Micro Total Bioassay System for Ingested Substances: Assessment of

Nov 23, 2010 - E-mail: [email protected]. Fax: +81-3-5841-8027., † .... A multi-throughput multi-organ-on-a-chip system on a plate for...
0 downloads 8 Views 2MB Size
Anal. Chem. 2010, 82, 9983–9988

Micro Total Bioassay System for Ingested Substances: Assessment of Intestinal Absorption, Hepatic Metabolism, and Bioactivity Yuki Imura,† Kiichi Sato,*,†,‡ and Etsuro Yoshimura† Department of Applied Biological Chemistry, School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo, Tokyo 113-8657, Japan, and Center for NanoBio Integration, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-8656, Japan Oral medicines and food constituents are absorbed in the intestine and metabolized in the liver, after which they exhibit their activity toward a target tissue. Micromodels of human tissues were developed to mimic these processes and bioactivities. By integrating the micromodels, we realized a micro total bioassay system for oral substances; this system comprised a microintestine, microliver, and the target components. The microchip was composed of a slide glass and polydimethylsiloxane (PDMS) sheets with microchannels fabricated by photolithography. Caco-2 cells were cultured in the intestine component, and HepG2 cells, in the liver component. The human breast carcinoma MCF-7 cells were cultured in the target component, and the activities of anticancer agents and estrogen-like substances were successfully assayed. By using this system, the overall properties of the ingested cyclophosphamide, epirubicin, 17-β estradiol, and soy isoflavone, i.e., their intestinal absorption, hepatic metabolism, and bioactivity toward target cells, could be assayed with operative ease. Further, the assay time and cell consumption were reduced compared to those in conventional in vitro bioassay systems. A bioassay using cultured cells is one of the most important analytical methods in the search for new drugs, functional food constituents, etc. In conventional bioassays, however, only one bioactivity can be examined. Oral medicines and food constituents show their effects only after intestinal absorption and hepatic metabolism, but no in vitro bioassay system is available that can simultaneously test both bioavailability and bioactivity. Evaluation of intestinal absorption is an important factor in the fields of pharmacology and nutrition. Most research has been conducted using Caco-2 cells as in vitro intestine models.1-3 These cells have tight junctions and form a single layer that allows selective permeation.4 In the field of pharmacokinetics, assessment of * Corresponding author. E-mail: [email protected]. Fax: +813-5841-8027. † Department of Applied Biological Chemistry. ‡ Center for NanoBio Integration. (1) Konishi, Y. Biosci. Biotechnol. Biochem. 2003, 67, 2297–2299. (2) Alsenz, J.; Haenel, E. Pharm. Res. 2003, 20, 1961–1969. (3) Marino, A. M.; Yarde, M.; Patel, H.; Chong, S. H.; Balimane, P. V. Int. J. Pharm. 2005, 297, 235–241. (4) Hilgers, A.; Conradi, R.; Burton, P. Pharm. Res. 1990, 7, 902–910. 10.1021/ac100806x  2010 American Chemical Society Published on Web 11/23/2010

hepatic metabolism is very important. Primary cultured cells,5 liver slices,6,7 and liver carcinoma cell lines (HepG2 cells,8 for example) are used for in vitro experiments. Although scientists have found the existing systems easy to apply, these systems are unnaturally large given the scale of the cells; a large amount of reagents, cells, and media are needed, and testing takes a very long time. 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 this requirement is miniaturization technology. Micro total analysis systems enable highly efficient analyses in parallel.9 Studies have reported some methods in which cells are cultured on a microchip under microflow conditions.10-14 In addition to this technique, other techniques are available in which all processes required for the bioassay can be conducted in a single device.15,16 We previously developed a micro system to evaluate intestinal absorption.17 In this study, microsystems to evaluate hepatic metabolism and bioactivity to breast cancer cells were developed, and we propose a fundamental concept of a micro total bioassay system for orally administered biologically active substances. By integrating three componentssmicrointestine, microliver, and target cellssinto a microchip, we could realize a preliminary micro (5) Li, A. P.; Reith, M. K.; Rasmussen, A.; Girsju, J. C.; Hall, S. D.; Xu, L.; Kaminski, D. L.; Cheng, L. K. Chem.-Biol. Interact. 1997, 107, 17–30. (6) Worboys, P. D.; Bradbury, A.; Houston, J. B. Drug Metab. Dispos. 1995, 23, 393–397. (7) Drahushuk, A. T.; Mcgarrigle, B. P.; Tai, H. L.; Kitareewan, S.; Goldstein, J. A.; Olson, J. R. Toxicol. Appl. Pharmacol. 1996, 140, 393–403. (8) Yoshitomi, S.; Ikemoto, K.; Takahashi, J.; Miki, H.; Namba, M.; Asahi, S. Toxicol. in Vitro 2001, 15, 245–256. (9) Manz, A.; Effenhauser, C. S.; Burggraf, N.; Harrison, D. J.; Seiler, K.; Fluri, K. J. Micromech. Microeng. 1994, 4, 257–265. (10) Takayama, S.; McDonald, J. C.; Ostuni, E.; Liang, M. N.; Kenis, P. J. A.; Ismagilov, R. F.; Whitesides, G. M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 5545–5548. (11) Takayama, S.; Ostuni, E.; LeDuc, P.; Naruse, K.; Ingber, D. E.; Whitesides, G. M. Chem. Biol. 2003, 10, 123–130. (12) Leclerc, E.; Sakai, Y.; Fujii, T. Biomed. Microdevices 2003, 5, 109–114. (13) Davidsson, R.; Boketoft, A.; Bristulf, J.; Kotarsky, K.; Olde, B.; Owman, C.; Bengtsson, M.; Laurell, T.; Emneus, J. Anal. Chem. 2004, 76, 4715–4720. (14) Tanaka, Y.; Sato, K.; Yamato, M.; Okano, T.; Kitamori, T. J. Chromatogr., A 2006, 1111, 233–237. (15) Tokuyama, T.; Fujii, S.; Sato, K.; Abo, M.; Okubo, A. Anal. Chem. 2005, 77, 3309–3314. (16) Goto, M.; Sato, K.; Murakami, A.; Tokeshi, M.; Kitamori, T. Anal. Chem. 2005, 77, 2125–2131. (17) Imura, Y.; Asano, Y.; Sato, K.; Yoshimura, E. Anal. Sci. 2009, 25, 1403– 1407.

Analytical Chemistry, Vol. 82, No. 24, December 15, 2010

9983

Figure 1. Schematic illustration of a microchip-based total bioassay system. The figure is not to scale.

total bioassay and assay the overall features, namely, intestinal absorption, hepatic metabolism, and bioactivity of various chemicals, including oral medicines, food constituents, and environmental endocrine disrupters (Figure 1). EXPERIMENTAL SECTION Reagents and Cells. The human intestinal model cell line Caco-2, human hepatocellular carcinoma cells HepG2, and human breast carcinoma cells MCF-7 were obtained from Riken BioResource Center (Ibaraki, Japan). Minimal essential medium (MEM) and Dulbecco’s modified Eagle’s medium (DMEM) were purchased from Sigma-Aldrich (St. Louis, MO). RPMI 1640 medium, fetal bovine serum (FBS), MEM nonessential amino acid (NEAA) solution (100×), and trypsin-EDTA solution for cell detachment were from Gibco, Invitrogen (Carlsbad, CA). Fibronectin for cell adhesion was from Asahi Glass (Tokyo, Japan). Calcein AM to stain living cells was purchased from Dojindo Laboratories (Kumamoto, Japan). Cyclophosphamide (CPA), epirubicin (EPI), 17-β estradiol (E2), and soy isoflavone (IF) were purchased from Wako Pure Chemicals (Osaka, Japan) as model samples for bioassays. Other reagents were obtained from Wako Pure Chemicals, unless otherwise specified. Microchip and Fluidics. The detailed procedures of microchip fabrication are as described previously.17 The microchip was composed of polydimethylsiloxane (PDMS) and a glass slide.18,19 It comprised two PDMS sheets, an upper and a lower sheet, both of which had microchannels fabricated by photolithography.20 The dimensions of the PDMS sheets were 75 × 25 × 3 mm (L × W × T). All the microchannels were 1.5 mm in width. A glass mold was fabricated with the SU-8 2100 photoresist (MicroChem; Newton, MA) in order to create microchannels with a height of 200 µm. To fabricate a microchamber with a height of 1 mm, a thin Teflon sheet (1-mm thickness) was manually cut to the appropriate size and placed on the glass mold instead of SU-8. Vertical microchannels connecting the upper and lower sides of the lower PDMS sheet were also manually fabricated. Teflon tubes of o.d. 1.6 mm were placed vertically on the glass mold, and PDMS prepolymer was cast onto it to a height of 3 mm. After curing, the Teflon tubes were removed, and vertical channels of i.d. 1.6 mm were obtained. Injection ports were set up in the upper PDMS sheet to enable introduction of the medium and cells into the microchip. The ports were made of Teflon tubes (i.d., 1 mm; o.d., 2 mm; length, 10 mm), which were covered with silicone tubes (i.d., 1 mm; o.d., 3 mm; length, 3 mm) so that they could be fixed (18) McDonald, J. C.; Whitesides, G. M. Acc. Chem. Res. 2002, 35, 491–499. (19) Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Anal. Chem. 1998, 70, 4974–4984. (20) Lorenz, H.; Despont, M.; Fahrni, N.; Brugger, J.; Vettiger, P.; Renaud, P. Sens. Actuators, A 1998, 64, 33–39.

9984

Analytical Chemistry, Vol. 82, No. 24, December 15, 2010

Figure 2. (a) Assembly drawing and (b) photograph of a total bioassay microchip. The figure is not to scale.

in the upper PDMS sheet. The injection port was connected to a fused silica capillary via an air-bubble trap.17 Bottom membranes of BioCoat fibrillar collagen type I, six-well plate 1.0-µm inserts (BD Biosciences; San Jose, CA) were cut to the appropriate size and used for cell culture support to create a microintestine with Caco-2 cells.17 First, the lower PDMS sheet containing microchannels was attached to a glass slide. Next, the collagen-coated membrane was placed on the microchamber in the lower PDMS sheet. The upper PDMS sheet was activated with oxygen plasma and then laminated with the lower PDMS sheet and a glass base (Figure 2). To enable easy handling, the microchip was placed in an autoclaved plastic box, which was kept in a CO2 incubator. All solutions were pumped with a microsyringe pump (KDS230; KD Scientific; Holliston, MA), for which a fused silica capillary was connected to a syringe needle and bonded with an epoxy adhesive. Cell Introduction into the Microchip. The cell suspension was introduced with a Gel-Saver 1-200 µL Gel Loading Tip (USA Scientific; Orlando, FL). The tip was inserted into the injection port, and the suspension was gradually injected. The suspension was allowed to spread in the microchamber. In order to locate cells in a culture chamber appropriately, inlet and outlet ports were set on either side of each chamber. After the cells were introduced into a chamber, all ports other than those at its ends were capped with a silicon tube whose other end was plugged with PDMS. Microintestine. The microintestine component was fabricated and maintained as described previously.17 Caco-2 cells were cultured in a 60-mm cell culture dish. The cells were maintained in MEM supplemented with 10% FBS and 1% NEAA solution and incubated at 37 °C in a humidified atmosphere of 5% CO2. Once the cells reached 70% confluence, they were treated with a trypsin-EDTA solution before fresh medium was added. The

resulting cell suspension was introduced into the microchamber and cultured on the collagen-coated membrane in the microchip. The culture was maintained with DMEM supplemented with 10% FBS and 1% NEAA solution at 37 °C in a humidified atmosphere of 5% CO2. The medium was periodically replaced. Microliver. HepG2 cells were cultured in a 60-mm cell culture dish. They were maintained in DMEM supplemented with 10% FBS and grown at 37 °C in a humidified atmosphere of 5% CO2. When they reached 70% confluence, they were trypsinized, and the resulting suspension was used for cell seeding. For cell culturing on the glass surface in the microchip, the suspension was introduced into a microchamber whose surface was pretreated with fibronectin solution. To create a beads-based microliver, the cells were cultured on Cytodex-3 microcarriers (GE Healthcare; Buckinghamshire, UK).21 The dry spheres were hydrated overnight (1 g of Cytodex with 50 mL of PBS) and then washed and sterilized by autoclaving. Before seeding, the sterilized microcarriers were washed twice with DMEM supplemented with 10% FBS, and 300 µL of swollen Cytodex per milliliter of DMEM was prepared. Cytodex (100 mg) in 5 mL of DMEM was added to a 60-mm nontreated dish, into which 1 × 105 HepG2 cells were seeded. The cells were grown at 37 °C in a humidified atmosphere of 5% CO2. After 5 d, the microcarrier beads with the cultured cells were packed in the microchannels. Target Cells. Human breast carcinoma, MCF-7,22 was selected as a target. The cells were cultured in a 60-mm cell culture dish in RPMI 1640 medium supplemented with 10% FBS at 37 °C in a humidified atmosphere of 5% CO2. After 70% confluence was reached, they were recovered with trypsin-EDTA, and fresh medium was added. Then, the cell suspension was introduced into the microchip and cultured on the fibronectin-coated glass surface of the microchannel overnight. The cells were cultured in DMEM supplemented with 10% FBS at 37 °C in a humidified atmosphere of 5% CO2. Cell Viability Test. As an indicator of cell growth and viability, resazurin-converting activity was determined using the CellTiterBlue cell viability assay reagent (Promega; Madison, WI). The reagent diluted with the medium (20% (v/v)) was applied to the cells at a flow rate of 0.4 µL/min. The resulting solution was collected, and the fluorescence intensity of the reaction product resorufin was monitored (ex: 570 nm, em: 585 nm).23 The viability of MCF-7 cells cultured in the microchip for 1 d was first measured, and the cells were then cultured for 2 d with the sample compound for bioassay dissolved in the medium. The viability of the treated MCF-7 cells was again measured, and the bioactivity of the sample was estimated from the ratio of the obtained value to the original activity. RESULTS AND DISCUSSION Microintestine. Caco-2 cells were cultured on a collagencoated membrane in the microchamber. The microchamber in which the Caco-2 cells were cultured was of dimensions 4 × 1.5 × 1 mm (L × W × D). A confluent Caco-2 cell sheet was obtained (21) Visvikis, A.; Goergen, J. L.; Oster, T.; Bagrel, D.; Wellman, M.; Marc, A.; Engasser, J. M.; Siest, G. Clin. Chim. Acta 1990, 191, 221–232. (22) Ciocca, D. R.; Fuqua, S. A. W.; Locklim, S.; Toft, D. O.; Welch, W. J.; McGuire, W. L. Cancer Res. 1992, 52, 3648–3654. (23) O’Brien, J.; Wilson, I.; Orton, T.; Pognan, F. Eur. J. Biochem. 2000, 267, 5421–5426.

Figure 3. (a-c) Cross-sectional illustration of microlivers. (a) Microliver cultured on a glass surface modified with fibronectin solution. Beads-based microliver packed in (b) a horizontal chamber and (c) a vertical chamber. Inlet ports were closed with silicone caps after cell introduction. Figure is not to scale. (d) Time course of the viability of the microlivers measured with a resazurin reagent. The viability was evaluated from the fluorescence intensity of the reaction product, resorufin, measured at 570 nm (ex.) and 585 nm (em.) (n ) 3).

after 3-d incubation. The optimal concentration of the seeded cell suspension was 5 × 106 cells/mL. The medium was exchanged every second day. A permeation test was conducted using Lucifer yellow (LY), which is known to be unable to permeate the intestinal epithelial cell layer, and CPA, which can permeate it well. LY could not pass through the cell layer in the microchip, and the permeation rate of CPA was the same as that reported previously.17 The results supported the good selective barrier function of the microintestine. Microliver. Two types of microlivers were developed: an onglass microliver and a beads-based microliver (Figure 3). The chamber for on-glass cell culture had dimensions of 4 × 1.5 × 0.2 mm (L × W × D). Typically, there were 2 × 104 cells in the chamber. For the beads-based microliver, microcarrier beads were packed in a horizontal microchamber (4 × 1.5 × 0.2 mm; L × W × D) or vertical microchamber (1.6-mm or 3.2-mm diameter, 3-mm length). On average, each bead had 1.5 × 102 cells on its surface. The number of cells cultured in the horizontal beads-based microliver was calculated at 1.5 × 104. The vertical microlivers of diameters 1.6 and 3.2 mm contained 2 × 105 cells and 9 × 105 cells, respectively. The cell viability test was conducted with these four kinds of microlivers to evaluate the viability of the cells in the livers (Figure Analytical Chemistry, Vol. 82, No. 24, December 15, 2010

9985

Figure 4. (a) Cross-sectional illustration of microliver and target cell components, and (b) viability changes of the target MCF-7 cells treated with cyclophosphamide (CPA) metabolized by various microlivers. CPA in medium (2 mM) was introduced at a flow rate of 0.4 µL/min into the microchip. The bioactivity was estimated from the ratio of the viability of MCF-7 cells incubated with CPA for 2 d to the viability before treatment (n ) 3).

3d). The viability of the HepG2 cells on the glass surface was relatively high at first, but it decreased within 24 h. This rapid decrease resulted from cell detachment, because attachment of HepG2 cells on the fibronectin-coated glass surface was not strong. Low viability of the horizontal beads-based microliver resulted from cell damage during bead introduction into the horizontal channel. On the other hand, the decrease in the viability of the cells on the microcarrier packed in the horizontal microchannels was not significant, and the cells showed relatively high and constant metabolic activity. The difference in the cell viability resulted from a difference in the volume of the microchambers and the ununiformity of the bead packing. Medium supply to the tightly packed area was limited, and on the contrary, fast flow of the medium brought down high shear stress to kill the cells. The difference in the cell viability change resulted from these factors. Under the examined conditions, the vertical microliver of diameter 3.2 mm showed the most stable liver activity. Target Cells. MCF-7 cells grew well on the fibronectin-treated glass surface in the microchamber (4 × 1.5 × 0.2 mm; L × W × D). The optimum cell suspension density for cell seeding was 5 × 105 cells/mL, and typically, 2 × 103 cells were grown in the chamber. The bioactivity of the sample compound was evaluated by the change in MCF-7 cell viability. Liver-Target Microchip. A microchip composed of the microliver and the target breast cancer cells was developed (Figure 4a). The microchambers for the HepG2 and MCF-7 cells were connected with a microchannel. The HepG2 cells and MCF-7 cells for on-glass culture were seeded a day before the bioassay, and HepG2 cells cultured on beads were packed in the chip just before the bioassay. CPA dissolved in the medium (2 mM) was metabo9986

Analytical Chemistry, Vol. 82, No. 24, December 15, 2010

lized by the HepG2 cells and then sent to the MCF-7 cells. CPA is a well-known prodrug-type anticancer agent, which kills breast cancer cells after hepatic metabolism. In the control experiments, i.e., without CPA and HepG2, the MCF-7 cell viability increased by 37% after 2-d culture because of proliferation (Figure 4b). The MCF-7 cell viability increased by the same level after 2-d culture with HepG2 cells, and effects of the coculture with HepG2 cells were not observed. Because the medium flow rate was high, loss of nutrients in the medium was thought to be low. MCF-7 cells cultured in the medium containing HepG2metabolized CPA showed weaker viability. In particular, the beadsbased microliver in the vertical microchamber (3.2-mm diameter) strongly affected CPA activation. From these results, we concluded that the 3.2-mm vertical microchamber was most suitable as a microliver, because its high metabolic activity enabled effective determination of bioactivity of the sample compound. Micro Total Bioassay System. The micro total bioassay chip has two microchannels (Figure 5a). The upper microchannel mimics a human intestinal lumen, and the lower one mimics a blood vessel. Caco-2 cells were introduced into the microchip 3 d before the bioassay. MCF-7 cells were seeded 1 d before the assay. HepG2 cells were cultured off-chip on Cytodex beads, which were packed in the chip just before the bioassay (Figure 5). Before the bioassay, the viability of the MCF-7 cells was measured with the resazurin reagent. DMEM with a sample was introduced into the upper channel (apical side) and normal DMEM flowed through the lower one (basal side), both at 0.4 µL/min. Using this system, anticancer agents and estrogen-like compounds were assayed as model samples. Whereas CPA is orally administered, EPI24 is an anticancer agent taken intravenously. E2 is a major estrogen compound in humans and is known as an endocrine disrupter. IF is naturally present in soybeans and exhibits estrogen-like activity. These estrogen-like compounds are known to stimulate the proliferation of breast cancer cells.25-27 In the current bioassay, 2 mM CPA, 10 µM EPI, 2 nM E2, or 3 µM IF was used as the sample, which was dissolved in the medium and applied to the upper channel of the microintestine at a flow rate of 0.4 µL/min for 2 d. The concentrations of the sample compounds were decided based on the results of the conventional bioassay using MCF-7 cells cultured in a microplate. If a sample could permeate the Caco-2 layer, it entered the lower channel; this implies that it can be absorbed into the human body. If the sample could not permeate the Caco-2 layer, it would be excreted. From the microintestine, the absorbed sample was brought to the microliver containing Cytodex beads with HepG2 cells. Finally, the sample reached the target MCF-7 cells and exerted its bioactivity on them. After 2-d culture, the total bioactivity was evaluated by measuring the viability of the target cells. After the bioassay, LY permeability was checked to confirm that the Caco-2 cell layer was not damaged by the sample. We (24) Yamamoto, D.; Tanaka, K.; Nakai, K.; Baden, T.; Inoue, K.; Yamamoto, C.; Takemoto, H.; Kamato, K.; Hirata, H.; Morikawa, S.; Inubushi, T.; Hioki, K. Breast Cancer Res. Treat. 2002, 72, 1–10. (25) Green, S.; Walter, P.; Kumar, V.; Krust, A.; Bornert, J. M.; Argos, P.; Chambon, P. Nature 1986, 320, 134–139. (26) Morito, K.; Hirose, T.; Kinjo, J.; Hirakawa, T.; Okawa, M.; Nohara, T.; Ogawa, S.; Inoue, S.; Muramatsu, M.; Masamune, Y. Biol. Pharm. Bull. 2001, 24, 351–356. (27) Hua, P.; Tsai, W.; Kuo, S. M. Biochim. Biophys. Acta 2003, 1627, 63–70.

Figure 5. (a) Cross-sectional illustration of the micro total bioassay chip. Samples exhibit their bioactivity in the target cell component (breast cancer cells, in this case) after intestinal absorption and hepatic metabolism. Each solution was introduced with a microsyringe pump at a flow rate of 0.4 µL/min. No figure is to scale. (b) Fluorescent micrograph of Caco-2 cells cultured on a membrane support in a microchamber. Living cells were fluorescently stained with calcein-AM. (c) Micrograph of a Cytodex bead with HepG2 cells on its surface. (d) Micrograph of MCF-7 cells cultured on the fibronectin-modified bottom surface of the microchamber.

found that the layer had retained good barrier function. In order to determine the properties of the sample compound, experiments without Caco-2 or HepG2 cells were also conducted. When Caco-2 cells are absent, the sample freely passes through the membrane and enters the lower channel by diffusion. Without HepG2 cells, the sample is not metabolized. Figure 6 shows the results of the micro total bioassay. In the control experiments without samples, the viability of MCF-7 cells increased by 23% after 2-d culture, because of cell proliferation. In the test experiments, EPI exhibited high anticancer activity in the absence of Caco-2 cells, which implied that it could not cross the intestinal cell layer. Because HepG2 cells seemed to have little effect on bioactivity, we concluded that hepatic metabolism might be less important for EPI. CPA showed high activity in the presence of HepG2 cells, implying that it was activated by hepatic metabolism. On the other hand, Caco-2 cells seemed to have little effect on the CPA activity, and the drug could easily cross the microintestine layer. E2 and IF stimulated MCF-7 cell proliferation. Their activity was higher in the absence of HepG2 cells. These estrogen-like compounds might be metabolized in the microliver, where their activity might be weakened. Further, Caco-2 cells seemed to have very little influence on E2 and IF, and like CPA, they could easily cross the microintestine layer. Most results of the micro total bioassay are consistent with the properties of the sample compounds. EPI, which is usually used intravenously, is known not to be able to cross the intestinal wall. The results obtained correspond to the properties of EPI. The microliver also worked well, judging from the bioassay results for CPA and the estrogen-like compounds. CPA is a typical

Figure 6. Viability changes of the target MCF-7 cells treated with (a) anticancer agents and (b) estrogen-like compounds measured by the micro total bioassay system. At a flow rate of 0.4 µL/min, samples were introduced into the microchip that had a microintestine, microliver, and target cell component, as shown in Figure 5. The bioactivity was estimated from the ratio of the viability of MCF-7 cells incubated with the samples for 2 d to the viability before treatment (n ) 3). The samples assayed were 2 mM cyclophosphamide (CPA), 10 µM epirubicin (EPI), 2 nM 17-β estradiol (E2), or 3 µM soy isoflavone (IF) dissolved in medium. Analytical Chemistry, Vol. 82, No. 24, December 15, 2010

9987

prodrug, showing strong activity only after hepatic metabolism.28 E2 and IF are known to be metabolized in the liver29 and cleared from the body gradually. As described above, this micro total bioassay system correctly evaluated the permeation and metabolism of these known substances. The micro total bioassay system described here can control successive processes with the simple operation of a microsyringe pump. This enabled not only the elimination of troublesome operations like pipetting, but also facilitated the assay of a small amount of sample metabolites with a short half-life, because the metabolites were immediately transferred from the microliver to the target cells. For example, CPA metabolites are known to have a short half-life, and their activity is difficult to assay,30 but the assay was successful in this study. Moreover, this system reduced the use of cells, samples, and reagents; the use of Caco-2 and MCF-7 cells was reduced to 20%, and that of samples and most reagents was reduced to 10% compared with the conventional in vitro experiments. However, the bioassay time required for the target cells was not reduced, because the cell response to chemical substances is a biological process and cannot be controlled by the device. However, all processes were performed in parallel in the micro total bioassay system. Therefore, the total assay time could be reduced from the 3 d required for the conventional in vitro assay to 2 d. The bioassay produced qualitative, rational results that were consistent with the known properties of the samples. However, (28) Chang, T. K. H.; Weber, G. F.; Crespi, C. L.; Waxman, D. J. Cancer Res. 1993, 53, 5629–5637. (29) Lee, A. J.; Cai, M. X. X.; Thomas, P. E.; Conney, H.; Zhu, B. T. Endocrinology 2003, 144, 3382–3398. (30) Graham, M. I.; Shaw, I. C.; Souhami, R. L.; Sidau, B.; Harper, P. G.; McLean, A. E. M. Cancer Chemother. Pharmacol. 1983, 10, 192–193.

9988

Analytical Chemistry, Vol. 82, No. 24, December 15, 2010

quantitative evaluation of bioactivity seems difficult. Precise control of the bioactivity of the micro-organs is difficult, and model cells have different activities from those in an actual human body. It is important to optimize the scale of micro-organs and the microflow in the chip for each biological process, especially in the case of the microliver. Alternatively, it might be better to use more sophisticated liver models, primary cultures, or liver slices to improve the microsystem. CONCLUSIONS A micro total bioassay system was developed to evaluate the intestinal absorption, hepatic metabolism, and antitarget cell bioactivity of orally ingested drugs, foods, and chemicals. We concluded that the developed system could be used to efficiently evaluate the overall properties of oral medicines and food constituents. ACKNOWLEDGMENT This work was partially supported by the Food Nanotechnology Project, The Ministry of Agriculture, Forestry and Fisheries, Japan, and a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan. NOTE ADDED AFTER ASAP PUBLICATION This paper was published ASAP on November 23, 2010 with errors in the caption of Figure 3. The corrected version was reposted on November 29, 2010. Received for review March 29, 2010. Accepted November 3, 2010. AC100806X