Interfacial processes that modulate the kinetics of lipase-mediated catalysis using porous silica host particles

Abstract

Surface immobilised lipases are important bioactive materials that have a wide range of applications in the biotechnology, chemical and pharmaceutical industries. However, the interfacial mechanism of action and the interplay between material characteristics and lipase activity are not well understood. A quartz crystal microbalance with dissipation (QCM-D) was used to elucidate interfacial processes between lipases and lipid films deposited on silica surfaces with varying wettabilities. Adsorption of triglycerides onto a hydrophilic support at multilayer coverage resulted in fast lipolysis kinetics, while adsorption onto a hydrophobic support hindered lipase activity and delayed lipolysis, characterised by changes in frequency and the release of free fatty acids from the QCM-D cell. In parallel, porous silica carriers with varying hydrophilicities/phobicities were used to confine lipid substrate molecules and manipulate lipase-mediated reactions, i.e. hydrolysis and esterification of triglycerides, as monitored using a pH-stat titrator. The surface chemistry of the carrier particles played a critical role on lipase action, whereby hydrophilic silica particles promoted catalysis of the hydrolysis reaction and hydrophobic particles promoted the reverse, esterification reaction. Physical observations of lipid film hydrolysis in combination with biomaterial design of lipid loaded porous silica particles provided advancements in understanding the mechanism of lipase action, which can be harnessed to tailor the delivery of poorly water-soluble molecules and improve the synthesis process of organic esters.

---

Introduction

Lipases are surface active proteins whose physiological function in the gut is to catalyze the hydrolysis of dietary fats into free fatty acids (FFA), di- or mono-glycerides and glycerol (Scheme 1). The digestion and absorption of lipids is controlled by the ability for pancreatic lipase to access and adsorb to the triglyceride droplet-in-water interface, facilitating the breakdown into digestion products that can be readily absorbed into the bloodstream.¹ Consequently, the rate of solubilisation and absorption of bioactive lipophilic compounds encapsulated within a lipid-based system (e.g. food or pharmaceutical products) is also controlled by the ability for lipase to digest the lipid in the gastrointestinal tract.² It is therefore possible to manipulate the bioavailability of poorly water-soluble drugs or vitamins by engineering lipid systems that regulate pancreatic lipase action.³

Scheme 1 A typical lipase-mediated digestion reaction of triglycerides.

Whilst the biological role of lipases is to cleave ester bonds for fat absorption, they are also utilised in other biotechnology and chemical processes due to their ability to catalyse the reverse, esterification reaction, e.g. triglyceride synthesis from alcohols and fatty acids.⁴ The equilibrium of a lipase-catalyzed reaction is governed by the aqueous content of the reaction medium, as water is a product of the esterification reaction.^5,6 The catalysis of esterification reactions by lipolytic enzymes presents a number of advantages over conventional chemical synthesis due to the enzymes' high stability and activity, mild reaction conditions, high catalytic efficiency and inherent selectivity, resulting in purer products.⁷ In contrast, chemical synthesis of glycerides requires harsh processing conditions and high temperatures, resulting in the formation of undesirable bi-products.⁸ Consequently, lipases have enormous industrial potential and value due to their improved performance compared to conventional synthesis techniques and there is growing demand for organic esters in biotechnological applications.^9–11

Due to the nature of their lipid substrates, lipase-mediated reactions occur at the oil-in-water interface. Thereby, enzymatic activity is controlled by the concentration and composition of the substrate at the interface,^12,13 which can be manipulated through microstructural changes,¹⁴ such as in surface area and bioaccessibility.¹⁵ A unique and essential catalytic feature of most lipase molecules that sets them apart from other esterases is a surface loop or lid domain, which protects the hydrophobic active site from the aqueous environment.¹⁶ The closed-lid conformation of a lipase molecule is inactive and unable to bind to lipid molecules. Upon interaction with an interface, the lid domain opens which then reduces the polarity of the exposed surface, increasing the affinity of the enzyme to the substrate.¹⁷

This “interfacial activation” mechanism has led to the immobilisation of lipases in a wide range of mesoporous colloidal carriers,¹⁸ most commonly porous silica particles.^19–22 By loading lipase in porous particles it is possible to control the bioaccessibility of the lipid substrate,^22,23 whilst maximising the catalytic performance and removing limitations associated with bulk enzyme solutions.²⁴ A subtle relationship exists between the interface to which lipase adsorbs and its activity.²⁵ Reis et al.²⁶ demonstrated that lipases adsorbed on hydrophilic silica favour hydrolysis reactions, whereas lipases bound to hydrophobic surfaces promote esterification reactions. This behaviour was attributed to the lipase adsorbing to the surface in a different orientation or conformation, which alters the interaction and affinity the active site has for lipid molecules. Orientating lipase molecules through surface chemical control was confirmed recently using time-of-flight secondary ion mass spectrometry (ToF-SIMS), whereby clear differences in lipase orientation at surfaces with different hydrophilicities/phobicities were observed through variations in mass spectral fragmentation patterns.²⁷ This introduces the ability to tailor-make and engineer carriers that enhance lipase performance for a desired application.¹⁵

Recently, porous silica particles were used as novel hosts for medium-chain length triglycerides (MCTs), in which lipid hydrolysis was altered by varying the nanostructure of the host silica particles and by manipulating the surface coverage of lipid.¹⁴ Additionally, lipid adsorption kinetics, monitored by quartz crystal microbalance with dissipation (QCM-D), and molecular orientation, monitored by ToF-SIMS, were reported to influence lipid digestibility through changes in lipid microenvironment, thus altered affinity with lipase.²⁸

QCM-D monitors changes in mass of an adsorbed layer through changes in vibrational frequency.²⁹ Snabe et al.³⁰ demonstrated the ability to physically observe lipase-mediated digestion by monitoring the mass change of a deposited lipid layer. Specifically, the role of pH and calcium ions was investigated to determine their influence on lipolytic activity and the mechanism of hydrolysis.³¹ Lipid digestion kinetics via QCM-D was characterised by a lag phase and mass ejection model, whereby addition of lipase corresponded to an increase in adsorbed mass until the charged product build up caused a rapid increase in surface potential and thus, FFA and monoglycerides were ejected. The time delay from enzyme addition until detection of mass ejection was found to correlate well with lipase activity.³¹

In this study, QCM-D was used to physically observe lipolytic activities of two lipase enzyme extracts upon interaction with multilayers of triglycerides adsorbed onto silica supports with varying surface chemistries. Candida antarctica lipase A (CalA), a fungal lipase used in organic ester synthesis, and porcine pancreatic lipase (PPL), a crude mammalian lipase extract used for in vitro studies that mimics human pancreatic juices, were selected for this study. The ability to manipulate lipase activity through surface chemistry was then used to engineer biomaterials, consisting of lipid or glycerol loaded silica particles with varying hydrophilicities/phobicities, in order to control the kinetics of lipase-mediated catalysis. This study provides further insights into the mechanistic mode of action for industrial and physiological lipases which can be used for smarter design of bioactive delivery carriers, as well as new and improved processing technologies for organic ester synthesis.

Materials and methods

Materials

Candida antarctica lipase A (CalA) was purchased from Sigma-Aldrich (Australia) and porcine pancreatic lipase (PPL) extract was supplied by Chem-Supply (Gillman, Australia). Hydrophilic porous silica particles, PS-1 (Aerosil 200, surface area 200 m² g⁻¹) and hydrophobic porous silica particles, PS-2 (Aerosil R812, surface area 220 m² g⁻¹, hydrophobized with hexamethyldisilazane (HMDS) by the manufacturer)³² were supplied by Evonik (Essen, Germany). Tributyrin (TB) was supplied by Acros Organics, ThermoFisher Scientific (Scoresby, Australia). Glycerol, oleic acid, chlorotrimethylsilane, 4-bromophenyl boronic acid (4-BBA) phosphate buffered saline (PBS) tablets, sodium hydroxide pellets and cyclohexane were all purchased from Sigma-Aldrich (Australia). Hexane (AR) grade was supplied by BDH Merck (Sydney, Australia). QCM-D sensor crystals (5 MHz), reactively sputter-coated with 50 nm of silicon dioxide, were purchased from Q-Sense (Gothenburg, Sweden). All chemicals were of analytic grade and used as received. High purity (Milli-Q) water was used throughout the study.

Quartz crystal microbalance with dissipation (QCM-D) investigation

QCM-D instrumentation. QCM-D measurements were performed on silicon dioxide coated QCM-D crystals (25 mm, 5 MHz) using a Q-Sense E4 system (Q-Sense, Inc., Biolin Scientific AB, Sweden). The QCM-D technique, detailed elsewhere,²⁹ provides information on the amount of adsorbed mass (including coupled water) through changes in the vibrational frequency, f, and the degree of frictional (viscous) losses in the adsorbed layer through changes in dissipation, D.

The Sauerbrey equation relates measured frequency shift (Δf) to the change in adsorbed mass per unit area (Δm).³³ This estimation is only suitable for rigid layers, i.e. layers with ΔD ≤ 10⁻⁶.²⁹ The mass calculated from the frequency change is only an approximation because the frequency change is due to the adsorbed material and any trapped or associated water molecules. Since the current work operates with a non-rigid film approximately 20–40 nm in thickness which varies over time due to enzymatic digestion, the Sauerbrey equation cannot be used for lipid film hydrolysis. Instead, a viscoelastic model was required to estimate layer thickness, d_f and coupled mass, m. The frequency and dissipation data was modelled with a Voigt-based model (incorporated in QTools, Q-Sense, Inc., Biolin Scientific AB) using the third (n = 3, 15 MHz) and fifth overtones (n = 5, 25 MHz) with a layer density of 1.00 × 10³ kg m⁻³. Detailed theory of the Voigt-based model can be found elsewhere.³⁴

Preparation of hydrophobic SiO₂ crystals. Silica coated QCM-D crystals were hydrophobized by immersion in a chlorotrimethylsilane (20 mM) in cyclohexane solution for 12 hours at ambient room temperature (25 °C), followed by rinsing with Milli-Q water prior to measurements. Sessile drop measurements were performed to determine contact angles of the hydrophobic QCM-D crystals. In all cases, the contact angles were greater than 90°.

Lipase adsorption on bare silica surfaces. Hydrophilic and hydrophobic silica coated crystals were mounted into the QCM-D unit and flushed with PBS solution (25 mM, pH 7.5) to establish baseline values with liquid. The crystal chamber temperature was maintained at 25 ± 0.1 °C. Enzyme (CalA or PPL) solutions (10 µg mL⁻¹) in PBS (25 mM) were introduced into the cell (1 mL min⁻¹) allowing lipase to adsorb onto the crystal surface under flow conditions for approximately 10 min. The cell was flushed with buffer to determine final changes in crystal frequency and dissipation. Measurement data for f and D were acquired at several harmonics simultaneously.

Lipid film hydrolysis. Changes in adsorbed mass associated with lipase-mediated digestion were monitored by firstly adsorbing a TB film onto the hydrophilic and hydrophobic crystal surfaces prior to exposing the cell to lipase solutions. Briefly, TB solution (0.1 M in 1 : 3 methanol : water solution) was slowly introduced into the QCM-D cell (1 mL min⁻¹), allowing lipid to adsorb onto the crystal surface under flow conditions for approximately 10 min, before stabilizing the lipid layer in PBS solution on both hydrophilic and hydrophobic surfaces. CalA or PPL solutions were introduced into the cell in an equivalent manner to above and changes in f and D were continuously monitored during lipid digestion. Ejected solutions were collected and analysed for FFA content by titration with NaOH (0.2 M) to neutralise pH to 7.5.

Substrate loaded porous silica particles

Preparation of lipid/glycerol loaded porous silica particles. Lipid (TB) was loaded into porous silica particles following the protocol developed in previous work,¹⁴ i.e. by immersing silica (1.50 g) in a TB (2.50 g) solution in 50 mL of hexane. The formed suspensions were tumbled for 24 h, after which excess solvent and unbound TB was removed by filtration with a 0.2 µm membrane. The filtered solid was washed with 25 mL of hexane to remove the TB bound to the external silica surface and refiltered twice. The solid powder was heated to 30 °C for 3 h to remove excess solvent. Glycerol was loaded in PS-1 and PS-2 in an equivalent manner using methanol as the organic solvent to dissolve the glycerol and wash the silica particles.

Submicrometer oil-in-water emulsions (10 wt% TB or glycerol) were prepared by homogenisation at a pressure of 1000 bar for 5–6 cycles (Avestin EmulsiFlex-C5 Homogenizer). An emulsifier was not used in this study. As a result, all catalytic reactions were performed immediately after preparing emulsions.

Characterisation of substrate loaded porous silica particles. The amount of TB or glycerol loaded in the porous silica particles was determined by thermogravimetric analysis (TGA). The particles were heated at a scanning rate of 10 °C min⁻¹ from 20 to 600 °C under nitrogen purging. The TB and glycerol completely decomposed below 500 °C, whilst the silica remained thermally stable under these conditions. The amount of loaded material within the silica was determined by the weight loss (accounting for trace amounts corresponding to residual solvent and water moisture). Particle sizing of silica particles, before and after loading, was conducted via laser diffraction (Malvern Mastersizer). The particle sizes were independent of the loading concentration. The surface morphology and porous nanostructure of PS-1 and PS-2 was characterised using scanning electron microscopy (SEM).

Lipase-mediated hydrolysis studies

Preparation of lipase solutions. CalA or PPL (1.0 g) were stirred in 10 mL PBS solution (25 mM) for 15 min, followed by centrifugation (4500 rpm, 4 °C) for 20 min. The supernatant was collected and refrigerated until use.

Lipid hydrolysis kinetics studies. Lipid digestion kinetics were monitored using a pH stat titrator (TIM854 Titration Manager, Radiometer, Copenhagen, Denmark). A known quantity of lipid loaded porous silica (equivalent to 200 mg TB) was dispersed in 20 g PBS solution (25 mM) by stirring continuously for 10 min at 600 rpm in a thermostatted glass reaction vessel (37 °C). The pH of digestion medium was re-adjusted with 0.1 M NaOH or HCl to 7.50 ± 0.01. Digestion of TB was initiated by the addition of lipase solution (0.5 mL, ∼500 TBU). Free fatty acids (FFA) produced in the reaction vessel were immediately titrated with 0.6 M NaOH via an auto-burette to maintain a constant pH in the digestion medium at the pre-set value of 7.50 ± 0.01. The base consumption was monitored for 60 min. All digestion studies were repeated 3 times and showed good reproducibility. It is noteworthy that the silica particles do not alter the pH through ionization during digestion. It was also assumed that lipid digestion in PS-1L and PS-2L occurred within the silica pores as previous studies demonstrated insignificant lipid desorption over the digestion period.¹⁴

Lipase-mediated reactions can be described with the following reaction:

where, k_H is the hydrolysis rate constant and k_E is the esterification rate constant. Lipid digestion kinetics were modelled using the pseudo-first-order model below:

% H = H_max(1 − e^−k_Ht) (2)

where % H is the percent lipid hydrolysis, H_max is the maximum achievable lipolysis, k_H is the pseudo-first-order hydrolysis rate constant and t is time. Since an excess amount of lipase was added to the reaction vessel, the rate of reaction was dependent on the quantity of substrate and thus, the hydrolysis reaction was considered a pseudo-first-order process.

Lipase-mediated esterification studies

Esterification of glycerol and oleic acid followed the protocol developed by Yesiloglu et al.⁷ A known quantity of glycerol loaded porous silica particles (equivalent to ∼400 mg glycerol) was dispersed in 20 g hexane. Milli-Q water (200 mg) was added to the solution to ensure lipase functionality. The pH of digestion medium was re-adjusted with 0.1 M NaOH or HCl to 7.5 ± 0.01. Oleic acid (400 mg) was added to the solution, which was stirred for 10 min in a thermostatted glass vessel (37 °C). Esterification was initiated by the addition of lipase solution (0.5 mL). The solution pH was continuously monitored throughout the reaction. At various times, aliquots (200 µL) of reaction mixture were withdrawn (and replaced with PBS) and inhibited by the addition of 2 µL of 0.5 M 4-BBA. FFA in each sample was titrated with 0.1 M NaOH to 7.5 ± 0.01 and the degree of esterification was determined as the percent of FFA consumed in the reaction mixture. Esterification kinetics were also fitted by a pseudo-first-order model:

% E = E_max(1 − e^−k_Et) (3)

where % E is the percent esterification at time t, and maximum achievable esterification, E_max, to determine the pseudo-first-order esterification rate constant, k_E.

Results and discussion

QCM-D analysis of lipase adsorption and lipid film hydrolysis

Role of surface chemistry on lipase adsorption to silica surfaces. The rate and extent of lipase adsorption was enhanced for both Candida antarctica lipase A (CalA) and porcine pancreatic lipase (PPL) on hydrophobic silica, compared to hydrophilic silica (Fig. 1). Changes in frequencies plateaued at 9.80 and 15.1 Hz for adsorption of CalA and PPL, respectively, on hydrophilic silica compared to 27.2 and 33.4 Hz for hydrophobic silica. Previous studies have shown that lipases irreversibly adsorb to hydrophobic surfaces due to strong hydrophobic interactions between the non-polar domains of the lipase molecules and the methylated surface.²⁷ This was supported by the 2–3 fold increase in frequency change for both enzymes when adsorbed to hydrophobic silica. In contrast, lipase adsorption on hydrophilic surfaces was considered reversible whereby an equilibrium surface concentration exists.

Fig. 1 Lipase adsorption profiles for (A) CalA and (B) PPL on hydrophilic (blue ●) and hydrophobic (red ■) silica supports. Lipase concentrations were 10 mM in PBS (25 mM) solution. Frequency shifts are given in terms of the fifth overtone (25 MHz).

Lipid film hydrolysis using QCM-D. Lipolytic activity was observed by monitoring changes in frequency for a deposited lipid film on hydrophilic and hydrophobic silica surfaces using QCM-D (Fig. 2). Frequency and dissipation changes were converted to adsorbed mass changes (ESI Fig. S1†), which indicated that lipase-mediated digestion of such lipid films was rapid and complete in less than 5 min.

Fig. 2 Frequency change as a function of time during lipase-mediated hydrolysis of a tributyrin film on a hydrophilic (blue line) and hydrophobic surface (red line) for (A) CalA and (B) PPL. Lipase injection occurred at (i) and flow rates remained constant at 10 µg mL⁻¹ for all samples.

The extent of lipid digestion, measured by the percent change in frequency due to lipid digestion, was greatest for a tributyrin (TB) film loaded on hydrophilic silica (82% compared to 26% mass desorption for lipid adsorbed to hydrophobic silica using CalA) (Fig. 3). However, the limitations and shortcomings of interpreting QCM-D data are well-known and can be misleading.³⁵ Consequently, the concentration of FFA released from the QCM-D cell was measured and correlated to the change in Sauerbrey mass. A linear correlation between the concentration of FFA in ejection media and mass desorption was observed (Fig. 3), highlighting the potential for QCM-D to be used as a tool to quantify lipase activity.

Fig. 3 Concentration of FFA in the ejection media as a function of Sauerbrey mass desorption due to the lipase-mediated hydrolysis of a tributyrin film on (blue ●) hydrophilic silica and (red ■) hydrophobic silica (SD, n = 3).

The digestion of all four samples comprised of the following phases: (i) lipase molecule adsorption to the lipid film; (ii) enzymatic activity initiation, leading to a build-up of charged digestion products and the formation of clusters (including encapsulated triglyceride substrate and products) at the lipid interface, and finally; (iii) mass ejection of clusters of FFA and glycerides as micelles and vesicles, due to a significant increase in hydration of soluble products (Fig. 4). This mechanism was supported by observed decreases in frequency for phase (i) due to the adsorption of lipase molecules, leading to a steady-state period for phase (ii), followed by a rapid increase in frequency as a result of mass ejection of digestion products for phase (iii). The time course of each event varied depending on the surface chemistry of the binding support and the type of lipase used. These findings were consistent with the QCM-D mass ejection model presented by Snabe et al.,^30,31 who demonstrated that lipase-mediated hydrolysis of a lipid film is a dynamic process that is quickly inhibited by the accumulation of digestion products at the lipid-in-water interface. In their studies, the course of lipid digestion was monitored for 20 min, but it was evident that the hydrolysis reactions studied here were complete/inhibited within 5 min. The difference in reaction rate/inhibition was attributed to differences in experimental variables (e.g. lipase and lipid types and initial lipid coverage).

Fig. 4 Schematic representation of lipid film hydrolysis. (i) Lipase adsorbs to lipid-in-water interface. (ii) Lipase catalyses the hydrolysis of triglycerides to FFA and monoglycerides which either partitions to the lipid-in-water interface or disperses within the aqueous phase. (iii) Digestion products form mixed micelles (or vesicles) which disperse in the aqueous phase, whilst other digestion products inhibit lipase adsorption to the lipid-in-water interface.

Role of surface chemistry on lipid film hydrolysis. Upon lipase injection, adsorbed mass increased for both lipid coated hydrophilic and hydrophobic silica surfaces due to the adsorption of lipase molecules to the lipid film. Adsorbed mass and layer thickness increased to approximately 100 ng cm⁻² and 2–3 nm, respectively, upon CalA and PPL injection for TB loaded on hydrophilic silica. Since the theoretical adsorbed monolayer of CalA has a mass loading of ∼83 ng cm⁻² and the hydrodynamic diameter of CalA molecules is approximately 4.5 nm,^22,36 it suggests an approximate monolayer of lipase is likely to have adsorbed to the lipid film. The increase in adsorbed mass was 5–7 fold greater for TB adsorbed on hydrophobic silica, indicating a difference in the affinity of charged products for the lipid interface.

The lag phase, given as the time delay between lipase injection and the detection of mass ejection,³¹ was also significantly shorter for both CalA and PPL when a hydrophilic silica substrate was employed. Mass ejection was detected 50 s and 62 s after lipase addition for CalA and PPL, respectively, when TB was adsorbed on hydrophilic silica, compared to 148 s and 181 s for the hydrophobic surface (Fig. 2). The ejection of digestion products from the lipid-in-water interface into the aqueous phase is dependent on a sufficient increase of surface polarity and potential, causing the adsorption of clusters at the lipid interface to become unfavourable. Consequently, the lag time provided an indication of the rate of free fatty acid build up within the system, and therefore lipase activity.³¹

For both hydrophilic and hydrophobic silica, lipid film hydrolysis was promptly inhibited after the initial rapid increase in frequency. The final Sauerbrey masses adsorbed on the hydrophilic surface, post-digestion were 550–650 ng cm⁻², which was significantly greater than the amount of lipase that adsorbed to the bare hydrophilic silica (Fig. 1). Consequently, despite the enhanced digestion kinetics on the hydrophilic surface, it is predicted that lipase activity was inhibited and the mass that remained on the silica surface consisted of undigested triglycerides, lipase molecules and digestion products. The inhibition of both CalA and PPL was clearly evident for TB adsorbed on hydrophobic silica as only a small portion of the adsorbed lipid (26% and 7%, respectively) was removed from the surface through hydrolysis.

The inhibition of lipase activity was attributed to the interference effect of digestion products.³⁷ Pafumi et al.³⁸ demonstrated the complete inhibition of gastric lipolysis by long-chain length FFA after 60 min digestion due to the accumulation of FFA clusters at the lipid-in-water interface. FFA reorganised into peripheral particles that encapsulated gastric lipase at the emulsion droplet interface, reducing its mobility. Consequently, it is hypothesised that the rapid build-up of digestion products resulted in the inhibition of both CalA and PPL action through an equivalent mechanism (Fig. 4(iii)), as evidenced by the small increase in adsorbed mass post-hydrolysis (Fig. 2). This was most evident for PPL-mediated digestion on hydrophobic silica, where a rapid increase in adsorbed mass (∼400 ng cm⁻²) was observed post digestion (Fig. 2). This highlights the importance of bile salts on pancreatic lipase action, which act by removing surface active products from emulsion droplets during physiological lipid digestion.³⁹

Tributyrin and glycerol loaded porous silica particles

TB and glycerol were adsorbed at a theoretical submonolayer coverage in hydrophilic porous silica particles (PS-1) and hydrophobic porous silica particles (PS-2) (Table 1) with surface areas of ∼200–220 m² g⁻¹.³² The porous nanostructure and surface morphology of both particles were equivalent (Fig. 5). The silica nanoparticles in PS-1 and PS-2 cluster into 50 nm–5 µm aggregates in a disordered, random fashion in the dry powder. The size of the voids in the aggregates (2–7 nm) was estimated theoretically as ranging between the pore size in hexagonally and tetrahedrally packed particles.⁴⁰ The mass of lipid loaded for PS-1L and PS-2L was 7.7 and 8.9 wt%, which corresponds to surface coverages of 0.61 and 0.64 equivalent monolayer, respectively. PS-1 and PS-2 were also used as hosts for glycerol molecules at a multilayer coverage to promote catalysis of the esterification reaction.

Table 1 Characterization of lipid- and glycerol-loaded porous silica particles with different surface chemistries^a

TB loaded sample	Mass of lipid loaded (wt%)	Lipid surface coverage	Loading efficiency (%)
a Mass of TB/glycerol loading was relative to silica weight. The surface coverage of lipid and glycerol in each sample was calculated based on the theoretical monolayer in each particle. The TB theoretical close packed monolayer is 12 and 14 wt% for PS-1 and PS-2, respectively, and the glycerol theoretical monolayer is 3.8 and 4.3 wt%, respectively. It was assumed that the lipid and glycerol was evenly distributed and could access all of the silica pores.
PS-1L	7.7 ± 0.5	0.61	4.6
PS-2L	8.9 ± 1.1	0.64	5.3

---

Glycerol loaded sample	Mass of glycerol loaded (wt%)	Glycerol surface coverage	Loading efficiency (%)
PS-1G	5.1 ± 0.3	1.3	3.0
PS-2G	7.0 ± 0.6	1.6	4.2

---

Fig. 5 SEM images of (A) hydrophilic, PS-1 and (B) hydrophobic, PS-2 porous silica particles. Scale bars represent 5 µm.

Lipase-mediated hydrolysis of lipid in porous silica host particles

Influence of surface chemistry. The hydrolysis kinetics (catalysed by either CalA or PPL) for a submicrometer TB emulsion (200 nm droplet size) and TB loaded in PS-1L and PS-2L are given in Fig. 6. For both enzyme extracts, the rate and extent of digestion were enhanced when a partial TB monolayer was loaded in PS-1L and inhibited in PS-2L, compared to the lipid droplets. Pseudo-first-order model fits to hydrolysis data are shown in Fig. 5, which precisely described the lipolysis of all systems over the digestion period (Table 2).

Fig. 6 Lipase-mediated hydrolysis kinetics for tributyrin emulsion droplets (green ▲), loaded in PS-1L (blue ●), and loaded in PS-2L (red ■) for (A) CalA and (B) PPL (SD, n = 3). Pseudo-first-order fits are represented by the lines for each curve.

Table 2 Pseudo-first-order kinetic analysis parameters for the hydrolysis of lipid systems^a

Lipid sample	Enzyme	First-order rate constant, kH (min^−1 × 10^−2)	Extent of hydrolysis, % H60
a The quality of fit (R^2) for all systems was greater than 0.98.
Emulsion	CalA	3.1 ± 0.4	84 ± 3
PS-1L	CalA	4.2 ± 0.3	91 ± 2
PS-2L	CalA	1.2 ± 0.1	51 ± 2
Emulsion	PPL	2.2 ± 0.7	73 ± 3
PS-1L	PPL	5.4 ± 0.9	96 ± 4
PS-2L	PPL	0	0

---

Enhanced lipase activity for PS-1L was evidenced by first-order rate constant values of 4.2 × 10⁻² and 5.4 × 10⁻² min⁻¹ for CalA and PPL, respectively, compared to values of 3.1 × 10⁻² and 2.2 × 10⁻² min⁻¹ for the digestion of the submicrometer emulsion. The increased rate and extent of lipid digestibility within PS-1L was attributed to a combination of the following mechanisms: (i) the surface area of lipid was enhanced 4-fold, increasing its bioaccessibility; (ii) lipid adsorbed onto hydrophilic silica in an orientation favourable for lipase interaction;²⁸ (iii) lipase adsorbed onto bare hydrophilic silica surfaces is in its active, open-lid conformation;²⁷ and (iv) digestion products are effectively removed from the lipid interface due to a potential electrostatic repulsion with the silica, reducing their interference effect (Fig. 7).¹⁴

Fig. 7 Schematic representation demonstrating the role surface chemistry and lipase type on lipolysis in lipid loaded porous silica particles using (A) Candida antarctica lipase A and (B) porcine pancreatic lipase extract.

In contrast, lipase activity was significantly inhibited when lipid was loaded in hydrophobic porous silica particles for both lipases. This was evidenced by a first-order rate constant value of 1.2 × 10⁻² min⁻¹ for CalA and complete inhibition for PPL. The catalytic domains covered by the lid domain of lipase are nonpolar and form strong hydrophobic interactions with other nonpolar interfaces, such as an emulsion droplet.⁴¹ Whilst this adsorption to a hydrophobic interface is reversible, the desorption of lipase molecules has been shown to be highly improbable.^42,43 Consequently, it is hypothesised that lipase adsorbed to the bare surface of PS-2 in an orientation that reduced the interaction between the active site and lipid molecules (Fig. 7).⁴³ Previous studies have also demonstrated that submonolayer levels of lipid on hydrophobic silica are orientated to restrict the rate of lipolysis.²⁸

Influence of lipase type. The difference in k_H and % H₆₀ for PS-1L were insignificant between the two lipase extracts (Table 2). This signifies that an equivalent mechanism of action was likely present for the two extracts in hydrophilic silica particles. In contrast, the lipase extract type significantly altered the lipase:lipid interaction within the hydrophobic porous silica particles. The PPL-mediated digestion of lipid encapsulated within PS-2L was completely inhibited. Whereas, CalA-mediated digestion of PS-2L resulted in k_H and % H₆₀ values of 1.2 × 10⁻² min⁻¹ and 51 ± 2%, respectively, demonstrating CalA was still active in hydrophobic silica particles despite being inhibited compared to PS-1L and lipid droplets.

Unlike CalA which consists of pure lipase molecules, PPL contains a number of surface active components, such as other proteins and enzymes, bile salts and a number of small charged species.⁴⁴ The lipase content responsible for lipolytic action in PPL accounts for less than 1% of the crude extract.⁴⁵ It is therefore hypothesised that the additional surface active components interfere with pancreatic lipase action during hydrolysis of lipid encapsulated in hydrophobic porous silica particles (Fig. 7). Wickham et al.⁴⁶ demonstrated a 5–10 fold inhibition in in vitro lipolysis at physiological concentrations of bile salts due to their high “surfactancy” and ability to adsorb to insoluble interfaces. This inhibitory effect is naturally overcome by co-lipase, a non-enzymatic protein co-factor that binds to pancreatic lipase to facilitate adsorption to bile salt-saturated lipid interfaces.¹ Since co-lipase is also present in PPL it is unlikely that bile salts are solely responsible for this complete inhibition of PPL action in PS-2L. However, Wickham et al.⁴⁷ also determined that bile salts in the continuous phase bind to pancreatic lipase molecules forming complexes away from the lipid interface. Several other studies have also demonstrated the potential for amphiphilic compounds, such as surfactants^48,49 and proteins,^50–52 to significantly inhibit lipase activity. In some scenarios, surface active molecules formed vesicles that encapsulated lipase, completely inhibiting its accessibility to the lipid interface.³⁸ Therefore, a mechanism must exist in PS-2L, which does not exist in PS-1L, whereby PPL extract components interfere with lipase action. It is hypothesised that lipase molecules adsorb to the hydrophobic silica surface in its inactive conformation, coinciding with adsorption of surface active compounds, which form clusters encapsulating the enzyme and preventing adsorption to the lipid interface (Fig. 6).

Furthermore, PPL has sn-1,3 regiospecificity, meaning it hydrolyses triglycerides at the 1 and 3 positions, producing two fatty acids and one 2-monoglyceride.⁵³ In contrast, CalA is the only lipase molecule that hydrolyses lipids at the sn-2 position.^13,36 Therefore, enhanced CalA activity in PS-2L compared to PPL may also be due to the difference in adsorption mechanism of both lipases, whereby CalA can still adsorb to lipids orientated in an unfavourable conformation and can cleave ester bonds at all three positions.

Lipase-mediated esterification of glycerol in porous silica host particles

Influence of surface chemistry. The rate and extent of esterification for free glycerol droplets and glycerol loaded in PS-1G and PS-2G are given in Fig. 8. The esterification kinetics were enhanced in hydrophobic porous silica host particles, when compared to hydrophilic particles, as quantified by pseudo-first-order kinetics. First-order rate constants for PS-1G of 0.28 × 10⁻² min⁻¹ and 0.34 × 10⁻² min⁻¹ were derived for CalA and PPL, respectively, compared to values for PS-2G of 1.39 × 10⁻² min⁻¹ and 0.47 × 10⁻² min⁻¹.

Fig. 8 Lipase-mediated esterification kinetics with pseudo-first-order fits for glycerol free in solution (green ▲), loaded in PS-1G (blue ●), and loaded in PS-2G (red ■) for (A) CalA and (B) PPL (SD, n = 3). Pseudo-first-order fits are represented by the lines for each curve. Esterification was performed in a 2 wt% oleic acid in hexane solution. Water (0.1 wt%) was added to the solution for lipase functionality.

The water content of the reaction medium affects the equilibrium as it is a product of the esterification reaction.⁷ It is thereby hypothesised that the hydrophobicity of the silica surface alters the localised water concentration at the silica and adsorbed glycerol interfaces. In addition to this, previous studies have demonstrated that a lipase's orientation and conformation is dependent on the surface chemistry of the interface to which it adsorbs.⁴³ Therefore, conformational changes in lipase tertiary structure at the solid–liquid interface may have also altered the accessibility of the lipase active site to the glycerol substrate. It is therefore predicted that esterification kinetics are controlled by the following factors: (i) localised water content; (ii) orientation of substrate molecules; and (iii) lipase conformation on the bare silica interface.

Influence of lipase type. The rate and extent of esterification in porous silica particles was dependent on the lipase extract used. The order of esterification extent by CalA was PS-2G (55 ± 3%) > free glycerol (45 ± 6%) > PS-1G (15 ± 4%), compared to free glycerol (38 ± 4%) > PS-2G (25 ± 5%) > PS-1G (19 ± 5%) for PPL. A decrease in esterification kinetics was observed for PS-2G when PPL was used. This demonstrates that the structure and composition of lipase has significant impact on its catalytic esterification activity. In a similar manner to PPL action in PS-2L, it is hypothesised that the surface active extract components in PPL had an inhibitory effect on the enzyme's activity by adsorbing to lipase molecules and the hydrophobic surface, reducing the ability for lipase adsorption to the glycerol substrate.⁴⁶

Influence of silica surface chemistry on esterification/lipolysis equilibrium

Kinetic analysis confirmed that the surface chemistry of the carrier particles manipulated the course of lipase-mediated catalysis when lipid or glycerol molecules were encapsulated within the porous silica matrix (Fig. 9). The equilibrium constant, K (k_H/k_E), describes the favoured direction of reaction for given conditions. When K > 1, the forward, hydrolysis reaction was favoured, whereas when K < 1, the reverse, esterification reaction was favoured. K values for PS-1 were 15.0 and 15.9 for CalA and PPL, respectively, compared to PS-2 values of 0.9 and 0. That is, when adsorbed to a hydrophilic surface, hydrolysis of triglycerides was enhanced and when adsorbed to a hydrophobic surface, glyceride synthesis from alcohols and free fatty acids (i.e. esterification) was enhanced. This was in accordance with previous studies that demonstrated hydrophobic matrices promoted esterification, whereas hydrophilic matrices promoted hydrolysis when lipase molecules were immobilised in porous silica.^25,26 Therefore, a subtle interrelationship exists between silica surface and both the lipid substrate and lipase molecules. Harnessing this relationship between material characteristics and lipase activity introduces the ability to engineer biomaterials that optimise the lipase-mediated reaction kinetics.

Fig. 9 Hydrolysis (left axis, blue bars) and esterification (right axis, red squares) pseudo-first-order rate constants as a function of porous silica particles and lipase type.

Conclusions

QCM-D studies of deposited lipid layers provided significant insight into the interfacial processes that control lipolysis kinetics. Changes in Sauerbrey adsorbed mass correlated linearly with FFA concentration expelled from the surface, highlighting the potential for QCM-D to be used to observe hydrolysis of adsorbed triglycerides. Surface chemistry of the silica binding support controlled the rate and extent of lipase-mediated catalysis. Confining lipid substrate molecules within the pores of porous silica particles with varying wettabilities presented a novel method for manipulating lipase activity. In agreement with QCM-D findings, lipolysis kinetics were enhanced when triglycerides were adsorbed in hydrophilic porous silica particles, compared to hydrophobic particles. In contrast, hydrophobic surfaces promoted lipase-mediated esterification when glycerol was loaded in silica particles. Thus, the surface chemistry of host particles can be altered to optimise catalysis of lipase-mediated reactions for specific applications, such as drug delivery/functional food systems and ester synthesis.

Acknowledgements

This work has been supported by the Australian Research Council (ARC) Discovery grant scheme (DP120101065). The University of South Australia is acknowledged for the PhD scholarship of Paul Joyce.

References

1. M. E. Lowe, The triglyceride lipases of the pancreas, J. Lipid Res., 2002, 43, 2007–2016.
2. 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, 673–691.
3. A. Tan, A. Martin, T.-H. Nguyen, B. J. Boyd and C. A. Prestidge, Hybrid Nanomaterials that Mimic the Food Effect: Controlling Enzymatic Digestion for Enhanced Oral Drug Absorption, Angew. Chem., Int. Ed., 2012, 51, 5475–5479.
4. A. Pandey, S. Benjamin, C. R. Soccol, P. Nigam, N. Krieger and V. T. Soccol, The realm of microbial lipases in biotechnology, Biotechnol. Appl. Biochem., 1999, 29, 119–131.
5. P.-Y. Stergiou, A. Foukis, M. Filippou, M. Koukouritaki, M. Parapouli, L. G. Theodorou, E. Hatziloukas, A. Afendra, A. Pandey and E. M. Papamichael, Advances in lipase-catalyzed esterification reactions, Biotechnol. Adv., 2013, 31, 1846–1859.
6. A. L. Paiva, V. M. Balcão and F. X. Malcata, Kinetics and mechanisms of reactions catalyzed by immobilized lipases, Enzyme Microb. Technol., 2000, 27, 187–204.
7. Y. Yesiloglu and I. Kilic, Lipase-catalyzed esterification of glycerol and oleic acid, J. Am. Oil Chem. Soc., 2004, 81, 281–284.
8. J. Tantrakulsiri, N. Jeyashoke and K. Krisanangkura, Utilization of rice hull ash as a support material for immobilization of Candida cylindracea lipase, J. Am. Oil Chem. Soc., 1997, 74, 173–175.
9. B. Andualema and A. Gessesse, Microbial Lipases and Their Industrial Applications: Review, Biotechnology, 2012, 11, 100.
10. A. Houde, A. Kademi and D. Leblanc, Lipases and their industrial applications: an overview, Appl. Biochem. Biotechnol., 2004, 118, 155.
11. E. Rogalska, I. Douchet and R. Verger, Microbial lipases: structures, function and industrial applications, Biochem. Soc. Trans., 1997, 25, 161.
12. 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, pp. 265–304.
13. 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, 20271.
14. 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, 2779–2788.
15. P. Joyce, C. P. Whitby and C. A. Prestidge, Bioactive Hybrid Particles from Poly(D,L-lactide-co-glycolide) Nanoparticle Stabilized Lipid Droplets, ACS Appl. Mater. Interfaces, 2015, 7, 17460–17470.
16. P. Reis, K. Holmberg, H. Watzke, M. E. Leser and R. Miller, Lipases at interfaces: a review, Adv. Colloid Interface Sci., 2009, 147–148, 237–250.
17. C. Eydoux, S. Spinelli, T. L. Davis, J. R. Walker, A. Seitova, S. Dhe-Paganon, A. De Caro, C. Cambillau and F. Carrière, Structure of human pancreatic lipase-related protein 2 with the lid in an open conformation, Biochemistry, 2008, 47, 9553–9564.
18. E. Serra, Á. Mayoral, Y. Sakamoto, R. M. Blanco and I. Díaz, Lipase immobilization in ordered mesoporous materials, Elsevier Science & Technology, 2007, pp. 897–900.
19. Z. Tao, Mesoporous silica-based nanodevices for biological applications, RSC Adv., 2014, 4, 18961–18980.
20. M. Nikolic, V. Srdic and M. Antov, Immobilization of lipase into mesoporous silica particles by physical adsorption, Biocatal. Biotransform., 2009, 27, 254.
21. R. M. Blanco, P. Terreros, M. Fernández-Pérez, C. Otero and G. Díaz-González, Functionalization of mesoporous silica for lipase immobilization: characterization of the support and the catalysts, Journal of Molecular Catalysis B: Enzymatic, 2004, 30, 83–93.
22. H. Gustafsson, C. Thörn and K. Holmberg, A comparison of lipase and trypsin encapsulated in mesoporous materials with varying pore sizes and pH conditions, Colloids Surf., B, 2011, 87, 464–471.
23. Y. Kang, J. He, X. D. Guo, X. Guo and Z. H. Song, Influence of pore diameters on the immobilization of lipase in SBA-15, Ind. Eng. Chem. Res., 2007, 46, 4474–4479.
24. K. J. Dussan, C. A. Cardona, O. H. Giraldo, L. F. Gutiérrez and V. H. Pérez, Analysis of a reactive extraction process for biodiesel production using a lipase immobilized on magnetic nanostructures, Bioresour. Technol., 2010, 101, 9542–9549.
25. C. Otero, E. Pastor, V. Fernández and A. Ballesteros, Influence of the support on the reaction course of tributyrin hydrolysis catalyzed by soluble and immobilized lipases, Appl. Biochem. Biotechnol., 1990, 23, 237–247.
26. P. Reis, K. Holmberg, T. Debeche, B. Folmer, L. Fauconnot and H. Watzke, Lipase-catalyzed reactions at different surfaces, Langmuir, 2006, 22, 8169–8177.
27. 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.
28. 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, 10198–10207.
29. M. Rodahl, F. Hook, A. Krozer, P. Brzezinski and B. Kaesmo, Quartz crystal microbalance setup for frequency and Q-factor measurements in gaseous and liquid environments, Rev. Sci. Instrum., 1995, 66, 3924–3930.
30. T. Snabe, M. T. Neves-Petersen and S. B. Petersen, Enzymatic lipid removal from surfaces—lipid desorption by a pH-induced “electrostatic explosion”, Chem. Phys. Lipids, 2005, 133, 37–49.
31. T. Snabe and S. B. Petersen, Lag phase and hydrolysis mechanisms of triacylglycerol film lipolysis, Chem. Phys. Lipids, 2003, 125, 69–82.
32. E. I. Ag, Aerosil, http://https://www.aerosil.com/product/aerosil/en/Pages/default.aspx, 2013.
33. G. Sauerbrey, Verwendung von Schwingquarzen zur Wägung dünner Schichten und zur Mikrowägung, Z. Phys., 1959, 155, 206–222.
34. F. Höök, B. Kasemo, T. Nylander, C. Fant, K. Sott and H. Elwing, Variations in coupled water, viscoelastic properties, and film thickness of a Mefp-1 protein film during adsorption and cross-linking: a quartz crystal microbalance with dissipation monitoring, ellipsometry, and surface plasmon resonance study, Anal. Chem., 2001, 73, 5796.
35. C. Wiener, R. Weiss, C. White and B. Vogt, Limitations in interpretation of Quartz Crystal Microbalance (QCM) beyond the rigid (Sauerbrey) to viscoelastic (lossy) transition, American Physical Society March Meeting 2014, Denver, Colorado, 2014.
36. D. J. Ericsson, A. Kasrayan, P. Johansson, T. Bergfors, A. G. Sandström, J.-E. Bäckvall and S. L. Mowbray, X-ray Structure of Candida antarctica Lipase A Shows a Novel Lid Structure and a Likely Mode of Interfacial Activation, J. Mol. Biol., 2008, 376, 109–119.
37. P. Reis, K. Holmberg, R. Miller, J. Krägel, D. O. Grigoriev, M. E. Leser and H. J. Watzke, Competition between lipases and monoglycerides at interfaces, Langmuir, 2008, 24, 7400–7407.
38. 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, 28070–28079.
39. H. Mu and C.-E. Høy, The digestion of dietary triacylglycerols, Prog. Lipid Res., 2004, 43, 105.
40. R. G. Avery and J. D. F. Ramsay, The sorption of nitrogen in porous compacts of silica and zirconia powders, J. Colloid Interface Sci., 1973, 42, 597–606.
41. K. Wannerberger and T. Arnebrant, Adsorption of Lipase to Silica and Methylated Silica Surfaces, J. Colloid Interface Sci., 1996, 177, 316–324.
42. F. X. Malacata, C. G. Hill and C. H. Amundson, Use of a lipase immobilized in a membrane reactor to hydrolyze the glycerides of butteroil, Biotechnol. Bioeng., 1991, 38, 853–868.
43. 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.
44. R. Birner-grünberger, H. Scholze, K. Faber and A. Hermetter, Identification of various lipolytic enzymes in crude porcine pancreatic lipase preparations using covalent fluorescent inhibitors, Biotechnol. Bioeng., 2004, 85, 147–154.
45. R. L. Segura, L. Betancor, J. M. Palomo, A. Hidalgo, G. Fernández-Lorente, M. Terreni, C. Mateo, A. Cortés, R. Fernández-Lafuente and J. M. Guisán, Purification and identification of different lipases contained in PPL commercial extracts: a minor contaminant is the main responsible of most esterasic activity, Enzyme Microb. Technol., 2006, 39, 817–823.
46. M. Wickham, M. Garrood, J. Leney, P. D. Wilson and A. Fillery-Travis, Modification of a phospholipid stabilized emulsion interface by bile salt: effect on pancreatic lipase activity, J. Lipid Res., 1998, 39, 623.
47. M. Wickham, P. Wilde and A. Fillery-Travis, A physicochemical investigation of two phosphatidylcholine/bile salt interfaces: implications for lipase activation, Biochim. Biophys. Acta Mol. Cell Biol. Lipids, 2002, 1580, 110–122.
48. 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, 770–781.
49. D. J. McClements, E. A. Decker and Y. Park, Controlling Lipid Bioavailability through Physicochemical and Structural Approaches, Crit. Rev. Food Sci. Nutr., 2009, 49, 48–67.
50. S. Mun, E. A. Decker, Y. Park, J. Weiss and D. J. McClements, Influence of Interfacial Composition on in vitro Digestibility of Emulsified Lipids: Potential Mechanism for Chitosan's Ability to Inhibit Fat Digestion, Food Biophys., 2006, 1, 21–29.
51. Y. Gargouri, R. Julien, G. Pieroni, R. Verger and L. Sarda, Studies on the inhibition of pancreatic and microbial lipases by soybean proteins, J. Lipid Res., 1984, 25, 1214.
52. Y. Gargouri, G. Pieroni, C. Rivière, A. Sugihara, L. Sarda and R. Verger, Inhibition of lipases by proteins. A kinetic study with dicaprin monolayers, J. Biol. Chem., 1985, 260, 2268.
53. 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, 20263.

---

Footnote

† Electronic supplementary information (ESI) available: Additional figures relating to QCM-D and lipolysis studies. See DOI: 10.1039/c6ra08934h

---