Giant unilamellar vesicles containing Rhodamine 6G as a marker for immunoassay of bovine serum albumin and lipocalin-2

Analytical Biochemistry 505 (2016) 66e72

Contents lists available at ScienceDirect

Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio

Giant unilamellar vesicles containing Rhodamine 6G as a marker for immunoassay of bovine serum albumin and lipocalin-2 Misato Sakamoto a, Atsushi Shoji b, Masao Sugawara a, * a b

Department of Chemistry, College of Humanities and Sciences, Nihon University, Setagaya, Tokyo 156-8550, Japan School of Pharmacy, Tokyo University of Pharmacy and Life Sciences, Hachioji, Tokyo 192-0392, Japan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 January 2016 Received in revised form 8 April 2016 Accepted 18 April 2016 Available online 23 April 2016

Functionalized giant unilamellar vesicles (GUVs) containing a fluorescence dye Rhodamine 6G is proposed as a marker in sandwich-type immunoassay for bovine serum albumin (BSA) and lipocalin-2 (LCN2). The GUVs were prepared by the electroformation method and functionalized with anti-BSA antibody and anti-LCN2 antibody, respectively. The purification of antibody-modified GUVs was achieved by conventional centrifugation and a washing step in a flow system. To antigen on an antibody slip, antibody-modified GUVs were added as a marker and incubated. After wash-out of excess reagents and lysis of the bound GUVs with Triton X-100, the fluorescence image was captured. The fluorometric immunoassays for BSA and LCN2 exhibited lower detection limits of 4 and 80 fg ml
1, respectively. © 2016 Elsevier Inc. All rights reserved.

Keywords: Giant unilamellar vesicles Sandwich-type immunoassay Rhodamine 6G Lipocalin-2 BSA

Vesicular bilayer membranes such as uni- and multilamellar liposomes, often functionalized with receptors, have found a variety of applications in biosensing studies [1e4]. Encapsulation of a dye into the inner aqueous phase of liposomes, the lysis of the liposomes with a detergent, and successive measurements of spectroscopic signal from the released dye are the common procedures in the so-called sandwich-type liposome immunoassay. The larger the amount of dyes encapsulated, the larger the spectroscopic signal after the lysis will be as far as the signal transduction step is appropriately designed. Giant unilamellar vesicles (GUVs) of 3e100 mm in diameter can be formed by several protocols [5], including electroformation on either platinum wires [6e8] or indium tin oxide-coated glasses [9],

Abbreviations used: GUV, giant unilamellar vesicle; LUV, large unilamellar vesicle; SLI, sandwich-type liposome immunoassay; BSA, bovine serum albumin; LCN2, lipocalin-2; DPhPC, 1,2-diphytanoyl-sn-glycero-3-phosphocholine; DOPE, 1,2dioleoyl-sn-glycero-3-phosphoethanolamine; B-cap-PE, 1,2-dioleoyl-sn-glycero-3phosphoethanolamine-N-(cap biotinyl); EDC, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride; Chol, cholesterol; DMSO, dimethyl sulfoxide; NHS, Nhydroxysuccinimide; ELISA, enzyme-linked immunosorbent assay; R6G, Rhodamine 6G; MTS, 3-mercaptopropyltrimethoxysilane; SPDP, N-succinimidyl 3-(2pyridyldithio)-propionate; NEM, N-ethylmaleimide; VPP, Vesicle Prep Pro; ITO, indium tin oxide; S/N, signal/noise; HSA, human serum albumin. * Corresponding author. E-mail address: [email protected] (M. Sugawara). http://dx.doi.org/10.1016/j.ab.2016.04.011 0003-2697/© 2016 Elsevier Inc. All rights reserved.

swelling in solution [10,11], solvent exchange [12], and other methods [13e16]. The electroformation in high concentration of sugars is most common because of its easiness of GUV preparation. Because GUVs can be imaged using a variety of optical microscopic approaches, the manipulation of GUVs is relatively easy. In addition, because of their large volume that allows encapsulating a large amount of dyes, GUVs will be attractive as a marker in immunoassay. Rough calculation indicates that GUVs of 10 mm in diameter prepared in the presence of a 1-mmol L
1 marker encapsulate approximately 108 molecules. This amount is much larger than that (200 molecules) for large unilamellar vesicles (LUVs; 100 nm in diameter). However, to our knowledge, GUVs have not been used for designing sandwich-type liposome immunoassay (abbreviated as SLI), seemingly because the recovery of functionalized GUVs from the reaction mixtures is difficult. In general, the recovery of antibody-modified liposomes requires the removal of coupling agents, unreacted antibody, and unreacted liposomes. For this purpose, antibody is often linked to liposomes immobilized on a solid support, that is, heterogeneous modification [2,3]. This approach has the advantage that both unreacted antibody and a coupling reagent are easy to be washed out, but the immobilization step is unsuitable for recovering antibody-modified liposomes. Alternatively, in homogeneous modification, antibody-modified liposomes are isolated by size exclusion chromatography [17e19]. However, the size exclusion

M. Sakamoto et al. / Analytical Biochemistry 505 (2016) 66e72

chromatography is capable of separating only liposomes smaller than 300e500 nm [20,21]. The density gradient centrifugation [17,22,23] and ultracentrifugation [24] are also used for recovering antibody-modified unilamellar liposomes, but these methods necessitate specialized instruments and training. In addition, mechanically stress is likely to cause antibody aggregation [25]. Consequently, lower speed centrifugation is simple and will be a more preferable choice for recovering antibody-modified GUVs. In this article, we propose the use of GUVs containing a fluorescence dye Rhodamine 6G as a marker in SLI. The GUVs, which are prepared by the electroformation method, are modified with anti-BSA (bovine serum albumin) and anti-LCN2 (lipocalin-2), respectively, in a homogeneous protocol. The unreacted GUVs are removed by centrifugation under almost the same conditions as those used for multilamellar liposomes, followed by a washing step in a flow system. The proposed assay is applied for detecting lipocalin-2, which is a neutrophil gelatinase-associated lipocalin [26,27], in highly diluted human serum. The quantification of lipocalin-2 in human serum and/or urine is important for biomarker of various diseases [26,28e31]. Materials and methods Materials 1,2-Diphytanoyl-sn-glycero-3-phosphocholine (DPhPC, 10 mg ml
1 chloroform solution), 1,2-dioleoyl-sn-glycero-3phosphoethanolamine (DOPE, 10 mg ml
1 chloroform solution), and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(cap biotinyl) (sodium salt) (powder, B-cap-PE) were purchased from Avanti Polar Lipids (Alabaster, AL, USA). 1-Ethyl-3-(3dimethylaminopropyl)carbodiimide hydrochloride (EDC) was obtained from Dojindo Laboratories (Kumamoto, Japan). Cholesterol (Chol), D(
)-sorbitol, D-(þ)-glucose, dimethyl sulfoxide (DMSO, dehydrated), and N-hydroxysuccinimide (NHS) were obtained from Wako Chemicals. Chol was recrystallized three times from methanol. Anti-BSA antibody (polyclonal, rabbit) (anti-BSA) was obtained from Funakoshi (Tokyo, Japan). Albumin from bovine serum (BSA, >97%), lipocalin-2 (human recombinant), human serum (male AB), transferrin human, haptoglobin from pooled human plasma (lyophilized powder), and sucrose were obtained from SigmaeAldrich Chemical (St. Louis, MO, USA). A human lipocalin-2/NGAL enzyme-linked immunosorbent assay (ELISA) kit was obtained from R&D Systems (Minneapolis, MN, USA). Antilipocalin-2 antibody (monoclonal) was obtained from R&D Systems. Rhodamine 6G (R6G) was obtained from Tokyo Kasei (Tokyo, Japan). 3-Mercaptopropyltrimethoxysilane (MTS, >99.9%) was obtained from ShineEtsu Chemical (Tokyo, Japan). N-Succinimidyl 3-(2-pyridyldithio)-propionate (SPDP) and N-ethylmaleimide (NEM) were obtained from Thermo Scientific (Rockford, IL, USA). All other chemicals used were of analytical reagent grade. Milli-Q water (Millipore reagent water system, Bedford, MA, USA) was used throughout the experiments. Micro cover glasses (diameter 15 mm, thickness 0.12e0.17 mm) were obtained from Matsunami Glass Industries (Tokyo, Japan).

67

Preparation of R6G-encapuslating GUVs GUVs were prepared from a lipid mixture consisting of 4.2 mg DPhPC, 0.22 mg cholesterol, and 0.021 mg DOPE in chloroform at a molar ratio of 9:1:0.050 by the electroformation method. Here, 20 ml of the lipid mixture in chloroform was put on an indium tin oxide (ITO)-coated slide glass and air-dried to a lipid film. After setting a silicon O-ring on it, 0.30 ml of 1 mol L
1 sorbitol containing 1.0 mM R6G was added. Then, an electric field (5 Hz, ac 3 V) was applied through two ITO-coated slide glasses for 2 h with VPP. The GUV suspension (abbreviated as R6GeGUVs) was recovered and stored in a microcentrifuge tube (1.5 ml). Preparation of immuno-GUVs The functionalization of GUVs with protein (anti-BSA or antiLCN2) was carried out as follows (see Fig. S1a in online supplementary material). A 10-ml portion of 16.5 mg ml
1 antiBSA containing 22 mM NHS and 55 mM EDC in 1 mol L
1 sorbitol was mixed with 0.10 ml of R6GeGUVs and incubated at room temperature for 15 min. The suspension in a plastic centrifugation tube (1.5 ml) was centrifuged at 14,000 g for 30 min at 20  C. After centrifugation, a 50-ml portion of the suspension from the top surface was taken out with a micropipette and its fraction was discarded. Then, the remaining part of the suspension (i.e., 50 ml) was collected with a micropipette into a plastic centrifuge tube. The collected GUVs hereafter are abbreviated as antiBSAeR6GeGUVs. The anti-BSAeR6GeGUVs were stored at 4  C until use. Preparation of GUVs modified with an anti-LCN2 was performed in the same manner as described for anti-BSAeR6GeGUVs. The immuno-GUVs hereafter are abbreviated as antiLCN2eR6GeGUVs. Preparation of anti-BSA or anti-LCN2 modified slips An antibody slip used for immobilizing antibody was prepared as follows (see Fig. S1b in supplementary material). First, cover slips were cleaned in 1 M NaOH for 3 h and washed thoroughly with Milli-Q water. The one-side surface of the cover slip was treated with 0.10 ml of 50% (v/v) MTS in toluene for 60 min at room temperature. The MTS-modified cover slip was washed thoroughly with toluene and dried at room temperature. On the other hand, a 0.10-ml portion of an anti-BSA antibody (not fragmented) or an anti-LCN2 antibody (not fragmented) solution (1 mg ml
1) was mixed with 1.25 ml of 20 mmol L
1 SPDP in DMSO (anhydrous) and incubated at room temperature for 1 h. Unreacted SPDP was removed with a dye removal column (Thermo Scientific) and replaced by 0.15 mol L
1 NaCl solution containing 0.1 mol L
1 NaH2PO4/NaOH buffer (pH 7.2) (a PBS buffer). Then, 0.10 ml of the activated antibody (0.1 mg ml
1) was placed on an MTS-modified slip and incubated for 60 min at 4  C. After washing with Milli-Q water, the slip was incubated with 53 mg ml
1 NEM in a PBS buffer in order to block unreacted sulfhydryl sites. The antibody slip was washed with 6 ml of Milli-Q water and stored in a PBS buffer until use.

Apparatus Procedure for immunoassay All fluorometric images of liposomes on cover slips were obtained with a FluorImager 595 (Molecular Dynamics, Sunnyvale, CA, USA). Vesicle Prep Pro (VPP; Nanion Technologies, Germany) was used for electroformation of GUVs. A Denki Kagaku Keiki (Tokyo, Japan) glass electrode pH meter (model IOL30) was used for pH measurements.

The procedure described here is that for the assay of BSA. However, the assay of lipocalin-2 was performed in the same manner as described for BSA except that anti-LCN2eR6GeGUVs were diluted 300 times with 10 mmol L
1 NaH2PO4 containing 10 mmol L
1 NaCl buffer (pH 7.4, a PB buffer) before its use.

M. Sakamoto et al. / Analytical Biochemistry 505 (2016) 66e72

An anti-BSA slip (diameter 15 mm, thickness 0.12e0.17 mm), just fitting into a flow chamber, was set in the chamber containing a PB buffer. A known amount of BSA in a PB buffer (15 ml) was added and incubated for 15 min. After running a PB buffer for 3 min at a flow rate of 0.2 ml min
1, a 15-ml portion of anti-BSAeR6GeGUVs suspension was added and incubated for 30 min (40 times dilution; Fig. S2). Excess GUVs and coupling agents were washed off by running a PB buffer for 15 min at a flow rate of 0.2 ml min
1. Finally, the chamber solution was removed with a micropipette, and then 0.60 ml of a PB buffer was added. The chamber along with the buffer solution and the anti-BSA slip was transferred on a stage of a FluorImager 595 imager. First, the fluorometric images of the slip were captured with a FluorImager 595 (excitation 514 nm, emission 570 nm). Then, a 10-ml portion of 20% (v/v) Triton X-100 in a PB buffer was added and mixed gently with a micropipette. The fluorometric images were captured again. The captured images were viewed and analyzed by an Image Quant version 5.0. The mean fluorescence intensity for the whole solution area except for the edge of the chamber was obtained. The quantification of BSA was based on

DF ¼

Ft
F0  100; F0

(1)

where Ft is the mean fluorescence intensity after Triton X-100 treatment and F0 is the mean fluorescence intensity before Triton X-100 treatment. Assay of LCN2 in human serum The sensitivity of the current assay was high enough to allow 5  104 times dilution of serum with a PB buffer. At first, an 80-ml portion of human serum (male type AB) was mixed with 920 ml of a PB buffer. The mixture (20 ml) was further diluted with 980 ml of a PB buffer. Finally, 10 ml of the diluted serum and 10 ml of 80 pg ml
1 LCN2 (spiked concentration: 1.0 pg ml
1) were mixed. For a nonspiked case, 10 ml of a PB buffer was added instead of spiked LCN2. For assay, a 15-ml portion of the serum sample was transferred on an anti-LCN2 slip in a chamber containing 585 ml of a PB buffer. After incubation for 15 min, the slip was washed with a PB buffer for 3 min at a flow rate of 0.2 ml min
1. Then, a 2-ml portion of antiLCN2eR6GeGUVs (see above) was added and treated as described above for BSA. Results and discussion

A

20 μm

B

Relative fluorescence (%)

68

50 40 30 20 10 0

Triton X-100

buffer

Fig.1. Encapsulation of R6G into GUVs consisting of DPhPC/Chol/DOPE (9:1:0.05, mol/ mol). (A) Photo of R6G-encapsulating GUVs in 1 mol L
1 D(
)-sorbitol prepared by the electroformation method. (B) Relative fluorescence intensity obtained with a FluorImager 595 after rapturing biotinylated GUVs consisting of DPhPC/Chol/DOPE/B-capPE (9:1:0.05:0.025, mol/mol) on avidin slips with Triton X-100.

fluorescence from R6G (Fig. 1B). These results show that R6G molecules at self-quenching concentration were successfully entrapped in the GUVs.

Encapsulation of R6G into GUVs Functionalization and collection of GUVs The formation of GUVs of DPhPC/Chol/DOPE (9:1:0.05, mol/ mol) was performed by the electroformation method in the presence of 1 mol L
1 sorbitol containing R6G at self-quenching concentration (i.e., 1.0 mmoL L
1). Fig. 1A shows a photo of R6Gentrapped GUVs after 10 times dilution with a PB buffer. The circular spot is due to R6G entrapped in the GUVs. The color intensity of the solution (i.e., the outer solution) was very weak, indicating that most of the R6G molecules were encapsulated in the GUVs. Because the removal of R6G from the outer solution by the conventional centrifugation was difficult (see below), we investigated the extent of R6G encapsulation in the GUVs by using biotinylated GUVs consisting of DPhPC/Chol/DOPE/B-cap-PE> (9:1:0.05 :0.025, mol/mol) and immobilizing the GUVs on an avidin slip, where the outer solution was easily exchanged in a flow system (see Fig. S3 in supplementary material). The addition of Triton X100 caused the immediate rapture of R6G-entrapped GUVs, thereby releasing R6G into the outer solution and emitting

After the introduction of anti-BSA on the surface of R6GeGUVs by the amine coupling method, the recovery of antibody-modified GUVs from the reaction mixture was necessary. We investigated the recovery of antibody-modified GUVs by centrifugation (14,000 g, 30 min) under similar conditions as those for the conventional multilamellar liposomes. The modification of R6GeGUVs with anti-BSA was evaluated using the relative fluorescence intensity caused by the capture of anti-BSAeR6GeGUVs on an antiBSA slip in the presence of BSA. Although the fraction of R6GeGUVs without modification of antibody was slightly rich in the upper part of the supernatant, antibody-modified GUVs were preferentially gathered into the lower fraction of the supernatant (Fig. 2). Therefore, the half-fraction (50 ml) from the top of the supernatant was discarded, and then only the remaining part (50 ml) of the supernatant was collected. The collected fraction contained anti-BSAeR6GeGUVs. Importantly, in this protocol,

M. Sakamoto et al. / Analytical Biochemistry 505 (2016) 66e72

A

upper fraction

lower fraction

69

A 0 min

6 min

20 μm

B

B upper fraction lower fraction

without Triton X-100 with Triton X-100

30

20

C 10

0

1200

anti-BSA-modified

unmodified

Fig.2. Recovery of anti-BSA-modified GUVs by centrifugation. (A) Photos of GUVs. Left: Upper fraction after centrifugation. Right: Lower fraction after centrifugation. (B) Fluorescence intensities obtained with a FluorImager 595 (excitation 514 nm, emission 570 nm) using anti-BSA slips. First bar: Upper fraction (anti-BSA-modified GUVs). Second bar: Lower fraction (anti-BSA-modified GUVs). Third bar: Upper fraction (unmodified GUVs). Fourth bar: Lower fraction (unmodified GUVs). Each anti-BSA slip was incubated with 0.1 mg mL
1 BSA solution.

excess of the amine coupling agents, unreacted antibody, nonencapsulated R6G, and unreacted R6GeGUVs still remained in the collected fraction. However, a washing step after the incubation of anti-BSAeR6GeGUVs removed these compounds (see below). It is noted that when GUVs without modification of antibody were prepared in a 1.0 mol L
1 sucrose solution (d ¼ 1.13 at 0.988 mol L
1) and diluted with D-(þ)-glucose (d ¼ 1.06 at 0.944 mol L
1) under an isotonic condition, most of the GUVs were collected into the bottom fraction of the supernatant (Fig. S4). This indicates that antibody-modified GUVs are difficult to be purified by centrifugation if sucrose is used. Effect of excess reagents on assay performance Although antibody-modified R6GeGUVs were preferentially concentrated into the lower fraction of the supernatant, excess of amine coupling agents, unbound antibody, unbound R6GeGUVs, and non-encapsulated R6G were present in the mixture. Therefore, prior to the assay, we performed a washing step in a flow system (chamber volume 0.60 ml) to remove these compounds. As shown

Fluorescence intensity (a.u.)

Relative fluorescence (%)

40

800

400

0

without with Triton X-100

Fig.3. Effect of a washing step in assay using an anti-BSA slip. (A) Photos of an anti-BSA slip during the flow of a PB buffer, showing a color change from orange (non-encapsulated R6G) to yellow. (B) Fluorescence images for an anti-BSA slip after incubation with anti-BSAeR6GeGUVs and a washing step. Left: Before lysis. Right: After lysis with Triton X-100. (C) Relative fluorescence intensities before and after lysis in the absence of BSA. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

in Fig. 3A, running a PB buffer at a flow rate of 0.2 mL min
1 caused a gradual change of the color of a chamber solution, showing that excess of the dye was washed out by a buffer solution. In addition, the fluorescence intensities with and without treatment of Triton

70

M. Sakamoto et al. / Analytical Biochemistry 505 (2016) 66e72

X-100 after the washing step were expected ones (Fig. 3B and C). The successful washing was also indicated by the relative fluorescence intensity in the absence of an analyte (BSA) being less than 8% (Fig. S2), even though GUVs were adsorbed nonspecifically and adventitiously. Importantly, the specific interaction between the analyte and anti-BSAeR6GeGUVs was unaffected by washing (Fig. S2). Although the washing step was performed in a sorbitolfree PB buffer, the rapture of GUVs was not observed until 20 min (Fig. S5). The effect of the remaining coupling agents and anti-BSA on the relative fluorescence intensity in the absence of GUVs was also investigated as negative control. The DF value was nearly zero, indicating that the remaining components do not affect the assay. Consequently, the washing step just before the fluorescence measurement was effective for the removal of interfering compounds.

Concentration dependence for BSA The concentration dependence for BSA (69 kDa, pI ¼ 4.6e5.1) [32] in a PB solution with the current method is shown in Fig. 4, where an increase in the fluorescence intensity (i.e., DF) was plotted against BSA concentration in the range from 1.0  10
14 to 1.0  10
12 g ml
1 on a linear scale (Fig. 4A) and a logarithmic scale (Fig. 4B). The DF increased steeply with BSA concentration up to 1.0  10
14 g ml
1. On the other hand, on a logarithmic scale, the fluorescence intensity was linearly related to an increase in BSA concentration. The lower detection limit for BSA (signal/noise [S/ N] ¼ 3, n ¼ 3) was 4 fg ml
1 as calculated from the fluorescence intensity at three times the standard deviation of DF in the absence of BSA, and the quantification limit was 13 fg ml
1 (S/N ¼ 10). The slight increase in DF in the absence of BSA can be ascribed to nonspecific adsorption of anti-BSAeR6GeGUVs. The lower detection limit of the current method was far superior to those of the electrochemical methods, such as channel-based assay (1 ng ml
1) [33] and Ag electrochemistry (0.04 mg ml
1) [34], and the fluorescence-based methods, such as fluorescent anisotropic assay (13.8 mg ml
1) [35], counter current electrophoresis (60 ng ml
1) [36], and spectrofluorometric assay (4 mg ml
1) [37], but it was lower than that of the microcontact imprinting method (6.6 ag ml
1) [38]. The improved lower detection limit was simply because the number of markers encapsulated into the GUVs was significantly large. Although multiple recognition sites are present on the functionalized GUV surface, and hence a loss of sensitivity may

A

occur, the large volume (fl level) of GUVs enables encapsulating 108 molecules per a single GUV (see above). We defined a signal transduction factor as an amount (mol) of markers released into the solution after Triton X-100 treatment to that of the BSA (analyte) in the solution. The amount of the marker (R6G) was evaluated from a calibration curve for R6G in a solution. The signal transduction factor was calculated to be approximately 108 at 0.10 pg ml
1 BSA, indicating high sensitivity due to the signal transduction ability of the current system. Concentration dependence for LCN2 To demonstrate that the current approach is useful for sandwich-type assay of other compounds, we performed the assay of lipocalin-2 (25 kDa) using anti-LCN2eR6GeGUVs and an antiLCN2-modified slip. The assay procedure was the same as that for BSA except that 300 times diluted anti-LCN2eR6GeGUVs were used. The use of anti-LCN2eR6GeGUVs at 300 times dilution markedly lowered the background fluorescence (i.e., F0 in the absence of LCN2). The concentration dependence for LCN2 in a PB solution is shown in Fig. 5. The fluorescence intensity (i.e., DF) increased with LCN2 concentration in the range from 1.0  10
13 to 1.0  10
11 g ml
1 on a linear scale (Fig. 5A) and a logarithmic scale (Fig. 5B). The lower detection limit for LCN2 was 80 fg ml
1 (S/ N ¼ 3, n ¼ 3), and the quantification limit was 267 fg ml
1 (S/N ¼ 10, n ¼ 3). The lower detection limit of the current method for LCN2 is far superior to those of the conventional ELISA methods (3 or 44 pg ml
1) [39], graphene-based immunoassay (0.7 pg ml
1) [40], amperometric sensor (1 ng ml
1) [41], and particle-enhanced turbidimetric assay (9.43 mg ml
1) [42]. Selectivity The effect of human serum albumin (HSA) and g-globulin on the assay of LCN2 was investigated by quantifying LCN2 (1.0 and 10 pg ml
1) in the presence of HSA ranging from 0 to 50 mg ml
1 and g-globulin (17 mg ml
1). The fluorescence intensity DF remained nearly constant even in the presence of HSA up to 50 mg ml
1, indicating that the presence of HSA did not affect the quantification of LCN2 (Figs. S6a and S6b). Similarly, g-globulin (17 mg ml
1), transferrin (3 mg ml
1), and haptoglobin (2 mg ml
1) did not affect the assay of LCN2 of 1.0 pg ml
1.

B 40

Relative fluorescence (%)

Relative fluorescence (%)

40

30

20

10

0 0

0.5 BSA (pg/ml)

1

30

20

10

0

0

0.01 0.1 BSA (pg/ml)

1

Fig.4. Concentration dependence for BSA on a linear scale (A) and a logarithmic scale (B). The fluorometric images of the slips were captured (excitation 514 nm, emission 570 nm) after lysis with Triton X-100.

M. Sakamoto et al. / Analytical Biochemistry 505 (2016) 66e72

A

71

B 40

Relative fluorescence (%)

Relative fluorescence (%)

40

30

20

10

30

20

10

0

0 0

2

4

6

8

10

0

0.1

Lipocalin-2 (pg/ml)

1

10

Lipocalin-2 (pg/ml)

Fig.5. Concentration dependence for LCN2 on a linear scale (A) and a logarithmic scale (B). The fluorometric images of the slips were captured (excitation 514 nm, emission 570 nm) after lysis with Triton X-100.

Table 1 Concentrations of LCN2 in diluted human serum and recovery (n ¼ 3).

Present method Sample (1)d Sample (2)e

Dilution ratea

Human serumb (pg mL
1)

Human serum þ LCN2c (pg mL
1)

Recovery (%)

5  104 5  104

0.84 ± 0.12 0.89 ± 0.65

1.9 ± 0.28

106 ± 30

e

e

(ng mL
1) ELISAf Sample (2)e a b c d e f

60

0.25 ± 0.05

Human serum (male, type AB) was diluted with a PB buffer (final dilution rate). Final concentration of LCN2. LCN2 of 1 pg ml
1 was added to diluted human serum. Human serum sample 1 (male, type AB). Human serum sample 2 (male, type AB). Assay with a human LCN2/NGAL ELISA kit (R&D Systems).

Real sample testing LCN2 is a small secretory glycoprotein and important as a prognostic and diagnostic biomarker for many diseases. The LCN2 level is changed by many diseases and disorders through up- or down-regulation of LCN2 [39,40]. The LCN2 level in human serum has been reported to be in the range from 50 to 120 ng ml
1 [43e46]. Considering the dynamic range of the current assay and the normal level of lipocalin-2 in human serum, the sample (human serum, male type AB) was diluted 5  104 times and subjected to the assay. The results of assay are given in Table 1. The determination of LCN2 in the diluted human serum showed that the concentrations of LCN2 are 0.84 pg ml
1 for the nonspiked case and 1.9 pg ml
1 for the spiked one. The concentration of LCN2 for the nonspiked sample indicates the presence of endogenous LCN2 in human serum. The recovery of the spiked LCN2 was 106 ± 30% (n ¼ 3) with an acceptable value. The assay was compared with that obtained with an ELISA kit (Fig. S7). The assay results showed an interesting difference between the current assay and the ELISA method. The concentration of LCN2 (44 ng ml
1 before dilution) was larger than that obtained by the ELISA method (15 ng ml
1). This is probably explained by the fact that various forms of LCN2 (i.e., monomeric, homodimeric, and heterodimeric forms) affect the assay performance with commercial immunoassay kits [47,48]. Therefore, in our case as well, the antibody configuration of the functionalized GUVs and/or the

immobilized ones on the slips was different, affecting the assay results. Conclusions In this study, a highly sensitive sandwich-type immunoassay using functionalized giant unilamellar vesicles containing a fluorescent dye as a marker was developed. The purification of the antibody-functionalized GUVs was partially achieved by centrifugation, and a washing step in a flow system allowed removing interfering compounds. Although the dynamic range of the assay is limited, the assay is highly sensitive with a detection limit of tens at the femtogram (fg) level. The current SLI allows us to dilute samples by virtue of the high sensitivity, although the analyte loss by the dilution step may occur to some extent. The limitation of applying the method is that the dynamic range for LCN2 is narrow, and hence the choice of dilution factor is important. But the assay is simple and rapid, and there is the potential of applying it for detection of various compounds (e.g., toxins) that will exist at very low levels at early stages of the disease. Acknowledgments This work was financially supported by a Grant-in-Aid for Scientific Research (C) (21350045) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. Financial support

72

M. Sakamoto et al. / Analytical Biochemistry 505 (2016) 66e72

from a research grant from the College of Humanities and Sciences at Nihon University is also acknowledged. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.ab.2016.04.011. References [1] K.A. Edwards, A.J. Baeumner, Liposomes in analyses, Talanta 68 (2006) 1421e1431. € ro € s, B. St€ [2] M. Bally, K. Bailey, K. Sugihara, D. Grieshaber, J. Vo adler, Liposome and lipid bilayer arrays towards biosensing applications, Small 6 (2010) 2481e2497. [3] S.M. Christensen, D.G. Stamou, Sensing-applications of Surface-based Single Vesicle Arrays, Sensors 10 (2010) 11352e11368. Basel. [4] Q. Liu, B.J. Boyd, Liposomes in biosensors, Analyst 138 (2013) 391e409. [5] P. Walde, K. Cosentino, H. Engel, P. Stano, Giant vesicles: preparations and applications, ChemBioChem 11 (2010) 848e865. [6] M.I. Angelova, D.S. Dimitrov, Liposome electroformation, Faraday Discuss. Chem. Soc. 81 (1986) 303e311. l e ard, Giant unilamellar vesicle formation under [7] T. Pott, H. Bouvrais, P. Me physiologically relevant conditions, Chem. Phys. Lipids 154 (2008) 115e119. ~ i, L.A. Bagatolli, Giant unilamellar vesicles [8] L.R. Montes, A. Alonso, F.M. Gon electroformed from native membranes and organic lipid mixtures under physiological conditions, Biophys. J. 93 (2007) 3548e3554. [9] D.J. Estes, M. Mayer, Giant liposomes in physiological buffer using electroformation in a flow chamber, Biochim. Biophys. Acta 1712 (2005) 152e160. [10] J.P. Reeves, R.M. Dowben, Formation and properties of thin-walled phospholipid vesicles, J. Cell. Physiol. 73 (1969) 49e60. [11] K. Akashi, H. Miyata, H. Itoh, K. Kinoshita Jr., Preparation of giant liposomes in physiological conditions and their characterization under an optical microscope, Biophys. J. 71 (1996) 3242e3250. [12] J.T. Buboltz, G.W. Feigenson, A novel strategy for the preparation of liposomes: rapid solvent exchange, Biochim. Biophys. Acta 1417 (1999) 232e245. [13] A. Moscho, O. Orwar, D.T. Chiu, B.P. Modi, R.N. Zare, Rapid preparation of giant unilamellar vesicles, Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 11443e11447. [14] M. Abkarian, E. Loiseau, G. Massiera, Continuous droplet interface crossing encapsulation (cDICE) for high throughput monodisperse vesicle design, Soft Matter 7 (2011) 4610e4614. [15] A. Weinberger, F.C. Tsai, G.H. Koenderink, T.F. Schmidt, R. Itri, W. Meier, T. Schmatko, A. Schrӧder, C. Marques, Gel-assisted formation of giant unilamellar vesicles, Biophys. J. 105 (2013) 154e164. [16] I. Motta, A. Gohlke, V. Adrien, F. Li, H. Gardavot, J.E. Rothman, F. Pincet, Formation of giant unilamellar proteo-liposomes by osmotic shock, Langmuir 31 (2015) 7091e7099. [17] T.D. Heath, B.A. Macher, D. Papahadjopoulos, Covalent attachment of immunoglobulins to liposomes via glycosphingolipids, Biochim. Biophys. Acta 640 (1981) 66e81. [18] S. Shukla, H. Leem, M. Kim, Development of a liposome-based immunochromatographic strip assay for the detection of Salmonella, Anal. Bioanal. Chem. 401 (2011) 2581e2590. [19] X. Song, S. Shukla, S. Oh, Y. Kim, M. Kim, Development of fluorescence-based liposome immunoassay for detection of Cronobacter muytjensii in pure culture, Curr. Microbiol. 70 (2015) 246e252. [20] M. Ollivon, A. Walter, R. Blumenthal, Sizing and separation of liposomes, biological vesicles, and viruses by high-performance liquid chromatography, Anal. Biochem. 152 (1986) 262e274. [21] S. Lesieur, C. Grabielle-Madelmont, M.T. Paternostre, M. Ollivon, Size analysis and stability study of lipid vesicles by high-performance gel exclusion chromatography, turbidity, and dynamic light scattering, Anal. Biochem. 192 (1991) 334e343. [22] Y. Hashimoto, M. Sugawara, H. Endoh, Coating of liposomes with subunits of monoclonal IgM antibody and targeting of the liposomes, J. Immunol. Methods 62 (1983) 155e162. [23] K.K. Matthay, T.D. Heath, C.C. Badger, I.D. Bernstein, D. Papahadjopoulos, Antibody-directed liposomes: comparison of various ligands for association, endocytosis, and drug delivery, Cancer Res. 46 (1986) 4904e4910. sormeaux, J. Bestman-Smith, P. Gourde, M.J. Tremblay, [24] I. Dufresne, A. De M.G. Bergeron, Targeting lymph nodes with liposomes bearing anti-HLAeDR Fab' fragments, Biochim. Biophys. Acta 1421 (1999) 284e294.

[25] S. Goswami, W. Wang, T. Arakawa, S. Ohtake, Developments and challenges for mAb-based therapeutics, Antibodies 2 (2013) 452e500.  Mesquita, J.C. Sousa, M. Correia-Neves, N. Sousa, J.A. Palha, [26] A.C. Ferreira, S. Da F. Marques, From the periphery to the brain: lipocalin-2, a friend or foe? Prog. Neurobiol. 131 (2015) 120e136. [27] M.K. Jha, S. Lee, D.H. Park, H. Kook, K.G. Park, I.K. Lee, K. Suk, Diverse functional roles of lipocalin-2 in the central nervous system, Neurosci. Biobehav. Rev. 49 (2015) 135e156. [28] X. Leng, Y. Wu, R.B. Arlinghaus, Relationships of lipocalin 2 with breast tumorigenesis and metastasis, J. Cell. Physiol. 226 (2011) 309e314. [29] J. Barrera-Chimal, N.A. Bobadilla, Are recently reported biomarkers helpful for early and accurate diagnosis of acute kidney injury? Biomarkers 17 (2012) 385e393. [30] G. Wu, H. Li, Q. Fang, S. Jiang, L. Zhang, J. Zhang, X. Hou, J. Lu, Y. Bao, A. Xu, W. Jia, Elevated circulating lipocalin-2 levels independently predict incident cardiovascular events in men in a population-based cohort, Arterioscler. Thromb. Vasc. Biol. 34 (2014) 2457e2464. [31] G. Wang, Y.C. Weng, X. Han, J.D. Whaley, K.R. McCrae, W.H. Chou, Lipocalin-2 released in response to cerebral ischaemia mediates reperfusion injury in mice, J. Cell. Mol. Med. 19 (2015) 1637e1645. [32] S. Kuudu, H. Matsuoka, H. Seto, Zwitterionic lipid (DPPC)eprotein (BSA) complexes at the airewater interface, Colloid Surf. B 93 (2012) 215e218. [33] M. Nishio, A. Shoji, M. Sugawara, Planar lipid bilayers containing gramicidin A as a molecular sensing system based on an integrated current, Anal. Sci. 28 (2012) 661e667. [34] P.M.S. Tchekwagep, C.P. Nanseu-Njiki, E. Ngameni, T. Arnebrant, T. Ruzgas, Quantification of BSA concentration by using Ag electrochemistry in chloride solution: extension of the linear range, Electrochim. Acta 135 (2014) 351e355. [35] R.C. Groza, A. Calvet, A.G. Ryder, A fluorescence anisotropy method for measuring protein concentration in complex cell culture media, Anal. Chim. Acta 821 (2014) 54e61. [36] A.R. ter Avest, G. van Steenis, A.D.M.E. Osterhaus, A comparison of an enzymelinked immunosorbent assay and counter current electrophoresis for the detection of bovine serum albumin in virus vaccines, J. Biol. Stand 15 (1987) 245e250. [37] Y. Yu, Y. Lai, X. Zheng, J. Wu, Z. Long, C. Liang, Synthesis of functionalized CdTe/CdS QDs for spectrofluorimetric detection of BSA, Spectrochim. Acta A 68 (2007) 1356e1361. € m, B. Mattiasson, MicrocontacteBSA imprinted [38] G. Ertürk, D. Berillo, M. Hedstro capacitive biosensor for real-time, sensitive, and selective detection of BSA, Biotechnol. Rep. 3 (2014) 65e72. €fer, Development of a rapid sandwich [39] S.K. Vashist, M. Saraswat, H. Holtho enzyme linked immunoassay procedure for the highly sensitive detection of human lipocalin-2/NGAL, Proc. Chem. 6 (2012) 141e148. [40] S.K. Vashist, Graphene-based immunoassay for human lipocalin-2, Anal. Biochem. 446 (2014) 96e101. [41] P. Kannan, H.Y. Tiong, D.-H. Kim, Highly sensitive electrochemical determination of neutrophil gelatinase-associated lipocalin for acute kidney injury, Biosens. Bioelectron. 31 (2012) 32e36. [42] D. Xiang, H. Zhang, J. Bai, J. Ma, M. Li, W. Gao, X. Zhang, J. Gao, C. Wang, Particle-enhanced turbidimetric immunoassay for determination of serum neutrophil gelatinase-associated lipocalin on the Roche Cobas c501 analyzer, Clin. Biochem. 46 (2013) 1756e1760. [43] S. Yavuz, A. Anarat, S. Acartürk, A.C. Dalay, E. Kesiktas¸, M. Yavuz, T.O. Acartürk, Neutrophil gelatinase associated lipocalin as an indicator of acute kidney injury and inflammation in burned children, Burns 40 (2014) 648e654. [44] M. Zorlu, M. Kiskac, C. Karatoprak, S. Kesgin, M. Cakirca, K. Yildiz, C. Ardic, M.A. Cikrikcioglu, R. Erkoc, Assessment of serum apelin and lipocalin-2 levels in patients with subclinical hypothyroidism, Ann. Endocrinol. Paris. 75 (2014) 10e14. [45] A. Yndestad, L. Landrø, T. Ueland, C.P. Dahl, T.H. Flo, L.E. Vinge, T. Espevik, S.S. Frøland, C. Husberg, G. Christensen, K. Dickstein, J. Kjekshus, E. Øie, L. Gullestad, P. Aukrust, Increased systemic and myocardial expression of neutrophil gelatinase-associated lipocalin in clinical and experimental heart failure, Eur. Heart J. 30 (2009) 1229e1236. [46] L. Landrø, J.K. Damås, T.H. Flo, L. Heggelund, T. Ueland, G.E. Tjønnfjord, T. Espevik, P. Aukrust, S.S. Frøland, Decreased serum lipocalin-2 levels in human immunodeficiency virus-infected patients: increase during highly active anti-retroviral therapy, Clin. Exp. Immunol. 152 (2008) 57e63. [47] G. Lippi, G. Cervellin, Neurophil gelatinase-associated lipocalin: a more specific assay is needed for diagnosing renal injury, Clin. Chim. Acta 413 (2012) 1160e1161. [48] L. Cai, J. Borowiec, S. Xu, W.Y. Han, P. Venge, Assay of urine levels of HNL/NGAL in patients undergoing cardiac surgery and the impact of antibody configuration on their clinical performances, Clin. Chim. Acta 403 (2009) 121e125.