Detection of Membrane Biointeractions Based on Fluorescence

(24) Xia, W.; Rininsland, F.; Wittenburg, S. K.; Shi, X.; Achyuthan, K. E.;. McBranch, D. ..... (42) Cho, W.; Bittova, L.; Stahelin, R. V. Anal. Bioch...
7 downloads 0 Views 210KB Size
Download PDF
Langmuir 2008, 24, 4125-4131

4125

Detection of Membrane Biointeractions Based on Fluorescence Superquenching Reema Zeineldin,†,‡ Menake E. Piyasena,§ Larry A. Sklar,| David Whitten,*,⊥ and Gabriel P. Lopez*,†,§,⊥ Center for Biomedical Engineering, Department of Chemical & Nuclear Engineering, Department of Chemistry, and Cancer Center and Department of Pathology, UniVersity of New Mexico, Albuquerque, New Mexico 87131 ReceiVed NoVember 15, 2007. In Final Form: January 8, 2008 Assays for biointeractions of molecules with supported lipid bilayers using fluorescence superquenching are described. A conjugated cationic polymer was adsorbed on to silica microspheres, which were then coated with an anionic lipid bilayer. The lipid bilayer attenuated superquenching by acting as a barrier between the conjugated polymer and its quencher. Biointeractions of the lipid bilayer with a membrane lytic peptide, melittin, were detected and quantitated by superquenching of the conjugated polyelectrolyte in flow cytometric and microfluidic bioassays. A higher sensitivity for detecting melittin lysis of the lipid bilayer at lower concentrations and shorter times for melittin action was found using flow cytometry in this study in comparison to other existing methods. This study combined the sensitivity of superquenching and flow cytometry to detect biointeractions with a lipid bilayer, which serves as a platform for developing functional assays for sensor applications, lipid enzymology, and investigations of molecular interactions. In addition, this study demonstrated proof-of-concept for using superquenching detected as a result of lipid bilayer disruption in a microfluidic format.

Introduction There is interest in employing microspheres as supports for lipid bilayer assemblies for use in applications such as targeted delivery, proteomics, and biosensing.1-7 Recently, enhanced sensitivity in detection for microsphere-based assays has been achieved through coating the microspheres with superquenchable fluorescent polymers. The signal enhancement occurs by a combination of increased light harvesting by the strongly absorbing conjugated or dye-pendant polymers and their amplified sensitivity to fluorescence quenching by energy- or electrontransfer quenchers.8-14 Superquenching, which can arbitrarily * To whom correspondence should be addressed. E-mail: whitten@ unm.edu (D.W.) and [email protected] (G.P.L.). † Department of Chemical & Nuclear Engineering. ‡ Current affiliation: College of Pharmacy, University of New Mexico, Albuquerque, NM 87131. § Department of Chemistry. | Cancer Center and Department of Pathology. ⊥ Center for Biomedical Engineering. (1) Bayerl, T.; Bloom, M. Biophys. J. 1990, 58, 357-362. (2) Buranda, T.; Huang, J.; Ramamrao, G. V.; Ista, L. K.; Larson, R. S.; Ward, T. L.; Sklar, L. A.; Lopez, G. P. Langmuir 2003, 19, 1654-1663. (3) Buranda, T.; Huang, J.; Perez-Luna, V. H.; Schreyer, B.; Sklar, L. A.; Lopez, G. P. Anal. Chem. 2002, 74, 1149-1156. (4) Loidl-Stahlhofen, A.; Hartmann, T.; Schottner, M.; Rohring, C.; Brodowsky, H.; Schmitt, J.; Keldenich, J. J. Pharm. Res. 2001, 18, 1782-1788. (5) Loidl-Stahlhofen, A.; Schmitt, J.; Noller, J.; Hartmann, T.; Brodowsky, H.; Schmitt, W.; Keldenich, J. AdV. Mater. 2001, 13, 1829-1834. (6) Piyasena, M. E.; Buranda, T.; Wu, Y.; Huang, J.; Sklar, L. A.; Lopez, G. P. Anal. Chem. 2004, 76, 6266-6273. (7) Zeineldin, R.; Piyasena, M. E.; Bergstedt, T. S.; Sklar, L. A.; Whitten, D.; Lopez, G. P. Cytometry Part A 2006, 69A, 335-341. (8) Chen, L.; McBranch, D. W.; Wang, H.-L.; Helgeson, R.; Wudl, F.; Whitten, D. G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12287-12292. (9) Chen, L.; McBranch, D. W.; Wang, R.; Whitten, D. Chem. Phys. Lett. 2000, 330, 27-33. (10) Harrison, B. S.; Ramey, M. B.; Reynolds, J. R.; Schanze, K. S. J. Am. Chem. Soc. 2000, 122, 8561-8562. (11) Wang, J.; Wang, D.; Miller, E. K.; Moses, D.; Bazan, G. C.; Heeger, A. J. Macromolecules 2000, 33, 5153-5158. (12) Tan, C.; Pinto, M. R.; Schanze, K. S. Chem. Commun. 2002, 2002 (5), 446-447. (13) List, E. W. J.; Creely, C.; Leising, G.; Schulte, N.; Schlueter, A.; Scherf, U.; Muellen, K.; Graupner, W. Chem. Phys. Lett. 2000, 325, 132-138.

be defined as a system (fluorophore or polymer containing several fluorophores) having a quenching constant enhanced by a factor of 104 M-1 or more, compared to “static” Stern-Volmer quenchers for single molecule fluorophore-quencher combinations, can be observed for polymers in solution and in a variety of supported formats.7,15-18 Superquenching of a polymer fluorophore by an oppositely charged quencher has been shown to be due to a combination of two effects. The observed quenching constant is found to be approximately the product of a “collection constant” associated with binding of the small molecule quencher to the conjugated polyelectrolyte8 and an “amplification factor” associated with interaction of individual chromophores via conjugation or aggregation.8,14 Several sensing applications have been developed on the basis of superquenching, through energy transfer or electron transfer between small molecule quenchers and conjugated polymers anchored on polystyrene microspheres. These include assays for nucleic acids, protease enzyme activity, and kinase/phosphatase enzyme activity assays.19-24 The sensing polyelectrolyte is generally attached to a microsphere support via adsorption to an (14) Gaylord, B. S.; Wang, S.; Heeger, A. J.; Bazan, G. C. J. Am. Chem. Soc. 2001, 123, 6417-6418. (15) Lu, L.; Helgeson, R.; Jones, R. M.; McBranch, D.; Whitten, D. J. Am. Chem. Soc. 2002, 124, 483-488. (16) Jones, R. M.; Lu, L.; Helgeson, R.; Bergstedt, T. S.; McBranch, D. W.; Whitten, D. G. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 14769-14772. (17) Jones, R. M.; Bergstedt, T. S.; McBranch, D. W.; Whitten, D. G. J. Am. Chem. Soc. 2001, 123, 6726-6727. (18) Kim, K.; Webster, S.; Levi, N.; Carroll, D. L.; Pinto, M. R.; Schanze, K. S. Langmuir 2005, 21, 5207-5211. (19) Kushon, S. A.; Ley, K. D.; Bradford, K.; Jones, R. M.; McBranch, D.; Whitten, D. Langmuir 2002, 18, 7245-7249. (20) Kushon, S. A.; Bradford, K.; Marin, V.; Suhrada, C.; Armitage, B. A.; McBranch, D. W.; Whitten, D. G. Langmuir 2003, 19, 6456-6464. (21) Kumaraswamy, S.; Bergstedt, T. S.; Shi, X.; Rininsland, F.; Kushon, S.; Xia, W.; Ley, K.; Achyuthan, K.; McBranch, D.; Whitten, D. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 7511-7515. (22) Pinto, M.; Schanze, K. S. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 75057510. (23) Rininsland, F.; Xia, W.; Wittenburg, S.; Shi, X.; Stankewicz, C.; Achyuthan, K.; McBranch, D.; Whitten, D. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 1529515300.

10.1021/la703575r CCC: $40.75 © 2008 American Chemical Society Published on Web 02/27/2008

4126 Langmuir, Vol. 24, No. 8, 2008

oppositely charged microsphere, covalent linkage, or association of biotin binding protein. A biological receptor may also be attached to the microsphere (before or after the polymer is attached) by one of these three methods. Depending on the specific sensing application, direct or competition assays have been developed.19-24 These assays are based on increasing the effective concentration of the quencher in the vicinity of the fluorescent polymer upon specific biointeraction. We recently demonstrated superquenching on silica microspheres in flow cytometry and showed that lipid bilayer assemblies overcoating the polymer could attenuate superquenching.7 Microsphere-supported lipid bilayers have been used in biosensing applications after incorporating receptors, ion channels, or porins.25-29 Thus, our use of lipid bilayers as barriers to superquenching is important, as we aim to develop strategies in which interactions between lipid bilayers with either nonspecific bioactive lytic peptides or biospecific interactive peptides lead to either transport of quenchers across lipid bilayer assemblies or to disruption of the bilayer. One of the model peptides that interacts with lipid bilayers that has been extensively studied is melittin. Melittin (MLT) is a bee venom cationic membrane-lytic lipase that consists of a single polypeptide chain of 26 amino acids and that has a bent R-helical structure. Its interaction with a lipid bilayer is dependent on the peptide concentration and the lipid composition of the bilayer.30 MLT usually forms pores in zwitterionic lipid vesicles,31,32 whereas it exhibits a detergent-like action in disrupting a lipid bilayer composed of anionic lipids.32,33 In this study, we report the use of superquenching for detecting biointeractions between low concentrations of MLT and a microsphere-supported lipid bilayer. The assay is based on the quenching of a cationic conjugated polyelectrolyte poly(pphenyleneethynylene) (PPE) derivative by an electron-transfer quencher, 9,10-anthraquinone-2,6-disulfonic acid (AQS) (Scheme 1). Overcoating of microsphere-supported PPE by the bilayer results in attenuation of superquenching by AQS, while biointeractions of a lipid bilayer membrane with the lytic peptide melittin results in fluorescence quenching in assays employing either flow cytometry or microfluidic channels. In this study, we demonstrate the use of a sensitive detection format that employs the use of superquenching and flow cytometry to investigate real-time interactions between membrane-active peptides and lipid bilayer in microspheres, and thus serve as a biosensing format. Furthermore, we demonstrate the applicability of this system in a microfluidic assay. Experimental Section Materials. The cationic polyelectrolyte poly(p-phenyleneethynylene) derivative (PPE) and the quencher 9,10-anthraquinone-2,6disulfonic acid (AQS) (Scheme 1) were obtained as reported previously.7 Borosilicate glass microspheres (5 µm diameter) were purchased in dry form from Duke Scientific (Palo Alto, CA). 1,2Dimyristoyl-sn-glycero-3-[phosphor-rac-(1-glycerol)] sodium salt (24) Xia, W.; Rininsland, F.; Wittenburg, S. K.; Shi, X.; Achyuthan, K. E.; McBranch, D.; Whitten, D. Assay Drug DeV. Technol. 2004, 2, 328-339. (25) Bayley, H.; Cremer, P. S. Nature 2001, 413, 226-230. (26) Anrather, D.; Smetazko, M.; Saba, M.; Alguel, Y.; Schalkhammer, T. J. Nanosci. Nanotech. 2004, 4, 1-22. (27) Trojanowicz, M.; Mulchandani, A. Anal. Bioanal. Chem. 2004, 379, 347350. (28) Huang, W.; Yang, X.; Wang, E. Anal. Lett. 2005, 38, 3-18. (29) Schmidt, J. J. Mater. Chem. 2005, 15, 831-840. (30) Bechinger, B. Crit. ReV. Plant Sci. 2004, 23, 271-292. (31) Ladokhin, A. S.; Selsted, M. E.; White, S. H. Biophys. J. 1997, 72, 17621766. (32) Papo, N.; Shai, Y. Biochem. 2003, 42, 458-466. (33) Ladokhin, A. S.; White, S. H. Biochim. Biophys. Acta 2001, 1514, 253260.

Zeineldin et al. Scheme 1. Molecular Structures of the Fluorescent-Conjugated Cationic Polyelectrolyte Poly(p-phenyleneethynylene) Derivative (PPE) and the Anionic Quencher 9,10-Anthraquinone-2,6-disulfonic Acid (AQS)

(DMPG) was obtained from Avanti Polar Lipids, Inc (Alabaster, AL). Phosphate-buffered saline (PBS), Triton X-100, and synthetic and natural melittin were obtained from Sigma (St. Louis, MO). Coating Borosilicate Microspheres with Polymer. The fluorescent cationic PPE was coated onto borosilicate microspheres using sufficient polymer (based on an estimated molecular area and an extinction coefficient of 35 100 L/mol cm per polymer repeat unit7) to provide 1.2 times monolayer coverage. The silica beads were suspended in ultrapure water and stirred at room temperature for 30 min. Bead suspensions were separated from the solution by centrifugation, and the colorless supernatant was discarded. The PPE-coated microspheres (MS-PPE) were rinsed with ultrapure water by four cycles of rinse, centrifuge, decant, and resuspend. Preparation of Unilamellar Lipid Vesicles. Small unilamellar vesicles (SUVs) were prepared using a 2 mM solution of DMPG in chloroform. The lipid was dried by nitrogen gas followed by vaccum. The dried lipid was resuspended in PBS (pH 7.4), incubated at 37 °C for 10 min, and then sonicated to optical clarity in a sonication bath (Branson Cleaning Equipment Co., Shelton, CT). Preparation of Microsphere-Supported Lipid Bilayers. Lipid bilayers were assembled on microspheres as previously described.1,2 Briefly, the SUVs were incubated at 37 °C for 5 min, and MS-PPE were added to them, and the mixture was stirred at room temperature using a vortex mixer for 30 min, followed by incubation at 37 °C for 5 min without mixing. The lipid-coated MS-PPE were washed by suspending them in PBS followed by centrifugation, and the clear supernatant was decanted and discarded. The cycle resuspend, centrifuge, and decant was repeated four additional times, and then the lipid-coated MS-PPE were resuspended in PBS. Flow Cytometry. Bead suspensions of 2.5 × 105 DMPG-coated MS-PPE in 200 µL of PBS were analyzed using a FACScan flow cytometer (Becton-Dickinson, Sunnyvale, CA) with excitation at 488 nm. Fluorescence signals were acquired on FL-1 channel (525 nm) using log amplification and analyzed with the CellQuest software. Kinetic analysis of disruption of supported lipid bilayer by melittin was performed by acquiring real-time data with continuous injection of the samples (106 MS-PPE in 800 µL of PBS) while being mixed using a magnetic stirrer. As the sample was delivered in a continuous stream to the flow cytometer, the data were likewise collected in a continuous stream. The accumulated data represented a time-resolved data file, containing parameters measured every second, which was analyzed by the proprietary software IDLeQuery developed by Bruce Edwards (Cancer Center, University of New Mexico, Albuquerque, NM). The software detects the time-resolved data and allows graphing a dot plot of a selected parameter (such as side scatter, forward scatter, number of events, or fluorescence intensity in any selected

Detection of Membrane Biointeractions channel) vs time. The software determines the median fluorescence, which is then exported with time into an Excel (Microsoft Corp., Redmond, WA) spreadsheet to be plotted. Fluorimetry. Fluorescence measurements of bead suspensions were performed using a Wallac 1420 multilabel counter (PerkinElmer, Shelton, CT) by excitation at 485 nm and collection of emission at 535 nm using top counting mode. A 96-well plate was used, where 200 µL samples containing either PPE polymer in solution or MS-PPE with or without AQS were analyzed using a counting time of 2 s per well. Fabrication of Microchannels. PDMS microchannels were constructed using soft lithographic techniques adapted from the literature.34 The microfluidic channels were fabricated with weirs to hold the beads in place as described elsewhere.6 The dimensions of the microchannel were as follows: length, 2 cm; width, 250 µm; and height, 60-70 µm. In order to trap beads near the outlet, the depth of the channel was limited to 12-15 µm. The prepared channel was irreversibly sealed on to a glass slide using an Ar plasma. Packing of Microchannels with MS-PPE. Microchannels were packed with DMPG-coated MS-PPE. Then 5 µL aliquots of bead solutions were injected into the column by applying a vacuum at the outlet. The length of the bead segment was about 5 mm. Beadpacked channels were kept wet with Tris buffer (100 mM Tris, pH 7.5, 150 mM NaCl) that was allowed to continuously percolate through the column under gravity until ready for use. Disruption Assays in Microchannels. The bead-packed microchannel was mounted onto a motorized vertical translational stage located in the sample holder space of a Model Fluorolog-3 SPEX fluorometer (Instruments S.A.). The bead segment was irradiated with 488 nm laser excitation. The inlet of the column was connected to a buffer reservoir, while the outlet was connected to a vacuum source. Several microliters of Tris buffer were passed through the microchannel before the injection of the sample. While applying the vacuum at the outlet, 10 µL of 120 mM quencher was injected directly into the column through the inlet silicone tubing using a 10 µL Hamilton syringe. After 55 min, 10 µL of a 1:1 mixture of 309 µM melittin and 120 µM quencher was injected. The interactions of the melittin and quencher were monitored as the change in the original intensity of fluorescence signal of PPE-coated beads at 520 nm.

Langmuir, Vol. 24, No. 8, 2008 4127

Figure 1. Stern-Volmer plot for quenching of PPE in solution (2) and of MS-PPE (9) by AQS as detected by fluorimetry. AQS was added to 200 µL of 9 × 105 fmol of PPE in solution or to 200 µL of 6.4 × 106 MS-PPE in suspension (equivalent to 9 × 105 fmol of PPE in solution). The error bars represent the standard deviation (SD) for fluorescence ratios obtained for four replicates.

(where Io and I represent the fluorescence intensities in the absence and in presence of the AQS). However, the adsorption of PPE to the microspheres led to a higher KSV. The KSV obtained for PPE in solution was 0.29 × 106 M-1, whereas for MS-PPE it was 0.81 × 106 M-1. This finding is consistent with other reports of enhanced superquenching of conjugated polymers when collected on nano- and microparticles.16 The increased KSV of PPE-MS indicates higher sensitivity in detection over PPE in solution. The KSV values for PPE and MS-PPE are lower than

that of other fluorescent conjugated polyelectrolytes on beads with energy transfer quenchers but are very comparable to the KSV obtained for a structurally related polyelectrolyte, cationic poly(p-phenylene-co-thiophene), when quenched by AQS in aqueous solution.35 Superquenching in a Flow Cytometric Assay for Biointeraction of Melittin with Microsphere-Supported Lipid Bilayers. We recently demonstrated that flow cytometry can be used as a sensitive and quantitative method for the detection of superquenching of the fluorescence of MS-PPE and also that formation of lipid bilayers around MS-PPE could attenuate superquenching in assays carried out by flow cytometry.7 Evaluation of different lipids led us to select DMPG as the lipid of choice for mediating the quenching of MS-PPE, where it blocked the quenching of MS-PPE by 10 µM AQS.7 Disrupting the DMPG lipid bilayer by adding 0.25% (w/v) Triton X-100 in the presence of 10 µM AQS results in reduction of fluorescence to a level comparable to that obtained in the absence of a lipid bilayer, which corresponds to ∼18% of the fluorescence of DMPG-coated MS-PPE in the absence of quencher.7 To assess the potential of using superquenching as a detector in assays for biospecific interactions with lipid bilayers, we adapted superquenching of MS-PPE for detecting disruption of a supported lipid bilayer by melittin. MLT is a bee venom cationic membrane-lytic peptide whose interaction with a lipid bilayer is dependent on the peptide concentration and the lipid composition of the bilayer.30 MLT usually forms pores in zwitterionic lipid vesicles,31,32 whereas with anionic lipid vesicles it exhibits a detergent-like action in disrupting the lipid bilayer.32,33 Since we used an anionic lipid to form a lipid bilayer around MS-PPE, we expect that MLT will disrupt the DMPG lipid bilayer by this means. This was tested by adding 10 µM AQS to DMPG-coated MS-PPE followed by treatment with varying concentrations of synthetic MLT (sMLT) or natural MLT (nMLT) and immediately reading the fluorescence by flow cytometry. Adding either sMLT or nMLT resulted in quenching of the MS-PPE in a concentrationdependent manner (Figure 2), indicating disruption of the anionic lipid bilayer, as suggested by other studies.33 These experiments demonstrated the use of superquenching in flow cytometric assays for establishing the curves for interactions of sMLT and nMLT with DMPG supported on MS-PPE. From Figure 2, the association constant (corresponding to the reciprocal of the point of half-

(34) Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Anal. Chem. 1998, 70, 4974-4984.

(35) Ramey, M. B.; Hiller, J.; Rubner, M. F.; Tan, C.; Schanze, K. S.; Reynolds, J. R. Macromolecules 2005, 38, 234-243.

Results and Discussion Stern-Volmer Quenching Constants by AQS for Free PPE and PPE-Coated Microspheres (MS-PPE). Silica microspheres (5 µm average diameter) were coated with ∼1.2 monolayer of PPE, which was equivalent to 0.14 fmol of polymer repeat unit per MS-PPE, as determined previously by fluorimetry.7 In order to compare the Stern-Volmer constant (KSV) for PPE polymer in solution to that of MS-PPE, fluorescence was determined after adding AQS at different concentrations to 200 µL of either 9 × 105 fmol of PPE in solution or to its equivalent of MS-PPE in suspension (6.4 × 106 microspheres) (Figure 1). The quenching of both PPE and MS-PPE followed a conventional Stern-Volmer equation:

Io/I ) 1 + KSV[Q]

4128 Langmuir, Vol. 24, No. 8, 2008

Figure 2. Superquenching as a result of disruption of lipid bilayer by MLT. DMPG-coated MS-PPE (2.5 × 105) in 200 µL of PBS were analyzed by flow cytometry, and 10 µM AQS was added, followed by treatment with Triton X-100 (TX-100), sMLT, or nMLT at final concentrations of 4 mM (0.25% w/v), 3.8 µM, or 1.6 µM, respectively. Fluorescence readings were determined 40 s after the addition of the TX-100 or MLT. (A) Representative fluorescence intensity distribution histograms of MS-PPE obtained by flow cytometry. The y-axis represents the number of fluorescent events (counts), and the x-axis represents the mean channel fluorescence intensity. (B) Fluorescence normalized to that of sample +AQS. The error bars represent the SD of the medians of normalized histograms obtained for four replicates.

saturation) for nMLT and sMLT with DMPG was 5 × 106 and 1.7 × 106 M-1, respectively. These association constants are comparable to the association constant of 2 × 106 M-1 reported for association of sMLT with immobilized anionic DMPG membranes using surface plasmon resonance.36 Lower concentrations of nMLT, in comparison to sMLT, were required to disrupt the DMPG lipid bilayer. This might be attributed to the fact that nMLT contains some phospholipase A2 (PLA2) as a contaminant. PLA2 has a high affinity for anionic phospholipids, with an association constant of 0.5 × 1010 M-1,37 which is ∼3000fold higher than that of MLT. Although PLA2 activity requires calcium, which is not present in our system, it has been reported that even upon addition of EDTA to natural MLT, PLA2 maintains its activity, probably due to tight complexing calcium with PLA2.38 Thus, our results suggesting that the small amount of PLA2 present with nMLT is participating in bilayer disruption are not surprising. We selected the two concentrations of sMLT and nMLT that caused equivalent normalized fluorescence and compared that outcome to disruption of the lipid bilayer by Triton X-100. This was tested by adding 10 µM AQS to DMPG-coated MS-PPE followed by treatment with Triton X-100, sMLT, or nMLT and immediately reading the fluorescence by flow cytometry. Figure 3A shows the fluorescence intensity distribution histograms obtained by flow cytometry for the DMPG-coated MS-PPE with the different treatments. These histograms resemble the ones reported previously for MS-PPE, without a lipid coating, in the absence and presence of AQS.7 Figure 3B shows that adding either sMLT or nMLT resulted in quenching of the MS-PPE to a level comparable to that occurring on addition of Triton X-100, but at very low concentrations of sMLT or nMLT. These results confirm that addition of small amounts of MLT lead to anionic lipid bilayer disruption. (36) Lee, T. H.; Mozsolits, H.; Aguilar, M. I. J. Pept. Res. 2001, 58, 464-476. (37) Kim, Y.; Lichtenbergova, L.; Snitko, Y.; Cho, W. Anal. Chem. 1997, 250, 109-116.

Zeineldin et al.

Existing assays for MLT that employ anionic unilamellar vesicles and conventional quenchers detect MLT lytic activity by fluorometric, calorimetric, or surface plasmon resonance (SPR) methods at concentrations ranging from 10 to 140 µM,33,38,39 and in one study incubating with MLT was for 20 h.33 Only one study evaluated interactions of low concentrations of sMLT (range 0.015-0.3 µM) with negatively charged membranes using SPR.32 In our study, the increased sensitivity as a result of using superquenching and the use of flow cytometry7 improved the sensitivity for detecting MLT lysis of lipid bilayer at lower concentrations than reported in the literature and without the need for prolonged incubation of MLT with the membranes or waiting for injected material to reach bilayers for interactions as in SPR. In this study, we evaluated the binding of MLT to lipid bilayers and determined association constants employing the sensitivity of flow cytometry and superquenching. This system could also be used with other membrane-lytic or even pore-forming polypeptides with possible modifications in the used superquenchable polyelectrolyte or the quencher. Such changes would be based on adaptations to the composition of the lipid bilayer required for interaction and the size of formed pores within the lipid bilayers. For example, inclusion of positively charged phopholipids in the lipid bilayer, because of polypeptides that interact with such lipids, may require using an anionic polyelectrolyte to coat the microspheres. In addition, detecting interactions of pore-forming polypeptides within the lipid bilayer will require using quenchers with a small enough size to pass through the formed pores to allow their interactions with a superquenchable polyelectrolyte coating the microspheres. Such adaptation would be dependent on the polypeptide studied and its special requirements. The kinetics of disruption of supported lipid bilayer by MLT was monitored by measuring real-time changes in fluorescence intensity by flow cytometry (Figure 4). Suspensions of 106 DMPG-coated MS-PPE in 800 µL of PBS were continuously injected and treated with AQS at a final concentration of 10 µM and then with either sMLT, Triton X-100, or no treatment. The green trace in Figure 4 (top graph) represents the control that is obtained for DMPG-coated MS-PPE in absence of AQS and without any treatments. Although it displays a slight increase in normalized fluorescence, it is negligible and is considered to display a near-constant normalized fluorescence. Treatment with sMLT in the absence of AQS led to a slight increase in normalized fluorescence that was time-dependent (Figure 4, top pink trace). This can be attributed to disruption of the lipid bilayer by sMLT and is consistent with a reversal of the slight quenching of fluorescence caused by the formation of a DMPG lipid bilayer over MS-PPE reported earlier.7 Addition of AQS alone to DMPG-coated MS-PPE (Figure 4, red trace) led to a 20% reduction in fluorescence, as reported earlier.7 Addition of Triton X-100 (second dip in purple trace) to DMPG-coated MS-PPE after the addition of AQS (first dip in purple trace) caused immediate decrease in fluorescence to ∼10% of its original value. In contrast, addition of sMLT, after the addition of AQS (Figure 4, blue trace), led to a gradual reduction of fluorescence intensity to ∼20% of the original value. Triton X-100 forms micelles at the concentration used in this study. On the other hand, although MLT in aqueous solutions forms aggregates at high concentrations,38 it is monomeric at the low concentrations used in this study. This suggests that the low concentrations (38) Dempsey, C. E. Biochim. Biophys. Acta 1990, 1031, 143-161.

Detection of Membrane Biointeractions

Langmuir, Vol. 24, No. 8, 2008 4129

Figure 3. Interaction curve for MLT-disruption of supported lipid bilayer by nMLT (concentrations of 0, 0.06, 0.12, 0.48, 0.96, 1.92, and 3.84 µM) and sMLT (concentrations of 0, 0.5, 0.1, 0.2, 0.4, 0.8, and 1.6 µM). Median fluorescence intensity was determined 40 s after adding MLT and was normalized to that without MLT as determined by flow cytometry. The error bars represent the SD of the medians of normalized histograms obtained for three replicates.

of MLT used in this study caused disruption of the bilayer by perhaps a different mechanism than that produced by Triton X-100. Detection of Lipid Bilayer Disruption by MLT in a Microfluidic Assay by Superquenching. Biomolecular assemblies on microspheres have been used to develop new microfluidic-based bioassay techniques.3,6,40,41 We investigated the feasibility of using the superquenching technique as a method of detection in microfluidic bioassays. The setup we used is shown in Figure 5A. The continuous exposure of the fluorescent MS-PPE segment to the laser can cause slight photobleaching, as seen in the first 90 min in Figure 5B. Injection of AQS into the microchannel results in a decrease of the fluorescence intensity, which corresponds to the change in the slope, between 90 and 120 min, in Figure 5B. This decrease in fluorescence is consistent with that observed by flow cytometry (Figure 4, trace C). The delayed response time (∼38 min) matches the time that AQS takes to reach the bead segment in the microchannel from the point of injection. The response time is injection point dependent and can be further decreased by injecting the samples at a point that is closer to the bead column. Upon adding AQS along with nMLT (added ∼40 min after adding AQS alone), there is a slight rise in fluorescence intensity followed by a significant decrease. The initial rise in fluorescence intensity is probably due to disruption of the lipid bilayer by nMLT, similar to the flow cytometric observation as in Figure 4 (purple trace), whereas the decrease in fluorescence corresponds to superquenching of MSPPE by AQS after disruption of DMPG bilayer by nMLT. There is no appreciable difference in the diffusion rates of MLT and (39) Constantinescu, I.; Lafleur, M. Biochim. Biophys. Acta 2004, 1667, 2637. (40) Sato, K.; Tokeshi, M.; Kimura, H.; Kitamori, T. Anal. Chem. 2001, 73, 1213-1218.

AQS. However, since MLT is positively charged, whereas AQS is negatively charged, there is a possibility that AQS, but not MLT, is initially repelled by the anionic bilayer, which results in initial interaction of MLT with the bilayer causing an increase in fluorescence. When the lipid bilayer is disrupted as a result of this interaction, AQS can interact with the polymer, resulting in superquenching. On the other hand, the rise in fluorescence before the decrease may be due to the fact that the increase may come slowly and before the bilayer has been totally removed. To verify that the rise was due to MLT, we next injected nMLT alone to a fresh microcolumn, and after ∼20 min we injected AQS, as shown in Figure 6. In this case, the time delay between the point of sample injection and the packed beads was ∼25 min. The initial decline in fluorescence during the first 20 min was due to photobleaching. After injecting nMLT, by ∼25 min we detected a rise in fluorescence that was followed by a significant decrease in fluorescence ∼25 min after injecting AQS. This suggests that the rise in fluorescence is indeed due to disruption of DMPG bilayer on MS-PPE, as seen in Figure 4 (purple trace). Our previous report demonstrated that formation of a DMPG lipid bilayer over MS-PPE caused slight quenching of PPE fluorescence.7 The present study suggests that disrupting the lipid bilayer reverses this quenching until AQS is added, which leads to a significant decrease in fluorescence, corresponding to superquenching of MS-PPE by AQS. In the microfluidic studies, we used a higher concentration of nMLT than we did in the flow cytometry studies. The concentration of nMLT was selected on the basis of giving a response equivalent to that obtained with Triton X-100. Our microfluidic experiments indicate the utility of superquenching as a detector of biointeractions in microfluidic channels, although when compared to flow cytometry it involves a delay time between injection and detection of a response. This is due to the differences in the rate

4130 Langmuir, Vol. 24, No. 8, 2008

Figure 4. Kinetics of supported lipid bilayer disruption by MLT. DMPG-coated MS-PPE (2.5 × 105) in 200 µL of PBS were analyzed by flow cytometry. Samples were treated with 10 µM AQS, followed by treatment with either Triton X-100 (TX-100) or sMLT at final concentrations of 4 mM (0.25% w/v) or 0.23 µM, respectively (bottom graph). Controls (top graph) included one without treatments, another with AQS only, and a third with sMLT only. Median florescence intensity for each trace was normalized to its first reading of fluorescence intensity. The first arrow on the traces in the bottom graph indicates addition of AQS, and the second arrow indicates addition of TX-100 or sMLT.

of mass transport of AQS to the membrane-coated microspheres in the two assay formats, which is expected, given the physical mixing characteristics of the two assay formats. On the other hand, the advantages of using microfluidics include lower consumption of reagents and the ability to design several reaction segments within the microchannel, thus allowing multiplexing. In this study, we demonstrated proof-of-concept that membrane disruption can be detected by superquenching, even when the lipid-bilayer-coated microspheres are closely packed in a microfluidic channel. Further development is needed to optimize a packed microcolumn assay.

Zeineldin et al.

Figure 5. Detecting MLT biointeraction with DMPG supported on MS-PPE in a microfluidic channel. (A) A schematic of the microfluidic channel used. Channel dimensions were typically 2 cm, 250 µm, 60-70 µm in length, breadth, and depth, respectively, and the length of the segment packed with DMPG-supported MSPPE (the inset in the schematic) was about 5 mm. The microspheres segment was irradiated with 488 nm laser excitation, and emission is detected at 520 nm. The inlet of the column was connected to a buffer reservoir, while the outlet was connected to a vacuum source. (B) Effect of adding AQS and nMLT on MS-PPE. Concentration of the first injected AQS was 120 µM. AQS+nMLT represents injecting a mixture of equal volumes of AQS and nMLT with final mixture concentrations of 60 and 154.5 µM, respectively. Fluorescence intensity is measured in arbitrary units. The arrows indicate injections. The time delay between the point of sample injection and the packed beads was ∼38 min.

Conclusion This study established the use of superquenching of fluorescent conjugated polyelectrolytes to detect biospecific interactions of a cationic peptide with a lipid bilayer supported on silica microspheres in flow cytometry and in microfluidic channels. Existing methods for studying binding of membrane-active peptides or polypeptides usually employ either unilamellar vesicles, in suspension or immobilized in chromatography columns, or supported lipid mono- or bilayers.42-44 Such methods (41) Sato, K.; Tokeshi, M.; Odake, T.; Kimura, H.; Ooi, T.; Nakao, M.; Kitamori, T. Anal. Chem. 2000, 72, 1144-1147. (42) Cho, W.; Bittova, L.; Stahelin, R. V. Anal. Biochem. 2001, 296, 153161. (43) Liu, X.-Y.; Nakamura, C.; Yang, Q.; Miyake, J. Anal. Biochem. 2001, 293, 251-257.

Figure 6. Effect of MLT and AQS on MS-PPE in the microfluidic channel. Concentrations of injected nMLT and AQS were 309 and 120 µM, respectively. Fluorescence intensity is measured in arbitrary units. The arrows indicate injections. The time delay between the point of sample injection and the packed beads was ∼25 min.

can be time-consuming and may require long incubation times or the use of high concentrations of peptide/polypeptide. In this study, we used the enhanced sensitivity of superquenching in (44) Mozsolits, H.; Lee, T.-H.; Wirth, H.-J.; Perlmutter, P.; Aguilar, M.-I. Biophys. J. 1999, 77, 1428-1444.

Detection of Membrane Biointeractions

addition to microsphere-based assays employing either flow cytomery or microfluidic channels for detecting lipid bilayer disruption by detergents such as Triton X-100 and for determining the activity of a membrane-lytic peptide. The advantages of these techniques are increased sensitivity, the ability to determine association constants by flow cytometry, and the lower consumption of reagents in microfluidic channels. In addition, both techniques have the potential for use in high-throughput screening. In this study, the use of the detergent Triton X-100 and the lytic peptide melittin disrupted the lipid bilayer, resulting in bringing the quencher to the proximity of the fluorescent polymer. This approach should be adaptable for sensing of pore-formation or disruption of lipid bilayers by certain integral membrane proteins, such as channel- and pore-forming proteins. Furthermore, sensing of small or large molecules that may directly disrupt

Langmuir, Vol. 24, No. 8, 2008 4131

or damage the lipid bilayer membrane or that may target molecules within the bilayer to cause a similar effect is feasible. This requires drawing correlations between the type and size of the quencher and the mode of membrane interruption by these molecules. The results of this study suggest the potential for developing a sensitive functional assays for sensor applications, lipid enzymology, and investigations of molecular interactions. Acknowledgment. This work was supported by the National Science Foundation through the NIRT (EEC0210835) and SENSORS (CTS0332315) programs. We are grateful to the Defense Threat Reduction Agency (Contract W911NF-07-10079) for partial support of this research. LA703575R