Article pubs.acs.org/Langmuir
Water-Binding Phospholipid Nanodomains and Phase-Separated Diacylglycerol Nanodomains Regulate Enzyme Reactions in Lipid Monolayers Teruyoshi Nagashima* and Shogo Uematsu Department of Physical Biochemistry, Graduate School of Pharmaceutical Sciences, University of Shizuoka, 52-1 Yada, Suruga Ward, Shizuoka City, Shizuoka 422-8526, Japan S Supporting Information *
ABSTRACT: Phospholipid 1-palmitoyl-2-oleoyl-sn-glycero-3phosphocholine (POPC) nanodomains covered with bound water as well as diacylglycerol 1-palmitoyl-2-oleoyl-sn-glycerol (POG) nanodomains separated from a lipid membrane were studied, using monolayer surfaces of POPC hydrolyzed by phospholipase C (PLC). The investigation was based on the analysis of compression isotherms and on atomic force microscope (AFM) observations of Langmuir−Blodgett (LB) films and Langmuir−Schaefer (LS) films. The results included reaction rate constants obtained by kinetic analysis of phosphocholine at surface pressures from 0.1 to 31 mN/m and determined by a luminol-enhanced chemiluminescence method. Monolayer elastic modulus values and fluorescence microscopic images confirmed that hydrolysis by PLC progressed in the intermediate monolayer between a liquid-expanded (L1) film and a liquid-condensed (L2) film at 2−17 mN/m. Furthermore, the intermediate film was confirmed to consist of L1 film and the POPC nanodomains in the L2 state are covered with bound water, conclusions based on the following AFM results: (1) nanodomains in POPC LS films were catalyzed by PLC, (2) POG nanodomains extended out from LB films of mixed POPC/ POG 9/1 (mol/mol) monolayers, and (3) POPC LS films were covered with bound water, as indicated by cross-sectional analysis. At the optimal surface pressure of 10 mN/m, when POPC nanodomains (L2), with internal diameters of ∼75 nm, were hydrolyzed by PLC, they shrank down into pockets of the same size as those that appeared with POG. The resulting pocket sizes on LS films were in agreement with POG nanodomain sizes on LB films. This study demonstrated that PLC reacted with POPC nanodomains (L2) dispersed in L1/L2 mixed phase monolayers selectively and that POG nanodomains were phase-separated from the monolayer as hydrolysis proceeded.
1. INTRODUCTION Lipolytic enzymes possess substrate specificity, and most react in lipid surfaces. Lipases and phospholipases are typical examples of hydrolases that react at the water/oil interfaces. Their substrate specificity is based not only on substrate chemical structures but also on lipid membrane structures that solubilize substrates. For example, to clarify a function of the second messenger diacylglycerol (DAG), which is produced from phospholipid hydrolysis by phospholipase C (PLC), it is necessary to first determine the physicochemical properties of the lipid membrane.1 Various models of enzyme reactions in lipid monolayers have been proposed by researchers for clarification of membrane functions. Verger and co-workers have analyzed the decomposition rates of phospholipid monolayers using phospholipase A2 (PLA2) and have proposed a kinetic model of an enzymatic hydrolysis reaction on lipid surfaces.2,3 Grainger et al. have observed by epifluorescence microscopy that PLA2 reacts at the boundaries of liquidexpanded (L1) film and liquid-condensed (L2) films in monolayers of phosphatidylcholine 1,2-dipalmitoyl-sn-glycero3-phosphocholine (DPPC).4 As the DPPC phase transition © 2015 American Chemical Society
temperature was greater than the measuring temperature, they were able to observe by epifluorescence microscopy the structure of the L2 film separated from the L1 film. Rao and Damodaran have proposed that protein adsorptions and lipolytic enzyme activities in lipid monolayers are related to interfacial water activity.5−8 It is unclear whether the above suggested models are applicable to other lipid monolayers that have optically invisible phase separations. For example, a 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) monolayer with a −2 °C phase transition temperature is interpreted as a L1 film at room temperature. Under these conditions, epifluorescence measurements using fluorescent phospholipid probes indicate a single homogeneous fluorescent phase9 and show no phase separation.10 Conversely, based on X-ray diffraction studies, the monolayer structure and phase transition were theoretically elucidated using a viewpoint of tilted hydrophobic acyl chains and ordering of the hydrophilic Received: October 2, 2014 Revised: December 5, 2014 Published: January 7, 2015 1479
DOI: 10.1021/la503906m Langmuir 2015, 31, 1479−1488
Article
Langmuir
Figure 1. (a) Hydrolysis reactions in POPC monolayers by PLC. Solid lines indicate changing surface pressure after PLC injection. Rates of hydrolysis analyzed by phosphocholine concentrations: closed triangles (10 mN/m, initial surface pressure), open triangles (17 mN/m), and closed squares (20 mN/m). Dotted curved lines are results of simulation of the pseudo-first-order reaction. The arrow and dotted vertical line shows PLC injection time (5 min). (b) π−A isotherm and E values in a monolayer: solid curved line, π−A isotherm in monolayer; closed circles, E values in monolayer; open circles, phase transition points; points A−D, phase transition based on inflection points of E values (A′−D′).
headgroups.11 However, as the POPC monolayers described above were in a liquid state at the measuring temperature, X-ray diffraction results of POPC monolayers have not yet been reported. In this report, we investigated hydrolysis reactions in POPC monolayers by PLC in terms of the nanoscale ordering of hydrophilic headgroups. Examination of the quantities of reaction-product phosphocholine showed that the reaction was dependent on the monolayer surface pressure and its kinetic reaction constant. The reasons for PLC activity changes with monolayer packing were investigated by determining the physicochemical properties of the monolayers. On the basis of the atomic force microscopic (AFM) image of Langmuir− Blodgett films of mixed monolayers of both POPC and DAG, DAG nanodomains were verified to phase-separate from the hydrophobic surface on a lipid membrane. Furthermore, AFM images from cross-sectional analysis of Langmuir−Schaefer (LS) films confirmed the presence of staircase-state bound water on the hydrophilic surface of POPC membranes. Moreover, AFM images of LS films of hydrolyzed monolayers confirmed that LS films showed water-binding POPC nanodomains in L1 phase and that nanodomains were hydrolyzed by PLC on the hydrophilic surfaces of POPC membranes. We conclude that monolayers were hydrolyzed by PLC in a mixed membrane composed of both L1 and L2 phases and that PLC reacted at the L2 phase of a nanosized disk or domain.
microscope (FM) imaging. The E values of monolayers at a given surface pressure were obtained from π−A data as12 dπT = −
E dA T Aπ
where Aπ is the area per molecule at the indicated surface pressure π. The fluorescent probe octadecyl rhodamine B (R18) was used for FM imaging. An additional probe, 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI), was used for the verification of the monolayer that was transferred onto substrates by Langmuir−Blodgett (LB) and Langmuir−Schaefer (LS) methods. LB films were prepared by depositing monolayers onto freshly cleaved mica substrates. Mica sheets were vertically extracted through the films. A cover glass alkylated with octadecyltrichlorosilane (OTS) was used for the preparation of the LS films. Films were prepared by horizontally contacting and removing monolayer sections with OTS-coated glass. The LB and LS films were dried overnight in a desiccator with dry silica gel at room temperature before the observation using atomic force microscopy (AFM). All monolayer measurements were performed on an air/water interface at 25.0 ± 0.1 °C, except for FM at 25.0 ± 0.5 °C. A standard pH 7.4 buffer composed of 30 mM Tris-HCl, 138 mM NaCl, and 5 mM CaCl2 was used as the subphase. All AFM images were observed using tapping mode conditions. Details of the materials and methods are provided in the Supporting Information (see Figure S1).
3. RESULTS 3.1. Kinetic Analysis of PLC Hydrolysis Reactions in Several Phases of POPC Monolayers. Under hydrolysis conditions, in which a fixed quantity of PLC was added to monolayers prepared from fixed quantities of POPC, kinetic reaction rates were examined at initial surface pressures of 2, 5, 10, 15, 17, and 20 mN/m. Both surface pressure changes and phosphocholine amounts were measured for 90 min (Figure 1a). Rate constant examination clearly confirmed the progression of the PLC reaction. It may be expected that, in investigating membrane-enzyme kinetics, the zero-order trough technique would be used to separate the reaction products from the substrate membrane.2,3,6−8 However, it is difficult to apply the zero-order trough method to the lipid system used in this study because of the properties of POG, including high lipophilicity, desorption stability when using β-cyclodextrin,13 and strong PLC adhesion to POG membranes. The mechanism of phosphocholine production from POPC monolayers was
2. EXPERIMENTAL SECTION POPC, DAG, 1-palmitoyl-2-oleoyl-sn-glycerol (POG), and PLC from Bacillus cereus were used in this study. For each POPC monolayer, PLC kinetic reaction rate constants were calculated for a variety of initial surface pressures based on the amount of hydrolysis product phosphocholine generated by the luminol-enhanced chemiluminescence method. Surface pressure (π) was measured by the Wilhelmy plate method and is defined as π = γo − γ where γo is the surface tension of the air/water interface and γ the surface tension in the presence of a lipid monolayer compressed to various packing densities. The state of the acquired surface pressure− area (π−A) isotherms was analyzed from both elastic modulus (E) values calculated from the isotherm as well as from fluorescent 1480
DOI: 10.1021/la503906m Langmuir 2015, 31, 1479−1488
Article
Langmuir Table 1. Hydrolysis Reaction and State of POPC Monolayer surface pressure (mN/m) monolayer properties a
initial concentration, % k, ×10−2 min−1 tlag, min π after tlag, mN/m POG concentration at tlag, % E, mN/m phase phase transition point a
6 × 10−2
1.3
2
5
10
15
17
20
31
0 0.4 G A
25 G/L1 B
0 NDb ND ND ND 28 L1
75 5.8 10 4.1 6.6 33 L1/L2
100 10 10 7.8 12.3 40 L1/L2
95 6.0 11 13 3.0 47 L1/L2
92 3.8 40 15 5.0 50 L2 C
0 ND >90 ND ND 58 L2
0 ND ND ND ND 76 collapse D
Percentage in contrast with POPC hydrolysis at 10 mN/m. bND, not detected.
Figure 2. FM images of POPC monolayers that included 0.1 mol % R18 at varying surface pressures: (a) 0.9 mN/m (G/L1 phase), (b) 2.1 mN/m (L1 phase), (c) 10.3 mN/m (L1/L2 mixed phase), and (d) 21.4 mN/m (L2 phase). Scale bar, 200 μm.
the above three pressures and the time-lag was ∼40 min. However, POPC hydrolysis was >90% and the reaction rate decreased only a little. At 20 mN/m, the time-lag exceeded 90 min and hydrolysis did not begin. Furthermore, hydrolysis at 2 mN/m was not detected. On the basis of the hydrolysis rate results, the hydrolyzed POPC amounts and simulated values resembled each other after the time-lag. The two steps of the reaction mechanism observed in monolayers after enzyme penetration were considered to be a primary and a secondary reaction. The primary reaction was the reaction preceding the time-lag. At the end of the primary reaction, POPC molecules were ∼10% hydrolyzed and surface pressures decreased a little from the initial pressures. The secondary reaction then proceeded after the time-lag as a pseudo-first-order reaction. In this reaction, there was a burst of hydrolysis soon after the time-lag although the decreasing surface pressures were similar to those of the primary reaction. To clarify the reason why the enzyme reaction depended on the initial surface pressure, π−A isotherm measurements were examined in monolayers. As the phase transition was not clear in a POPC monolayer’s smooth isotherm (Figure 1b, solid curved line), monolayer phases were next assessed by the analysis of monolayer elasticity using E values (Figure 1b, black dots). Four inflection points in the E values were initially labeled as phase transition points as follows: (A′) G phase to G/L1 phase, (B′) G/L1 phase to L1 phase, (C′) L1 phase to L2 phase, and (D′) collapse, where the G phase is the gas phase, L1 phase the liquid-expanded phase, and L2 phase the liquid-condensed phase. The points A′−D′ in the E values corresponded with points A−D in the isotherm. The E values ranged from 0.4 to 76 mN/m from A′ to D′. To verify the labeled phases, POPC monolayer structures were observed by FM. In FM images, bright areas indicated both the presence of the R18 probe and monolayer fluidity (Figure 2), with G and L1 phases coexisting in the monolayer (Figure 2a). Homogeneous bright images were observed in monolayers at
explained from penetration (eq 1) to complexation (eq 2) to hydrolysis (eq 3) as follows: k1
PLC XooY PLCP
(1)
k −1
k2
POPC + PLCP XooY POPC−PLCP k −2
(2)
POPC−PLCP + H 2O k3
XooY POG + phosphocholine + PLCP k −3
(3)
The first process represents that PLC in the subphase penetrates into the monolayer as “PLCP,” the second process that the PLCP and POPC form intramembranous complexes as “POPC−PLCP”, and the third process that the POPC−PLCP complex decomposes. When the phosphocholine production rate depends on a rate-limiting hydrolysis process, the reaction is considered to be a pseudo-first-order reaction as follows: a = a0[1 − e−k(t − tlag)]
Here, a is the simulated phosphocholine concentration, a0 is the initial POPC concentration, k is the kinetic reaction rate constant, and t and tlag are the elapsed time and the time-lag of the PLC reaction, specifically, the time until the pseudo-firstorder reaction starts after enzyme addition. The two dotted curved lines in Figure 1a show the simulated phosphocholine concentration. The kinetic reaction rate constants and initial lipid concentrations are shown in Table 1. Monolayer surface pressures slightly increased after PLC addition, indicating that PLC penetrated into the monolayer, following the first process (Figure 1a). At initial surface pressures of 5, 10, and 15 mN/m, all three time-lags were ∼10 min (Table 1). At 10 mN/m, both the k value and hydrolyzed POPC amounts were at their highest values. At 17 mN/m, the beginning of the pseudo-first-order reaction was slower than 1481
DOI: 10.1021/la503906m Langmuir 2015, 31, 1479−1488
Article
Langmuir
Figure 3. AFM images of the hydrophobic surface on (a) POPC LB film at the prepared surface pressure of 10 mN/m (areas of 500 × 500 nm2) and (b) POPC/POG 9/1 mixed LB film at 18.5 mN/m (1 × 1 μm2). Height scales at right. Both films were visualized with 0.2 mol % DiI.
surface pressures >2 mN/m (Figure 2b−d). Under these experimental conditions, a phase transition was not observed in either the L1 or L2 phases. The homogeneous bright images were similarly confirmed using the probes of DiI or 1,2dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(7-nitro-21,3-benzoxadiazol-4-yl) (ammonium salt) and were the same as that of previous studies.9,10 On the basis of FM imaging, the motility of POPC acyl chains was considered not to have decreased and probe solubility in the monolayers was a constant at 25 °C. On the basis of the above results, detailed phases were assigned to the L1 and L2 phases. First, the L1 phase was divided in two phases. As mentioned above, PLC hydrolyzed monolayers at initial surface pressures of 5, 10, and 15 mN/m, with hydrolysis not detected at 2 mN/m. Therefore, the L1 phase was divided as two different subphases. Interestingly, because of hydrolysis at the initial surface pressure of 5 mN/m, the pressure decreased to
