Decyl Methacrylate Liquid Polymer Matrix and Ion-Exchange-Based Pot

decyl methacrylate as a matrix. A general scheme for the polymerization and introduction of sensing components creates a matrix that allows for the ut...
0 downloads 0 Views 277KB Size
Download PDF
Anal. Chem. 2001, 73, 2221-2228

Fluorescent Nanosensors for Intracellular Chemical Analysis: Decyl Methacrylate Liquid Polymer Matrix and Ion-Exchange-Based Potassium PEBBLE Sensors with Real-Time Application to Viable Rat C6 Glioma Cells Murphy Brasuel,† Raoul Kopelman,*,† Terry J. Miller,‡ Ron Tjalkens,‡ and Martin A. Philbert‡

Department of Chemistry and Department of Environmental Health Sciences, The University of Michigan, Ann Arbor, Michigan 48109

Fluorescent spherical nanosensors, or PEBBLEs (probes encapsulated by biologically localized embedding), in the 500 nm-1 µm size range have been developed using decyl methacrylate as a matrix. A general scheme for the polymerization and introduction of sensing components creates a matrix that allows for the utilization of the highly selective ionophores used in poly(vinyl chloride) and decyl methacrylate ion-selective electrodes. We have applied these optically silent ionophores to fluorescence-based sensing by using ion-exchange and highly selective pH chromoionophores. This allows the tailoring of selective submicrometer sensors for use in intracellular measurements of important analytes for which selective enough fluorescent probes do not exist. The protocol for sensor development has been worked out for potassium sensing. It is based on the BME-44 ionophore (2-dodecyl-2methyl-1,3-propanediylbis[N-[5′nitro(benzo-15-crown-5)4′-yl]carbamate]). The general scheme should work for any available ionophore used in PVC or decyl methacrylate ion-selective electrodes, with minor adjustments to account for differences in ionophore charge and analyte binding constant. The reversible and highly selective sensors developed have a subsecond response time and an adjustable dynamic range. Applications to live C6 glioma cells demonstrate their utility; the intracellular potassium activity is followed in real time upon extracellular administration of kainic acid. Rapid advances in the biomedical field pose new challenges to analytical chemistry in the field of chemical sensors: real-time, noninvasive analysis of chemical processes inside live cells and their subcompartments. A traditional strength of chemical sensors (optical, electrochemical, etc.) is the minimal chemical interference between sensor and sample, achieved with the use of inert, “biocompatible” matrixes or interfaces. However, when it comes to penetrating individual live cells, the physical size of the sensor results in biological damage and resultant biochemical conse† ‡

Department of Chemistry. Department of Environmental Health Sciences.

10.1021/ac0012041 CCC: $20.00 Published on Web 04/10/2001

© 2001 American Chemical Society

quences. The past decade has seen significant progress in the miniaturization of electrodes and fiber-optic “optodes”.1,2 However, even reductions to tips with a 100-nm radius or below are not completely satisfactory.2 The physiologist’s approach to analysis in live cells has focused on “molecular probes”. These ensembles of individual sensor molecules consist of units small enough to avoid physical damage to the cell. However, there is practically no protection from chemical interference between sample and probe, including differential sequestration of probe molecules inside the cell or leakage out of it.3-6 A recent development in sensor design attempts to combine the advantages of sensor tips and molecular probes, i.e., avoid both physical and chemical interference between sensor (probe) and sample (cell or organelle). These new sensors are nanoparticle or nanobead sensors and are called PEBBLEs (probes encapsulated by biologically localized embedding),7-10 to emphasize that noninvasive delivery to the intracellular location is of major importance and needs to be part and parcel of an adequate design.7 Optimal delivery has been effected by “gene gun” or liposomes (ibid). To date PEBBLEs consist of an indicator dye (“molecular probe”), and often a reference dye, all protected inside an inert cross-linked, acrylamide matrix. Chemical interference, such as protein binding to the indicator dye, is prevented only when the dye is encapsulated inside the inert matrix. Thus, it is not satisfactory to use ready-made nanobeads with indicator dyes attached to their surface. A PEBBLE can work with as little as a (1) Buhlmann, P.; Pretsch, E.; Bakker, E. Chem. Rev. 1998, 98, 1593-1647. (2) Tan, W.; Kopelman, R.; Barker, S. L. R.; Miller, M. T. Anal. Chem. 1999, 71, 606A-612A. (3) Herman, B. Fluorescence Microscopy, 2nd ed.; Springer: New York, 1998. (4) Mason, W. T. Biological Techniques; Academic Press: San Diego, CA, 1993. (5) Nuccitelli, R. Methods in Cell Biology; Academic Press: San Diego, CA, 1994. (6) Slavik, J. Fluorescent Probes in Cellular and Molecular Biology; CRC Press: Boca Raton, FL, 1994. (7) Clark, H. A.; Barker, S. L. R.; Brasuel, M.; Miller, M. T.; Monson, E.; et al. Sens. Actuators B 1998, 51, 12-16. (8) Clark, H. A.; Hoyer, M.; Philbert, M. A.; Kopelman, R. Anal. Chem. 1999, 71, 4831-4836. (9) Clark, H. A.; Kopelman, R.; Tjalkens, R.; Philbert, M. A. Anal. Chem. 1999, 71, 4837-4843. (10) Clark, H. A.; Hoyer, A.; Parus, S.; Philbert, M.; Kopelman, R. Mikrochim. Acta 1999, 131, 121-128.

Analytical Chemistry, Vol. 73, No. 10, May 15, 2001 2221

single indicator dye per PEBBLE, provided that the number of PEBBLEs used is sufficient for generating a satisfactory signal/ noise ratio. If only a single PEBBLE is used, it has to contain a sufficient number of indicator and reference dyes (on the order of 1000) to give satisfactory signal/noise signal. While the use of fluorescent indicator molecules, in either free3-6 or encapsulated8-10 form, has proven valuable in the study of a number of intracellular analytes, there are many ions for which no fluorescent indicator dye is sufficiently selective or even available. A more sophisticated class of optical sensors, with much higher selectivity, relies on chemical equilibrium or steady state among different components. One example is the “bulk optode” or ion-selective optode (ISO), where the matrix (hydrophobic liquid polymer) contains a selective lipophilic ionophore (“optically silent”), a chromoionophore, and an ionic additive.1,11-14 The operation of the entire system is based on having a thermodynamic equilibrium that controls ion exchange (for sensing cations) or ion coextraction (for sensing anions), i.e., an equilibrium-based correlation between different ion species. To achieve sensor miniaturization, fluorescence rather than absorbance has been utilized. An analogous approach uses fluorescent voltage-sensitive dyes rather than the chromoionophore.15-17 We have explored the first approach to develop nanosensors for ions that do not have satisfactory fluorescent dyes.7 Potassium sensing was chosen as a starting point for the development of liquid polymer PEBBLE nano-optodes because of this ion’s predominant intracellular concentration and the body of literature surrounding its detection by a number of different sensing schemes. Common methods for the determination of intracellular potassium concentration include patch clamping and the use of microelectrodes. These methods suffer from the same limitations experienced with pulled fiber-optic sensors. The detection of one analyte in a given section of a cell is possible, but not that of multiple analytes at multiple locations in a viable cell. In contrast, PEBBLEs are a new tool designed to spatially resolve many analytes in a single viable cell, all simultaneously. However, due to the absence of fluorescent indicators selective enough for potassium, the standard (acrylamide) PEBBLE approach is not adequate. The same goes for available molecular probessnone of them is selective enough (for potassium over sodium) for quantitative intracellular applications. Thus, our first utilization of the liquid polymer PEBBLE concept has been for a potassium sensor. The first challenge of this work is to determine whether the calibration of these “nanosized” systems can be described by the accepted thermodynamic equations. The second challenge relates to the nanofabrication process. For fiber-optic or other film-based optodes, it is relatively easy to switch from one kind of matrix to the other. For instance, plasticized PVC optodes are prepared by (11) Bakker, E.; Simon, W. Anal. Chem. 1992, 64, 1805-1812. (12) Morf, W. E.; Seiler, K.; Lehmann, B.; Behringer, C.; Hartman, K.; Simon, W. Pure Appl. Chem. 1989, 61, 1613-1618. (13) Kurihara, K.; Ohtsu, M.; Yoshida, T.; Abe, T.; Hisamoto, H.; Suzuki, K. Anal. Chem. 1999, 71, 3558-3566. (14) Suzuki, K.; Ohzora, H.; Tohda, K.; Miyazaki, K.; Watanabe, K.; Inoue, H.; Shirai, T. Anal. Chim. Acta 1990, 237, 155-164. (15) Huber, C.; Werner, T.; Krause, C.; Wolfbeis, O. S.; Leiner, M. J. P. Anal. Chim. Acta 1999, 398, 137-143. (16) Krause, C.; Werner, T.; Wolfbeis, O. S. Anal. Sci. 1998, 14, 163-167. (17) Krause, C.; Werner, T.; Huber, C.; Wolfbeis, O. S. Anal. Chem. 1999, 71, 1544-1548.

2222

Analytical Chemistry, Vol. 73, No. 10, May 15, 2001

a quick and simple dipping method, whether they are macroscopic or microscopic.18-21 However, for making PVC PEBBLEs, a grinding process was explored,22 resulting in supermicrometer sizes and nonspherical shapes. The nanoemulsion process for preparing PEBBLEs is subtle, and there is no universal method for making hydrophilic, hydrophobic, and ampiphilic nanospheres that contain the right matrix and right chemical components, in the proper proportions. Thus, switching from single dye-containing hydrophilic polyacrylamide nanospheres to multicomponent, hydrophobic, liquid polymer sensors is not yet a routine procedure. One solution is described in this report. Instead of using the typical ion-selective electrode (ISE) (or ISO) matrix, of plasticized PVC, we used the decyl methacrylate (DMA) matrix. Recently, long-chain methacrylates have proven their utility in ionsensitive field-effect transistor (ISFET) membranes, in ISE membranes, and in membranes for optical polyion probes.23-25 We have thus applied a cross-linked DMA to the development of liquid polymer nano-optodes. The properties of this matrix are highly adjustable. One can change either the plasticizer content or the cross-linking ratios in order to gain the desired properties.24,25 We have found that the DMA lends itself well to polymerization in an emulsion, enabling the fabrication of uniform spherical particles. The development of DMA emulsion polymerization was also based on existing technologies of drug delivery. The protocol uses poly(ethylene glycol) (PEG) to prevent the aggregation of nanoparticles.26-28 In this work, we describe the preparation of DMA/PEG PEBBLEs and the characterization of potassium nanosensors based on this design. The resulting nanosphere sensors obey thermodynamic relations for calibration, with the dynamic range, selectivity, and reversibility typical of macroscopic “bulk optodes”. In addition, these decyl methacrylate PEBBLE nanosensors have been delivered into cells with minimal damage to cell or PEBBLE, and utilized for real-time chemical analysis in viable rat C6 glioma cells. EXPERIMENTAL SECTION Reagents. Bis(2-ethylhexyl) sebacate (DOS), chromoionophore III (or 9-(diethylamino)-5-[(2-octoyldecyl)imino]benzo[a]phenoxazine (ETH 5350)), potassium ionophore III (or 2-dodecyl2-methyl-1,3-propanediylbis[N-[5′nitro(benzo-15-crown-5)-4′yl]carbamate] (BME-44)), potassium tetrakis-[3,5-bis(trifluoromethyl)phenyl borate (KTFPB),poly(ethylene glycol) 5000 monomethyl ether, poly(ethylene glycol) 6000, poly(ethylene glycol) 20 000, tetrahydrofuran (THF), and potassium peroxodisulfate (18) Barker, S. L. R.; Thorsrud, B. A.; Kopelman, R. Anal. Chem. 1998, 70, 100104. (19) Kopelman, R.; Miller, M. T.; Brasuel, M.; Clark, H. A.; Hoyer, M.; Philbert, M. Proc. SPIE 1999, 3540, 198-205. (20) Shortreed, M. R.; Dourado, S.; Kopelman, R. Sens. Actuators B 1997, 3839, 8-12. (21) Tan, W.; Shi, Z.-Y.; Smith, S.; Birnbaum, D.; Kopelman, R. Science 1992, 258, 778-781. (22) Clark, H. A. Ph.D. Dissertation, University of Michigan, Ann Arbor, MI, 1999. (23) Levichev, S. S.; Bratov, A. V.; Vlasov, Y. G. Sens. Actuators B 1994, 1819, 625-628. (24) Ambrose, T. M.; Meyerhoff, M. E. Electroanalysis 1996, 8, 1095-1100. (25) Ambrose, T. M.; Meyerhoff, M. E. Anal. Chem. 1997, 69, 4092-4098. (26) Peraccjia, M. T.; Vauthier, C.; Puisieux, F.; Couvreur, P. J. Biomed. Mater. Res. 1997, 34, 317-326. (27) Slomkowski, S. Prog. Polym. Sci. 1998, 23. (28) Ishizu, K.; Tahara, N. Polymer 1996, 37, 1729-1734.

Table 1. Composition of PEBBLEs Polymerization Solution Made with PEG 5000 Monoethyl Ether as Surfactant and Concentration of Components Added during THF Swelling matrix 1 matrix 2 matrix 3 matrix 4 matrix 5 Feed Composition during Polymerization wt % cross-linker 48.06 43.88 39.16 34.51 wt % DOS 43.78 43.87 43.26 43.02

25.85 42.17

Concentration of Components Added during Swelling mmol/kg BME-44 18.51 17.04 18.64 18.29 17.84 mmol/kg ETH 5350 9.37 9.37 9.68 9.53 9.71 mmol/kg KTFPB 10.50 8.96 10.84 9.57 10.37

were obtained from Fluka (Ronkonkoma, NY) and used without further purification. Decyl methacrylate was obtained from Pfaltz & Bauer (Waterbury, CT). Hexanediol dimethacrylate (HDDMA) and kainic acid were purchased from Aldrich (Milwaukee, WI). Standard solutions of NaCl, KCl, (Aldrich) and 2-amino-2(hydroxymethyl)propane 1,3-diol TRIS (Fluka) were prepared in 18 M water, Barnstead I Thermolyne Nanopure II system (Dubuque, IA). PEBBLE Preparation. A batch of PEBBLE sensors typically consists of 210 mg of decyl methacrylate, 180 mg of hexanediol dimethacrylate, 300 mg of dioctyl sebacate, with 10-30 mmol/ kg each of ionophore, chromoionophore, and ionic additives added after spherical particle synthesis. The spherical particles are prepared by dissolving decyl methacrylate, hexanediol dimethacrylate, and dioctyl sebacate in 2 mL of hexane. To a 100-mL roundbottom flask, in a water bath on a Corning pc-351 hot plate stirrer, 75 mL of pH 2 HCl was added along with 1793 mg of PEG 5000 monomethyl ether, PEG 6000, or PEG 20 000, and the resultant mixture was stirred and degassed. The hexane-dissolved monomer cocktail was then added to the reaction flask (under nitrogen), stirred at full speed, and the water bath temperature was raised to 80 °C over 30-40 min. A 6-mg aliquot of potassium peroxodisulfate was then added to the reaction, and stirring was reduced to medium speed. The temperature was kept at 80 °C for an additional 2 h, and then the reaction was allowed to return to room temperature and stir for 8-12 h. The resulting polymer was suction filtered through a Fisherbrand glass microanalysis vacuum filter holder with a Whatman Anodisc filter (0.2-µm pore diameter). The polymer was rinsed three times with water and three times with ethanol to remove excess PEG and unreacted polymer. THF was then used to leach out the DOS, and then the PEBBLEs were again filtered and rinsed. They were allowed to dry in a 70 °C oven overnight. Dry polymer was then weighed out, and DOS, BME-44, ETH5350, and KTFPB were added to this dry polymer, so that the resulting polymer would have 40% DOS, 20 mmol/kg BME-44, 10 mmol/kg ETH 5350, and 10 mmol/kg KTFPB. See Table 1 for actual PEBBLE feed compositions and for final sensor component composition. Enough THF was added to this mixture to just wet the PEBBLEs. The PEBBLEs were allowed to swell for 8 h and then the THF was removed by rotoevaporation. The resulting PEBBLE sensors were rinsed with double-distilled water and allowed to air-dry. Imaging and Optics. The PEBBLEs were suspended in a 50/ 50 water/ethanol solution. A few drops of this suspension were evaporated on a glass cover slip and sputtered with gold. Then

the SEM images were taken on a Hitachi S-3200N scanning electron microscope. Suspension data on PEBBLE fluorescence was obtained on an Olympus inverted fluorescence microscope, IMT-II (Lake Success, NY), using Nikon 50-mm f/1.8 camera lenses to project the image into an Acton 150-mm spectrograph (Acton, MA) with spectra read on a Princeton Instruments, liquid nitrogen-cooled, 1024 × 256 CCD array (Trenton, NJ). The PEBBLEs were excited using a mercury arc lamp with excitation and emission light filtered by an Olympus green dichroic mirror unit (module IMT2-DMG, excitation filter 425-550 nm, long-pass emission filter 575 nm. Images of PEBBLE-loaded cells were obtained using an Olympus FluoView 300 scanning confocal microscope system equipped with an Ar-Kr and He-Ne laser. Data were acquired using the Olympus FluoView package. Cell images were taken with 488-nm excitation of the PEBBLEs and a 580-nm long-pass filter. Cell imaging used the confocal system; all spectral data were acquired using the INT-II Olympus inverted fluorescent microscope. PEBBLE Calibration. To calibrate potassium-sensitive PEBBLEs, 20 mL of 1.5 g of PEBBLEs/L of 0.05 TRIS buffer (pH 7.2) was prepared. The suspension was suction filtered through a quantitative Whatman No. 2 cellulose filter to remove aggregates. Five milliliters of the resulting suspension was then pipetted into each of three 20-mL scintillation vials. A blank was prepared using blank PEBBLE polymer, also in TRIS buffer. Spectra of the blank were subtracted from the data obtained for the PEBBLE suspensions. Aliquots of 2 M KCl, 0.05 M TRIS (pH 7.2) were added to the vials to obtain different KCl concentrations. A period of 1-5 min was allowed for suspension equilibration between spectra. Three spectra were obtained at each calibration point, for each vial, with 0.5 s of acquisition time and 0.2 s between spectra. The pH of the suspension was monitored simultaneously with a glass pH electrode. For NaCl titration, aliquots of 4 M NaCl, 0.05 M TRIS (pH 7.2) were added until a concentration of 1 M NaCl was obtained; the pH of the suspension was then changed by adding measured aliquots of NaOH. This was done because as the optode is K+ selective it takes a much higher concentration of Na+ (∼1000 times more) to elicit the same type of response at pH 7.2. However, at concentrations above 1 M, the activity calculation becomes more complicated. On the other hand, because the response depends on log aX+/aH+, the entire response curve can be obtained even without going much above 1 M NaCl, by simply changing the pH with NaOH. The results of the separate solution method were checked against measured K+ response of the sensors in a constant 0.5 M background of Na+, buffered at pH 7.2. PEBBLE Response Time and Reversibility. The response time of the PEBBLE sensors was obtained using an initial suspension prepared in the same manner as described above. The suspension, in the scintillation vial, was placed on the microscope. The gating software for the CCD was set to take continuous spectra with 0.1-s acquisitions. The PEBBLE suspension was continually illuminated with 514-nm excitation. First, a KCl solution was added, and then after the signal stabilized, additional TRIS buffer (pH 7.2) was added to test the response to decreasing K+ concentration. To test the reversibility of the PEBBLE sensors, the same setup as above was used with the CCD software set to take seven spectra at each point, standard additions of buffer and Analytical Chemistry, Vol. 73, No. 10, May 15, 2001

2223

KCl solution were added to the scintillation vial, to oscillate the concentration between 45 and 12 mM K+, and spectra were taken immediately after mixing and after 30 s or more to ensure complete equilibrium. The time for the spectra was recorded with time zero being the initial mixing time of the PEBBLEs into the 45 mM K+, TRIS (pH 7.2) buffer. PEBBLE Usable Lifetime. The small size of the PEBBLE sensors suggests possible problems with leaching of membrane components. For fiber-based optodes with a 2-µm thickness based on the same components but using PVC as the liquid polymer matrix, the lifetime in aqueous solution was ∼45 min.29 To test the usable lifetime of the potassium sensors, seven pH-buffered (TRIS 10 mM, pH 7.2) solutions of varying potassium concentrations were prepared (1000, 800, 500, 200, 100, and 10 mM KCl). Spectra of PEBBLEs, soaked in each solution, were taken at various times in order to follow any changes in signal resulting from the leaching of membrane components out of the decyl methacrylate matrix. Of the membrane components. BME-44 is the least hydrophobic. It will be the first component to leach. The useful sensor lifetime thus depends on the residence time of BME44 in the membrane.29 Cell Culture. Rat C6 glioma cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 400 mg/L Dglucose, 2 mM L-glutamine, 10% fetal bovine serum, 0.3% penicillin, streptomycin, and neomycin and incubated at 37 °C in a 5% CO2 environment. Cells were released from culture dishes by trypsin treatment 1 day prior to experiments and plated at a density of 150 000 cells/plate on uncoated 22-mm glass cover slips in 35mm culture dishes. PEBBLE Delivery and Intracellular Measurements. PEBBLEs were delivered into the intracellular environment using a BioRad (Hercules, CA) Biolistic PDS-1000/He system. A thin film of PEBBLEs suspended in water was dried on the target membrane. The cells (on glass cover slip) were removed from the culture medium and rinsed with 36 °C 1× Hanks buffered salt solution and placed in a microscope cell. The microscope cell was placed in gene gun and a 15-in. Hg vacuum was applied to the system. The PEBBLEs were delivered successfully using a firing pressure of 650 psi. Firing pressures of greater than 800 psi removed cells from the cover slip. Following PEBBLE delivery, the cells were rinsed three times with Hanks 1× buffered salt solution and incubated in the same solution during analysis. Immediately following PEBBLE delivery, cells were placed on the INT-II Olympus inverted fluorescent microscope. The gating software for the CCD was set to take continuous spectra at 1.3-s intervals. After 20 s, and after 30 s, 50 µL of 0.4 mg/mL kainic acid was injected into the microscope cell. Kainic acid is known to stimulate cells by causing the opening of ion channels. RESULTS AND DISCUSSION Emulsion and Preparation Parameters. Initial PEBBLE production attempted to incorporate all sensing components into the matrix during polymerization. UV and thermal radical initiation were attempted. In both cases, the chromoionophore component of the K+ sensor was damaged. This led to the development of the “addition by swelling” step. Attempts were made to allow the (29) Shortreed, M.; Bakker, E.; Kopelman, R. Anal. Chem. 1996, 68, 26562662.

2224 Analytical Chemistry, Vol. 73, No. 10, May 15, 2001

Table 2. Diameter of PEBBLEs As Determined from SEM Images size

percentage

340 ( 40 nm 400 ( 40 nm 500 ( 40 nm 600 ( 40 nm 700 ( 40 nm

8 5 48 37 2

Figure 1. SEM of gold-coated decyl methacrylate PEBBLEs produced in emulsion polymerization with PEG 5000 monoethyl ether as surfactant. Size distribution of 500 ( 40 nm ) 48% and 600 ( 40 nm ) 37% of the PEBBLEs produced (see Table 2).

monomer to form droplets in an emulsifier-free polymerization utilizing potassium peroxodisulfate as initiator.27 Irregular polymer spheres were produced in this manner. It was found that PEG is an excellent surfactant in oil-in-water emulsions26 because it is very soluble in water, and the possibility of attaching PEG to the surface of the polymer particles holds some advantages for in vivo applications, based on its suppression of protein binding to nanoparticles.30,31 Initial attempts to create cross-linked decyl methacrylate PEBBLEs resulted in a wide distribution of sizes with poor suspension characteristics and extreme aggregation in aqueous suspension. The addition of PEG to the polymerization solution resulted in a more uniform shape and size distribution of particles and reduced the aggregation of the finished PEBBLE sensors in aqueous suspension. The resultant size distribution as measured by SEM can be seen in Table 2. Figure 1 shows a SEM micrograph of the PEBBLEs. The bulk of the PEBBLEs have diameters between 500 and 600 nm with 98% of the PEBBLEs having a diameter smaller than 700 nm. PEG 5000 monomethyl ether, PEG 6000, and PEG 20 000 were all used in this manner. We note that in many cases PEG has been used to increase the particle circulation time for in vivo applications. This is accomplished by reducing protein binding to the particle, which slows the removal of particles by the phagocytic cells of the reticuloendothelial system (RES).30,31 For our nanoparticles, it was found that the type of PEG used did not affect the binding amount of bovine serum albumin (BSA) to the PEBBLEs. This was (30) Ogris, M.; Brunner, S.; Schuller, S.; Kircheis, R.; Wagner, E. Gene Ther. 1999, 6, 595-605. (31) Kingshott, P.; Griessar, H. J. Curr. Opin. Solid State Mater. Sci. 1999, 4, 403-412.

Figure 2. Calibration of matrix 1 with 48.06% cross-linker (0) and matrix 5 with 25.85% cross-linker (O) based K+ PEBBLE sensors. Solid curves are theoretical lines from eq 2. The solid lines delimit values for log(aK+/aH+) found in intracellular media, and the dashed lines delimit the zone typical for extracellular ratios.33

determined using the method of Panagi et al. for protein binding on spherical liposomes for drug delivery.32 There was no statistical difference in the amount of protein left in a 3 mg/mL BSA solution after incubation with PEBBLEs with PEG 5000 monomethyl ether, PEG 6000, and PEG 20 000, as compared to incubation without PEBBLEs (all solutions cleaned by syringe filtration to remove PEBBLEs after incubation). For all solutions tested, ∼2 mg/mL BSA remained in the solution after incubation. The control indicates that most of the protein loss was a result of nonspecific binding to the incubation vials and the filter. Initial studies indicated that the response time of the PEBBLEs was on the order of 30 min or greater for K+ PEBBLEs coated with PEG 20 000 but significantly less for PEG 5000 monomethyl ether and PEG 6000. On the basis of these results, PEG 5000 monomethyl ether was chosen as the stabilizer for the PEBBLE fabrication. Adjustment of the cross-linker-to-monomer ratios changes the diffusion characteristics of ions in the sensors. Matrixes feed compositions are shown in Table 1. It was found that increasing the cross-linker concentration shifts the dynamic response range to higher values of log aK+/aH+ or higher potassium concentrations at a given pH. This is consistent with the findings of Ambrose and Meyerhoff.25 While in the work of Ambrose and Meyerhoff the effect is due to changes in polyion diffusion rate in the decyl methacrylate matrix of heparin-sensitive probes, our probe’s response is based on bulk equilibrium, and thus, the effect is due to the influence of the matrix on the complex-forming constant of BME-44 or the pKa of the chromoionophore. Figure 2 shows the calibration data for matrixes 1 and 5 (48 and 25% cross-linker, respectively). The solid curves represent theoretical responses of the sensor based on matrix composition (covered in detail below). It can be seen that almost doubling the cross-linker content shifts the dynamic range 1 order of magnitude. Dashed lines delimit typical extracellular activity ratios and the solid lines delimit the intracellular levels (log(aK+/aH+)).33 While changing the cross-linker content allows adjustment of the dynamic range, the response of sensors with higher than 26% cross-linker is less (32) Panagi, Z.; Avgoustakis, K.; Evangelatos, G.; Ithakissios, D. S. Int. J. Pharm. 1999, 176, 203-207. (33) Ammann, D. Ion-Selective Microelectrodes; Springer: Berlin, 1986.

reproducible from experiment to experiment and thus all further data come from matrix 5. Potassium PEBBLE Response and Calibration. In this study, the fluorescence response scheme of the DMA-based PEBBLE sensors follows closely previous work on PVC-based fiber potassium-selective ion sensors.20,29 The hydrogen ionselective chromoionophore ETH5350 competes with the ionophore BME-44 as cations enter the liquid polymer matrix. The lipophilic additive, KTFPB, maintains ionic strength in the matrix and aids in preventing the coextraction of anions. TFPB- acts as a counterion to the H+ and K+ ions in the membrane. This allows for charge neutrality in the membrane without negative counterions being brought in from solution into the membrane. Data analysis of the ETH 5350 spectra obtained from the PEBBLE calibration was based on the theoretical treatment of ion-exchange sensors developed by Simon, Bakker, and collaborators.1,11,12,29 For the incorporation of a selective neutral ionophore into a matrix along with a selective chromoionophore for indirect ion monitoring (ion-exchange sensors), the metal ion activity aiv+ in solution (see eq 2) is a function of the hydrogen ion activity aH+ in solution, interfering cations ajz+ (where Kopt ij is the selectivity coefficient) and the constants [Ltot], [Ctot], and [Rtot-], which are total ionophore (ligand) concentration, total chromoionophore concentration, and total lipophilic charge site concentration, in the membrane. Note that [CH] is the protonated chromoionophore concentration and [C] is the free base concentration. It is assumed that all components added during the matrix PEBBLE swelling procedure go into the matrix. The parameter Π has been defined18,29 as the relative portion of the protonated chromoionophore, Π ) [CH]/[Ctot]. It follows that

aiv+ + Kopt ij ajZ+ ) 1 Kexch

(

)

(1 - Π)aH+ Π

((

v

))

[Rtot] - (Π)[Ctot] 1 v [Ltot] - {[Rtot] - (Π) [Ctot]} v

(1)

which for primary and interfering ions with charges of 1 simplifies here to

aiv+ + Kopt ij ajZ+ )

1 Kexch

(

(Π-1 -1)aH+

[Ltot]

[Rtot - Π[Ctot]]

)

-1

-1

(2) This work takes advantage of an indicator with two fluorescence emission maximums (λ) that gives a relative intensity that changes with the degree of protonation (Π). This degree of protonation, Π, can be evaluated in terms of the fluorescence intensity ratio, Fλ2/Fλ1, given by the protonated chromoionophore intensity Fλ2 and the deprotonated chromoionophore intensity Fλ1 (See Figure 3 for spectra).29 The fluorescent intensity ratio is also C C Fλ2 g(ΦC2 [C] + ΦCH Π(ΦCH 2 [CH]) 2 - Φ2 ) + Φ 2 ) ) Fλ1 g(ΦC[C] + ΦCH[CH]) Π(ΦCH - ΦC) + ΦC 1

1

1

1

(3)

1

where g combines the geometrical factor of the instrumentation, light intensity of excitation, and molar absortivity of the dye at Analytical Chemistry, Vol. 73, No. 10, May 15, 2001

2225

Figure 3. Normalized emission spectra from suspended K+ PEBBLE sensors using the pH chromoionophore ETH 5350 for ion correlation spectroscopy in tandem with BME-44 (matrix composition 4 from Table 1) for the monitoring of K+ activity (PEBBLEs excited with mercury arc lamp and green dichroic mirror). (A) 10, (B) 50, (C) 200, and (D) 500 mM KCl and (E) 2.0 M KCl, all buffered at pH 7.2 with 10 mM Tris buffer.

the excitation wavelength (assumes molar absortivity of the protonated and deprotonated forms is equal at the excitation wavelength), ΦC2 and ΦCH 2 are the fluorescence “quantum yield” at λ2 for the deprotonated and protonated forms of the dye, respectively, and ΦC1 and ΦCH 1 are the same as above except for at λ1. In an ideal situation, where the fluorescence peaks of the protonated and deprotanted form of the dye are well resolved (ΦC2 ) 0 and ΦCH 1 ) 0) eq 3 simplifies to

Fλ2 ) Fλ1

(ΦCH 2 Π) C (Φ1 (1 - Π))

Π )Κ 1-Π

(4)

where the constant kappa (Κ) can be determined experimentally at the point where Π ) 0.5 (concentrations of protonated and deprotonated chromoionophores are equal). In the nonideal situation, where there is overlap in the fluorescence of protonated and deprotonated forms, this simplification is not possible. Then the experimentally obtained calibrations are normalized to Π by solving eq 3 in terms of Π and using data normalized to the isoemmisive point of the dye, giving eq 5 (superscripts P and D denote the completely protonated state and completely deprotonated state of the chromoionophore, respectively; lack of superscript denotes intermediate points):

FD λ2 Π)

FD λ2 FD λ1

-

-

FD λ1 FPλ2 + FD λ1

Fλ2 Fλ1

( )

Fλ2 FPλ1 -1 Fλ1 FD λ1

(5)

The response of the potassium PEBBLE sensors corresponding to matrix 5 in Table 2 is given in Figure 4. The data points for potassium and sodium responses are plotted along with the corresponding theoretical curves based on eq 2. The constant Kexch is determined from a line fit to the experimental data. Then the 2226

Analytical Chemistry, Vol. 73, No. 10, May 15, 2001

Figure 4. (A) Response of BME-44-based decyl methacrylate PEBBLEs to potassium (O) and sodium (4), along with theoretical curves. Theoretical curves are constructed solving eq 3 for aIv+, (I+ being K+ and Na+ in this case), for a given value of Π. The solid lines delimit values for log(aK+/aH+) found in intracellular media, and the dashed lines delimit the typical extracellular ratios.33 (B) Response of K+ PEBBLEs to standard additions of KCl in TRIS buffer (O) compared to a similar experiment run in a constant background of 0.5 M Na+ (b). The theoretical lines are drawn by eq 3 with log Kijopt ) -3.3.

theoretical curve is plotted using the experimentally determined Kexch and the constants Rtot, Ctot, and Ltot to find the expected aI+ for a given value of Π. Dashed lines delimit typical extracellular activity ratios, and the solid lines delimit the intracellular levels (log(aK+/aH+)).33 We find that the response matches well with the theory. The dynamic range at pH 7.2 extends from 0.63 mM to 0.63 M aK+. The log of the selectivity for potassium versus sodium, determined by measuring the horizontal separation of the response curves at Π ) 0.5, is -3.3. This value, when used to plot the expected response in constant 0.5 M Na+ interference, matches the experimental data obtained (see Figure 4). This value indicates a selectivity similar to or better than that obtained for other matrixes incorporating BME-44, e.g., -3.1 in PVC-based fiber-optic work and -3.0 in PVC-based microelectrodes,20 and exactly matches the value given in the review by Buhlmann et al.1 The selectivity should be more than sufficient for measurement in intracellular media where potassium concentration33 is ∼89 mM and sodium is between 5 and 15 mM.

Figure 5. Response time of K+-sensitive PEBBLEs to added KCl and added buffer solutions. It can be seen that response in the forward direction is ∼0.5 s (0-40 mM KCl) (A) and the reverse (see inset) is slightly longer (40-20 mM KCl, B), 0.8 s.

PEBBLE Lifetime. We note that in all the described characterization experiments we never experienced problems with the suspension of K+ PEBBLEs bleaching. The lifetime of the PEBBLE nanosensors was found to be 30 min (PEBBLE recipes 3-5 in Table 2), due to component leaching from the liquid polymer membrane. This is consistent with the lifetime of PVCbased optodes of the same composition.29 After 30 min, the sensor response can deviate up to 7% from the initial calibration data at lower K+ concentration. After 90 min, the deviation is up to 13% at lower K+ concentrations. The deviations are smaller at larger K+ concentrations. The small size of our sensors, beneficial in many ways, compounds the problem of leaching, because even a small loss of ionophore can drastically change the matrix component ratios. However, as the PEBBLEs are single-use sensors made for quick measurements inside in vitro cells that only survive a short period of time, this is acceptable. PEBBLE Response Time and Reversibility. The ratio of the protonated chromoionophore to free base was analyzed versus time in the response time measurements. It was found (see Figure 5) that in going from log aK+/aH+ ) 3.6 to 5.7 the response time (10-90% signal change) was ∼0.5 s (for a concentration change of over 2 decades). In the reverse direction, the response was ∼0.8 s (see Figure 5). This fast, subsecond response time of the PEBBLEs is a direct result of their small size. Diffusion in decyl methacrylate is in the range of 10-8 cm2/s, with small variations depending on cross-linker content.24 Thus, for a PEBBLE radius of ∼300 nm, we expect a diffusion time of ∼10-3 s. This is consistent with the experimental values, which are upper limit values due to the solution mixing times. The sensors also prove to be fully reversible. Figure 6 shows the PEBBLEs, starting in an initial solution of 45 mM KCl, which is diluted to 13 mM and brought back to 45 mM by standard addition of buffer and of 2 M KCl solution, respectively. Intracellular Study. Confocal microscopy was used to determine the localization of the PEBBLE sensors after gene gun delivery. Figure 7 shows the confocal fluorescent image of the PEBBLEs overlaid with a Nomarski differential interference contrast image of the cells. The image indicates that the PEBBLE sensors are localized in the cytoplasm of the glioma cells. Figure 8 shows the

Figure 6. Reversibility of K+-sensitive PEBBLEs. The fluorescence ratio measured was plotted vs the experimental time (determined by stopwatch). The PEBBLEs started in 45 mM K+, buffered at pH 7.2 with TRIS. The solution was then diluted with buffer to a concentration of 12 mM K+; enough standard KCl solution was then added to bring the concentration of the solution back to 45 mM K+.

Figure 7. Confocal image of PEBBLE fluorescence, overlaid with Nomarski image of rat C6 glioma cells. 488-nm excitation, 580-nm long-pass filter.

Figure 8. Ratio data of K+ PEBBLEs in C6 glioma cells during the addition of kainic acid (50 µL of 0.4 mg/mL) at 20 s and at 60 s. Ratios were converted to log(aK+/aH+) using solution calibration of the PEBBLEs. Log(aK+/aH+) is seen to increase after kainic acid addition (and subsequent K+ channel openings).

Analytical Chemistry, Vol. 73, No. 10, May 15, 2001

2227

PEBBLE sensors inside the cells responding to kainic acid addition to the cell medium at 20 and 60 s. Log (aK+/aH+) increases, indicating either an increase in K+ concentration or a decrease in H+ concentration (increase in pH). The amount of kainic acid added is not known to affect the pH of cells in culture. Thus, the change is likely due to increasing intracellular K+, the expected trend. The membrane of C6 glioma cells can initiate an inward rectifying K+ current indicated by specific K+ channels, a documented role in the control of extracellular potassium.34 Thus, when stimulated with a channel opening agonist, the K+ concentration within the glioma cells was expected to increase. While this work has used the signal of many PEBBLEs, we have successfully obtained signal with sufficient signal/noise ratio from a single PEBBLE entrapped in an acrylamide gel. The ability to obtain signal from single PEBBLEs7,35 will be important when data are obtained simultaneously for many analytes using multiplexed PEBBLEs in single cells. CONCLUDING REMARKS Decyl methacrylate-co-hexanediol dimethacrylate has been established as a platform for the construction of nano-optical fluorescent probes (PEBBLEs). The theory describing ion exchange in micrometer-sized polymer films also works for nanometer-sized liquid polymer PEBBLEs. In particular, PEBBLE performance in suspension has been demonstrated. Using PEG in the formulation gave a more uniform sensor diameter of 500600 nm as compared to a range of several micrometers to 500 nm in our previous methods. The PEBBLEs proved to be both selective and reversible. The PEBBLEs are comparable to the PVC-based ion-exchange optodes in their response and selectivity for K+ over Na+. The DMA platform should prove to be very adaptable to selected applications and analyte concentration (34) Emmi, A.; Wenzel, H. J.; Schwartzkroin, P. A. J. Neurosci. 2000, 20, 39153925. (35) Sasaki, K.; Shi, Z.-Y.; Kopelman, R.; Masuhara, H. Chem. Lett. 1996, 141142.

2228

Analytical Chemistry, Vol. 73, No. 10, May 15, 2001

ranges. We have demonstrated the ability to shift the response to K+ by adjusting the cross-linker content. A general design for liquid polymer PEBBLEs has been established. The flexibility in composition of the matrix and sensor components permits this design to be used for a number of different analytes and analyte concentrations. The small size enables novel intracellular studies. Initial studies have shown that the new PEBBLEs can be delivered to viable cells and these cells can be observed in a timeresolved manner, responding to changes in their environment. The sensitivity of these sensors to pH changes will require further consideration when working in the intracellular environment. Fortunately, the small size of PEBBLEs sensors will allow the insertion of more than one type of sensor. The second sensor could be either a pH PEBBLE or a sensor for another common cation (spectrally resolved in either emission or excitation from the K+ PEBBLE sensors). In this manner, one could either determine the intracellular K+ concentration directly, knowing pH, or determine, for example, the ratio of K+/Na+, in situations where the ratio of two cations to each other is more important than the actual concentration.19 ACKNOWLEDGMENT We thank Professor Mark E. Meyerhoff and Dr. Theresa M. Ambrose for help in the initial stages of utilizing DMA and Dr. Maria J. Moreno for assistance with the data normalization formalism. We also thank the University of Michigan Electron Microbeam Analysis Laboratory (funded in part by NSF grant EAR-9628196) for use of the SEM. We acknowledge NIH grants R01-GM50300-04A1 (R.K.) and R01-ES08846 (M.A.P.) and NCI contract N01-CO-07013 for funding.

Received for review October 11, 2000. Accepted February 27, 2001. AC0012041