- Email: [email protected]
A 3D tumor spheroid chip with the pharmacokinetic drug elimination model developed by balanced droplet dispensing
Sensors and Actuators B 174 (2012) 436–440
Contents lists available at SciVerse ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
A 3D tumor spheroid chip with the pharmacokinetic drug elimination model developed by balanced droplet dispensing Taeyoon Kim 1 , Il Doh, Young-Ho Cho ∗ Cell Bench Research Center, KAIST, 373-1 Guseong-dong, Yuseong-gu, Daejeon 305-701, Republic of Korea
a r t i c l e
i n f o
Article history: Received 1 March 2012 Received in revised form 9 July 2012 Accepted 21 August 2012 Available online 6 September 2012 Keywords: Pharmacokinetic drug elimination model Balanced droplet dispensing 3D tumor spheroid chip
a b s t r a c t This paper presents a three-dimensional (3D) tumor spheroid chip where the pharmacokinetic (PK) drug elimination model is developed by balanced droplet dispensing. The PK model should reflect the exponential decrease of drug concentration in a body to improve the reliability of cell-based screening assay. The previous cell chips, however, maintain the concentration of anti-cancer agents at constant or discretely decrease the concentration level by manual process. In the present chip where medium droplet dispensers are integrated on a well array, the small droplets of fresh medium and well waste are autonomously replaced each other, the exponential elimination of drug concentration in the well is provided. In the experimental study, we demonstrated that the present chip successfully develops the exponential elimination model using fluorescein isothiocyanate (FITC) and cisplatin. Although the exponential decay of cisplatin in the present chip did not show significant effect biologically due to the current protocol where 3 days of constant concentration pre-treatment is performed, we observed different growth pattern of H358 spheroids depending on peak concentration of 0, 1, 10, and 100 M. The present chip has potential for analyzing the response of 3D tumor spheroids in the pharmacokinetically relevant drug elimination model for high-reliability spheroid-based drug screening chip. © 2012 Elsevier B.V. All rights reserved.
1. Introduction High-reliability anti-cancer drug screening is important to select optimized drug candidates among numerous agents for effective cancer chemotherapy. In screening process, the cellular response to various anti-cancer drugs and its combination is analyzed. In order to improve the reliability of the in vitro cell-based anti-cancer drug screening, the pharmacokinetic (PK) parameters of the drug should be considered so that the screening results can precisely reflect the time course effects of drugs to tumor [1]. The understanding of the PK processes in a body [2] is essential to design the optimized anti-cancer drug regimen such as dosage and medication intervals of drugs for cancer patients. Therefore, the PK model of drugs should be applied in the cell-based drug screening to predict the physiological efficacy to cancer patients. In general PK models of drugs, body consists of several compartments [2] and the drug concentration in plasma or serum is generally regarded as the exponentially decreasing model due to the reaction of a drug within or between the compartments as shown in Fig. 1 (solid line). The previous cell-based drug screening chips [3–11], however, characterized the cellular response to
∗ Corresponding author. Tel.: +82 42 350 8691; fax: +82 42 350 8690. E-mail address: [email protected] (Y.-H. Cho). 1 Process Methodology Development Group, Samsung Electromechanics Co., Ltd. 0925-4005/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.snb.2012.08.041
anti-cancer agents without any chemical dilution of drug. The drug concentration in these chips [3–11] shows the constant value throughout the drug screening process as the dashed line in Fig. 1. In the some other cell-based screening methods [12,13], the drugs were manually diluted by the periodic medium renewal. However, the manual medium renewal process is time-consuming and results in the well contamination and large deviation of measured results caused by human factors. In addition, the medium renewal at the specific interval results in the discrete reduction of drug concentration (dotted line), different from the continuous exponential profile in the PK model (solid line). In this work, we present 3D spheroid chip capable to apply the exponential drug elimination model in Fig. 1 (solid line) based on the balanced droplet dispensing. Since the small droplets of fresh medium and well waste are autonomously replaced each other, the exponential decrease of drug concentration in the well is obtained. This paper focuses on the design and experimental demonstration of the 3D tumor spheroid chip for analyzing the response of spheroids to the anti-cancer drug in the PK model of exponential drug elimination. 2. Theory and design The present chip, illustrated in Fig. 2, consists of two polydimethylsiloxane (PDMS) layers: a well layer and a droplet dispensing layer. Previously, we had developed a simple method capable to supply a fresh medium droplet, named as the balanced
T. Kim et al. / Sensors and Actuators B 174 (2012) 436–440
437
Assuming that the medium droplet supplied from the droplet dispenser layer results in an uniform drug concentration, C, in the well, we derive the first-order differential equation of Q ˙ + C(t) = 0 C(t) V
(1)
where Q and V are the perfusion rate and the volume of drug solution in the well, respectively. By solving the differential equation, the drug concentration, C, is expressed as a exponential function of
Q
C(t) = C0 exp −
Fig. 1. Comparison of pharmacokinetic (PK) drug elimination models in the previous and present chips.
droplet dispensing [14]. In this method, the hydraulic-head difference, h, between the well-inlet and drain-outlet, cause the removal of a waste droplet at the drain. The decreased pressure in the well due to volume expansion provokes a fresh medium droplet supply from the dispensing layer. Therefore, the fresh medium droplet (Fig. 2) is autonomously supplied to the well without any pumps in the exact amount of the removed waste droplet. The well layer has an array (4 × 8) of the well (diameter: 6 mm). All experiments are done in an array format. For the generation of 3D tumor spheroids, we used a hanging drop method as shown in Fig. 3. We seed cells in the well array and place the chip upside down so that the cells aggregate each other due to the gravitational force in the suspended medium at the center of medium curvature by hanging drop method [15–17]. The wells of the present chip are initially supplied with a drug at the specific concentration of C0 . Then the drug solution in the well array is gradually diluted from the initial concentration, C0 , to 0, based on the balanced medium droplet dispensing of Fig. 3.
V
t
(2)
where C0 is the initial drug concentration at the well. As a result, the balanced droplet dispensing is capable to develop the equivalent one-compartmental PK model illustrated in Fig. 3b. In the PK model in Fig. 4b, the body is depicted as a kinetically homogeneous unit with the assumption that the drug achieves instantaneous distribution throughout the body following the drug administration. As a result, the drug concentration to time profile of the well and the one-compartmental PK model shows the identical exponential response of Fig. 1 (solid line). Here, the half-life time, t1/2 , required to reduce the drug concentration to one half its initial value, is derived as t1/2 =
V ln 2 Q
(3)
Therefore, the half-life time, t1/2 , of the anti-cancer drug can be controlled by adjusting the perfusion rate, Q, and the volume, V, of drug solution in the well. 3. Materials and methods 3.1. Fabrication process The droplet dispenser layer was fabricated from the molding and bonding processes of two PDMS plates. We fabricated the
Fig. 2. The present chip utilizing the balanced droplet dispensing for exponential elimination profile of drug concentration.
Fig. 3. Spheroid formation in the well layer using hanging drop method.
438
T. Kim et al. / Sensors and Actuators B 174 (2012) 436–440
4. Experimental results and discussion 4.1. Drug elimination
Fig. 4. The relative fluorescein isothiocyanate (FITC) concentration as a function of the perfusion duration in the present chip; inserted figures show the fluorescence images of the FITC solution during the medium perfusion.
8 mm-thick top plate of the droplet dispenser layer by curing PDMS pre-polymer mixture (curing agent-to-PDMS ratio of 1:10, Sylgard 184, Dow Corning) in an acrylic jig for 12 h at 75 ◦ C. The 3 mmthick bottom plate of the droplet dispenser layer was obtained by curing 24 g PDMS pre-polymer mixture on a 4 in. bare silicon wafer for 2 h at 85 ◦ C. We bonded the top and bottom PDMS plates of the droplet dispenser layer by treating the bonding surfaces with O2 plasma for 30 s. Then we plugged a 2 mm-long polypropylene tip at the center of each well bottom after perforating a 1mm-diameter hole by a puncher. The well layer was fabricated from the similar process for the droplet dispenser layer, excepting that the 4 mm-thick bottom plate is fabricated by curing 30 g PDMS pre-polymer on the drain channel mold for 2 h at 85 ◦ C. The fabricated droplet dispenser layer and well layer were sterilized by an autoclave and dried overnight. The bottom surfaces of the wells were treated with 2 wt.% bovine serum albumin (BSA) solution for 1 h at room temperature to prevent the cell adhesion. Finally, we stacked the PDMS layers and tightly sealed them using an acrylic jig with bolted joint. From the fabricated channel dimension, we obtained the fluidic resistance of the main and branch drain channels as 3.13 × 1011 Pa s/m3 and 2.24 × 1014 Pa s/m3 , respectively, and determined the hydraulichead difference, h, of 38 mm for generating the perfusion rate, Q, of 0.1 l/min. 3.2. Cell preparation In the experimental study, we used H358 human non-small cell lung cancer (NSCLC) cell line which is stably transfected by green fluorescence protein (GFP) to facilitate the cell image analysis. We cultured the H358 cells in complete media consisting of RPMI-1640 (Gibco) supplemented with 10% (v/v) fetal bovine serum (FBS; Gibco), 0.2% (v/v) G418 solution (Sigma Aldrich), and 1% (v/v) penicillin–streptomycin (Invitrogen). The cells were cultured in a humidified 37 ◦ C and 5% CO2 incubator and routinely passaged at 80% confluence. We collected the cells by treating the culture dish with 0.25% trypsin/EDTA (Invitrogen) for 3 min at 37 ◦ C, then neutralized the cell solution with 10% RPMI-1640 medium (Gibco). After spinning down the cells with a centrifuge at 1500 rpm for 3 min, we resuspended the cells in fresh medium. We adjusted the number of cells by measuring the cell concentration with a hemocytometer (Marienfeld), and seeded the cells in the chip by using a micropipette (Eppendorf).
To demonstrate the capability for the exponential drug elimination, we first measured the concentration profile of fluorescein isothiocyanate (FITC) gradually diluted from the balanced medium droplet dispensing. We supplied 140 l of FITC solution in the wells, and perfused the medium at the rate, Q, of 0.1 l/min using the balanced droplet dispensing. Then we collected the FITC solution from the drain outlet and measured the fluorescence intensity at the intervals of 1 day for the total duration of 5 days. Fig. 4 shows the measured relative FITC concentration as a portion of the initial value during the medium perfusion for 5 days. In the graph of Fig. 4, the average and standard deviation of the FITC concentration were obtained from three measurements. Based on Eq. (3), the half-life, t1/2 , of the FITC is calculated as 16 h for the well solution volume, V, of 140 l and the perfusion rate, Q, of 0.1 l/min. In the previous pharmacologic study [18], the half-life of in vivo cisplatin administration was measured as 21.7 ± 14.1 h. Therefore, further modification of the well solution volume, V, and the perfusion rate, Q is required to precisely adjust the present half-life to that of in vivo profile. In the measured FITC dilution graph of Fig. 5, the in vivo cisplatin profile [18] is also illustrated to clarify the difference between the in vivo and experimental profiles. As a result, we observed that the measured FITC concentration profile agrees well with the exponential function of Eq. (3) within the error of less than 14.8%. Therefore, we experimentally verified that the present chip implements the exponential drug elimination in the one-compartmental PK model using the balanced droplet dispensing. 4.2. 3D tumor spheroid growth pattern We formed 376.7 ± 23.4 m-diameter 3D spheroids from 5 × 103 H358 cells for 5 days and supplied the anti-cancer drug, cisplatin, with the four different concentrations of 0 (Control), 1, 10, and 100 M into the wells and incubated the spheroids for 3 days in 37 ◦ C and 5% CO2 . Then we implemented the exponential cisplatin elimination using the medium perfusion at the rate, Q, of 0.1 l/min. Based on the previously established protocols [13] for the drug treatment of tumor spheroids, we incubated the spheroids at the constant cisplatin concentration for 3 days ahead of the exponential drug elimination for comparison. During the exponential cisplatin elimination, we refilled the droplet dispenser layer with fresh medium solution once per 36 h. Fig. 5 shows the time-lapse fluorescent spheroid images captured by an inverted fluorescent microscope. The diameter of the 3D spheroids was measured from the capture images using MATLAB® image processing toolbox. Fig. 6 shows the measured spheroid growth pattern. In the graph of Fig. 6, the average and standard deviation of the spheroid diameter were measured from three different spheroids. We observed that the 3D spheroids steadily grow without the cisplatin treatment (C0 = 0 M). However, the 3D spheroids treated with cisplatin showed different growth pattern. In the cisplatin elimination model with the peak concentration, 1 M, the H358 spheroids initially followed the growth pattern similar to that of control (without cisplatin treatment) but shrank after the 3 days of drug elimination. The cisplatin elimination model with higher concentration of 10 M showed the rapid shrinkage of the spheroid after the drug treatment. At the C0 of 100 M, we could not even measure the spheroid diameter since the spheroids were completely disaggregated directly after the drug administration. Although the exponential decay of cisplatin in the present devices did not show significant effect biologically due to the current protocol where 3 days of constant concentration pre-treatment is performed, these
T. Kim et al. / Sensors and Actuators B 174 (2012) 436–440
439
Fig. 5. Time-lapse microscopic fluorescence images of 3D spheroids, formed by H358 cell number of 5 × 103 , cultured for 15 days in the cisplatin elimination model with the four different peak concentrations.
cells in the 3D spheroid model was much higher than 2D model in all cisplatin concentration. Fig. S1 also showed that some cells in the 3D spheroid model were still alive. This indicated that the drug cannot penetrate into the core of spheroid and cells in this region could survive. PDMS we used for the fabrication of chip is a hydrophobic material with a non-specific absorption characteristic. Hydrophobic drugs may not follow an exponential decay in the present chip for a long time culture due to the drug absorption at the PDMS surface. For the use of various drugs, other materials with good anti-absorption characteristics should be considered. Due to the simplicity of the structure, the micromachining of these materials would not be a big concern. 5. Conclusions
Fig. 6. Diameter of 3D spheroids, formed by H358 cell number of 5 × 103 , cultured for 15 days in the cisplatin elimination model with the three different peak concentrations.
results suggest that the H358 spheroids shows different growth response in the cisplatin elimination model with different peak concentrations. For further biological study on the effect of exponential drug decay, cell responses with and without exponential decay should be examined. For a comparison of 2D monolayer and 3D spheroid models, we measured the cell viability from these models using the identical number (5 × 103 ) of H358 cells treated by cisplatin at varying static concentration for 3 days (Supplementary information, Fig. S1). While the results of Fig. 5 shows the simple growth pattern of 3D spheroids, the cell viability of Fig. S1 measured from the fluorescence staining method (Supplementary information, Fig. S2) showed the different responses between 2D monolayer and 3D spheroid models against the cisplatin treatment. The viability of
We designed, fabricated, and characterized the 3D tumor spheroid chip for the drug–response analysis of spheroids in the PK model of drug elimination developed by balanced droplet dispensing. We experimentally demonstrated the 3D tumor spheroid chip successfully provides the exponential drug elimination model within the error less than 14.8%. In the cisplatin elimination model with the different peak concentrations of 0, 1, 1, 10, and 100 M, we observed different growth patterns for the H358 spheroids. For further biological claims on the effect of exponential drug decay, cell responses with and without exponential decay should be examined. The present chip has potential for analyzing the response of 3D tumor spheroids in the pharmacokinetically relevant drug elimination model for high-reliability spheroid-based drug screening chip. Acknowledgement This research was supported by the Converging Research Center Program funded by the Ministry of Education, Science and Technology (Project No. 2011K000864).
440
T. Kim et al. / Sensors and Actuators B 174 (2012) 436–440
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.snb.2012.08.041. References [1] S. Rockwell, In vivo–in vitro tumour cell lines: characteristics and limitations as models for human cancer, British Journal of Cancer Supplement 4 (1980) 118–122. [2] M.M. Ames, G. Powis, J.S. Kovach, Pharmacokinetics of Anticancer Agents in Humans, Elsevier, 1983. [3] Y. Torisawa, H. Shiku, T. Yasukawa, M. Nishizawa, T. Matsue, Three-dimensional micro-culture system with a silicon-based cell array device for multi-channel drug sensitivity test, Sensors and Actuators B: Chemical 108 (2005) 654–659. [4] J. Pihl, J. Sinclair, E. Sahlin, M. Karlsson, F. Petterson, J. Olofsson, O. Orwar, Microfluidic gradient-generating device for pharmacological profiling, Analytical Chemistry 77 (2005) 3897–3903. [5] P.J. Lee, P.J. Hung, V.M. Rao, L.P. Lee, Nanoliter scale microbioreactor array for quantitative cell biology, Biotechnology and Bioengineering 94 (2006) 5–14. [6] M. Yamada, T. Hirano, M. Yasuda, M. Seki, A microfluidic flow distributor generating stepwise concentrations for high-throughput biochemical processing, Lab on a Chip 6 (2006) 179–184. [7] C.-W. Li, J. Yang, M. Yang, Dose-dependent cell-based assays in V-shaped microfluidic channels, Lab on a Chip 6 (2006) 921–929. [8] Z. Wang, M.-C. Kim, M. Marquez, T. Thorsen, High-density microfluidic arrays for cell cytotoxicity analysis, Lab on a Chip 7 (2007) 740–745. [9] S. Sugiura, J. Edahiro, K. Kikuchi, K. Sumaru, T. Kanamori, Pressure-driven perfusion culture microchamber array for a parallel drug cytotoxicity assay, Biotechnology and Bioengineering 100 (2008) 1156–1165. [10] K. Ziolkowska, E. Jedrych, R. Kwapiszewski, J. Lopacinska, M. Skolimowski, M. Chudy, PDMS/glass microfluidic cell culture system for cytotoxicity tests and cells passage, Sensors and Actuators B: Chemical 145 (2010) 533–542. [11] M.-H. Wu, Y.-H. Chang, Y.-T. Liu, Y.-M. Chen, S.-S. Wang, H.-Y. Wang, C.-S. Lai, T.-M. Pan, Development of high throughput microfluidic cell culture chip for perfusion 3-dimensional cell culture-based chemosensitivity assay, Sensors and Actuators B: Chemical 155 (2011) 397–407. [12] L.A. Kunz-Schughart, Multicellular tumor spheroids: intermediates between monolayer culture and in vivo tumor, Cell Biology International 23 (1999) 157–161. [13] J. Friedrich, C. Seidel, R. Ebner, L.A. Kunz-Schughart, Spheroid-based drug screen: considerations and practical approach, Nature Protocols 4 (2009) 309–324. [14] T. Kim, Y.-H. Cho, A pumpless cell culture chip with the constant medium perfusion-rate maintained by balanced droplet dispensing, Lab on a Chip 11 (2011) 1825–1830. [15] J.M. Kelm, N.E. Timmins, C.J. Brown, M. Fussenegger, L.K. Nielsen, Method for generation of homogeneous multicellular tumor spheroids applicable to a wide variety of cell types, Biotechnology and Bioengineering 83 (2003) 173–180.
[16] D. Wlodkowic, S. Faley, J. Skommer, D. McGuinness, J.M. Cooper, Biological implications of polymeric microdevices for live cell assays, Analytical Chemistry 81 (2009) 9828–9833. [17] Y.-C. Tung, A.Y. Hsiao, S.G. Allen, Y.-S. Torisawa, M. Ho, S. Takayama, Highthroughput 3D spheroid culture and drug testing using a 384 hanging drop array, Analyst 136 (2011) 473–478. [18] P.J. O’Dwyer, F. LaCreta, M. Hogan, N. Rosenblum, J.L. O’Dwyer, R.L. Comis, Pharmacologic study of etoposide and cisplatin by the intraperitoneal route, Journal of Clinical Pharmacology 31 (1991) 253–258.
Biographies Taeyoon Kim received a B.S. degree from the School of Mechanical Engineering at Yonsei University in 2004 and a M.S. degree from the Department of Biosystems at the Korea Advanced Institute of Science and Technology (KAIST) in 2006; and a Ph.D. degree from the Department of Bio and Brain Engineering at the KAIST in 2012. In March 2012, Dr. Kim moved to Samsung Electro-Mechanics Co., Ltd., where he is currently a senior engineer. His research interests are focused on micro/nano fabrication, micro sensors and actuators. Il Doh received a B.S. degree summa cum laude from the Department of Mechanical Engineering at Ajou University in 2002 and a M.S. degree from the Department of Mechanical Engineering at the Korea Advanced Institute of Science and Technology (KAIST) in 2003; and a Ph.D. degree from the Department of Bio and Brain Engineering at the Korea Advanced Institute of Science and Technology (KAIST) in 2009. His research interests are focused on biomicrofluidic microsystems for point-of-care and lab-on-a-chip applications. Young-Ho Cho received a B.S. degree summa cum laude from Yeungnam University, Daegu, Korea, in 1980; a M.S. degree from the Korea Advanced Institute of Science and Technology (KAIST), Seoul, Korea, in 1982; and a Ph.D. degree from the University of California at Berkeley for his electrostatic actuator and crab-leg microflexure research completed in December 1990. From 1982 to 1986 he was a Research Scientist of CAD/CAM Research Laboratory, Korea Institute of Science and Technology (KIST), Seoul, Korea. He worked as a post-graduate researcher (1987–1990) and a post-doctoral research associate (1991) of the Berkeley Sensor and Actuator Center (BSAC) at the University of California at Berkeley. In August 1991, Dr. Cho moved to KAIST, where he is currently professor in the Departments of Bio and Brain Engineering & Mechanical Engineering as well as the director of the National Research Convergence Center for Circulating Tumor Cells (CTCRC). In Korea, he has served as the chair of MEMS Division in Korean Society of Mechanical Engineers, the chair of Steering Committee in Korea National MEMS Programs, and the Committee of National Nanotechnology Planning Board. Dr. Cho has also served for international technical society as the general co-chair of IEEE MEMS 2003 Conference, general chair of PowerMEMS 2011 conference, the Program Committee of IEEE Optical MEMS Conference, the chief delegate of the Republic of Korea in World Micromachine Summit. Dr. Cho is a member of IEEE and ASME.
Contents lists available at SciVerse ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
A 3D tumor spheroid chip with the pharmacokinetic drug elimination model developed by balanced droplet dispensing Taeyoon Kim 1 , Il Doh, Young-Ho Cho ∗ Cell Bench Research Center, KAIST, 373-1 Guseong-dong, Yuseong-gu, Daejeon 305-701, Republic of Korea
a r t i c l e
i n f o
Article history: Received 1 March 2012 Received in revised form 9 July 2012 Accepted 21 August 2012 Available online 6 September 2012 Keywords: Pharmacokinetic drug elimination model Balanced droplet dispensing 3D tumor spheroid chip
a b s t r a c t This paper presents a three-dimensional (3D) tumor spheroid chip where the pharmacokinetic (PK) drug elimination model is developed by balanced droplet dispensing. The PK model should reflect the exponential decrease of drug concentration in a body to improve the reliability of cell-based screening assay. The previous cell chips, however, maintain the concentration of anti-cancer agents at constant or discretely decrease the concentration level by manual process. In the present chip where medium droplet dispensers are integrated on a well array, the small droplets of fresh medium and well waste are autonomously replaced each other, the exponential elimination of drug concentration in the well is provided. In the experimental study, we demonstrated that the present chip successfully develops the exponential elimination model using fluorescein isothiocyanate (FITC) and cisplatin. Although the exponential decay of cisplatin in the present chip did not show significant effect biologically due to the current protocol where 3 days of constant concentration pre-treatment is performed, we observed different growth pattern of H358 spheroids depending on peak concentration of 0, 1, 10, and 100 M. The present chip has potential for analyzing the response of 3D tumor spheroids in the pharmacokinetically relevant drug elimination model for high-reliability spheroid-based drug screening chip. © 2012 Elsevier B.V. All rights reserved.
1. Introduction High-reliability anti-cancer drug screening is important to select optimized drug candidates among numerous agents for effective cancer chemotherapy. In screening process, the cellular response to various anti-cancer drugs and its combination is analyzed. In order to improve the reliability of the in vitro cell-based anti-cancer drug screening, the pharmacokinetic (PK) parameters of the drug should be considered so that the screening results can precisely reflect the time course effects of drugs to tumor [1]. The understanding of the PK processes in a body [2] is essential to design the optimized anti-cancer drug regimen such as dosage and medication intervals of drugs for cancer patients. Therefore, the PK model of drugs should be applied in the cell-based drug screening to predict the physiological efficacy to cancer patients. In general PK models of drugs, body consists of several compartments [2] and the drug concentration in plasma or serum is generally regarded as the exponentially decreasing model due to the reaction of a drug within or between the compartments as shown in Fig. 1 (solid line). The previous cell-based drug screening chips [3–11], however, characterized the cellular response to
∗ Corresponding author. Tel.: +82 42 350 8691; fax: +82 42 350 8690. E-mail address: [email protected] (Y.-H. Cho). 1 Process Methodology Development Group, Samsung Electromechanics Co., Ltd. 0925-4005/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.snb.2012.08.041
anti-cancer agents without any chemical dilution of drug. The drug concentration in these chips [3–11] shows the constant value throughout the drug screening process as the dashed line in Fig. 1. In the some other cell-based screening methods [12,13], the drugs were manually diluted by the periodic medium renewal. However, the manual medium renewal process is time-consuming and results in the well contamination and large deviation of measured results caused by human factors. In addition, the medium renewal at the specific interval results in the discrete reduction of drug concentration (dotted line), different from the continuous exponential profile in the PK model (solid line). In this work, we present 3D spheroid chip capable to apply the exponential drug elimination model in Fig. 1 (solid line) based on the balanced droplet dispensing. Since the small droplets of fresh medium and well waste are autonomously replaced each other, the exponential decrease of drug concentration in the well is obtained. This paper focuses on the design and experimental demonstration of the 3D tumor spheroid chip for analyzing the response of spheroids to the anti-cancer drug in the PK model of exponential drug elimination. 2. Theory and design The present chip, illustrated in Fig. 2, consists of two polydimethylsiloxane (PDMS) layers: a well layer and a droplet dispensing layer. Previously, we had developed a simple method capable to supply a fresh medium droplet, named as the balanced
T. Kim et al. / Sensors and Actuators B 174 (2012) 436–440
437
Assuming that the medium droplet supplied from the droplet dispenser layer results in an uniform drug concentration, C, in the well, we derive the first-order differential equation of Q ˙ + C(t) = 0 C(t) V
(1)
where Q and V are the perfusion rate and the volume of drug solution in the well, respectively. By solving the differential equation, the drug concentration, C, is expressed as a exponential function of
Q
C(t) = C0 exp −
Fig. 1. Comparison of pharmacokinetic (PK) drug elimination models in the previous and present chips.
droplet dispensing [14]. In this method, the hydraulic-head difference, h, between the well-inlet and drain-outlet, cause the removal of a waste droplet at the drain. The decreased pressure in the well due to volume expansion provokes a fresh medium droplet supply from the dispensing layer. Therefore, the fresh medium droplet (Fig. 2) is autonomously supplied to the well without any pumps in the exact amount of the removed waste droplet. The well layer has an array (4 × 8) of the well (diameter: 6 mm). All experiments are done in an array format. For the generation of 3D tumor spheroids, we used a hanging drop method as shown in Fig. 3. We seed cells in the well array and place the chip upside down so that the cells aggregate each other due to the gravitational force in the suspended medium at the center of medium curvature by hanging drop method [15–17]. The wells of the present chip are initially supplied with a drug at the specific concentration of C0 . Then the drug solution in the well array is gradually diluted from the initial concentration, C0 , to 0, based on the balanced medium droplet dispensing of Fig. 3.
V
t
(2)
where C0 is the initial drug concentration at the well. As a result, the balanced droplet dispensing is capable to develop the equivalent one-compartmental PK model illustrated in Fig. 3b. In the PK model in Fig. 4b, the body is depicted as a kinetically homogeneous unit with the assumption that the drug achieves instantaneous distribution throughout the body following the drug administration. As a result, the drug concentration to time profile of the well and the one-compartmental PK model shows the identical exponential response of Fig. 1 (solid line). Here, the half-life time, t1/2 , required to reduce the drug concentration to one half its initial value, is derived as t1/2 =
V ln 2 Q
(3)
Therefore, the half-life time, t1/2 , of the anti-cancer drug can be controlled by adjusting the perfusion rate, Q, and the volume, V, of drug solution in the well. 3. Materials and methods 3.1. Fabrication process The droplet dispenser layer was fabricated from the molding and bonding processes of two PDMS plates. We fabricated the
Fig. 2. The present chip utilizing the balanced droplet dispensing for exponential elimination profile of drug concentration.
Fig. 3. Spheroid formation in the well layer using hanging drop method.
438
T. Kim et al. / Sensors and Actuators B 174 (2012) 436–440
4. Experimental results and discussion 4.1. Drug elimination
Fig. 4. The relative fluorescein isothiocyanate (FITC) concentration as a function of the perfusion duration in the present chip; inserted figures show the fluorescence images of the FITC solution during the medium perfusion.
8 mm-thick top plate of the droplet dispenser layer by curing PDMS pre-polymer mixture (curing agent-to-PDMS ratio of 1:10, Sylgard 184, Dow Corning) in an acrylic jig for 12 h at 75 ◦ C. The 3 mmthick bottom plate of the droplet dispenser layer was obtained by curing 24 g PDMS pre-polymer mixture on a 4 in. bare silicon wafer for 2 h at 85 ◦ C. We bonded the top and bottom PDMS plates of the droplet dispenser layer by treating the bonding surfaces with O2 plasma for 30 s. Then we plugged a 2 mm-long polypropylene tip at the center of each well bottom after perforating a 1mm-diameter hole by a puncher. The well layer was fabricated from the similar process for the droplet dispenser layer, excepting that the 4 mm-thick bottom plate is fabricated by curing 30 g PDMS pre-polymer on the drain channel mold for 2 h at 85 ◦ C. The fabricated droplet dispenser layer and well layer were sterilized by an autoclave and dried overnight. The bottom surfaces of the wells were treated with 2 wt.% bovine serum albumin (BSA) solution for 1 h at room temperature to prevent the cell adhesion. Finally, we stacked the PDMS layers and tightly sealed them using an acrylic jig with bolted joint. From the fabricated channel dimension, we obtained the fluidic resistance of the main and branch drain channels as 3.13 × 1011 Pa s/m3 and 2.24 × 1014 Pa s/m3 , respectively, and determined the hydraulichead difference, h, of 38 mm for generating the perfusion rate, Q, of 0.1 l/min. 3.2. Cell preparation In the experimental study, we used H358 human non-small cell lung cancer (NSCLC) cell line which is stably transfected by green fluorescence protein (GFP) to facilitate the cell image analysis. We cultured the H358 cells in complete media consisting of RPMI-1640 (Gibco) supplemented with 10% (v/v) fetal bovine serum (FBS; Gibco), 0.2% (v/v) G418 solution (Sigma Aldrich), and 1% (v/v) penicillin–streptomycin (Invitrogen). The cells were cultured in a humidified 37 ◦ C and 5% CO2 incubator and routinely passaged at 80% confluence. We collected the cells by treating the culture dish with 0.25% trypsin/EDTA (Invitrogen) for 3 min at 37 ◦ C, then neutralized the cell solution with 10% RPMI-1640 medium (Gibco). After spinning down the cells with a centrifuge at 1500 rpm for 3 min, we resuspended the cells in fresh medium. We adjusted the number of cells by measuring the cell concentration with a hemocytometer (Marienfeld), and seeded the cells in the chip by using a micropipette (Eppendorf).
To demonstrate the capability for the exponential drug elimination, we first measured the concentration profile of fluorescein isothiocyanate (FITC) gradually diluted from the balanced medium droplet dispensing. We supplied 140 l of FITC solution in the wells, and perfused the medium at the rate, Q, of 0.1 l/min using the balanced droplet dispensing. Then we collected the FITC solution from the drain outlet and measured the fluorescence intensity at the intervals of 1 day for the total duration of 5 days. Fig. 4 shows the measured relative FITC concentration as a portion of the initial value during the medium perfusion for 5 days. In the graph of Fig. 4, the average and standard deviation of the FITC concentration were obtained from three measurements. Based on Eq. (3), the half-life, t1/2 , of the FITC is calculated as 16 h for the well solution volume, V, of 140 l and the perfusion rate, Q, of 0.1 l/min. In the previous pharmacologic study [18], the half-life of in vivo cisplatin administration was measured as 21.7 ± 14.1 h. Therefore, further modification of the well solution volume, V, and the perfusion rate, Q is required to precisely adjust the present half-life to that of in vivo profile. In the measured FITC dilution graph of Fig. 5, the in vivo cisplatin profile [18] is also illustrated to clarify the difference between the in vivo and experimental profiles. As a result, we observed that the measured FITC concentration profile agrees well with the exponential function of Eq. (3) within the error of less than 14.8%. Therefore, we experimentally verified that the present chip implements the exponential drug elimination in the one-compartmental PK model using the balanced droplet dispensing. 4.2. 3D tumor spheroid growth pattern We formed 376.7 ± 23.4 m-diameter 3D spheroids from 5 × 103 H358 cells for 5 days and supplied the anti-cancer drug, cisplatin, with the four different concentrations of 0 (Control), 1, 10, and 100 M into the wells and incubated the spheroids for 3 days in 37 ◦ C and 5% CO2 . Then we implemented the exponential cisplatin elimination using the medium perfusion at the rate, Q, of 0.1 l/min. Based on the previously established protocols [13] for the drug treatment of tumor spheroids, we incubated the spheroids at the constant cisplatin concentration for 3 days ahead of the exponential drug elimination for comparison. During the exponential cisplatin elimination, we refilled the droplet dispenser layer with fresh medium solution once per 36 h. Fig. 5 shows the time-lapse fluorescent spheroid images captured by an inverted fluorescent microscope. The diameter of the 3D spheroids was measured from the capture images using MATLAB® image processing toolbox. Fig. 6 shows the measured spheroid growth pattern. In the graph of Fig. 6, the average and standard deviation of the spheroid diameter were measured from three different spheroids. We observed that the 3D spheroids steadily grow without the cisplatin treatment (C0 = 0 M). However, the 3D spheroids treated with cisplatin showed different growth pattern. In the cisplatin elimination model with the peak concentration, 1 M, the H358 spheroids initially followed the growth pattern similar to that of control (without cisplatin treatment) but shrank after the 3 days of drug elimination. The cisplatin elimination model with higher concentration of 10 M showed the rapid shrinkage of the spheroid after the drug treatment. At the C0 of 100 M, we could not even measure the spheroid diameter since the spheroids were completely disaggregated directly after the drug administration. Although the exponential decay of cisplatin in the present devices did not show significant effect biologically due to the current protocol where 3 days of constant concentration pre-treatment is performed, these
T. Kim et al. / Sensors and Actuators B 174 (2012) 436–440
439
Fig. 5. Time-lapse microscopic fluorescence images of 3D spheroids, formed by H358 cell number of 5 × 103 , cultured for 15 days in the cisplatin elimination model with the four different peak concentrations.
cells in the 3D spheroid model was much higher than 2D model in all cisplatin concentration. Fig. S1 also showed that some cells in the 3D spheroid model were still alive. This indicated that the drug cannot penetrate into the core of spheroid and cells in this region could survive. PDMS we used for the fabrication of chip is a hydrophobic material with a non-specific absorption characteristic. Hydrophobic drugs may not follow an exponential decay in the present chip for a long time culture due to the drug absorption at the PDMS surface. For the use of various drugs, other materials with good anti-absorption characteristics should be considered. Due to the simplicity of the structure, the micromachining of these materials would not be a big concern. 5. Conclusions
Fig. 6. Diameter of 3D spheroids, formed by H358 cell number of 5 × 103 , cultured for 15 days in the cisplatin elimination model with the three different peak concentrations.
results suggest that the H358 spheroids shows different growth response in the cisplatin elimination model with different peak concentrations. For further biological study on the effect of exponential drug decay, cell responses with and without exponential decay should be examined. For a comparison of 2D monolayer and 3D spheroid models, we measured the cell viability from these models using the identical number (5 × 103 ) of H358 cells treated by cisplatin at varying static concentration for 3 days (Supplementary information, Fig. S1). While the results of Fig. 5 shows the simple growth pattern of 3D spheroids, the cell viability of Fig. S1 measured from the fluorescence staining method (Supplementary information, Fig. S2) showed the different responses between 2D monolayer and 3D spheroid models against the cisplatin treatment. The viability of
We designed, fabricated, and characterized the 3D tumor spheroid chip for the drug–response analysis of spheroids in the PK model of drug elimination developed by balanced droplet dispensing. We experimentally demonstrated the 3D tumor spheroid chip successfully provides the exponential drug elimination model within the error less than 14.8%. In the cisplatin elimination model with the different peak concentrations of 0, 1, 1, 10, and 100 M, we observed different growth patterns for the H358 spheroids. For further biological claims on the effect of exponential drug decay, cell responses with and without exponential decay should be examined. The present chip has potential for analyzing the response of 3D tumor spheroids in the pharmacokinetically relevant drug elimination model for high-reliability spheroid-based drug screening chip. Acknowledgement This research was supported by the Converging Research Center Program funded by the Ministry of Education, Science and Technology (Project No. 2011K000864).
440
T. Kim et al. / Sensors and Actuators B 174 (2012) 436–440
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.snb.2012.08.041. References [1] S. Rockwell, In vivo–in vitro tumour cell lines: characteristics and limitations as models for human cancer, British Journal of Cancer Supplement 4 (1980) 118–122. [2] M.M. Ames, G. Powis, J.S. Kovach, Pharmacokinetics of Anticancer Agents in Humans, Elsevier, 1983. [3] Y. Torisawa, H. Shiku, T. Yasukawa, M. Nishizawa, T. Matsue, Three-dimensional micro-culture system with a silicon-based cell array device for multi-channel drug sensitivity test, Sensors and Actuators B: Chemical 108 (2005) 654–659. [4] J. Pihl, J. Sinclair, E. Sahlin, M. Karlsson, F. Petterson, J. Olofsson, O. Orwar, Microfluidic gradient-generating device for pharmacological profiling, Analytical Chemistry 77 (2005) 3897–3903. [5] P.J. Lee, P.J. Hung, V.M. Rao, L.P. Lee, Nanoliter scale microbioreactor array for quantitative cell biology, Biotechnology and Bioengineering 94 (2006) 5–14. [6] M. Yamada, T. Hirano, M. Yasuda, M. Seki, A microfluidic flow distributor generating stepwise concentrations for high-throughput biochemical processing, Lab on a Chip 6 (2006) 179–184. [7] C.-W. Li, J. Yang, M. Yang, Dose-dependent cell-based assays in V-shaped microfluidic channels, Lab on a Chip 6 (2006) 921–929. [8] Z. Wang, M.-C. Kim, M. Marquez, T. Thorsen, High-density microfluidic arrays for cell cytotoxicity analysis, Lab on a Chip 7 (2007) 740–745. [9] S. Sugiura, J. Edahiro, K. Kikuchi, K. Sumaru, T. Kanamori, Pressure-driven perfusion culture microchamber array for a parallel drug cytotoxicity assay, Biotechnology and Bioengineering 100 (2008) 1156–1165. [10] K. Ziolkowska, E. Jedrych, R. Kwapiszewski, J. Lopacinska, M. Skolimowski, M. Chudy, PDMS/glass microfluidic cell culture system for cytotoxicity tests and cells passage, Sensors and Actuators B: Chemical 145 (2010) 533–542. [11] M.-H. Wu, Y.-H. Chang, Y.-T. Liu, Y.-M. Chen, S.-S. Wang, H.-Y. Wang, C.-S. Lai, T.-M. Pan, Development of high throughput microfluidic cell culture chip for perfusion 3-dimensional cell culture-based chemosensitivity assay, Sensors and Actuators B: Chemical 155 (2011) 397–407. [12] L.A. Kunz-Schughart, Multicellular tumor spheroids: intermediates between monolayer culture and in vivo tumor, Cell Biology International 23 (1999) 157–161. [13] J. Friedrich, C. Seidel, R. Ebner, L.A. Kunz-Schughart, Spheroid-based drug screen: considerations and practical approach, Nature Protocols 4 (2009) 309–324. [14] T. Kim, Y.-H. Cho, A pumpless cell culture chip with the constant medium perfusion-rate maintained by balanced droplet dispensing, Lab on a Chip 11 (2011) 1825–1830. [15] J.M. Kelm, N.E. Timmins, C.J. Brown, M. Fussenegger, L.K. Nielsen, Method for generation of homogeneous multicellular tumor spheroids applicable to a wide variety of cell types, Biotechnology and Bioengineering 83 (2003) 173–180.
[16] D. Wlodkowic, S. Faley, J. Skommer, D. McGuinness, J.M. Cooper, Biological implications of polymeric microdevices for live cell assays, Analytical Chemistry 81 (2009) 9828–9833. [17] Y.-C. Tung, A.Y. Hsiao, S.G. Allen, Y.-S. Torisawa, M. Ho, S. Takayama, Highthroughput 3D spheroid culture and drug testing using a 384 hanging drop array, Analyst 136 (2011) 473–478. [18] P.J. O’Dwyer, F. LaCreta, M. Hogan, N. Rosenblum, J.L. O’Dwyer, R.L. Comis, Pharmacologic study of etoposide and cisplatin by the intraperitoneal route, Journal of Clinical Pharmacology 31 (1991) 253–258.
Biographies Taeyoon Kim received a B.S. degree from the School of Mechanical Engineering at Yonsei University in 2004 and a M.S. degree from the Department of Biosystems at the Korea Advanced Institute of Science and Technology (KAIST) in 2006; and a Ph.D. degree from the Department of Bio and Brain Engineering at the KAIST in 2012. In March 2012, Dr. Kim moved to Samsung Electro-Mechanics Co., Ltd., where he is currently a senior engineer. His research interests are focused on micro/nano fabrication, micro sensors and actuators. Il Doh received a B.S. degree summa cum laude from the Department of Mechanical Engineering at Ajou University in 2002 and a M.S. degree from the Department of Mechanical Engineering at the Korea Advanced Institute of Science and Technology (KAIST) in 2003; and a Ph.D. degree from the Department of Bio and Brain Engineering at the Korea Advanced Institute of Science and Technology (KAIST) in 2009. His research interests are focused on biomicrofluidic microsystems for point-of-care and lab-on-a-chip applications. Young-Ho Cho received a B.S. degree summa cum laude from Yeungnam University, Daegu, Korea, in 1980; a M.S. degree from the Korea Advanced Institute of Science and Technology (KAIST), Seoul, Korea, in 1982; and a Ph.D. degree from the University of California at Berkeley for his electrostatic actuator and crab-leg microflexure research completed in December 1990. From 1982 to 1986 he was a Research Scientist of CAD/CAM Research Laboratory, Korea Institute of Science and Technology (KIST), Seoul, Korea. He worked as a post-graduate researcher (1987–1990) and a post-doctoral research associate (1991) of the Berkeley Sensor and Actuator Center (BSAC) at the University of California at Berkeley. In August 1991, Dr. Cho moved to KAIST, where he is currently professor in the Departments of Bio and Brain Engineering & Mechanical Engineering as well as the director of the National Research Convergence Center for Circulating Tumor Cells (CTCRC). In Korea, he has served as the chair of MEMS Division in Korean Society of Mechanical Engineers, the chair of Steering Committee in Korea National MEMS Programs, and the Committee of National Nanotechnology Planning Board. Dr. Cho has also served for international technical society as the general co-chair of IEEE MEMS 2003 Conference, general chair of PowerMEMS 2011 conference, the Program Committee of IEEE Optical MEMS Conference, the chief delegate of the Republic of Korea in World Micromachine Summit. Dr. Cho is a member of IEEE and ASME.