using the 10 s. Grids were air-dried until needed.
Stokes–Einstein equation:
kB T
RH ¼ 3. Results and discussion
6pgD
where kB is the Boltzmann constant, T the absolute temperature and Small angle neutron scattering employing fully deuterated fatty
g the viscosity of the solvent. The polydispersity index (PDI) of the acids to enable contrast variation, and small angle X-ray scattering
size distribution was determined from the second cumulant: together provide complementary approaches to elucidate structure
and map the location of the fatty acid in mixed micelles. Since
l
PDI ¼  22 neutrons scatter from the atomic core, the replacement of hydro-
C gen with deuterium, which has a much greater scattering length
where l2 is the second cumulant and C  the mean of the inverse density, strongly enhances the contrast in SANS. The scattering
decay time. contrast can be modified by the choice of solvent (i.e. D2O/H2O)
Depolarized dynamic light scattering (DDLS) was used to confirm or the selective deuteration of molecules. This study combines
the presence of anisotropic particles observed by SANS and SAXS. both contrast variation methods. Fully deuterated fatty acids in
DDLS measurements were performed on an ALV-5022F spectrome- combination with non-deuterated monoglyceride, bile salt and
ter, using a vertically polarized Helium Neon laser (wavelength phospholipid were used. Using D2O as a solvent resulted in scatter-
633 nm). The scattered light passed through a crossed polarizer ing arising primarily from the bile salt, phospholipid and mono-
which was carefully adjusted to achieve minimum scattered glyceride, whereas in H2O the scattering signal is primarily from
intensity. Correlation functions were collected at different scatter- the deuterated fatty acid. The X-ray and neutron scattering length
ing angles between 20° and 90°. At each angle, 10 measurements densities of the components are summarized in Table SI-1.

Please cite this article in press as: S. Phan et al., J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.11.026
4 S. Phan et al. / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

3.1. Effect of lipid chain length and pH on solubilization of lipids and Crystallinity was not present when the sample temperature was
structure of bile salt mixed micelles increased to 37 °C, where SAXS measurements did not reveal Bragg
peaks for any of the chain lengths studied in H2O (Fig. 2B). This
The size and shape, and presence of crystallized of components suggests that the fatty acid is molten and solubilized in the mixed
from mixed bile salt micelles, upon the incorporation of the diges- bile salt micelles at physiological pH and temperature, which is
tion products of C8, C12 and C14 triglycerides were studied. With required to enable absorption of lipid, although the melting points
increasing lipid chain length, the size of the micelles increased as of C12 and C14 fatty acid are higher, at 43.5 °C and 54.4 °C respec-
might be expected. tively [29]. The results suggest that the solvation of the fatty acids
While there was no evidence of crystallization in the C8 and C12 in the mixed bile salts micelles is sufficient to overcome the greater
system, crystallization was observed in the C14 system at pH 6.5, enthalpy of fusion for the longer chain saturated fatty acids.
representative of duodenal pH. The Bragg peaks at q = 0.15, 0.16, Cryo-TEM was used to visualize the structures in the C14 system
0.19 and 0.33 Å1 result from the presence of fatty acid/soap crys- in H2O at pH 6.5 (Fig. 4). The images show the presence of micelles
tals where the fatty acid is protonated (Fig. 2A). The peaks were and emulsified microemulsion droplets, the latter with dimensions
absent at higher pH values, as the fatty acid is fully deprotonated of a few hundred nanometers in diameter. Microemulsions have
and solubilized in the mixed bile salt micelles. been shown to occur in systems containing lipid, surfactant and
The positions of Bragg peaks are consistent with values for crys- water [51,52].
talline C14 fatty acid (Fig. 3A). The presence of crystalline material Coexisting plate-like structures with well-defined edges were
was confirmed with transmission electron microscopy in selected present, with dimensions from several hundreds of nanometers
area diffraction mode with the negative stain phosphotungstic acid to microns, indicative of crystalline lipid. It has previously been
(Fig. 3B). reported that the addition of long chain lipid digestion products
SAXS measurements and electron diffraction on the C14 samples to mixed bile salt micelle systems, leads to an increase in hydro-
prepared in D2O, provided consistent results. At intestinal pH, no phobicity and decrease in solubility, resulting in crystallization
peaks were evident, however at lower pH, a Bragg peak at [22].
q = 0.20 Å1 was present which persisted until the temperature These results can have important implications for the absorp-
was raised to 60 °C (Fig. SI-7). The crystallization behavior is attrib- tion of saturated fatty acids. Crystallization of fatty acid in the
uted to the protonated state of the C14 fatty acid at the low pH small intestine is undesirable and it has previously been shown
studied in these samples prepared in D2O compared to the samples that protonated fatty acids and long chain fatty acids (C12 and
prepared in H2O, and the differences in crystal polymorph may be longer) precipitate from intestinal contents and are subsequently
due to the solvent used. excreted from the body rather than being absorbed [30,31]. In a

Fig. 2. Increasing size was observed with increasing hydrocarbon chain length in C8, C12, C14 MG + dFA systems in H2O BS/PL micelles at varying pH values. Crystallinity was
present in the C14 system at pH 6.5. SAXS measurements were performed at (A) 27 °C and (B) 37 °C.

Fig. 3. (A) Bragg peaks in the C14 MG + dFA H2O system at pH 6.5 correlate to crystalline C14 fatty acid, as determined by SAXS measurements. (B) Diffraction from crystalline
material was confirmed with electron microscopy using the negative stain phosphotungstic acid.

Please cite this article in press as: S. Phan et al., J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.11.026
 S. Phan et al. / Journal of Colloid and Interface Science xxx (2014) xxx–xxx 5

Fig. 4. Cryo-TEM images of C14 MG + dFA system in H2O BS/PL micelles at pH 6.5 show (A) many plate-like structures with well-defined edges indicative of crystalline
material (B)–(D) large particles (indicated by arrows). Micelles were observed throughout the sample.

Fig. 5. A transition from elongated to spherical particles was observed in C12 MG + dFA systems in (A) D2O and (B) H2O BS/PL micelles, and C14 MG + dFA systems in (C) D2O
and (D) H2O BS/PL micelles. A decrease in the apparent size is observed with increasing pH. SANS measurements were performed at 27 °C.

Please cite this article in press as: S. Phan et al., J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.11.026
6 S. Phan et al. / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

similar manner, cholesterol-rich vesicles were shown to produce deprotonated and pack to form highly curved spherical micelles
cholesterol crystals which led to gallstone formation [20]. due to electrostatic repulsion and to minimize interactions of
hydrocarbon tails with polar solvent.
The radius of gyration (Rg) was calculated from the SANS curves
3.2. Localization of components of mixed bile salt micelles using SANS
and is presented in Fig. 6. The Rg for the mixed micelles decreased
with increasing pH at both contrasts (i.e. in D2O and H2O). The
SANS experiments in combination with the contrast matching
hydrodynamic radii (RH) from DLS measurements presented in
technique were used to study the location of components in the
the same figure confirm this trend.
mixed bile salt micelles.
Comparing the Rg values in D2O and H2O at low pH, larger
In all systems containing monoglyceride and fatty acid, an
values were found in D2O. In D2O, the scattering arises from bile
increase in the pH led to a decrease in the forward scattering in
salt/phospholipid micelles and monoglyceride, whereas in H2O, it
the D2O and H2O based systems (see Fig. 5).
arises primarily from the deuterated fatty acid. This disparity in
The SANS curves in Fig. 5 from the C12 MG + dFA systems show a
size at low pH therefore suggests that the bile salt, phospholipid
transition from elongated to spherical particles with decrease in
and monoglyceride are mainly located in a shell surrounding the
the apparent size with increasing pH. To confirm the presence of
core which is composed predominately of the deuterated fatty acid
elongated particles at low pH, DDLS measurements were per-
similar to the previous study in a shorter C8 chain system [36]. At
formed on the C12 system. At high pH the depolarized scattering
pH > 7, the disparity between the two values diminished, indicat-
was within the noise of the instrument, indicating that the parti-
ing an equal distribution of the fatty acid within the mixed
cles are spherical. However, at low pH, the depolarized signal
micelles, whereby the C12 fatty acid remains part of the mixed
was well above the noise, confirming the results from SANS that
micelles. This is in contrast to the previous finding on the C8 sys-
the particles are elongated.
tem where the fatty acid was redistributing from the micelles to
Changing the pH did not influence the size of bile salt/phospho-
the aqueous phase with increasing pH. This is consistent with
lipid micelles in the absence of fatty acid/monoglyceride (Fig. SI-8).
reports that medium chain lipids have reduced capacity to be
Thus the influence of pH on structure formation can be attributed
incorporated into, and swell, bile salt micelles [7].
to the ionization state of the incorporated fatty acid, and changes
The size and shape of BS/PL micelles was shown to vary with
in packing geometry of the fatty acids according to the critical
the addition of lipolysis products of C8, C12 and C14 triglycerides
packing parameter theory [33]. At low pH, the fatty acids are fully
under physiologically relevant pH conditions as summarized in
protonated and incorporation in the chain region of the existing
Fig. 7. With increasing lipid chain length, crystallization was
structures, with overall reduced mean curvature is preferred,
observed under low pH conditions, but at elevated pH, lipids are
favouring elongation. At elevated pH, the fatty acids are fully
solubilized in bile salt mixed micelles which may facilitate lipid
digestion and absorption by the body.

4. Conclusions

This study provides insight into likely structure formation

during lipid digestion, where lipolytic products are solubilized by
bile salts to form mixed micelles. These structures are important
to promote the absorption of dietary lipids and vitamins. Self-
assembly structure formation in bile salt/phospholipid micelles
with incorporated monoglyceride and fatty acid, was seen to
depend on composition, lipid chain length and pH; where increas-
ing lipid chain length and decreasing pH increased the apparent
size of the micelles. Crystallization of fatty acid in the C14 system
was observed, but this was dissipated at physiological temperature
and with increasing pH. This is important as varying colloidal
structures have different solubilization capacities for drug, and
Fig. 6. Rg and RH for C12 MG + dFA systems in D2O and H2O show a decrease in the thus understanding of structure in lipid systems is critical for
micellar size with increasing pH.
enabling rational design of lipid based formulations.

Fig. 7. Summary of phase behavior of BS/PL micelles with the addition of lipolysis products of C8, C12 and C14 triglycerides under physiologically relevant pH conditions.

Please cite this article in press as: S. Phan et al., J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.11.026
 S. Phan et al. / Journal of Colloid and Interface Science xxx (2014) xxx–xxx 7

Acknowledgments [18] K. Muller, Hepatology 4 (1984) S134.

[19] M.A. Long, E.W. Kaler, S.P. Lee, Biophys. J. 67 (1994) 1733.
[20] A.S. Luk, E.W. Kaler, S.P. Lee, Biochemistry 36 (1997) 5633.
The authors would like to thank Lynne Waddington for assis- [21] M.J. Monte, J.J.G. Marin, A. Antelo, J. Vazquez-Tato, World J. Gastroenterol. 15
tance with the cryo-TEM studies which were undertaken at CSIRO (2009) 804.
[22] M.C. Carey, D.M. Small, Am. J. Med. 49 (1970) 590.
Materials Science and Engineering, Victoria, Australia.
[23] A.F. Hofmann, B. Borgstrom, Fed. Proc. 21 (1962) 43.
This research was undertaken on the SAXS/WAXS beamline at [24] A.F. Hofmann, Biochem. J. 89 (1963) 57.
the Australian Synchrotron, Victoria. Funding is acknowledged [25] G.A. Kossena, W.N. Charman, B.J. Boyd, C.J.H. Porter, J. Pharm. Sci. 94 (2005)
481.
from the Australian Research Council under the Discovery Projects
[26] C.W. Pouton, Eur. J. Pharm. Sci. 29 (2006) 278.
scheme DP120104032. BJB is a recipient of an Australian Research [27] C.J.H. Porter, N.L. Trevaskis, W.N. Charman, Nat. Rev. Drug Discovery 6 (2007)
Council Future Fellowship (FT120100697). 231.
[28] J.B. German, C.J. Dillard, Am. J. Clin. Nutr. 80 (2004) 550.
[29] M. Bockisch, in: (Ed.)^(Eds.), AOCS Press, 1998.
Appendix A. Supplementary material [30] U. Bracco, Am. J. Clin. Nutr. 60 (1994) 1002S.
[31] D.M. Small, Annu. Rev. Nutr. 11 (1991) 413.
[32] J. Fallingborg, Dan. Med. Bull. 46 (1999) 183.
Supplementary data associated with this article can be found, in [33] J.N. Israelachvili, D.J. Mitchell, B.W. Ninham, J. Chem. Soc. Faraday Trans. 72
the online version, at http://dx.doi.org/10.1016/j.jcis.2014.11.026. (1976) 1525.
[34] S. Salentinig, L. Sagalowicz, O. Glatter, Langmuir 26 (2010) 11670.
[35] S. Salentinig, S. Phan, J. Khan, A. Hawley, B.J. Boyd, ACS Nano 7 (2013) 10904.
References
[36] S. Salentinig, S. Phan, T.A. Darwish, N. Kirby, B.J. Boyd, E.P. Gilbert, Langmuir 30
(2014) 7296.
[1] J.B. German, C.J. Dillard, Am. J. Clin. Nutr. 80 (2004) 550. [37] J. Aron-Wisnewsky, J. Dore, K. Clement, Nat. Rev. Gastroenterol. Hepatol. 9
[2] L. Cordain, S.B. Eaton, A. Sebastian, N. Mann, S. Lindeberg, B.A. Watkins, J.H. (2012) 590.
O’Keefe, J. Brand-Miller, Am. J. Clin. Nutr. 81 (2005) 341. [38] K.W.Y. Lee, C.J.H. Porter, B.J. Boyd, J. Pharm. Sci. 102 (2013) 565.
[3] H.M. Cao, K. Gerhold, J.R. Mayers, M.M. Wiest, S.M. Watkins, G.S. Hotamisligil, [39] C.J.H. Porter, A.M. Kaukonen, A. Taillardat-Bertschinger, B.J. Boyd, J.M.
Cell 134 (2008) 933. O’Connor, G.A. Edwards, W.N. Charman, J. Pharm. Sci. 93 (2004) 1110.
[4] M. Garaulet, F. Perez-Llamas, M. Perez-Ayala, P. Martinez, F.S. de Medina, F.J. [40] L. Sek, C.J.H. Porter, W.N. Charman, J. Pharm. Biomed. Anal. 25 (2001) 651.
Tebar, S. Zamora, Am. J. Clin. Nutr. 74 (2001) 585. [41] A.M. Kaukonen, B.J. Boyd, C.J.H. Porter, W.N. Charman, Pharm. Res. 21 (2004)
[5] S. Karki, P. Chakrabarti, G.R. Huang, H. Wang, S.R. Farmer, K.V. Kandror, Plos 245.
One 6 (2011). [42] A.F. Hofmann, Adv. Chem. Ser. (1968) 53.
[6] P. Toren, B.C. Mora, V. Venkateswaran, Lipid Insights 6 (2013) 37. [43] K. Kleberg, F. Jacobsen, D.G. Fatouros, A. Mullertz, J. Pharm. Sci. 99 (2010) 3522.
[7] G.A. Kossena, B.J. Boyd, C.J.H. Porter, W.N. Charman, J. Pharm. Sci. 92 (2003) [44] D.G. Fatouros, I. Walrand, B. Bergenstahl, A. Mullertz, Pharm. Res. 26 (2009)
634. 361.
[8] J.S. Patton, M.C. Carey, Science 204 (1979) 145. [45] L. Kalantzi, E. Persson, B. Polentarutti, B. Abrahamsson, K. Goumas, J.
[9] J.E. Staggers, O. Hernell, R.J. Stafford, M.C. Carey, Biochemistry 29 (1990) 2028. Dressman, C. Reppas, Pharm. Res. 23 (2006) 1373.
[10] M.W. Rigler, R.E. Honkanen, J.S. Patton, J. Lipid Res. 27 (1986) 836. [46] E.M. Persson, A.S. Gustafsson, A.S. Carlsson, R.G. Nilsson, L. Knutson, P. Forsell,
[11] O. Hernell, J.E. Staggers, M.C. Carey, Biochemistry 29 (1990) 2041. G. Hanisch, H. Lennernas, B. Abrahamsson, Pharm. Res. 22 (2005) 2141.
[12] A.F. Hofmann, B. Borgstrom, J. Clin. Invest. 43 (1964) 247. [47] E.P. Gilbert, J.C. Schulz, T.J. Noakes, Phys. B – Condens. Matter 385–386 (2006)
[13] A. Mullertz, D.G. Fatouros, J.R. Smith, M. Vertzoni, C. Reppas, Mol. Pharm. 9 1180.
(2012) 237. [48] D.I. Svergun, M.H.J. Koch, Rep. Prog. Phys. 66 (2003) 1735.
[14] G.A. Kossena, W.N. Charman, B.J. Boyd, D.E. Dunstan, C.J.H. Porter, J. Pharm. Sci. [49] N.M. Kirby, S.T. Mudie, A.M. Hawley, D.J. Cookson, H.D.T. Mertens, N.
93 (2004) 332. Cowieson, V. Samardzic-Boban, J. Appl. Crystallogr. 46 (2013) 1670.
[15] J. Borne, T. Nylander, A. Khan, Langmuir 17 (2001) 7742. [50] D.E. Koppel, J. Chem. Phys 57 (1972) 4814.
[16] G.A. Kossena, W.N. Charman, B.J. Boyd, C.J.H. Porter, J. Controlled Release 99 [51] D.O. Shah, Micelles, Microemulsions, and Monolayers: Quarter Century
(2004) 217. Progress at the University of Florida, 1998, pp. 1.
[17] J. Gustafsson, T. Nylander, M. Almgren, H. Ljusberg-Wahren, J. Colloid Interface [52] H.N. Prajapati, D.M. Dalrymple, A.T.M. Serajuddin, Pharm. Res. 29 (2012) 285.
Sci. 211 (1999) 326.

Please cite this article in press as: S. Phan et al., J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.11.026