Monoacyl phosphatidylcholine inhibits the formation

PHASCI-03808; No of Pages 9 European Journal of Pharmaceutical Sciences xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

European Journal of Pharmaceutical Sciences journal homepage: www.elsevier.com/locate/ejps

Monoacyl phosphatidylcholine inhibits the formation of lipid multilamellar structures during in vitro lipolysis of self-emulsifying drug delivery systems Thuy Tran a, Scheyla D.V.S. Siqueira a, Heinz Amenitsch b, Thomas Rades a, Anette Müllertz a,c,⁎ a b c

Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark Institute for Inorganic Chemistry, Graz University of Technology, Stremayrgasse 9/V, A-8010 Graz, Austria Bioneer: FARMA, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark

a r t i c l e

i n f o

Article history: Received 22 June 2016 Received in revised form 21 November 2016 Accepted 22 November 2016 Available online xxxx Keywords: Monoacyl phosphatidylcholine Self-emulsifying drug delivery systems In vitro lipolysis Colloidal structures Small-angle X-ray scattering Cryogenic transmission electron microscopy

a b s t r a c t The colloidal structures formed during lipolysis of self-emulsifying drug delivery systems (SEDDS) might affect the solubilisation and possibly the absorption of drugs. The aim of the current study is to elucidate the structures formed during the in vitro lipolysis of four SEDDS containing medium-chain glycerides and caprylocaproyl polyoxyl-8 glycerides (Labrasol), with or without monoacyl phosphatidylcholine (MAPC). In situ synchrotron small-angle X-ray scattering (SAXS) was combined with ex situ cryogenic transmission electron microscopy (cryo-TEM) and dynamic light scattering (DLS) to elucidate the generated structures. The SAXS scattering curves obtained during the lipolysis of MAPC-free SEDDS containing 43–60% w/w Labrasol displayed a lamellar phase peak at q = 2.13 nm−1 that increased with Labrasol concentration, suggesting the presence of multilamellar structures (MLS) with a d-spacing of 2.95 nm. However, SEDDS containing 20–30% w/w MAPC did not form MLS during the lipolysis. The cryo-TEM and DLS studies showed that MAPC-free SEDDS formed coarse emulsions while MAPC-containing SEDDS formed nanoemulsions during the dispersion in digestion medium. From the first minute and during the entire lipolysis process, SEDDS both with and without MAPC generated uni-, bi-, and oligo-lamellar vesicles. The lipolysis kinetics in the first minutes of the four SEDDS correlated with an increased intensity of the SAXS curves and the rapid transformation from lipid droplets to vesicles observed by cryo-TEM. In conclusion, the study elucidates the structures formed during in vitro lipolysis of SEDDS and the inhibitory effect of MAPC on the formation of MLS. © 2016 Published by Elsevier B.V.

1. Introduction Poorly water-soluble drugs often have low and variable oral absorption as a consequence of their limited solubility and slow dissolution rate in the aqueous environment of the gastrointestinal tract. The successful use of lipid-based formulations (LBF) to enhance the absorption of these compounds has been previously described (Müllertz et al., 2010). Among the LBF, self-emulsifying drug delivery systems (SEDDS) are of special interest since they have been shown to efficiently improve the oral bioavailability of several poorly-water soluble drugs (Kovarik et al., 1994; Larsen et al., 2008; Nielsen et al., 2008). SEDDS often contain synthetic surfactants, which may lead to irritation of the gastrointestinal mucosa (Hauss, 2007). It is therefore important to reduce the amount of synthetic surfactants in SEDDS (while maintaining their self-emulsification capacity) by using natural surfactants. Monoacyl phosphatidylcholine (MAPC) (Table 1) (also known as lyso⁎ Corresponding author at: Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark. E-mail address: [email protected] (A. Müllertz).

phosphatidylcholine) is a natural surfactant present in the human intestinal tract as a digestion product of phospholipids (van Hoogevest and Wendel, 2014). MAPC enhanced the lymphatic transport of α-tocopherol from an oil-in-water emulsion in rats (Koo and Noh, 2001). Soybean MAPC has been used in medium-chain (MC) glycerides-based SEDDS to limit the concentration of synthetic surfactants and reduce the nanoemulsion droplet sizes formed when the formulations were dispersed in simulated gastric and intestinal fluids media (Tran et al., 2016). One of the mechanisms by which LBF enhance drug solubilisation and oral absorption is stimulating the secretion of digestive juices thus increasing the concentration of bile salts, phospholipids and cholesterol in the intestinal lumen (Kossena et al., 2007). These endogenous substances combine with lipids and lipid digestion products (such as fatty acids and monoglycerides) to form complex colloidal structures (such as mixed-micelles, vesicles and liquid crystalline phases). These colloidal structures impact the distribution and solubilisation of drugs in the digestive environment (Kossena et al., 2004; Kossena et al., 2005). Consequently, the intermediate colloidal phases originating from lipid digestion play an important role in the solubilisation and possibly oral

http://dx.doi.org/10.1016/j.ejps.2016.11.022 0928-0987/© 2016 Published by Elsevier B.V.

Please cite this article as: Tran, T., et al., Monoacyl phosphatidylcholine inhibits the formation of lipid multilamellar structures during in vitro lipolysis of self-emulsifying..., European Journal of Pharmaceutical Sciences (2016), http://dx.doi.org/10.1016/j.ejps.2016.11.022

2

T. Tran et al. / European Journal of Pharmaceutical Sciences xxx (2016) xxx–xxx

Table 1 Principal components of the used excipients and their chemical structures. Excipient

Composition

Captex 300

TG of predominantly C8:0 (70%) and C10:0 FA.

Chemical structure of principal components

Capmul MCM EP MG (45–75%) and DG (20–50%) predominantly of C8:0 (50–90%) and C10:0 FA (10–50%).

Labrasol

PEG-8 monoester, diesters of C8:0/C10:0; MG, DG, TG of C8:0/C10:0 FA; Free PEG residue (Jannin et al., 2014)

Lipoid S LPC 80

MAPC (80%) and PC (13%) of C18:2 (48–56%), C16:0 (20–27%), C18:0 (5–8%), C18:1 (7–9%) FA.

Abbreviations: DG: diacylglycerides; FA: fatty acid; MAPC: monoacyl phosphatidylcholine; MG: monoacylglycerides; PC: phosphatidylcholine; PEG: polyethylene glycol; TG: triacylglycerides.

absorption of drugs. Thus, a thorough investigation into the colloidal systems formed during digestion is believed to be required for a rational lipid-based formulation design. Considering the role of lipolysis products in drug solubilisation, simple dissolution or dispersion tests are often inadequate to evaluate LBF performance (Larsen et al., 2011). In vitro lipolysis models have therefore been developed, simulating the conditions of lipid digestion in the human intestine, to study the generated colloidal structures and the drug solubilisation and partitioning between the different phases (Christensen et al., 2004). Fatouros et al. (2007a, 2007b) have used cryogenic transmission electron microscopy (cryo-TEM) and smallangle X-ray scattering (SAXS) to study the structural evolution of SEDDS containing sesame oil, Maisine 35-1, Cremophor RH40 and ethanol (30:30:30:10% w/w) during the in vitro lipolysis. In these studies, spherical and elongated unilamellar vesicles were observed under the cryo-TEM and lamellar and inverse hexagonal phases were identified by the SAXS (Fatouros et al., 2007a, 2007b). Even though these studies have shown the likely presence of nanostructures during lipolysis, the experiments were conducted ex situ. Consequently, it is difficult to draw definitive conclusions about these findings, especially because lipid digestion is a dynamic process with continuous evolution of the colloidal phases. In situ SAXS was used by Salentinig et al. (2011) to study the colloidal structures formed during the digestion of triolein, which transformed from oil droplets to micellar cubic, inverse hexagonal, and bicontinuous cubic liquid crystalline phases during digestion (Salentinig et al., 2011). Subsequently, synchrotron SAXS was applied to lipolysis studies of SEDDS due to its shorter acquisition time, allowing real-time monitoring of structural evolution (Phan et al., 2013, 2015; Warren et al., 2011). Considering the above-mentioned potential use of MAPC in SEDDS and the importance of the nature of the colloidal systems during lipolysis, the aim of the present work is to elucidate the structures formed during the lipolysis of SEDDS both with and without MAPC. For that, in situ synchrotron SAXS coupled to the in vitro lipolysis model and combined with ex situ cryo-TEM and dynamic light scattering (DLS) were used.

containing 98.0% PC) were kindly donated by Lipoid GmbH (Ludwigshafen am Rhein, Germany). Captex 300 (Captex) (glyceryl tricaprylate/tricaprate) and Capmul MCM EP (Capmul) (glyceryl monocaprylate) were obtained from Abitec (Columbus, OH, USA). Labrasol (PEG-8 caprylocaproyl glycerides) was kindly provided by Gattefossé (Saint-Priest, France). Pancreatin from porcine pancreas (8 × USP specifications), sodium taurodeoxycholate (NaTDC) hydrate (N95% pure), tris(hydroxymethyl)aminomethane (Tris), maleic acid and 4-bromophenylboronic acid (4-BPB) (≥95.0% pure) were purchased from Sigma-Aldrich (St Louis, MO, USA). Sodium chloride was purchased from VWR (Radnor, PA, USA). Sodium hydroxide pellets were obtained from Merck (Darmstadt, Germany). Water was purified by an SG Ultraclear water system (SG Water GmbH, Barsbüttel, Germany). 2.2. Methods 2.2.1. Preparation of formulations The chemical structures and the amount of excipients used in each investigated SEDDS are shown in Tables 1 and 2. The two MAPC-containing SEDDS are designated F20 and F30 (the number corresponds to the percentage of LPC in the SEDDS). The two MAPC-free SEDDS are designated F0 and F30*; F0 contains a Labrasol amount equal to the total amount of Labrasol and LPC in F20 or F30, whereas F30* contains the same glycerides:Labrasol ratio as F30. Capmul was melted at 50 °C and homogenized prior to use. All formulations were prepared by weighing and mixing all components in a glass vial for 4 h at 45 °C. The formulations were allowed to equilibrate at room temperature overnight before further experiments to ensure excipient homogeneity. 2.2.2. In vitro lipolysis model The experimental set-up consisted of a pH-stat apparatus (Metrohm AG, Herisau, Switzerland), containing a Titrando 842, an 804 Ti Stand, an Table 2 Composition of the SEDDS used in the in vitro lipolysis experiments.

2. Materials and methods

Formulation

2.1. Materials

F0 F20 F30 F30*

Lipoid S LPC 80 (LPC) (from soybean, containing 80.8% MAPC and 13.2% phosphatidylcholine (PC)) and Lipoid S PC (from soybean,

Excipients (g) Captex 300

Capmul MCM EP

Labrasol

Lipoid S LPC 80

TOTAL

0.2 0.2 0.2 0.2

0.2 0.2 0.2 0.2

0.6 0.4 0.3 0.3

– 0.2 0.3 –

1.0 1.0 1.0 0.7



802 stirrer, a glass pH electrode (iUnitrode), an 800 Dosino dosing unit and a 10 mL autoburettes. The apparatus was operated using the Tiamo 2.0 software (Metrohm AG, Herisau, Switzerland). The in vitro lipolysis protocol was based on the one described by Tran et al. (2016) with modifications in digestion medium volume and enzyme activity to fit the sensitivity of the SAXS measurements. Briefly, before adding lipase extract, each SEDDS was dispersed in 27 mL digestion medium (2.95 mM NaTDC, 0.26 mM PC, 2.0 mM Tris, 2.0 mM maleic acid, 73.0 mM sodium chloride, 1.4 mM calcium chloride, adjusted to pH 6.5) for 5 min in a thermostated vessel maintained at 37 °C. The amount of formulations in each experiment was 1.0 g for F0, F20, and F30, and 0.7 g for F30* (0.7 g of F30* contain the same amount of Capmul, Captex, and Labrasol as 1.0 g of F30) (Table 2). The pH of the dispersion was adjusted to 6.5. The addition of 3 mL fresh pancreatic lipase extract (pH 6.5) to obtain an enzyme concentration of 660 USP units/mL initiated the lipolysis reaction. During the lipolysis, the pH was maintained at 6.5 for 60 min by adding 0.6 mM NaOH solution. Independent lipolysis experiments were performed to determine the lipolysis kinetics and the formed structures monitored by in situ SAXS, cryo-TEM and DLS (described below). 2.2.3. In situ SAXS In situ SAXS measurements were performed at the Austrian SAXS beamline of the ELETTRA synchrotron radiation facility (Trieste, Italy) using a photon energy of 8 keV and a sample-to-detector distance of 995.7 mm (Amenitsch et al., 1998). The sample-to-detector distances have been calibrated with silver behenate (d-spacing = 5.8376 nm) (Amenitsch et al., 1998). The scheme of the experimental setup is presented in Fig. 1. Using a peristaltic pump, the samples were pumped from the digestion vessel through a quartz capillary for real-time SAXS monitoring during the in vitro lipolysis experiments and then returned to the digestion vessel (Ågren et al., 1999). The 2D SAXS patterns were collected by a Pilatus3 1M detector (Dectris, Baden-Dättwil, Switzerland) and integrated in the azimuthal direction. All SAXS data presented in the Results and discussion section have been normalized and subsequently, background subtracted (SAXS profile of digestion buffer), using Igor Pro software version 6.3.6.4 (Wavemetrics Inc., Lake Oswego, OR, USA). The d-spacing, or the lattice parameter, of the lamellar phase was calculated using the equation d = 2π/q where q was the location of the first lamellar phase peak. The detected lamellar phase

3

peaks were fitted to Lorentz functions using the Igor Pro software to estimate the peak intensity. 2.2.4. Cryo-TEM studies Samples were taken before the addition of enzyme extract, and at 1 and 60 min of lipolysis and treated by adding lipase inhibitor (5 μL of 1.0 M 4-BPB (in methanol) per mL of sample). The samples were vitrified by a Vitrobot automated vitrification device (FEI, Eindhoven, The Netherlands) prior to cryo-TEM observation. Three microliter of each sample was applied on discharged Lacy formvar/silicon monoxide film grids (Ted Pella Inc., CA, US). The grids were blotted in the vitrification device under controlled environmental conditions (4 °C, 100% relative humidity), then automatically plunged into liquid ethane to rapidly freeze the samples and transferred to liquid nitrogen (approximately − 174 °C). The frozen samples were then transferred to a Gatan 626 cryoholder (Gatan Inc., Warrendale, PA, USA) coupled to a FEI Tecnai G2 transmission electron microscope under low-dose conditions at −176 °C (FEI, Eindhoven, The Netherlands). Images were recorded by a FEI Eager 4k CCD camera (FEI, Eindhoven, The Netherlands). 2.2.5. Particle size measurements using DLS Samples of the dispersed and digested formulations were measured by DLS (37 °C, 173° backscattering angle, 0,686 cP sample viscosity) using a Zetasizer Nano ZS (Malvern, Worcestershire, UK). The digested samples were taken after 1, 30 and 60 min of digestion and immediately inhibited by adding 5 μL of 1.0 M 4-BPB (in methanol) per mL of sample. Subsequently, the digested samples were centrifuged for 4 min at 13,400 rpm (≈ 19,100 g) and 37 °C in a 5417R Eppendorf centrifuge (Hauppauge, NY, USA), and the supernatant was harvested for DLS measurements. For monodisperse samples, the mean particle size was reported as z-average value (i.e. mean particle size based on the signal intensity). For samples of high polydispersity (polydispersity index (PdI) N 0.5), size distribution by intensity was reported with the width and mean value of the largest particle size population. 3. Results and discussion In vitro lipolysis of the four SEDDS containing different amounts of Labrasol and MAPC was investigated using in situ SAXS combined with ex-situ cryo-TEM and DLS. The three techniques provide complementary

Fig. 1. Schematic representation of the in vitro lipolysis setup coupled to a SAXS flow-through cell.


4


approximately half of the total amounts of NaOH consumed after 60 min. Subsequently, a slow consumption of NaOH was observed in all lipolysis profiles because the fatty acids released from the initially fast digestion accumulate at the oil/water interface and inhibit the enzyme activity (Golding and Wooster, 2010). After 60 min of lipolysis, F0, F20, F30 and F30* consumed 2.05 ± 0.04, 1.74 ± 0.02, 1.56 ± 0.01, and 1.83 ± 0.01 mmol NaOH, respectively. In F20 and F30, replacing Labrasol by MAPC reduced the consumption of NaOH. When comparing the lipolysis rates of F30* and F30, it was suggested that adding MAPC to F30* inhibited the digestion of the formulation. The inhibitory effect of MAPC on the digestion of MC glycerides and Labrasol in SEDDS correlates with the result reported by Tran et al. (2016), where a different formulation concentration and enzyme activity were used, and Tsuzuki et al. (2004), where MAPC was shown to inhibit pancreatic lipase activity on long and medium-chain lipid emulsions. The total amount of NaOH consumed after 60 min lipolysis of the four SEDDS corresponded to the amount of digestible lipid substrates, i.e., MC glycerides and Labrasol.

Fig. 2. In vitro lipolysis profiles of F0, F20, F30 and F30* showing the amounts of titrant consumed over 60 min. Data are presented as mean ± SD (n = 3). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

information elucidating the influence of MAPC on the colloidal structures of the generated systems. The observed structural evolutions were matched to the corresponding amount of consumed titrant NaOH during the lipolysis to analyse the relationship between the lipolysis kinetics and the formation of colloidal species. 3.1. Lipolysis kinetics of the investigated SEDDS The lipolysis kinetics of the four SEDDS was determined by the in vitro lipolysis model and is shown in Fig. 2. The initial lipolysis of the four systems was rapid; after 4 min, all systems had consumed

3.2. Formation of multilamellar structures during the lipolysis of MAPC-free SEDDS monitored by in situ SAXS The effect of Labrasol concentration on the formation of colloidal structures was studied by comparing the structural evolution during the in vitro lipolysis of F0 and F30* using SAXS. The dispersion of F0 or F30* in digestion medium (1.0 or 0.7 g in 30 mL) resulted in a dispersion containing 1.3% w/v MC glycerides and 2.0 or 1.0% w/v Labrasol, respectively. The SAXS intensity versus scattering vector profiles during the lipolysis of F0 and F30* are presented in Fig. 3. The two systems formed vesicles within the first minute of digestion, indicated by the increased scattering intensity of the SAXS curves (Fig. 3A and B). Differences between the two digesting systems occurred after 10 min of lipolysis, when the lipolysis of F0 resulted in a single peak at scattering vector q = 2.13 nm−1, which corresponds to multilamellar structures (MLS) with a lattice parameter of 2.95 nm (Fig. 3A and C). This lattice parameter is in agreement with a study using in situ SAXS conducted by

Fig. 3. In situ SAXS profiles during 60 min of in vitro lipolysis of F0 (A) and F30* (B) with the SAXS profiles shifted vertically as a function of time (The black arrows indicate the course of the digestion). SAXS profiles at single time points of F0 (C) and F30* (D) during dispersion and digestion (i.e., at 1, 10, 15, 40 and 60 min of lipolysis).



Phan et al. (2015), who reported that digested C8 and C10 triglycerides formed MLS with lattice dimensions of 2.5 and 2.99 nm, respectively. The intensity of the lamellar phase peaks from the digested F0 sample increased as the lipolysis continued (Fig. 3A), showing that the formation of MLS did not reach a plateau after 60 min of lipolysis. More MLS might have been produced if the digestion had continued. The observed MLS were possibly generated by the continuous accumulation of the released digestion products on the previously formed lamellar structures such as uni-, bi- and oligo-lamellar vesicles (Fatouros and Müllertz, 2008). For F30*, the SAXS profiles started displaying a small peak also at q = 2.13 nm−1 after about 40 min of lipolysis, which corresponds to MLS with an identical lattice dimension as the MLS of F0 (Fig. 3C and D). However, the intensity of the lamellar phase peak of F30* was noticeably lower and increased more slowly than the intensity of the lamellar phase peak of F0. This comparison suggests that MLS were formed at a lower rate and concentration, but with identical lattice dimensions in SEDDS containing less Labrasol. On the one hand, the MLS from both F0 and F30* had consistent lattice dimension with increasing concentration over digestion time, possibly because this structure comprised similar molecules, which are digestion products of Labrasol and MC glycerides, i.e., ionized and unionized MC fatty acid, monoglycerides, and PEG-monoesters. Both MC monoglycerides and hydrated fatty acid: fatty acid soap (1:1) can form bilayers under high dilution conditions in water (Cistola et al., 1986; Larsson, 2007). On the other hand, as the intensity of the lamellar phase peak increased with the Labrasol concentration, one could expect that the digestion products only released from Labrasol, i.e., PEG-monoesters and PEG, may play an important role in MLS formation. The contribution of PEG-monoesters may be explained by their packing parameters with large hydrophilic head-group areas, which modify the average critical packing parameter of the digesting systems to a favourable value for MLS formation (Israelachvili, 2011). The SAXS peak intensity recorded at q = 2.13 nm−1 and the corresponding Lorentz curves fitted by the Igor Pro software are present in Fig. 4A and B. The height of the fitted Lorentz curves, representing the lamellar phase peak intensity, was plotted together with the recorded lipolysis kinetics to study the relationship between the amounts of digestion products and the structural evolution over time (Fig. 4C). The early fast stage of lipolysis kinetics was in agreement with the fast structural evolution of the two systems observed from the SAXS data, which suggests a rapid formation of vesicles. Subsequently, a slow consumption of NaOH was observed, corresponding to the gradual evolution in the SAXS data with the consistent presence of vesicles and the increasing concentration of MLS. However, the NaOH consumption and the peak intensity of F0 and F30* increased at different rates (Fig. 4C). During the lipolysis, the lamellar phase peak occurred after about 10 min for

5

F0 and 40 min for F30*, which corresponds to different amounts of NaOH consumed (i.e., 1.32 mmol NaOH for F0 and 1.71 mmol NaOH for F30*). After 60 min, the lipolysis of F30* generated a low-intensity peak and consumed 1.83 ± 0.01 mmol NaOH, corresponding to the amount of NaOH consumed by F0 after 36 min with a higher peak intensity. The lamellar phase peak intensity of F30* after 60 min was as low as the peak after 13 min of F0 lipolysis. In general, the lamellar phase peak was lower in F30* compared to F0 when the two formulations consumed the same amount of NaOH. The absolute amount of titrant consumed, or the amount of fatty acids released and titrated, therefore, did not linearly reflect the formation and the concentration of MLS. This can be explained by the hypothesis that the titrated fatty acids were not the only component of the MLS; and PEG-glycerides, which were present at a lower concentration in F30*, were possibly responsible for the reduced presence of MLS in F30* compared to F0. 3.3. Effect of MAPC on lamellar phase formation during the lipolysis of MAPC-containing SEDDS monitored by in situ SAXS The lipolysis of F20 and F30 was studied to investigate the effect of adding MAPC to F30*, or gradually replacing a part of Labrasol in F0 by MAPC, on the structural evolution. MAPC concentration reached 0.53% w/v (~10.5 mM) or 0.8% w/v (~15.7 mM) when dispersing F20 or F30 in the digestion medium, respectively (Table 2). Similarly to the SAXS data obtained for F0 and F30*, a fast formation of vesicles occurred from 1 min of the lipolysis of F20 and F30, reflected in the changes of the SAXS curves (Fig. 5). The SAXS data showed a slight decrease of the q value for the hump in the vesicles scattering profile over time, suggesting an increase of vesicle sizes for both formulations. After 60 min of lipolysis, the lamellar phase peaks at q = 2.13 nm−1 observed for F0 and F30* were not present for F20 and F30 (Figs. 3 and 5). The absence of lamellar phase peaks in the scattering curves of F20 and F30 suggested that MAPC prevented the formation of MLS from the digestion products. The majority of MAPC probably stayed intact during the lipolysis because MAPC was found to be poorly digested by porcine pancreatin (Tran et al., 2016). Based on the theory that the critical packing parameter determines the colloidal structures of a lipid system, the presence of MAPC could change the mean critical packing parameter of the digesting SEDDS to an unfavourable value for MLS formation. It is important to note that MAPC is an amphiphilic lipid with a large hydrophilic head group, forming large zwitterionic micelles in aqueous media (Small, 1968). When incorporated into unilamellar lipid vesicles, MAPC promotes curvature change and vesicle fission; as a result, small unilamellar vesicles bud off from the larger vesicles formed initially (Tanaka et al., 2004). The reason for the inhibition effect of MAPC on MLS formation might be: (1) MAPC generated micelles or vesicles

Fig. 4. Analysis of the peak intensity at q = 2.13 nm−1 from the SAXS data and lipolysis kinetics over time during lipolysis of F0 and F30* using curve fitting. Fitted curves (in black in (A), and in red in (B)) for lamellar phase peaks of F0 (A) and F30* (B) (the SAXS intensity profiles were shifted vertically as a function of time). Lipolysis profile of F0 (black line) and F30* (red line); and evolution of the lamellar phase peak intensity (q = 2.13 nm−1) of F0 (black dots) and F30* (red dots) during 60 min of lipolysis (C) (from 0 to 40 min, no peak was observed for F30*). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)


6


Fig. 5. In situ SAXS profiles during 60 min of in vitro lipolysis of F20 and F30. Evolution of SAXS profiles of F20 (A) and F30 (B) with the SAXS profiles shifted vertically as a function of time (the black arrows indicate the course of digestion time). SAXS profiles at single time points of F20 (C) and F30 (D) during dispersion and lipolysis (i.e., at 1, 30, and 60 min of lipolysis).

with PEG-glycerides from Labrasol and therefore reduced the level of PEG-glycerides available for forming MLS; or (2) MAPC molecules, with a long fatty acid chain, disturbed the ordered structure of MLS generated by MC fatty acids. In terms of correlation between structural evolution and lipolysis kinetics, the fast initial lipolytic reaction correlated well with the abrupt changes of SAXS scattering curves at 1 min. At 60 min of lipolysis, the total amounts of NaOH consumed for F20 and F30 were 1.74 ± 0.02 and 1.56 ± 0.01 mmol, respectively. While these amounts of NaOH correspond to no lamellar phase peak observed for F20 and F30, they equal the amounts of NaOH consumed at 30 and 19 min of lipolysis with noticeable lamellar phase peaks for F0. This supports the suggestion in the previous section that the concentration of titrated fatty acid was not a straightforward indicator for the presence of a lamellar phase peak in the SAXS data. In the literature, a general agreement between digestion kinetics and increase of the lamellar phase peak was found during the digestion of a SEDDS containing Maisine 35-1, sesame oil, Cremophor RH40 and ethanol (30:30:30:10% w/w) (Warren et al., 2011). Phan et al. (2013, 2015) also reported a correlation between the formation of a lamellar phase peak and the titration profiles when studying the digestion of MC triglycerides. Complementary to the aforementioned studies, the present study demonstrates that the titration profile of the investigated SEDDS correlated well with the general structural evolution, but not with the kinetics of MLS formation. The influence of colloidal structure on the solubilisation and absorption of lipophilic drugs has previously been studied (Kossena et al., 2005; Pham et al., 2016). Lamellar systems consisting of C8, C12 or C18:1 fatty acid and monoglycerides provided higher solubilisation capacity for cinnarizine and stimulated paracellular transport of rat intestinal membranes better than micellar/vesicular systems consisting of the same components (Kossena et al., 2005). In the study of Kossena et al., however, it was difficult to identify if the difference in solubilisation

capacity was attributed to the lipid compositions or the structures of the phases. Moreover, the dynamics of colloidal species may affect drug absorption. For example, the formation of rigid cubic phases from a mixture of phytantriol and tributyrin during digestion resulted in prolonged gastric retention and absorption of cinnarizine compared to a micellar solution of digested tributyrin (Pham et al., 2016). Although the current study focuses on the colloidal structure development during the digestion of placebo formulations, it is important to investigate the effect of lamellar phases using MAPC on the solubilisation and absorption of drug in future studies. 3.4. Characterisation of the nanostructures formed before and during digestion by cryo-TEM and DLS The cryo-TEM images of the dispersed F0 and F30 before the lipolysis and after 1 and 60 min of lipolysis are presented in Figs. 6 and 7, respectively. In total, 93 cryo-TEM images from F0 and 87 cryo-TEM images from F30 were evaluated. The dispersed systems from F0 in digestion medium were oil-inwater emulsions (Fig. 6A and B) with the oil droplets inadequately emulsified which did not disperse in the aqueous medium in the cryoTEM grid holes, but attached to the lipophilic grid walls before freezing. F0 dispersed to a coarse emulsion with droplet sizes in the micrometer range and unsuitable for DLS measurement. After 1 min of lipolysis, oil droplets were unobservable under the cryo-TEM, possibly due to attachment to the lipophilic grid, whereas bilayers generated from digested lipids were omnipresent (Fig. 6C-E). The distorted bilayers shown as rough lines in the images suggest ongoing lipolytic reactions. Small uni- or bilamellar vesicles of 50–200 nm co-existed with large, distorted and occasionally opened vesicles (indicated by dashed arrows). A wide distribution of the vesicle sizes was also suggested by the DLS data (PdI = 0.60). The DLS data showed that after 1 min of lipolysis of F0, the sizes of 70% of particles in the sample ranged from 190 to



7

Fig. 6. Cryo-TEM images of structures formed during dispersion (A, B), after 1 min (C-E) and 60 min of lipolysis (F-I) of F0. Black arrows indicate distorted bilayers; dashed arrows indicate opened vesicles; the black triangle indicates collagen structure from pancreatin; and the asterisk indicates an oligo-lamellar vesicle comprising 6 layers.

712 nm with a mean particle size of the main population of 417 nm. This size distribution agrees with the size of vesicles observed in cryo-TEM images (Fig. 6C-E). The evolution from coarse oil droplets to vesicles after 1 min of lipolysis was fast and in excellent agreement with the above-mentioned fast lipolysis kinetics and changes in the SAXS curves at the beginning of the lipolysis. This was also in line with the decrease of particle sizes after 1 min of lipolysis when more nano-sized particles were formed. After 60 min of lipolysis of F0, the majority of observed vesicles were uni- and bilamellar, occasionally elongated or opened (indicated by dashed arrows) (Fig. 6F-H). Oligo-lamellar vesicles were also present (indicated by asterisk) (Fig. 6I). DLS size distribution data suggest that the sample was polydisperse with PdI = 0.53 and 71% of the particles ranged from 91 to 615 nm with a mean particle size of the main population of 388 nm, which correlates with vesicle sizes observed in Fig. 6FH. A larger number of different types of vesicles were observed at 60 min compared to 1 min, possibly due to a higher level of digestion products present at 60 min. The level of distorted bilayers with rough surface and large, distorted vesicles decreased at 60 min of lipolysis suggesting that these structures were in an intermediate state between coarse oil droplets and nano-sized vesicles induced by lipolytic reactions. The diminished presence of these intermediate structures at 60 min correlates with the slower lipolysis kinetics

during later stages of the digestion (Figs. 2 and 6F-I). The long structure indicated by a black triangle in Fig. 6H was collagen from the pancreatin extract, which was observed in all digesting samples (Fatouros et al., 2007a). Remarkably, the MLS corresponding to the lamellar phase peak of the in situ SAXS curves could not be observed in the ex situ cryo-TEM study of F0. This disagreement in the results from these two methods might be explained by the different treatment and conditions that the samples experienced during ex situ cryo-TEM, such as the addition of lipolysis inhibitor solubilized in methanol and temperature variation before sample freezing. The small volume of sample employed for cryo-TEM analysis may not always represent the whole system; especially if the MLS are present in a low concentration. Due to these drawbacks of ex situ cryo-TEM, in situ SAXS seems to be a better technique for studying structures formed during the dynamic process of in vitro lipolysis. For both F20 and F30, the replacement of Labrasol by MAPC resulted in monodisperse nanoemulsions upon dispersion in the digestion medium. The nanoemulsions had z-average values of 300 ± 16 nm (PdI = 0.17) and 286 ± 3 nm (PdI = 0.23) for F20 and F30, respectively. In Fig. 7A and B, it can be seen that F30 emulsified better than F0 and dispersed in the grid cavity without attaching to the lipophilic grid walls before freezing. After 1 min of digestion, digested F20 was a polydisperse system with PdI = 0.58, while digested F30, with more MAPC


8


Fig. 7. Cryo-TEM images of structures formed during dispersion (A, B), after 1 min (C, D) and 60 min of in vitro lipolysis (E-I) of F30. The dashed arrow indicates opened vesicles.

incorporated, was a monodisperse system with PdI = 0.26 and z-average value decreased to 152 ± 5 nm. The particle size distribution of F30 measured by DLS was comparable to the small uni- and bilamellar vesicles observed in the cryo-TEM images (Fig. 7C and D). The different particle sizes between F30 and F20 at 1 min of lipolysis can be explained by the different amounts of MAPC incorporated. For F30, at 30 and 60 min of digestion, more fatty acids, monoglycerides and PEG-monoesters were released than at 1 min of digestion, which transformed the small vesicles to larger vesicles (Fig. 7F-I), and generated polydisperse systems (PdI = 0.60 (after 30 min) and 0.53 (after 60 min)). After 60 min of lipolysis of F30, the DLS size distribution showed that 70% of the particles have sizes ranging from 91 to 712 nm; the mean particle size of the largest population was 534 nm. Although the z-average could not be measured precisely by DLS due to the high polydispersity of the sample, the DLS size distribution correlates to the vesicle sizes observed in cryo-TEM images (Fig. 7E-I). The increase of vesicle sizes was also indicated by the SAXS data with decreasing q value for the hump in SAXS profiles of F30. In summary, at 60 min of lipolysis, F0 formed a system with uni- and bilamellar vesicles and oligo-lamellar vesicles of up to 6 layers. F30 formed a system with abundant large unilamellar vesicles, occasionally bilamellar and oligo-lamellar vesicles with equal or less than three layers. The structural properties of F0 and F30 before the lipolysis and after 60 min of lipolysis are summarized in Table 3.

Table 3 Morphological evolution of structures formed during dispersion and digestion of F0 and F30 based on in situ SAXS combined to ex situ cryo-TEM and DLS studies. The MLS in F0 at 60 min of lipolysis was found only with in situ SAXS.

+: occasionally present; ++: present; +++: abundantly present.



4. Conclusion This study elucidated the structural evolution of SEDDS during in vitro lipolysis in an attempt to better understand the effect of the natural surfactant MAPC on the formation of colloidal structures using in situ SAXS, ex situ cryo-TEM and ex situ DLS studies. Based on scattering curves, morphology and particle sizes, the structural evolutions were in agreement with the lipolysis kinetics when emulsion droplets rapidly developed to vesicles. The presence of MAPC reduced the size of emulsion droplets formed during dispersion, reduced the size of the self-assembled vesicles formed during in vitro lipolysis, and inhibited the formation of MLS. These effects might modify the solubilisation and oral absorption of drug co-administered with these SEDDS. It is therefore important to thoroughly investigate the influence of these colloidal structures on drug absorption in vivo. Funding sources The study was financially supported by the University of Copenhagen (Denmark), the Phospholipid Research Center (Heidelberg, Germany) and EU FP7 CALIPSO grant (no. 312284). Acknowledgments The SAXS study was performed at the Elettra synchrotron beamline facility (Trieste, Italy) with kind support from scientists and technicians of the facility. Financial support from the University of Copenhagen, the Phospholipid Research Center and EU FP7 CALIPSO grant (nr. 312284) is kindly acknowledged. We also thank Klaus Qvortrup and Michael Johnson from the Core Facility for Integrated Microscopy, University of Copenhagen for the support with cryo-TEM imaging. References Ågren, P., Lindén, M., Rosenholm, J.B., Schwarzenbacher, R., Kriechbaum, M., Amenitsch, H., Laggner, P., Blanchard, J., Schüth, F., 1999. Kinetics of cosurfactant-surfactant-silicate phase behavior. 1. Short-chain alcohols. J. Phys. Chem. B 103, 5943–5948. Amenitsch, H., Rappolt, M., Kriechbaum, M., Mio, H., Laggner, P., Bernstorff, S., 1998. First performance assessment of the small-angle X-ray scattering beamline at ELETTRA. J. Synchrotron Radiat. 5, 506–508. Christensen, J.Ø., Schultz, K., Mollgaard, B., Kristensen, H.G., Müllertz, A., 2004. Solubilisation of poorly water-soluble drugs during in vitro lipolysis of medium-and long-chain triacylglycerols. Eur. J. Pharm. Sci. 23, 287–296. Cistola, D.P., Atkinson, D., Hamilton, J.A., Small, D.M., 1986. Phase behavior and bilayer properties of fatty acids: hydrated 1: 1 acid-soaps. Biochemistry 25, 2804–2812. Fatouros, D.G., Müllertz, A., 2008. In vitro lipid digestion models in design of drug delivery systems for enhancing oral bioavailability. Expert Opin. Drug Metab. Toxicol. 4, 65–76. Fatouros, D.G., Bergenstahl, B., Müllertz, A., 2007a. Morphological observations on a lipidbased drug delivery system during in vitro digestion. Eur. J. Pharm. Sci. 31, 85–94. Fatouros, D.G., Deen, G.R., Arleth, L., Bergenstahl, B., Nielsen, F.S., Pedersen, J.S., Müllertz, A., 2007b. Structural development of self nano-emulsifying drug delivery systems (SNEDDS) during in vitro lipid digestion monitored by small-angle X-ray scattering. Pharm. Res. 24, 1844–1853. Golding, M., Wooster, T.J., 2010. The influence of emulsion structure and stability on lipid digestion. Curr. Opin. Colloid Interface Sci. 15, 90–101.

9

Hauss, D.J., 2007. Oral lipid-based formulations. Adv. Drug Deliv. Rev. 59, 667–676. Israelachvili, J.N., 2011. 20 - Soft and Biological Structures. In: Israelachvili, J.N. (Ed.), Intermolecular and Surface Forces, third ed. Academic Press, San Diego, pp. 535–576. Jannin, V., Rodier, J.D., Musakhanian, J., 2014. Polyoxylglycerides and glycerides: effects of manufacturing parameters on API stability, excipient functionality and processing. Int. J. Pharm. 466, 109–121. Koo, S.I., Noh, S.K., 2001. Phosphatidylcholine inhibits and lysophosphatidylcholine enhances the lymphatic absorption of α-tocopherol in adult rats. J. Nutr. 131, 717–722. Kossena, G.A., Charman, W.N., Boyd, B.J., Dunstan, D.E., Porter, C.J., 2004. Probing drug solubilization patterns in the gastrointestinal tract after administration of lipid-based delivery systems: a phase diagram approach. J. Pharm. Sci. 93, 332–348. Kossena, G.A., Charman, W.N., Boyd, B.J., Porter, C.J., 2005. Influence of the intermediate digestion phases of common formulation lipids on the absorption of a poorly water-soluble drug. J. Pharm. Sci. 94, 481–492. Kossena, G.A., Charman, W.N., Wilson, C.G., O'Mahony, B., Lindsay, B., Hempenstall, J.M., Davison, C.L., Crowley, P.J., Porter, C.J.H., 2007. Low dose lipid formulations: effects on gastric emptying and biliary secretion. Pharm. Res. 24, 2084–2096. Kovarik, J.M., Mueller, E.A., Van Bree, J.B., Tetzloff, W., Kutz, K., 1994. Reduced inter-and intraindividual variability in cyclosporine pharmacokinetics from a microemulsion formulation. J. Pharm. Sci. 83, 444–446. Larsen, A., Holm, R., Pedersen, M.L., Müllertz, A., 2008. Lipid-based formulations for danazol containing a digestible surfactant, Labrafil M2125CS: in vivo bioavailability and dynamic in vitro lipolysis. Pharm. Res. 25, 2769–2777. Larsen, A.T., Sassene, P., Müllertz, A., 2011. In vitro lipolysis models as a tool for the characterization of oral lipid and surfactant based drug delivery systems. Int. J. Pharm. 417, 245–255. Larsson, K., 2007. Kinetically stable aqueous lipid dispersions prepared by crystallization processes. J. Dispers. Sci. Technol. 28, 39–42. Müllertz, A., Ogbonna, A., Ren, S., Rades, T., 2010. New perspectives on lipid and surfactant based drug delivery systems for oral delivery of poorly soluble drugs. J. Pharm. Pharmacol. 62, 1622–1636. Nielsen, F.S., Petersen, K.B., Müllertz, A., 2008. Bioavailability of probucol from lipid and surfactant based formulations in minipigs: influence of droplet size and dietary state. Eur. J. Pharm. Biopharm. 69, 553–562. Pham, A.C., Hong, L., Montagnat, O., Nowell, C.J., Nguyen, T.-H., Boyd, B.J., 2016. In vivo formation of cubic phase in situ after oral administration of cubic phase precursor formulation provides long duration gastric retention and absorption for poorly watersoluble drugs. Mol. Pharm. 13, 280–286. Phan, S., Hawley, A., Mulet, X., Waddington, L., Prestidge, C.A., Boyd, B.J., 2013. Structural aspects of digestion of medium chain triglycerides studied in real time using sSAXS and cryo-TEM. Pharm. Res. 30, 3088–3100. Phan, S., Salentinig, S., Hawley, A., Boyd, B.J., 2015. How relevant are assembled equilibrium samples in understanding structure formation during lipid digestion? Eur. J. Pharm. Biopharm. 96, 117–124. Salentinig, S., Sagalowicz, L., Leser, M.E., Tedeschi, C., Glatter, O., 2011. Transitions in the internal structure of lipid droplets during fat digestion. Soft Matter 7, 650–661. Small, D.M., 1968. A classification of biologic lipids based upon their interaction in aqueous systems. J. Am. Oil Chem. Soc. 45, 108–119. Tanaka, T., Sano, R., Yamashita, Y., Yamazaki, M., 2004. Shape changes and vesicle fission of giant unilamellar vesicles of liquid-ordered phase membrane induced by lysophosphatidylcholine. Langmuir 20, 9526–9534. Tran, T., Xi, X., Rades, T., Müllertz, A., 2016. Formulation and characterization of selfnanoemulsifying drug delivery systems containing monoacyl phosphatidylcholine. Int. J. Pharm. 502 (1), 151–160. Tsuzuki, W., Ue, A., Nagao, A., Endo, M., Abe, M., 2004. Inhibitory effect of lysophosphatidylcholine on pancreatic lipase-mediated hydrolysis in lipid emulsion. Biochim. Biophys. Acta (BBA) – Mol. Cell Biol. Lipids 1684, 1–7. van Hoogevest, P., Wendel, A., 2014. The use of natural and synthetic phospholipids as pharmaceutical excipients. Eur. J. Lipid Sci. Technol. 116, 1088–1107. Warren, D.B., Anby, M.U., Hawley, A., Boyd, B.J., 2011. Real time evolution of liquid crystalline nanostructure during the digestion of formulation lipids using synchrotron small-angle X-ray scattering. Langmuir 27, 9528–9534.
