Microcirculation System with a Dialysis Part for Bioassays Evaluating

Dec 19, 2012 - In vitro Models and On-Chip Systems: Biomaterial Interaction ... Organ/body-on-a-chip based on microfluidic technology for drug discove...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Microcirculation System with a Dialysis Part for Bioassays Evaluating Anticancer Activity and Retention Yuki Imura,† Etsuro Yoshimura,† and Kiichi Sato*,‡ †

Department of Applied Biological Chemistry, School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo, Tokyo 113-8657, Japan ‡ Department of Chemistry and Chemical Biology, School of Engineering, Gunma University, 1-5-1 Tenjin-cho, Kiryu, Gunma 376-8515, Japan ABSTRACT: Medicines are distributed to the whole body and excreted over time. A micromodel of the circulation− excretion system was developed to mimic these processes. This system comprised a dialysis part, a microperistaltic pump, and a target tissue. This microcirculation system was created on a microchip composed of a glass slide and polydimethylsiloxane sheets with microchannels fabricated by photolithography. A dialysis membrane was settled between two channels to form the dialysis part, and a pneumatic peristaltic pump was used to make the solution flow. The excretion and half-life of solute substances absorbed to albumin were changed according to their affinity to the protein. MCF-7 human breast cancer cells were cultured as target cells for drug samples, and the activities of anticancer agents were assayed using our system. Our data demonstrated that the anticancer activity of docetaxel or thio-TEPA could be assayed on the microcirculation−excretion chip. This system may allow for reduced consumption of cells and reagents compared to those required for conventional in vitro bioassay systems.

B

highly efficient analyses in parallel.6 Some techniques are available in which all processes required for complex bioanalyses can be conducted using a single device.7,8 In the human body, the bloodstream is not a single stroke, but rather acts as a circular flow. While there are many in vitro fluidic cell culture systems that use single stroke technologies, few systems utilize circular flow.9,10 Indeed, circulation of the sample is important when the sample concentration changes over time, as occurs in the human body. There is a field of research about organs-on-a-chip applications: several micro models of human organs were constructed on a single device and evaluated through a single procedure.11 We have developed a micro total bioassay system to evaluate intestinal absorption, hepatic metabolism, and bioactivity in breast cancer cells, that is, the absorption and metabolism parameters in ADME.12−14 There are several other reports of a microsystem for absorption and metabolism9,15 but none for the distribution and excretion. Here, we propose a preliminary microsystem useful for the latter part of ADME, distribution and excretion. A dialysis part mimicking glomerular filtration by means of a dialysis membrane, a microperistaltic pump for circulation, and a target tissue were integrated into a microchip (Figure 1). The medium in the closed circuit was circulated via a pneumatic peristaltic pump,16 excreting solutes

ioassays with cultured cells are one of the most important analytical methods in the search for new drugs. Many cell lines are used to evaluate physiological functions.1−4 For example, in order to evaluate the activity of anticancer agents, cancer cells are used as a target. Anticancer activity is generally measured by determining cell death percentages. In conventional methods, only one aspect of the drug candidate can be analyzed, even though the usability/efficacy of such drugs is dependent on their total features. To select promising drug candidates, absorption, distribution, metabolism, and excretion (ADME) should be examined. Drug kinetics plays an important role in these processes, especially distribution and excretion. Binding to plasma proteins is one of the most significant characteristics affecting drug kinetics. Drugs exist in the blood in two forms, protein-bound and unbound. The unbound fraction exhibits pharmacologic effects and is subjected to metabolism or excretion. Thus, plasma protein binding can influence the biological half-life of the drug. In order to measure plasma protein binding, equilibrium dialysis is often performed.5 However, this method requires a significant amount of time. To evaluate distribution and excretion, animal experiments are usually conducted. Animal experiments are ideal; however, they are expensive and are associated with ethical problems. Moreover, the number of new compounds that have been discovered is very large. Therefore, evaluation methods with higher throughput, lower cost, and easier procedures are required. One of the ways to fulfill these requirements is miniaturization technology. Micro total analysis systems enable © 2012 American Chemical Society

Received: October 10, 2012 Accepted: December 19, 2012 Published: December 19, 2012 1683

dx.doi.org/10.1021/ac302938q | Anal. Chem. 2013, 85, 1683−1688

Analytical Chemistry

Article

Figure 1. Schematic illustration of the microcirculation−excretion system for bioassay. A pneumatic peristaltic pump creates the medium circulation, MCF-7 cells are exposed to samples (anticancer agents, in this case), and the dialysis membrane selectively excretes solutes according to their molecular weights. The figure is not to scale.

through a dialysis membrane. Therefore, the amount of a sample in the medium was decreased over time depending on the characteristics of the sample. This system was useful in the evaluation of plasma protein binding. The anticancer agents thio-TEPA (TESPA)17 and docetaxel (DTX)18 were used as model substrates for this system with MCF-7 breast cancer cells as the target tissue. The activities of these anticancer agents were evaluated, and our preliminary evaluation of drug persistence was successful.

■

EXPERIMENTAL SECTION Reagents and Cells. MCF-7 human breast cancer cells were obtained from Riken BioResource Center (Ibaraki, Japan). Dulbecco’s modified Eagle’s medium (DMEM), mineral oil, bovine serum albumin (BSA), and FITC-BSA were purchased from Sigma-Aldrich (St. Louis, MO). Trypsin-EDTA for cell detachment, lucifer yellow (LY), fetal bovine serum (FBS), and RPMI 1640 medium were from Invitrogen (Carlsbad, CA). Fibronectin for cell adhesion was from Asahi Glass (Tokyo, Japan). Model drug candidates for assays, such as warfarin (WRF), TESPA, and DTX, were purchased from Wako Pure Chemicals (Osaka, Japan). Microchip and Fluidics. Detailed procedures of microchip fabrication were described previously.12 The microchip was made of polydimethylsiloxane (PDMS) and a glass slide.19,20 It comprised three PDMS sheets, an upper, middle, and lower sheet, each of which contained microchannels fabricated by photolithography (Figure 2).21 Dimensions of the upper and middle PDMS sheets were 75 × 25 × 3 mm (L × W × T), and the dimensions of the lower sheet were 75 × 25 × 0.1 mm (L × W × T). An SU-8 on glass mold was fabricated with an SU-8 2015 photoresist (MicroChem, Newton, MA) to make 50−150 μm deep microchannels. The lower PDMS sheet was fabricated with a spin coater to a thickness of 0.1 mm (2000 rpm, 30 s).

Figure 2. (a) Assembly drawing and (b) photograph of the microcirculation−excretion chip. A: Microperistaltic pump, composed of four control channels. B: Dialysis part, composed of two parallel microchannels and the dialysis membrane. C: Cell culture component on the microchannel of the closed circuit. The figure is not to scale.

1684

dx.doi.org/10.1021/ac302938q | Anal. Chem. 2013, 85, 1683−1688

Analytical Chemistry

Article

recovered with trypsin-EDTA, and the fresh medium was added. Then, the cell suspension was introduced onto the microchip and cultured on the fibronectin-coated glass surface of the microchannel as described previously.13 In brief, the cell suspension (1.0 × 107 cells/mL) was introduced with a 1 mL syringe (Terumo, Tokyo, Japan). A silicon tube (2 × 4 mm; i.d. × o.d.) was connected to the syringe. The other side of the tube was connected to the cell inlet port, and the suspension was gradually injected. The suspension was allowed to spread into the cell culture chamber. In order to ensure that the cells settled in the appropriate chamber, inlet and outlet ports were set on opposite sides of the chamber. After the cells were introduced into the chamber, both ports were capped with a silicon tube that was plugged with PDMS. Permeation Test. Permeation tests of model samples were performed using the microchip. LY, WRF, and FITC-BSA were used as model samples. LY is a hydrophilic fluorescent substance with a low molecular weight. WRF is a hydrophobic substance with a low molecular weight that binds easily to albumin. FITC-BSA is a fluorescent substance with a high molecular weight. The sample solution was introduced into the lower microchannel with the microsyringe pump or the microperistaltic pump. Permeated analyte was run in the upper microchannel. The stream in the upper channel was facilitated by the syringe pump. The resulting solution from the upper microchannel was collected and analyzed. The permeability was evaluated by calculating the permeability coefficient.23 Determination of Analyte Amounts. The amount of analytes withdrawn from the upper and lower microchannels was determined by fluorometric measurement (LY: ex. 428 nm, em. 536 nm; WRF: ex. 320 nm, em. 380 nm; FITC-BSA: ex. 495 nm, em. 520 nm). The concentrations of LY in the microchannel were measured with a fluorescent microscope (Olympus IX 71, Tokyo, Japan; ex. 460−495 nm, em. 510−550 nm, dm. 505 nm). Cell Viability Test. Resazurin-converting activity was measured using a CellTiter-Blue cell viability assay reagent (Promega, Madison, WI) as described previously.13 The reagent, diluted to 20% (v/v) in medium, was applied to the cells. The fluorescence intensity of the reaction product, resorufin, was monitored (ex. 570 nm, em. 585 nm).24 The viability of MCF-7 cells cultured on the microchip for 1 day was first measured, and the cells were then cultured for 1 day with the sample compound dissolved in the medium. The viability of the treated MCF-7 cells was measured again, and the bioactivity of the sample was estimated from the ratio of the obtained bioactivity to the original activity.

Vertical microchannels connecting the upper and lower sides of the middle PDMS sheet were also manually fabricated as follows. Teflon tubes with 0.25 mm i.d. covered with silicon tubes were placed on the glass mold. Inside these tubes, fused silica capillaries (o.d. = 0.2 mm) were settled. The PDMS prepolymer was cast onto these tubes to a 3 mm height. After curing, the fused silica capillaries were removed, and 0.25 mm i.d. vertical channels were obtained. Injection ports were fabricated in the upper PDMS sheet to allow introduction of the medium and cells onto the microchip. The ports were made of Teflon tubes (1 × 2 × 10 mm; i.d. × o.d. × L), which were covered with silicone tubes (1 × 3 × 3 mm; i.d. × o.d. × L) to allow fixation on the upper PDMS sheets. The lower PDMS sheet was attached to a glass slide. The middle PDMS sheet was attached to the other side of the lower PDMS sheet. Next, a cellulose dialysis membrane (MWCO = 15 000) cut from a Visking tube was placed over the dialysis channel on the middle PDMS sheet. Finally, the upper PDMS sheet was attached to the other side of the middle PDMS sheet to sandwich the membrane between the upper and middle PDMS sheets. The glass slide and PDMS sheets were activated by oxygen plasma before attachment. For easy handling, the microchip was placed in an autoclaved plastic box and kept in a CO2 incubator. The outer flow of the dialysis part was pumped with a microsyringe pump (KDS230, KD Scientific, Holliston, MA). The fused silica capillary was connected to a syringe needle and bonded with an epoxy adhesive. Closed-Circuit Microchannel. A sample solution was infused into the closed-circuit microchannel through the injection port. After filling the circulation, the solution remaining in the injection port was removed, and the port was filled with mineral oil. This prevented contamination and evaporation from the injection port. Dialysis Part. For the dialysis part, a dialysis membrane was used to separate two microchannels in the same manner as the cell supporting membrane in our previous work.12 Buffer or medium representing primitive urine flowed in the upper channel. Sample solution representing the bloodstream flowed in the lower circulation channel. The dialysis membrane between the two channels allowed only small molecules to pass through. The substances that moved from the lower channel to the upper channel were collected in a microtube. BSA was dissolved in the sample solution, mimicking plasma proteins. Hydrophobic drug candidates were adsorbed to BSA and were less able to permeate the dialysis membrane. Microperistaltic Pump. Four microchannels (200 × 150 μm; W × T) were fabricated to act as a micropneumatic peristaltic pump.16 The pressure in the regulatory channels was controlled with PC-controlled air solenoid valves (V100 series, SMC; Tokyo, Japan). Compressed air was brought into the channels at 0.02 MPa and was then divaricated into a quartet solenoid valve. The solenoid valves were controlled with a digital switch (NI9477) and NI cDAQ-9172 using LabView software (National Instruments; Austin, TX). The microperistaltic pump was settled over the closed circuit and allowed sequential flow pressing in the lower channel. Target Cells. MCF-7 human breast cancer cells22 were selected as the target tissue. The cells were cultured at 37 °C in a 60 mm cell culture dish with RPMI 1640 medium supplemented with 10% FBS in a humidified atmosphere of 5% CO2. After reaching 70% confluence, the cells were

■

RESULTS AND DISCUSSION Dialysis Part. We conducted a permeation test on the microchip and calculated the permeability coefficient from the experimental data (Figure 3). In these experiments, the dimensions of the dialysis part were 4 × 0.5 × 0.05 mm (L × W × T). The flow rate of each solution was 0.2 μL/min, and 10 μM LY, 500 μM FITC-BSA, or 500 μM WRF was used as a sample solution. FITC-BSA could not permeate the dialysis membrane because it was too large to pass through the pore of the membrane. However, LY and WRF, which were small enough to cross through the pores of the membrane, permeated the dialysis membrane well when BSA was not included in the sample solution. The permeation coefficients were 53.0 × 10−6 1685

dx.doi.org/10.1021/ac302938q | Anal. Chem. 2013, 85, 1683−1688

Analytical Chemistry

Article

Figure 3. Permeability coefficients obtained with the microchip system (n = 3). The calculation method was described previously.12

and 33.3 × 10−6 cm/s, respectively. The permeation coefficient of WRF was decreased to 4.0 × 10−6 cm/s (12% of its original value) when 0.1% BSA was dissolved in the sample solution. On the other hand, LY maintained a permeation coefficient of 35.1 × 10−6 cm/s (67% of its original value), even when 0.1% BSA was included in the sample solution. The difference was likely attributable to differences in the affinities of these samples to BSA; that is, WRF was more likely to bind to BSA than LY. These results were consistent with the known features of these materials.25 Microperistaltic Pump. The dimensions of the horizontal microchannel of the closed circuit were 60 × 0.2 × 0.05 mm (L × W × T). The vertical microchannels measured 0.3 × 6 mm (i.d. × L), while the joints that connected the horizontal and vertical microchannels measured 1 × 0.2 mm (i.d. × L). Therefore, the interior volume of the entire closed circuit was estimated to be 1.2 μL. The approximate flow rate in the closed circuit was measured by particle image velocimetry using Fluoresbrite microbeads (Polysciences, Warrington, PA). The flow rate was about 11.8 nL/min, indicating that the solution traversed the entire circuit in 100 min. Microcirculation−Excretion System. Both the dialysis part and the microperistaltic pump described above were mounted onto 1 microchip (Figure 2). A part of the closed circuit was in contact with the dialysis membrane and ran parallel to the outer flow (upper microchannel) across the membrane. The interior volume of the closed circuit was 1.2 μL. The dimensions of the horizontal microchannel, excluding the dialysis part, were 50 × 0.2 × 0.05 mm (L × W × T), whereas those of the dialysis part were 4 × 0.5 × 0.05 mm (L × W × T). In order to confirm that the pump was able to circulate the sample solution throughout the entire circuit, including every inlet, outlet, and corner, LY remaining in the channel and excreted out of the channel was measured. LY solution (10 μM) was loaded as a sample solution. The results are shown in Figure 4a,b. When the pump did not work, the LY concentration in the closed circuit hardly changed, and little LY excretion was observed. The amount of LY excreted from the chip was 0.9 pmol; this value is almost the same as the amount of LY located in the dialysis part (10 μM; 4 × 0.5 × 0.05 mm). This indicated that LY did not diffuse through the microchannel and that LY located only in the dialysis part was excreted. On the other hand, when the microperistaltic pump was functioning to circulate the sample solution, the LY concentration in the closed circuit decreased, and amount of excreted LY increased over time, suggesting that the sample solution was circulating around the closed circuit and that most of the LY was successfully excreted.

Figure 4. (a,b) Time course of LY excretion measured with this novel system. (a) Amount of LY excreted to the outer stream. (b) Estimated amount of LY remaining in the closed circuit. (c) Excretion of LY and warfarin.

The microcirculation−excretion system was used to test the permeability of the chemical substances considering the effect of plasma protein binding. The excretion of substrates bound to BSA (at 0.1%) in the sample solution was measured over time. The results are shown in Figure 4c. LY, which was weakly bound to BSA, was excreted from the microchannel over time, with 70% of LY excreted in 2 h. On the other hand, the excretion of WRF was very slow because the majority of WRF was bound to BSA and could not pass through the pore of the dialysis membrane. These results were consistent with those of permeation tests and might indicate the properties of the drugs in a human body. The renal excretion of WRF is very small (less than 1%),26 and WRF has a long half-life in the human body.27 Therefore, it is suggested that this microcirculation− excretion system may be a useful tool for the study of drug kinetics. Bioassays Using the Novel Microcirculation−Excretion System. The scale of the closed circuit was designed for cell culture. The interior volume of the closed circuit was 2.9 μL, and the total dimensions of the horizontal microchannel were 50 × 0.3 × 0.1 mm (L × W × T). The dimensions of the vertical microchannel were 0.2 × 6 mm (i.d. × L), while those of the joints were 1 × 1 mm (i.d. × L). The dimensions of the dialysis part were 4 × 0.5 × 0.1 mm (L × W × T), and the flow rate in the closed circuit was estimated at 25 nL/min. 1686

dx.doi.org/10.1021/ac302938q | Anal. Chem. 2013, 85, 1683−1688

Analytical Chemistry

Article

microsyringe pump and a pneumatic peristaltic pump. This will not only enable the elimination of troublesome operations like pipetting but also allow for more in vivo-like bioassays. With some improvements, this system may be applicable to the analysis of pharmacologically important properties, such as halflife or area under the curve (AUC), which are difficult to measure with in vitro experiments. In this study, hydrophobic PDMS was used as a microchip material. Thus, nonspecific adsorption of the sample to the microchip likely occurred. To improve this problem, hydrophilic MPC coatings30 or other techniques may be useful. This system requires only 1.2 μL of a sample solution when the sample compounds are to be analyzed chemically and 2.9 μL of a solution for bioassays. These volumes are 2 or more orders of magnitude less than conventional methods. Additionally, this microfluidic system can be connected to other assay components, such as microintestines or microlivers, as we reported previously.13,14 Thus, this system has the potential to be applied to more complex and sophisticated analyses.

MCF-7 cells were cultured in a part of the closed circuit (Figure 2). The microchannels were filled with DMEM containing 10% FBS and 100 μM TESPA or 2 μM DTX. In order to prevent nutritional deficiency in the closed circuit, fresh DMEM was infused into the upper microchannel as an outer stream with a syringe pump. The flow rate of the upper microchannel was 0.05 μL/min. Figure 5 shows the results of bioassays using the microcirculation−excretion system. In control experiments, the same

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Fax: +81-277-30-1251. Notes

Figure 5. Cell death percentages in target MCF-7 cells treated with anticancer agents. The anticancer activity was estimated from the ratio of viable MCF-7 cells incubated with samples for 1 day to the number of viable MCF-7 cells before treatment (n = 3).

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported in part by a Grant-in-Aid for Scientific Research (KAKENHI; Grant Number 24350035) and a grant from the Food Nanotechnology Project, the Ministry of Agriculture, Forestry and Fisheries, Japan.

concentrations of drugs and 10% FBS were added to the outer stream. Under these conditions, drug samples were not excreted to the upper channel, and the drug concentration in the closed circuit was not decreased. The anticancer activity of TESPA was lower when the excretion system worked correctly than during the control experiment, which lacked proper excretion. The initial TESPA concentration was enough to show 50% cell death rate in conventional microchip-based bioassay, and less concentration showed smaller cell death rate (data not shown). Thus, the weak anticancer activity under the proper excretion was likely due to the decreased concentration of TESPA in the closed circuit. On the other hand, the activity of DTX did not decrease significantly, even when the excretion system worked correctly. This was likely because DTX was not excreted via the dialysis part and remained in the closed circuit. These bioassay results using the microcirculation−excretion system are consistent with the properties of sample compounds. Because the plasma protein binding ratio of TESPA is about 10%,28 TESPA is likely to be free from serum proteins and to permeate the dialysis membrane. In contrast, the plasma protein binding ratio of DTX is more than 90%.29 Therefore, most of the DTX was bound to serum proteins, such as albumin, and could not be excreted at the dialysis part. As described above, this microcirculation−excretion system successfully mimicked glomerular filtration in the renal excretion of these known substances.

■

REFERENCES

(1) Alley, M. C.; Scudiero, D. A.; Monks, A.; Hursey, M. L.; Czerwinski, M. J.; Fine, D. L.; Abbott, B. J.; Mayo, J. G.; Shoemaker, R. H.; Boyd, M. R. Cancer Res. 1988, 48 (3), 589−601. (2) Sakai, Y.; Shoji, R.; Kim, B. S.; Sakoda, A.; Suzuki, M. Environ. Monit. Assess. 2001, 70 (1−2), 57−70. (3) Hilgers, A. R.; Conradi, R. A.; Burton, P. S. Pharm. Res. 1990, 7 (9), 902−910. (4) Yoshitomi, S.; Ikemoto, K.; Takahashi, J.; Miki, H.; Namba, M.; Asahi, S. Toxicol. in Vitro 2001, 15 (3), 245−256. (5) Oravcova, J.; Bohs, B.; Lindner, W. J. Chromatogr., B 1996, 677 (1), 1−28. (6) Manz, A.; Effenhauser, C. S.; Burggraf, N.; Harrison, D. J.; Seiler, K.; Fluri, K. J. Micromech. Microeng. 1994, 4 (4), 257−265. (7) Tokuyama, T.; Fujii, S.; Sato, K.; Abo, M.; Okubo, A. Anal. Chem. 2005, 77 (10), 3309−3314. (8) Goto, M.; Sato, K.; Murakami, A.; Tokeshi, M.; Kitamori, T. Anal. Chem. 2005, 77 (7), 2125−2131. (9) Kimura, H.; Yamamoto, T.; Sakai, H.; Sakai, Y.; Fujii, T. Lab Chip 2008, 8 (5), 741−746. (10) Kim, H. J.; Huh, D.; Hamilton, G.; Ingber, D. E. Lab Chip 2012, 12 (12), 2165−2174. (11) Baker, M. Nature 2011, 471 (7340), 661−665. (12) Imura, Y.; Asano, Y.; Sato, K.; Yoshimura, E. Anal. Sci. 2009, 25 (12), 1403−1407. (13) Imura, Y.; Sato, K.; Yoshimura, E. Anal. Chem. 2010, 82 (24), 9983−9988. (14) Imura, Y.; Yoshimura, E.; Sato, K. Anal. Sci. 2012, 28 (3), 197− 199. (15) Gao, D.; Li, H. F.; Wang, N. J.; Lin, J. M. Anal. Chem. 2012, 84 (21), 9230−9237. (16) Unger, M. A.; Chou, H. P.; Thorsen, T.; Scherer, A.; Quake, S. R. Science 2000, 288 (5463), 113−116.

■

CONCLUSIONS In this study, we developed a microsystem to evaluate the time course of glomerular filtration and the bioactivity of drugs. We concluded that the developed system could be used to efficiently evaluate the properties of drugs. The microcirculation−excretion system described here can control successive processes with simple operation of a 1687

dx.doi.org/10.1021/ac302938q | Anal. Chem. 2013, 85, 1683−1688

Analytical Chemistry

Article

(17) van Maanen, M. J.; Smeets, C. J. M.; Beijnen, J. H. Cancer Treat. Rev. 2000, 26 (4), 257−268. (18) Morse, D. L.; Gray, H.; Payne, C. M.; Gillies, R. J. Mol. Cancer Ther. 2005, 4 (10), 1495−1504. (19) McDonald, J. C.; Whitesides, G. M. Acc. Chem. Res. 2002, 35 (7), 491−499. (20) Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Anal. Chem. 1998, 70 (23), 4974−4984. (21) Lorenz, H.; Despont, M.; Fahrni, N.; Brugger, J.; Vettiger, P.; Renaud, P. Sens. Actuators, A 1998, 64 (1), 33−39. (22) Ciocca, D. R.; Fuqua, S. A. W.; Locklim, S.; Toft, D. O.; Welch, W. J.; McGuire, W. L. Cancer Res. 1992, 52 (13), 3648−3654. (23) Alsenz, J.; Haenel, E. Pharm. Res. 2003, 20 (12), 1961−1969. (24) O’Brien, J.; Wilson, I.; Orton, T.; Pognan, F. Eur. J. Biochem. 2000, 267 (17), 5421−5426. (25) Pinkerton, T. C.; Koeplinger, K. A. Anal. Chem. 1990, 62 (19), 2114−2122. (26) Burton, M. E. Applied Pharmacokinetics & Pharmacodynamics: Principles of Therapeutic Drug Monitoring; Lippincott Williams & Wilkins: Baltimore, MD, 2006; pp 715−751. (27) Horton, J. D.; Bushwick, B. M. Am. Fam. Physician 1999, 59 (3), 635−646. (28) McDermott, B. J.; Double, J. A.; Bibby, M. C.; Wilman, D. E. V.; Loadman, P. M.; Turner, R. L. J. Chromatogr. 1985, 338 (2), 335−345. (29) Urien, S.; Barre, J.; Morin, C.; Paccaly, A.; Montay, G.; Tillement, J. P. Invest. New Drugs 1996, 14 (2), 147−151. (30) Goda, T.; Matsuno, R.; Konno, T.; Takai, M.; Ishihara, K. Colloids Surf., B 2008, 63 (1), 64−72.

1688

dx.doi.org/10.1021/ac302938q | Anal. Chem. 2013, 85, 1683−1688