Effects of Proteoliposome Composition and Draw Solution Types on

Jan 11, 2013 - Separation Performance of Aquaporin-Based Proteoliposomes: ... 1. INTRODUCTION. Aquaporins are a large family of water transport protei...
1 downloads 0 Views 1MB Size
Article pubs.acs.org/est

Effects of Proteoliposome Composition and Draw Solution Types on Separation Performance of Aquaporin-Based Proteoliposomes: Implications for Seawater Desalination Using Aquaporin-Based Biomimetic Membranes Yang Zhao,†,‡ Ardcharaporn Vararattanavech,†,§ Xuesong Li,†,‡ Claus HélixNielsen,∥,⊥ Thomas Vissing,∥ Jaume Torres,†,§ Rong Wang,†,‡ Anthony G. Fane,†,‡ and Chuyang Y. Tang*,†,∥ †

Singapore Membrane Technology Centre, ‡School of Civil and Environmental Engineering, and §School of Biological Sciences, Nanyang Technological University, Singapore 639798 ∥ Aquaporin A/S, Copenhagen Bio Science Park (COBIS), DK2200, Copenhagen, Denmark ⊥ DTU Physics, Technical University of Denmark, DK2800 Kgs. Lyngby, Denmark S Supporting Information *

ABSTRACT: Aquaporins are a large family of water transport proteins in cell membranes. Their high water permeability and solute rejection make them potential building blocks for high-performance biomimetic membranes for desalination. In the current study, proteoliposomes were prepared using AquaporinZ from Escherichia coli cells, and their separation properties were characterized by stopped-flow measurements. The current study systematically investigated the effect of proteoliposome composition (lipid type, protein-to-lipid ratio (PLR), and the addition of cholesterol) on water permeability and NaCl retention. Among the various lipids investigated, 1,2-dioleoyl-snglycero-3-phosphocholine (DOPC)-based proteoliposomes were found to have excellent osmotic water permeability and NaCl reflection coefficient values. Increasing the PLR of DOPC proteoliposomes up to 1:200 increased their osmotic water permeability. However, further increase in the PLR reduced the osmotic water permeability probably due to the occurrence of defects in the proteoliposomes, whereas the addition of cholesterol improved their osmotic water permeation likely due to defects sealing. The current study also investigated the effect of major dissolved ions in seawater (e.g., Mg2+ and SO42−) on the stability of proteoliposomes, and design criteria for aquaporin-based biomimetic membranes are proposed in the context of desalination.

1. INTRODUCTION Aquaporins are a large family of water transport proteins in cell membranes.1,2 These proteins have highly defined nanoscale pores that allow water molecules to rapidly pass through and that effectively retain dissolved solutes.3 For example, after incorporating aquaporins into the lipid bilayer, the permeability of the reconstituted bilayer can be enhanced by an order of magnitude compared to that of the original bilayer while maintaining high retention against solutes.4 Due to such an ideal combination of separation properties of aquaporins,2,5 there has been significant interest in synthesizing aquaporinbased high-performance biomimetic membranes, especially for desalination applications.5−11 Reverse osmosis (RO) membrane based desalination has experienced rapid growth over the last few decades. Current RO membranes are generally thin film composite (TFC) type, where a polyamide rejection layer of ∼200 nm is formed by interfacial polymerization of diamine and trimesoyl chloride monomers.12,13 Despite the significant improvements in membrane separation properties over the last 30 years, modern © 2013 American Chemical Society

seawater RO membranes have relatively low water permeability of ∼1−2 L/(m2·h·bar) (∼2.8 × 10−12 to 5.6 × 10−12 m/ (s·Pa)).14−16 Compared to conventional TFC RO membranes, a recent theoretical study estimated that biomimetic membranes with an aquaporin-containing rejection layer could be 2 orders of magnitude more permeable.5 Recent studies on aquaporin-based biomimetic membranes have been focusing on membrane structure designs,11 such as proteoliposomes embedded TFC structure,10 polymer tethered biolayer,17 pore-suspending,7,8 and biomembrane on aperture partition arrays.18−21 Meanwhile, both lipid bilayer systems22 and aquaporin incorporation4,23,24 have been systematically studied. Nevertheless, there is still a lack of systematic comparison of separation properties of different lipid−protein systems, particularly their water permeability and NaCl Received: Revised: Accepted: Published: 1496

March 20, 2012 January 2, 2013 January 3, 2013 January 11, 2013 dx.doi.org/10.1021/es304306t | Environ. Sci. Technol. 2013, 47, 1496−1503

Environmental Science & Technology

Article

(Merck) and 3% OG. The sample was subjected to five times lysis cycles in a microfluidizer followed by centrifugation to remove the insoluble materials. The supernatant was passed through a Q-Sepharose fast flow column (Amersham Pharmacia), and the flow through was collected. The flowthough fraction was topped up with 250 mM NaCl before it was loaded onto a pre-equilibrated Ni−NTA column (BioRad). The protein was allowed to bind to the resin with gentle shaking at 4 °C overnight. The nickel resin with bound fusion protein was washed with 20 column volumes of buffer containing 20 mM Tris pH 8.0, 300 mM NaCl, 25 mM imidazole, 2 mM β-mercaptoethanol, and 10% glycerol. The bound proteins were eluted with elution buffer containing 20 mM Tris pH 8.0, 300 mM NaCl, 300 mM imidazole, 2 mM βmercaptoethanol, 10% glycerol, and 30 mM OG. The fractions containing the fusion protein were checked by gel electrophoresis and concentrated to the concentration of 5−10 mg/ mL with an Amicon concentrator, with a membrane molecular weight cutoff (MWCO) of 10 000 Da (Milipore). The protein concentration of AqpZ was determined by measuring the UV absorbance at 280 nm (AqpZ extinction coefficient = 35 090 1/ (M·cm); molecular weight = 24 524 Da). The concentrated AqpZ was kept frozen at −80 °C until use. 2.3. Liposome and Proteoliposome Preparation. Lipid vesicles were prepared by the film rehydration method25,26 (Supporting Information, S2). A 10 mg sample of lipid dissolved in 0.5 mL of chloroform was dried under nitrogen gas to form a thin lipid film. In some experiments, a predetermined amount of cholesterol was mixed with a given lipid (DOPC in this study) to form a cholesterol-containing lipid film. In either case, the resulting film was kept in a vacuum desiccator for at least 2 h. A 1 mL phosphate-buffered saline (PBS) buffer solution (pH 7.4) was used to rehydrate the lipid film, followed by three cycles of freeze−thaw treatment.27 The resulting solution contained unilamellar lipid vesicles with a wide size distribution. Liposomes with uniform size (see Section 2.4.1) were obtained by extruding the solution through a 200 nm pore size polycarbonate filter for 21 times using an Avestin extruder (Canada).28 Proteoliposomes were prepared by incorporating AqpZ into liposomes by the dialysis method.4 Briefly, an AqpZ solution was mixed with a second solution containing 10 mg/mL lipid vesicle and 1% detergent OG at a desired PLR (ranging from 1:800 to 1:50 mol/mol), followed by incubating at room temperature for 1 h. Dialysis tubing (Spectrum laboratories, USA, with MWCO 12−14 kDa) was used to remove OG from the proteoliposome solution by dialyzing it against a PBS buffer solution at pH 7.4 with 1000 times the volume for 3 days under room temperature (23 °C). During this period, the dialysis PBS buffer solution was changed once every day. After the 3 day dialysis, AqpZ was successfully reconstituted into lipid vesicles. 2.4. Liposome and Proteoliposome Characterization. 2.4.1. Size and ζ Potential Characterization. The sizes of liposome and proteoliposome were determined using a Zetasizer Nano ZS (Malvern Instruments Limited, UK). The measured diameter was used for water permeability calculation of liposomes or proteoliposomes. In addition, the size determination was also used to monitor vesicle solution quality. In this study, the solution polydispersity index (PDI) was consistently smaller than 0.2, indicating a narrow size distribution of the vesicles.29 A uniform size distribution helps to minimize errors in water permeability determination (Section 2.4.2) of vesicles. ζ potential values of liposomes

retention properties. Furthermore, most of the prior studies on aquaporin lipid layers use simple solutes such as sucrose. In contrast, seawater in the typical desalination process contains many different types of ions, such as Na+, Mg2+, Cl−, SO42−, and so forth. Their effect on the separation performance of the reconstituted lipid layers has to been understood to achieve optimized membrane desalination performance. The objective of the current research was to systematically study the effects of the types of lipid, the protein-to-lipid ratio (PLR), and the presence of dissolved ions on the aquaporin− lipid bilayer properties. Results from the current study lead to design criteria for optimal aquaporin-based biomimetic membranes for desalination purpose.

2. MATERIALS AND METHODS 2.1. Chemicals. Unless mentioned otherwise, ultrapure water from a Milli-Q ultrapure water system (Milli-pore Singapore Pte Ltd.) with a resistivity of 18.2 MΩ·cm was used for preparing reagents in this study. Analytical grade NaCl, KCl, Na2HPO4, KH2PO4, MgCl2, MgSO4, and Na2SO4 with purities over 99% were purchased from Merck (Germany). Sucrose (ultrapure grade) was obtained from USB Corporation (Cleveland, OH, USA). Chemicals used in AquaporinZ (AqpZ) expression and purification, including ampicillin, chloramphenicol, isopropyl-βthiogalactoside (IPTG), Tris, β-mercaptoethanol, glycerol, and lysozyme, were obtained from Sigma−Aldrich and were either ACS (American Chemical Society) grade or SigmaUltra grade. Benzonase was purchased from Merck. Ni−NTA resin was purchased from Bio-Rad. n-Octyl-β-D-glucopyranoside (OG, ultrapure grade, Merck, Germany) was used as detergent during proteoliposome preparation. Lipids used in the current study include 1,2dioleoyl-sn-glycero-3-phosphocholine (DOPC), Escherichia coli extract lipid, 1,2-diphytanoyl-sn-glycero-3-phosphocholine (DPhPC), 1,2-dioleoyl-sn-glycero-3-phospho-(1′-rac-glycerol) (sodium salt) (DOPG), and 1,2-dioleoyl-3-trimethylammonium-propane (chloride salt) (DOTAP) (see the Supporting Information, S1, for their chemical structures). These lipids were provided in chloroform solutions (20 mg lipid/mL) by Avanti Polar Lipids (Alabaster, AL, USA). Lipids were kept at −20 °C until use. All chemicals were used without further purification. In some experiments, cholesterol (Avanti Polar Lipids) was used as an additive for proteoliposome preparation. 2.2. AqpZ Expression and Purification. AqpZ, an aquaporin found in the E. coli cell membrane, was chosen in this study due to its availability and its well-characterized properties. 4 Expression and purification of AqpZ was performed according to previously reported protocols.4,9 The pET3a plasmid containing the AqpZ gene was transformed into the E. coli competent cell strain C41-pLysS for protein overexpression. Cells from a single colony was picked to inoculate in Terrific Broth (TB) media with 100 μg/mL ampicillin and 34 μg/mL chloramphenicol and was grown overnight at 37 °C. Overnight cultures were diluted 100-fold into fresh TB broth and propagated to a density of about 1.2− 1.5 (OD at 600 nm). The cells were induced with 1 mM IPTG and grown at 37 °C for 3 h before centrifugation. AqpZ was purified by using ion-exchange chromatography followed by Ni−NTA affinity chromatography. The harvested cells were resuspended in anion-exchange binding buffer (20 mM Tris pH 8.0, 50 mM NaCl, 2 mM β-mercaptoethanol, 10% glycerol) in the presence of 0.4 mg/mL lysozyme, 50 units of bensonase 1497

dx.doi.org/10.1021/es304306t | Environ. Sci. Technol. 2013, 47, 1496−1503

Environmental Science & Technology

Article

and proteoliposomes were also measured by a Zetasizer Nano ZS. 2.4.2. Water Permeability and Solute Reflection Coefficient Evaluation. A SX20 stopped-flow spectrometer (Applied Photophysics, UK) was used to characterize the water permeability of liposomes and proteoliposomes (Supporting Information, S2). The osmolarities of all solution used in the stopped-flow tests were characterized by a Vapor Osmometer 5520 (Wescor, Inc., Logan, UT, USA). The fluorescence kinetic mode was chosen for all stopped-flow tests, with a light source of 500 nm wavelength. Rapid mixing of the sample solution and the draw solution was driven by 8 atm pressurized nitrogen gas with a dead time of 500 μs. In all the stopped-flow measurements, the temperature was maintained at 23 ± 1 °C (room temperature). Vesicle volume change rate was recorded as a function of time, based on the relationship between the fluorescence signal and the vesicle volume. Vesicle volume reduction was due to water transport outward, which will be affected by the osmolarity difference across the vesicle as well as the water permeability of the vesicle. In a typical stopped-flow experiment, both the draw solution and the solution inside the vesicles had an identical PBS buffer concentration, such that the water permeation was induced by the draw solute (e.g., sucrose or NaCl) concentration. The water permeability can be calculated by the following:4,5 Pf =

k

(

)

S *Vw*Δosm V0

(1)

In eq 1, k is the average rate constant for the stopped-flow fluorescence signal decay (which indicates the rate of vesicle volume reduction). Its value is obtained by curve fitting of stopped-flow results to a single order exponential (Fluorescence = a * e−k*t + c, where a and c are fitting parameters). S/ V0 is the surface area to initial volume ratio of the vesicles (refer to the Supporting Information, S3, for information on vesicle size and typical S/V0 values), Vw is the partial molar volume of water (0.018 L/mol), and Δosm is the difference in osmolarity between the intravesicular and extravesicular aqueous solutions. In addition to NaCl, other inorganic draw solutes such as MgCl2, MgSO4, and Na2SO4 were also evaluated due to the significant concentrations of Mg2+ and SO42− in typical seawater. Furthermore, additional stopped-flow measurements were also performed for mixtures of sucrose and MgCl2 (in PBS buffer). The reflection coefficient σ of a given solute can be determined by comparing the measured water permeability using the particular draw solute (Pf,solute) to that for a reference solute (Pf,reference) with the same Δosm. Sucrose was used as the reference solute in the current study, since it is a relatively large molecule with nearly complete retention by lipid vesicles.4 Thus, the apparent reflection coefficient of smaller draw solutes such as NaCl can be calculated by the following:30 σ=

Figure 1. Stopped-flow results of DOPC liposomes and proteoliposomes (PLR = 1:200). (a) Normalized light scattering curves using NaCl or sucrose as the draw solution. (b) The water permeability coefficient (based on sucrose draw solution) and the NaCl reflection coefficient (based on eq 2) of DOPC liposomes and proteoliposomes. Stopped-flow experimental conditions: both NaCl and sucrose solution had an osmolarity of 993 mosm/L, while the vesicles contained PBS buffer solution with osmolarity of 281 mosm/L. The stopped-flow measurements were performed by mixing an equal volume of the draw solution and the vesicles, which gave an osmolarity difference of 356 mosm/L across the vesicles. Error bars indicate the standard deviation in test results based on three to five independent samples.

mosm/L), the kinetic rate constant for the DOPC liposomes was ∼20.5 1/s. In comparison, a much higher rate constant (188 1/s) was observed for the reconstituted DOPC AqpZ proteoliposomes (prepared at a nominal PLR of 1:200). The corresponding water permeability of the proteoliposomes was 6.9 × 10−4 m/s, which is an order of magnitude higher than the corresponding DOPC liposomes. A similar trend was also observed when NaCl was used as the draw solution, confirming the high water permeability of AqpZ. On the basis of a PLR of 1:200, the water permeability per AqpZ monomer can be estimated based on the proteoliposome area coverage of AqpZ to be 3.2 × 10−20 m3/s,31,32 which is in reasonable agreement with reported literature data (∼4 × 10−20 m3/s)33. Figure 1b shows the NaCl reflection coefficients (eq 2, using sucrose as the reference solute5) of DOPC liposomes and proteoliposomes. In both cases, the apparent reflection coefficients for NaCl were close to unity. This suggests that

Pf,solute Pf,reference

(2)

3. RESULTS AND DISCUSSION 3.1. Separation Properties of DOPC Liposomes and Proteoliposomes. Figure 1a shows typical stopped-flow measurements of liposomes and proteoliposomes prepared from DOPC. By using sucrose as a draw solute (Δosm = 356 1498

dx.doi.org/10.1021/es304306t | Environ. Sci. Technol. 2013, 47, 1496−1503

Environmental Science & Technology

Article

DOPG- and DOTAP-based proteoliposomes were results of unfavorable interactions between AqpZ and these lipids. In the case of DPhPC, previous studies have shown that this lipid was not optimal for aquaporin reconstitution.34 With respect to DOPG and DOTAP, we note that both lipids were strongly charged (due to the presence of the phospho group and the trimethylammonium group, respectively, see Figure S1, Supporting Information). ζ potential values of the DOPG and DOTAP liposomes (measured in PBS buffer solution at pH 7.4 and 23 °C) were ∼−30 mV and ∼+30 mV, significantly more charged compared to other types of liposomes (Supporting Information, S4), leading to the speculation that the poor water permeabilities of the DOPG- and DOTAPbased proteoliposomes were partially due to poor protein incorporation and/or defect creation resulting from the unfavorable charge interaction between AqpZ and these lipids.35 Figure 3 presents the effect of PLR on the water permeability of E. coli extract lipids and DOPC-based proteoliposomes. For

both the DOPC lipid bilayer and AqpZ had good NaCl rejection. 3.2. Effect of Proteoliposome Composition on Separation Properties. 3.2.1. Effect of Proteoliposome Composition on Water Permeation. The effects of proteoliposome composition (lipid type, PLR, and the inclusion of cholesterol) on separation properties were investigated in the current study. Water permeability reported in this section was based on stopped-flow measurements using sucrose as the draw solute. Figure 2 presents the water permeabilities of liposomes

Figure 2. Water permeability coefficients of liposomes and proteoliposomes (PLR at 1:200) prepared from various lipids. Stopped-flow experiments were performed using sucrose solution as a draw solution and a net osmolarity difference of 356 mosm/L across the vesicles. Error bars indicate the standard deviation in test results based on three independent samples.

and proteoliposomes prepared from the various lipids, where the PLR of the proteoliposomes was fixed at 1:200. In general, all liposomes investigated exhibited low water permeability (