Porous nanostructure controls kinetics, disposition and self-assembly structure of lipid digestion products

Abstract

A combination of proton nuclear magnetic resonance (¹H NMR) and synchrotron small-angle X-ray scattering (sSAXS) was used to discriminate the speciation and structure evolution of lipolysis products for submicron lipid droplets and lipid loaded in porous silica particles. The free fatty acid (FFA)-to-glyceride ratio was controlled by confining medium-chain length triglycerides (MCT) in porous silica particles, which influenced the colloidal self-assembly structures formed within the lipolysis media. FFA and glycerides released during hydrolysis formed highly geometrically organised structures in a time-dependent manner. Structural transitions from coexisting emulsion droplets and micelles to lamellar structures, and finally to inverse hexagonal phase were observed during the digestion of lipid droplets. In contrast, when hosted in porous silica particles the digestion of lipids resulted in the self-assembly of a lamellar phase that was independent of digestion time. The evolution of structure during lipid digestion was dependent on lipolysis kinetics and the relative concentration of FFA to glycerides, which highlights important implications for the controlled delivery and absorption of lipophilic bioactives.

---

Introduction

Understanding the mechanism of lipase action and evolution of structure formed by lipid digestion products in the gastrointestinal tract (GIT) is of fundamental importance for the controlled delivery and uptake of lipophilic bioactive compounds.^1,2 Lipid digestion is initiated in the GIT by gastric and pancreatic lipases which hydrolyse ingested triglycerides (TG) to free fatty acids (FFA) and monoglycerides (MG) or diglycerides (DG).^3,4 Digestion products combine with bile salts and phospholipids to form mixed micelles and other colloidal structures, which further solubilise encapsulated poorly water-soluble compounds (e.g. vitamins and drugs) for enhanced absorption into the systemic bloodstream.^5,6 The self-assembled structure of the amphiphilic digestion products is an important factor in determining fat absorption and drug solubilisation capacity in the GIT,⁷ and thus the bioavailability of encapsulated components can be manipulated by altering the colloidal phases formed.⁸

Despite the importance of geometric organisation of lipolytic products on fat and drug absorption, there is little known concerning the interrelationship between key emulsion characteristics and lipid properties on the fate of digestion products. Phan et al.⁹ elucidated structures produced in real time during in vitro lipolysis of medium-chain length triglyceride (MCT) droplets using flow-through synchrotron small angle X-ray scattering (sSAXS) in conjunction with cryogenic transmission electron microscopy (cryo-TEM). Time-dependent structural transitions from micellar to lamellar phases were observed and correlated well with a discontinuity in the digestion profile, suggesting there was a link between the change in structure and the rate of lipolysis. This was further highlighted by the close correlation between the rate of evolution of the lamellar phase, that is the intensity of the first lamellar reflection, and the rate of reaction. Alternate studies have demonstrated that complex phase transitions, which are dependent on the relative concentration and nature of lipolytic products,¹⁰ impact the rate of lipolysis through the self-assembly of non-lamellar lipid liquid crystalline structures, including cubic and inverse hexagonal phases, at o/w interfaces.^11–13

Since, lipid digestion is an interfacial process that is influenced by the concentration and composition of TG at the oil-in-water (o/w) interface,^14–16 it may be possible to engineer materials that control colloidal structure through manipulating interfacial properties of the lipid, thus altering the lipolysis kinetics and nature of lipolysis products. Previously, we controlled lipase-mediated digestion kinetics by loading MCT in porous silica particles with distinctly different nanostructures.¹⁷ The rate and extent of lipid hydrolysis were enhanced in hydrophilic porous silica particles at submonolayer lipid coverages, compared to a submicron emulsion (Fig. 1), due to a combination of the following: (i) a 4-fold increase in interfacial surface area of lipid (160 m² g⁻¹ for porous silica particles compared to 35 m² g⁻¹ for the submicron emulsion), facilitating an increase in lipase adsorption to the o/w interface;^18,19 (ii) changes in orientation and conformation of lipid and lipase molecules on bare hydrophilic silica surfaces;^20,21 and (iii) the reduced interference of digestion products and lecithin (used as a stabiliser) on lipase for silica particles,^22,23 maintaining optimal lipase affinity for the lipid substrate.^24,25

Fig. 1 (A) Lipase-mediated digestion kinetics and (B) pseudo-first-order fit for a submicron o/w emulsion (blue ●) and lipid loaded porous silica particles (red ■), under fasted state conditions. Modified with permission from previous work.¹⁷ Copyright 2014 American Chemical Society.

Lipolysis kinetics were precisely described by a pseudo-first-order model for lipid loaded porous silica particles (Fig. 1(B)). In contrast, lipid digestion for the submicron emulsion was only described by a first-order model during the initial 4 min, resulting in the first-order rate constant, k₁, value of 0.31 min⁻¹ only describing the initial kinetics. This was due to the sharp decrease in reaction rate after ∼4 min, considered due to the interference effect of surface active digestion products (FFA, MG and DG) and lecithin on lipase activity.²⁶ Consequently, a biphasic pseudo-first-order model was required to precisely describe the lipolysis data of emulsion droplets,²⁷ with a derived secondary first-order rate constant (k₂) of 0.03 min⁻¹. This suggests a mechanism exists in porous silica particles that prevents digestion products interfering with lipase action. We hypothesised that the electrostatic repulsion between negatively charged FFA and the negatively charged silica surface reduced the affinity of the digestion products for the lipid-in-water interface, and thereby their competitive inhibition.²⁸ However, since lipolysis was monitored via NaOH titration, it was not possible to discriminate the speciation of lipolysis products.

In the current study, we relate the time-dependent nature and concentration of lipid digestion products to the formation of liquid crystalline structure in lipolysis media when porous silica particles were used as hosts for lipid molecules. The chemical nature of glycerides and FFA present in lipolytic media was elucidated by analysing specific signals of ¹H NMR spectra, which has been shown to be a powerful technique in evaluating the extent of hydrolysis for complex lipid mixtures.^29–34 Colloidal structures were characterized at different time points during digestion using sSAXS, providing further insights into the key parameters and material characteristics that influence geometric organisation of lipolytic products. This information can be used to develop strategies to control fat metabolism and optimise solubilisation and absorption of lipophilic bioactives in lipid-based delivery systems.

Experimental section

Materials

Hydrophilic porous silica particles, PS (Aerosil 380, surface area 380 m² g⁻¹) were supplied by Evonik (Essen, Germany). Medium-chain length triglycerides (MCT; Miglyol^® 812) were obtained from Hamilton Laboratories (Adelaide, Australia), hexane (AR Grade) and soybean lecithin from BDH Merck (Sydney, Australia). Materials used for the lipolysis study, including sodium taurodeoxycholate (NaTDC) 99%, trizma maleate, type X-E L-α-lecithin, porcine pancreatin extract, 4-bromophenyl boronic acid (4-BBA), calcium chloride dehydrate, dichloromethane and sodium hydroxide pellets, were purchased from Sigma-Aldrich (Australia). Deuterated chloroform (CDCl₃) was supplied by Cambridge Isotope Laboratories, Inc. All chemicals were of analytic grade and used as received. High purity (Milli-Q) water was used throughout the study.

Preparation and characterisation of lipid loaded silica particles

MCT was loaded into porous silica particles following the protocol established in previous work.¹⁷ The amount of lipid loaded in the porous silica particles was determined by thermogravimetric analysis (TGA). Particle sizing of silica particles, before and after lipid loading, was conducted via laser diffraction (Malvern Mastersizer). Submicron oil-in-water emulsions (0.6 w/w% soybean lecithin, 10 w/w% MCT) were homogenised under a pressure of 1000 bar for 5–6 cycles (Avestin EmulsiFlex-C5 Homogeniser). The size of the emulsion droplets were measured via dynamic light scattering (Zetasizer NanoZS, Malvern Instruments).

In vitro lipolysis studies

Fasted state mixed micelles and pancreatic extracts were prepared using the method previously described.³⁵ A known quantity of sample (∼200 mg lipid) was dispersed in 18 mL of buffered micellar solution by stirring continuously for 10 min in a thermostated glass reaction vessel (37 °C). Lipolysis was initiated by addition of 2 mL of pancreatin extract (∼2000 TBU). Aliquots were withdrawn periodically for analysis of digestion product composition and self-assembly structures and inhibited by the addition of 10 μL of 0.5 M 4-BBA. Aliquots taken from the digestion of lipid loaded silica particles were centrifuged to remove silica particles from the media. Pellets were discarded with no further analysis undertaken. The progress of lipid digestion was monitored for 30 min.

Proton nuclear magnetic resonance (¹H NMR) studies

¹H NMR sample preparation. Aliquots (2 mL) of in vitro lipolysis samples (at t = 0, 2, 5, 15 and 30 min) were subject to a liquid–liquid extraction, using dichloromethane in a proportion of 2 : 3 (v/v), according to previous studies.^31,33 Dichloromethane was selected due to its ability to extract lipids, its high volatility and its suitable polarity. The extraction was performed in triplicate to ensure the majority of lipid extracts partitioned to the organic phase. Dichloromethane was evaporated under reduced pressure, at room temperature to avoid lipid oxidation. The remaining lipid film was dissolved in CDCl₃ (800 μL) for ¹H NMR studies.

¹H NMR spectra acquisition. ¹H NMR spectra of digested lipids were obtained using a Bruker UltraShield™ 300. The acquisition parameters, selected based on previous work by Nieva-Echevarría et al.,³² were: spectral width 6410 Hz, relaxation delay 3 s, number of scans 64, acquisition time 4.819 s and pulse width 90°.

Quantification of lipolysis products and extent of lipid digestion from ¹H NMR spectral data. ¹H NMR signals were assigned to protons of specific lipid compounds (Table 1) in order to differentiate between molecules in the complex digestion media. Integration of the signal areas from the ¹H NMR spectra facilitated the relative molar concentrations of the various lipid components to be calculated. The signal areas in the spectra are proportional to the number of protons that generate them.³⁶ Consequently, the relative number of moles of the different molecules present can be calculated by the equations below, which were modified from Nieva-Echevarría et al.:³²where N is the relative number of moles for the corresponding compound, A is the area of the ¹H NMR spectra for the corresponding signals, labelled according to those in Table 1. The molar percentages of acyl groups supported on the different glyceride structures (TG, DG and MG), along with the molar percentage of FFA, were derived using the corresponding N values and eqn (5)–(8) below.where MG%, DG%, TG% and FFA% are the molar percentages of the different lipolytic molecules and N_total is total number of moles of digestion products:

N_total = 3N_TG + 2N_DG + N_MG + N_FFA (9)

Table 1 Chemical shift assignments of the ¹H NMR signals of the main protons present in the lipid mixtures formed during in vitro lipolysis. More detailed information can be found elsewhere.^29,31–33 The signal letters correlate to those in Fig. 2 and 3^a

Signal	Chemical shift (ppm)	Multiplicity	Compound(s)
a Abbreviations: d: doublet, t: triplet, m: multiplet.
(a)	0.88–0.89	t	Acyl groups and FA
(b)	1.19–1.42	m	Acyl groups and FA
(c)	1.61–1.72	m	Acyl groups and FA
(d)	2.26–2.38	m	Acyl groups and FA
(e)	2.37–2.44	m	Acyl groups in TG
(f1)	1.92–2.15	m	Acyl groups and FA
(f2)	2.77	t
(g)	3.65	d	Glyceryl group in 1-MG
(h)	3.73	d	Glyceryl group in 1,2-DG
(i)	3.84	d	Glyceryl group in 2-MG
(j)	4.10–4.18	dd	Glyceryl group in TG
	4.26–4.40	dd
(k)	4.93	m	Glyceryl group in 2-MG
(l)	5.08	m	Glyceryl group in 1,2-DG
(m)	5.13–5.28	m	Acyl groups and FA
(n)	5.24–5.35	m	Glyceryl group in TG

---

Synchrotron small-angle X-ray scattering (sSAXS) studies

sSAXS measurements were performed at the Australian Synchrotron. Lipolysis samples were placed in capillaries where they were inserted into a 37 ± 0.1 °C thermostated metal heating block controlled by a Peltier system. An X-ray beam of wavelength 1.1271 Å (11 keV) and a sample to detector distance of 1015 mm, covering a q-range of 0.014–0.65 Å⁻¹, were selected. The 2D SAXS patterns were collected using a Pilatus 1 M detector (active area 169 × 179 mm², pixel size 172 μm). Silver behenate was used for calibration. Data was normalised using ScatterBrain Analysis program. The scattering data was converted to a plot of intensity versus scattering vector, q, using the equation q = (4π/λ)sin θ/2. Lamellar and hexagonal space groups were determined by the relative positions of Bragg peaks in the scattering curves, which correspond to the reflections on planes defined by their (hkl) Miller indices.³⁷ Submicron emulsion droplets were characterised by their pronounced, broad correlation peak, which reflects the mean particle to particle distance.

Results and discussion

Lipase-mediated digestion of emulsion droplets versus lipid loaded porous silica particles

Quantification of lipolysis products generated during lipase-mediated digestion. Time-dependent ¹H NMR spectra of digested lipid samples from the lipolysis of submicron emulsion droplets (∼180 nm diameter) and medium-chain length triglycerides (MCT) loaded in porous silica particles (10 wt% lipid loading relative to silica; 0.43 equivalent monolayer surface coverage) are presented in Fig. 2 and 3, respectively. At t = 0 min for the submicron emulsion, ¹H NMR signals corresponding to protons supported by acyl groups (signals (a) to (e)) and those of the glyceryl backbone (signals (j) and (n)) of triglycerides (TG) were evident (Fig. 2). In contrast, peaks indicative of the glyceryl backbone of TG were mostly absent from the ¹H NMR spectra of the bulk digestion buffer for lipid loaded in porous silica particles (Fig. 3), highlighting that TG are encapsulated within the pores of the silica particles. The peaks observed in the spectra at t = 0 min, for the porous silica sample, such as the high intensity signal observed relating to the glyceryl backbone of 1,2-DG (signal (h)), correspond to the multitude of glyceridic compounds present in the mixed micellar solution. As digestion proceeded for MCT loaded in porous silica particles (i.e. t = 2 min), the intensity of peaks correlating to the TG glyceryl backbone increased slightly, suggesting a small portion of TG desorbed from the silica pores into the aqueous phase, prior to undergoing hydrolysis. This lead to a partial increase in the molar concentration of TG, coupled with a subsequent decrease in FFA concentration, in the bulk digestion media (Fig. 4).

Fig. 2 ¹H NMR spectra of lipid extracts from various time points during in vitro lipolysis of a submicron MCT emulsion. Inset: enlargements of key spectral regions. The assignment of signals correlate to those in Table 1.

Fig. 3 ¹H NMR spectra of lipid extracts from various time points during in vitro lipolysis of MCT loaded in porous silica particles. Inset: enlargements of key spectral regions. The assignment of signals correlate to those in Table 1.

Fig. 4 Relative molar percentages of various lipolytic products (TG, DG, MG and FFA) within the digestion media during in vitro lipolysis for (A) a submicron emulsion and (B) MCT loaded in porous silica particles.

The initiation of hydrolysis of lipid droplets was observed by a reduction in intensity, or the disappearance of peaks correlating to the glyceryl backbone of TG, and the formation of peaks relating to the glyceryl backbone of DG (signals (h) and (l)) and MG (signals (g), (i) and (k)). ¹H NMR signals indicative of 1,3-DG, which consist of multiplets between 4.05–4.21 ppm,^31,33 were absent from spectra over the course of digestion. Furthermore, 2-MG were the predominant isoform of MG present within the digestion media for the first 15 min of lipolysis. This is in agreeance with previous studies that have demonstrated the stereoselectivity of pancreatic lipase to hydrolyse the sn-1 and sn-3 positions of the glyceryl backbone^15,38 and further highlights the usefulness of ¹H NMR in analysing the stereoselectivity of lipases.³⁹ However, peaks relating to the glyceryl backbone of 1-MG (signal (g)) became apparent after 15 min. Åkesson et al.⁴⁰ showed that MG can undergo structural rearrangements through isomerisation at the sn-2 position to form 1-MG and 3-MG. Hence, it is hypothesised that the sn-1 isomers of MG are formed through isomerisation, and not due to the selective digestion of TG and DG at the sn-2 position.

A complex mixture of glycerides and free fatty acids (FFA) were released during the digestion of submicron lipid droplets, with a 76% reduction in the molar percentage of TG over the lipolysis period (Fig. 4(A)). In contrast, FFA were the predominant lipolytic product released during the digestion of lipid loaded in porous silica particles, whereby 100% FFA molar percentage was reached within 15 min of lipolysis (Fig. 4(B)). A spike in molar concentration of TG in the digestion buffer occurred at t ∼ 2 min due to TG desorption from the silica pores, but this was completely hydrolysed by t = 15 min. Signals indicative of DG and MG were absent from ¹H NMR spectra for the digestion of lipid loaded in porous silica particles. It was previously hypothesised that the accelerated rate of lipolysis for lipid encapsulated in porous silica particles was due to the reduced interference effect of digestion products on lipase action.¹⁷ Electrostatic repulsion between negatively-charged FFA molecules and the negatively-charged silica surface cause FFA to be expelled from the silica pores into the aqueous phase.²⁸ Thus, FFA desorb from the silica pores whereas MG and DG are likely to remain within the pores due to minimal electrostatic repulsion between the neutral molecules and the silica surface. Furthermore, loading lipophilic molecules in mesoporous silica is a common method employed to achieve supersaturation,^41–43 thus the solubility of glycerides are likely enhanced within the silica pores compared to the bulk aqueous solution. Since only the bulk aqueous phase of the digestion buffer was analysed for lipid digests, and not the pellet, peaks corresponding to glycerides were mostly absent from the lipid loaded silica sample.

Lipase-mediated digestion kinetics derived from lipolytic product evolution, using ¹H NMR, for a submicron emulsion and MCT loaded in porous silica particles correlated well with lipolysis observed using NaOH titration (Fig. 1). Hydrolysis of lipid droplets was well-characterised by a sharp decrease in the pseudo-first-order rate at t ∼4 min, which occurred simultaneously with a plateau in the molar concentrations of FFA and TG (Fig. 4). Furthermore, lipolysis titration indicated that 80.9% of TG were digested after 30 min for the submicron emulsion, which is consistent with changes in ¹H NMR spectra that showed a 76.7% decrease in TG molar concentration over 30 min. Thus, ¹H NMR not only discriminates between the concentrations and chemical nature of lipid digestion products, but also provides an accurate indication of the rate and extent of lipolysis.^31,33

Structural changes of lipolysis products during lipase-mediated digestion

Time-dependent X-ray scattering data for the digestion of a submicron emulsion and lipid loaded porous silica particles indicated clear structural differences in the self-assembly of colloidal phases (Fig. 5). Prior to digestion (t = 0 min) two broad maxima in the sSAXS demonstrated the presence of o/w droplets co-existing with bile salt micelles for the submicron emulsion. The initiation of digestion by lipase addition lead to the formation of FFA, MG and DG (Fig. 4). To begin with, micelles were not saturated with lipolysis products until ∼t = 6–8 min, whereby peaks indicative of a lamellar (L_α) phase, with q values of X and Y Å⁻¹ at ratios of 1 and 2 indicating a repeat distance of 30 Å, were evident. Thus, as digestion proceeded, a critical FFA and glyceride concentration existed where the micelles and/or residual oil droplets transformed to liposomes due to swelling and saturation of micellar structures with digestion products. A transition from emulsion droplets and swollen micellar structures to L_α structures for the digestion of MCT-in-water droplets was previously observed after ∼20 min of lipolysis.⁹ The rate of increase in the intensity of scattering from lamellar phase was dependent on lipid concentration and thus, the accelerated formation of the L_α phase in this study was considered due to the smaller lipid droplet size and increase in rate of lipase-mediated digestion, leading to enhanced FFA saturation of bile salts micelles.

Fig. 5 sSAXS profiles over time during digestion of (A) a submicron emulsion and (B) lipid loaded porous silica particles. The arrows show the identifiable Bragg peak positions for (A) H₂ and (B) L_α structures.

At ∼t = 10 min, peaks with spacing ratios of 1 : √3 : √4 indicative of the presence of a H₂ phase with q values of 0.215, 0.372 and 0.430 Å⁻¹ were observed (Fig. 5). This suggests that digestion products partitioned at the o/w interface or were solubilised within the lipid droplets, as H₂ phases are highly unstable in aqueous environments.¹¹ The lattice constant (a) of the H₂ phase decreased slightly from 34 to 33 Å during digestion (Fig. 6). Recently, Salentinig et al.^11,44 observed the internal restructuring and the formation of inverse hexagonal phases inside fat globules during the digestion of long-chain length triglycerides under simulated GIT conditions. Colloidal structures containing FFA are dependent on the relative concentrations of digestion products and are highly pH sensitive due to electrostatic interactions between the charged head groups.¹¹ The apparent pK_a for the formation of a H₂ phase between oleic acid and monoolein was between 6 and 7, depending on the concentration of FFA. At pH 7.5, a H₂ phase was favourable for a FFA : MG ratio of 4 : 1.¹⁰ Since oleic acid and octanoic acid (the most prevalent FFA formed in this study) have comparable isoelectric points (∼4.9), it was reasonable to suggest that octanoic acid forms H₂ structures in similar environments as in the case of oleic acid. According to the ¹H NMR spectra for the digestion of the submicron emulsion, at t > 8 min, the FFA : MG ratio was between 4 : 1 and 5 : 1. Thus, it is hypothesised that the conditions and the relative concentrations of lipolysis products were favourable for digestion products to transform the internal structure of the lipid droplets into a H₂ phase and subsequently increase the polarity of the lipid-in-water interface.¹¹

Fig. 6 Time dependence of lattice parameters (a) for structures formed by the lipolysis of a submicron MCT emulsion (blue markers) and MCT loaded porous silica particles (red markers). Square markers (■) correspond to a for a lamellar phase, whilst circular markers (●) correspond to a for an inverse hexagonal phase. Inset: derived plot for D spacing of the lattice from the peak positions versus the reciprocal spacing ratio for the 1 : 2 : 3 peaks corresponding to the L_α structure (■) and the 1 : √3 : √4 peaks corresponding to the H₂ structure (●).

In contrast to the behaviour of the silica-free emulsion, immediately after the addition of lipase for the digestion of lipid encapsulated in porous silica particles, equidistant peaks of a L_α structure at q = 0.169, 0.338 and 0.507 Å⁻¹ (a = 37 Å) were observed (Fig. 7). A multilamellar phase was evident initially, highlighted by additional peaks at q = 0.212 and 0.424 Å⁻¹, but disappeared after t > 2 min. After a small initial decrease from 37 to 36 Å, a remained constant throughout the digestion period (Fig. 6) and the L_α structure was unaltered (Fig. 5). Previous studies have shown that repeat distances of L_α phases increase with increasing FFA : MG concentration.¹⁰ One hypothesis to explain the formation of a larger bilayer during the digestion of lipid encapsulated in porous silica particles is that liposomes are enriched with FFA and depleted with MG due to their absence within the aqueous phase, which restricted their concentration within the formed liposomes. Alternatively, the change in a corresponds directly with a reduction in DG concentration, whereby the DG concentration is zero at t ≥ 2 min. Thus, the presence of DG in the lipolysis media may result in the formation of a larger bilayer due to the steric hindrance of the alkyl chains.

Fig. 7 Schematic representation of the transition of self-assembly structures by digestion products during the lipolysis of (A) a submicron emulsion and (B) lipid loaded porous silica particles.

The biphasic lipolysis kinetics for the digestion of the submicron emulsion (Fig. 1), and the lipolysis data evidently not following a singular rate expression, is predicted to be closely related to the complex phase transitions of digestion products. The reduction in the pseudo-first-order rate constant from 0.31 to 0.03 min⁻¹ for the digestion of the submicron emulsion at t ∼ 5 min correlated well with the transition from micelles, to L_α and H₂ phases. This confirms potential ‘fouling’ of the lipid-in-water interface by digestion products coincident with geometric structure formation and change in internal droplet microstructure.^11,12 Previous findings hypothesised that discontinuities in digestion profiles were caused by changes in the self-assembly structure of lipolysis products;⁹ however, lipid digestion kinetics increased after the formation of a lamellar phase, in contrast to the inhibited lipase activity observed upon lamellar evolution in the current study. Alternatively, inhibition of lipase activity may be caused by substrate depletion⁴⁵ and an increase in polarity due to the formation of a H₂ phase within the lipid droplet,¹¹ since lipolysis of emulsion droplets has been shown to be a self-regulated process.^45,46 It is assumed that the phase transition does not match the exact time of change in reaction rate (t = 4 min) due to the dynamic behaviour of static samples and the use of a lipase inhibitor (4-BBA) in this study, which has been shown to cause subtle differences in structures compared to real-time sSAXS analysis.⁹

In contrast, it is hypothesised that enhanced lipolysis kinetics for lipid loaded in porous silica particles facilitated the immediate and sustained formation of L_α structure throughout digestion. Since lipase action was uninhibited and well-described by pseudo-first-order kinetics in silica particles, it further suggests a mechanism exists that reduces the interference effect of digestion products, specifically FFA, on lipase action. We propose that the electrostatic repulsion between FFA and the silica surface discourages digestion products from residing at the o/w interface,¹⁷ evidenced by ¹H NMR spectra, and the 4-fold increase in interfacial surface area of lipid promotes the self-assembly of the lamellar structure within the aqueous phase, not within the lipid phase as is the case for the submicron emulsion.²⁴ Phan et al.⁹ demonstrated that the growth in intensity of a lamellar phase (i.e. the height of the first lamellar reflection) is indicative of the rate of digestion. The average intensity of the first reflection was 2-fold greater throughout digestion of lipid loaded porous silica particles (average peak height of 69 at q = 0.169 Å⁻¹) compared to the submicron emulsion (average peak height of 30 at q = 0.215 Å⁻¹).

Whilst, the exact mechanism for controlling the structural organisation of digestion products in complex lipid-based systems in the presence of porous silica particles is still not completely understood, the ability to manipulate the interaction between amphiphilic molecules formed during digestion is clear. This provides insights into the design and engineering of materials to optimise the lipid digestion, solubilisation and absorption of lipophilic bioactive compounds. Further work is required to differentiate between internal restructuring of lipid and the formation of colloidal phases in the aqueous environment, as well as to determine key parameters (e.g. interfacial surface area) that influence the structure of colloidal assemblies formed during lipid digestion.

Conclusions

¹H NMR and sSAXS were successfully employed to monitor the time-dependent disposition and speciation of lipolysis products, in conjunction with the structure evolution of colloidal phases assembled during the digestion of a submicron MCT emulsion and MCT loaded in porous silica particles. Hydrolysis of emulsion droplets in a micellar milieu formed a complex mixture of glyceride molecules and FFA, which self-assembled into a lamellar structure before transitioning to an inverse hexagonal phase. In contrast, a FFA-rich lamellar phase was the predominant structure formed during digestion of lipid loaded porous silica particles. Thus, using porous silica particles as hosts for lipid molecules presents a novel method to control lipase-mediated digestion kinetics, along with the release and resultant self-assembly of FFA and glycerides.

Acknowledgements

This work was undertaken on the SAXS/WAXS beamline at the Australian Synchrotron, Victoria. Funding was provided by the Australian Research Council (ARC) Discovery grant scheme (DP120101065). The University of South Australia is acknowledged for the PhD scholarship of Paul Joyce. We thank Stefan Salentinig, Stephanie Phan and Achal Bhatt for their assistance and operation of the sSAXS.

References

1. C. W. Pouton and C. J. H. Porter, Formulation of lipid-based delivery systems for oral administration: materials, methods and strategies, Adv. Drug Delivery Rev., 2008, 60(6), 625–637.
2. C. W. Pouton, Formulation of poorly water-soluble drugs for oral administration: physicochemical and physiological issues and the lipid formulation classification system, Eur. J. Pharm. Sci., 2006, 29(3–4), 278–287.
3. M. E. Lowe, The triglyceride lipases of the pancreas, J. Lipid Res., 2002, 43, 2007–2016.
4. M. Golding, T. J. Wooster, L. Day, M. Xu, L. Lundin and J. Keogh, Impact of gastric structuring on the lipolysis of emulsified lipids, Soft Matter, 2011, 7, 3513–3523.
5. C. J. H. Porter, C. W. Pouton, J. F. Cuine and W. N. Charman, Enhancing intestinal drug solubilisation using lipid-based delivery systems, Adv. Drug Delivery Rev., 2008, 60(6), 673–691.
6. C. J. H. Porter, N. L. Trevaskis and W. N. Charman, Lipids and lipid-based formulations: optimising the oral delivery of lipophilic drugs, National Review of Drug Discovery, 2007, 6(3), 231–248.
7. G. A. Kossena, B. J. Boyd, C. J. H. Porter and W. N. Charman, Separation and characterization of the colloidal phases produced on digestion of common formulation lipids and assessment of their impact on the apparent solubility of selected poorly water-soluble drugs, J. Pharm. Sci., 2003, 92(3), 634–648.
8. G. A. Kossena, W. N. Charman, B. J. Boyd and C. J. H. Porter, Influence of the intermediate digestion phases of common formulation lipids on the absorption of a poorly water-soluble drug, J. Pharm. Sci., 2005, 94(3), 481–492.
9. S. Phan, A. Hawley, X. Mulet, L. Waddington, C. A. Prestidge and B. J. Boyd, Structural aspects of digestion of medium chain triglycerides studied in real time using sSAXS and Cryo-TEM, Pharm. Res., 2013, 30(12), 3088.
10. S. Salentinig, L. Sagalowicz and O. Glatter, Self-assembled structures and pK_a value of oleic acid in systems of biological relevance, Langmuir, 2010, 26(14), 11670.
11. S. Salentinig, L. Sagalowicz, M. E. Leser, C. Tedeschi and O. Glatter, Transitions in the internal structure of lipid droplets during fat digestion, Soft Matter, 2011, 7(2), 650–661.
12. D. P. Chang, J. Barauskas, A. P. Dabkowska, M. Wadsäter, F. Tiberg and T. Nylander, Non-lamellar lipid liquid crystalline structures at interfaces, Adv. Colloid Interface Sci., 2015, 222, 135–147.
13. J. Barauskas, H. Anderberg, A. Svendsen and T. Nylander, Thermomyces lanuginosus lipase-catalyzed hydrolysis of the lipid cubic liquid crystalline nanoparticles, Colloids Surf., B, 2016, 137, 50–59.
14. S. Ransac, F. Carrière, E. Rogalska, R. Verger, F. Marguet, G. Buono, E. Pinho Melo, J. S. Cabral, M.-P. Egloff, H. Tübeurgh and C. Cambillau, The kinetics, specificities and structural features of lipases, in Molecular Dynamics of Biomembranes, ed. J. F. Op den Kamp, Springer, Berlin, Heidelberg, 1996, ch. 22, pp. 265–304.
15. E. Rogalska, S. Ransac and R. Verger, Stereoselectivity of lipases II. Stereoselective hydrolysis of triglycerides by gastric and pancreatic lipases, J. Biol. Chem., 1990, 265(33), 20271.
16. F. K. Winkler, A. D'Arcy and W. Hunziker, Structure of human pancreatic lipase, Nature, 1990, 343(6260), 771–774.
17. P. Joyce, A. Tan, C. P. Whitby and C. A. Prestidge, The role of porous nanostructure in controlling lipase-mediated digestion of lipid loaded into silica particles, Langmuir, 2014, 30(10), 2779–2788.
18. D. J. McClements, E. A. Decker, Y. Park and J. Weiss, Designing food structure to control stability, digestion, release and absorption of lipophilic food components, Food Biophys., 2008, 3(2), 219–228.
19. D. J. MCClements, E. A. Decker and Y. Park, Controlling lipid bioavailability through physicochemical and structural approaches, Crit. Rev. Food Sci. Nutr., 2009, 49(1), 48–67.
20. P. Joyce, I. Kempson and C. A. Prestidge, Orientating lipase molecules through surface chemical control for enhanced activity: a QCM-D and ToF-SIMS investigation, Colloids Surf., B, 2016, 142, 173–181.
21. P. Joyce, I. Kempson and C. A. Prestidge, QCM-D and ToF-SIMS investigation to deconvolute the relationship between lipid adsorption and orientation on lipase activity, Langmuir, 2015, 31(37), 10198–10207.
22. Z. Vinarov, S. Tcholakova, B. Damyanova, Y. Atanasov, N. D. Denkov, S. D. Stoyanov, E. Pelan and A. Lips, Effects of emulsifier charge and concentration on pancreatic lipolysis: 2. Interplay of emulsifiers and biles, Langmuir, 2012, 28(33), 12140.
23. S. Mun, E. A. Decker and D. J. McClements, Influence of emulsifier type on in vitro digestibility of lipid droplets by pancreatic lipase, Food Res. Int., 2007, 40(6), 770–781.
24. A. Malaki Nik, A. J. Wright and M. Corredig, Impact of interfacial composition on emulsion digestion and rate of lipid hydrolysis using different in vitro digestion models, Colloids Surf., B, 2011, 83(2), 321–330.
25. P. Joyce, C. P. Whitby and C. A. Prestidge, Nanostructuring biomaterials with specific activities towards digestive enzymes for controlled gastrointestinal absorption of lipophilic bioactive molecules, Adv. Colloid Interface Sci. , under review.
26. Y. Pafumi, D. Lairon, P. L. de la Porte, C. Juhel, J. Storch, M. Hamosh and M. Armand, Mechanisms of inhibition of triacylglycerol hydrolysis by human gastric lipase, J. Biol. Chem., 2002, 277.
27. A. Tan, S. Simovic, A. K. Davey, T. Rades, B. J. Boyd and C. A. Prestidge, Silica nanoparticles to control the lipase-mediated digestion of lipid-based oral delivery systems, Mol. Pharm., 2010, 7(2), 522–532.
28. P. Joyce, C. P. Whitby and C. A. Prestidge, Interfacial processes that modulate the kinetics of lipase-mediated catalysis using porous silica host particles, RSC Adv., 2016, 6(49), 43802–43813.
29. M. D. Guillén, I. Carton, E. Goicoechea and P. S. Uriarte, Characterization of cod liver oil by spectroscopic techniques. New approaches for the determination of compositional parameters, acyl groups, and cholesterol from 1H nuclear magnetic resonance and Fourier transform infrared spectral data, J. Agric. Food Chem., 2008, 56(19), 9072–9079.
30. A. Martínez-Yusta and M. D. Guillén, A study by 1H nuclear magnetic resonance of the influence on the frying medium composition of some soybean oil-food combinations in deep-frying, Food Res. Int., 2014, 55, 347–355.
31. B. Nieva-Echevarria, E. Goicoechea, M. J. Manzanos and M. D. Guillen, Usefulness of 1H NMR to study the food lipolysis during in vitro digestion, in Magnetic Resonance in Food Science: Defining Food by Magnetic Resonance, The Royal Society of Chemistry, 2015, pp. 31–38.
32. B. Nieva-Echevarría, E. Goicoechea, M. J. Manzanos and M. D. Guillén, A method based on 1H NMR spectral data useful to evaluate the hydrolysis level in complex lipid mixtures, Food Res. Int., 2014, 66, 379–387.
33. B. Nieva-Echevarría, E. Goicoechea, M. J. Manzanos and M. D. Guillén, Usefulness of 1H NMR in assessing the extent of lipid digestion, Food Chem., 2015, 179, 182–190.
34. N. P. Vidal, M. J. Manzanos, E. Goicoechea and M. D. Guillén, Quality of farmed and wild sea bass lipids studied by 1H NMR: Usefulness of this technique for differentiation on a qualitative and a quantitative basis, Food Chem., 2012, 135(3), 1583–1591.
35. L. Sek, C. J. H. Porter, A. M. Kaukonen and W. N. Charman, Evaluation of the in vitro digestion profiles of long and medium chain glycerides and the phase behaviour of their lipolytic products, J. Pharm. Pharmacol., 2002, 54(1), 29–41.
36. M. Balci, 5-Spin–Spin splitting to different Protons, in Basic 1H- and 13C-NMR Spectroscopy, Elsevier Science, Amsterdam, 2005, pp. 135–158.
37. S. Hyde, Handbook of Applied Surface and Colloidal Chemistry, John Wiley & Sons, Ltd., 2001, vol. 2.
38. S. Ransac, E. Rogalska, Y. Gargouri, A. M. Deveer, F. Paltauf, G. H. d. Haas and R. Verger, Stereoselectivity of lipases. I. Hydrolysis of enantiomeric glyceride analogues by gastric and pancreatic lipases, a kinetic study using the monomolecular film technique, J. Biol. Chem., 1990, 265(33), 20263.
39. I. C. Chandler, O. W. Howarth and D. H. G. Crout, Measuring stereoselectivity in lipase-catalyzed acidolysis reactions by ultra-high resolution 13C nuclear magnetic resonance, J. Am. Oil Chem. Soc., 2001, 78(9), 953–958.
40. B. Åkesson, S. Gronowitz and B. Herslöf, Stereospecificity of hepatic lipases, FEBS Lett., 1976, 71(2), 241–244.
41. M. Van Speybroeck, R. Mellaerts, R. Mols, T. D. Thi, J. A. Martens, J. Van Humbeeck, P. Annaert, G. Van den Mooter and P. Augustijns, Enhanced absorption of the poorly soluble drug fenofibrate by tuning its release rate from ordered mesoporous silica, Eur. J. Pharm. Sci., 2010, 41(5), 623–630.
42. R. Mellaerts, R. Mols, P. Kayaert, P. Annaert, J. Van Humbeeck, G. Van den Mooter, J. A. Martens and P. Augustijns, Ordered mesoporous silica induces pH-independent supersaturation of the basic low solubility compound itraconazole resulting in enhanced transepithelial transport, Int. J. Pharm., 2008, 357(1–2), 169–179.
43. R. Mellaerts, J. A. G. Jammaer, M. Van Speybroeck, H. Chen, J. V. Humbeeck, P. Augustijns, G. Van den Mooter and J. A. Martens, Physical state of poorly water soluble therapeutic molecules loaded into SBA-15 ordered mesoporous silica carriers: a case study with itraconazole and ibuprofen, Langmuir, 2008, 24(16), 8651–8659.
44. S. Salentinig, S. Phan, A. Hawley and B. J. Boyd, Self-assembly structure formation during the digestion of human breast milk, Angew. Chem., Int. Ed., 2015, 54(5), 1600.
45. P. Reis, K. Holmberg, R. Miller, M. E. Leser, T. Raab and H. J. Watzke, Lipase reaction at interfaces as self-limiting processes, C. R. Chim., 2009, 12, 8.
46. P. Reis, H. Watzke, M. Leser, K. Holmberg and R. Miller, Interfacial mechanism of lipolysis as self-regulated process, Biophys. Chem., 2010, 147(3), 93–103.

---

Footnote

† Electronic supplementary information (ESI) available: More details on the materials and methods is provided. See DOI: 10.1039/c6ra16028j

---