Biomimetic Microfluidic Device for in Vitro Antihypertensive Drug

Mar 27, 2014 - In this work, we developed a facile method for evaluating antihypertensive drugs using a microfluidic chip. This microfluidic chip was ...
1 downloads 0 Views 5MB Size
Article pubs.acs.org/molecularpharmaceutics

Biomimetic Microfluidic Device for in Vitro Antihypertensive Drug Evaluation Lei Li,† Xiaoqing Lv,‡ Serge Ostrovidov,§ Xuetao Shi,*,§ Ning Zhang,*,‡ and Jing Liu† †

Key Laboratory of Cryogenics, Technical Institute of Physics and Chemistry, Chinese Academy of Sciences & Beijing Key Laboratory of Cryo-Biomedical Engineering, Beijing100190, China ‡ Research Center of Basic Medic Science, Tianjin Medical University, Tianjin 300070, China § WPI Advanced Institute for Materials Research, Tohoku University, 2-1-1 Katahira, Aoba-ku, Sendai 980-8577, Japan S Supporting Information *

ABSTRACT: Microfluidic devices have emerged as revolutionary, novel platforms for in vitro drug evaluation. In this work, we developed a facile method for evaluating antihypertensive drugs using a microfluidic chip. This microfluidic chip was generated using the elastic material poly(dimethylsiloxane) (PDMS) and a microchannel structure that simulated a blood vessel as fabricated on the chip. We then cultured human umbilical vein endothelial cells (HUVECs) inside the channel. Different pressures and shear stresses could be applied on the cells. The generated vessel mimics can be used for evaluating the safety and effects of antihypertensive drugs. Here, we used hydralazine hydrochloride as a model drug. The results indicated that hydralazine hydrochloride effectively decreased the pressure-induced dysfunction of endothelial cells. This work demonstrates that our microfluidic system provides a convenient and cost-effective platform for studying cellular responses to drugs under mechanical pressure. KEYWORDS: microfluidic chip, antihypertensive drug evaluation, endothelial cells, mechanical pressure



INTRODUCTION High blood pressure (hypertension) is a major risk factor for cardiac and vascular-related diseases, including stroke, heart failure, and chronic kidney disease.1−3 Approximately 26% of the adult population throughout the world suffers from primary (with no obvious medical cause) or secondary (caused by other conditions) hypertension.4 Hypertension has become one of the most serious health problems in the world due to the prevalence of obesity and unhealthy dietary habits and lifestyle.5 Lifestyle changes, including dietary changes and weight loss, are the most highly recommended routes to lower blood pressure in people with prehypertension or high-normal blood pressure.6 However, treatment with antihypertensive drugs is necessary for patients who suffer from serious hypertension.7 In particular, a reduction in blood pressure by 5 mmHg will significantly decrease the risk of cardiac diseases, such as stroke (34% decrease).8 Endothelial cells (ECs), which directly contact blood or lymph, line the entire vascular system, from the capillaries to the heart.9 These cells play a variety of central and critical roles in harnessing vascular functions.10−12 Endothelial dysfunction is a fundamentally important marker in hypertensive patients and may serve as a prognostic sign of cardiovascular diseases.13 In serious pulmonary hypertension, endothelial cells exhibit disordered proliferation, along with concurrent neoangiogenesis, and then form glomeruloid structures.14 Disordered © 2014 American Chemical Society

endothelial cells also exhibit altered production of endothelial vasoactive mediators (e.g., NO, endothelin-1, and thromboxane).15 Therefore, evaluating the effects of antihypertensive drugs on the functions of endothelial cells would have great significance for the treatment of hypertension.16−18 Microfluidic devices have emerged as a revolutionary, novel platform for a range of applications in electronics and basic biological studies.19−21 Advances in microfluidic techniques have enabled the construction of biomimetic tissue/organ replacements by replicating the in vivo physiological microenvironment.22 These biomimetic tissue/organ platforms can be used for drug scanning, toxicity testing, and drug evaluation.23−25 Efforts are in progress to construct different platforms that simulate various tissues/organs, including bone, vessels, the kidney, the lung, and the heart.26−28 Several reports have described the development of artificial vessels using microfluidic devices or microfabrication from hydrogels.29,30 However, these platforms are typically used for tissue engineering purposes. Moreover, for the evaluation of Special Issue: Engineered Biomimetic Tissue Platforms for in Vitro Drug Evaluation Received: Revised: Accepted: Published: 2009

January 18, 2014 March 25, 2014 March 27, 2014 March 27, 2014 dx.doi.org/10.1021/mp5000532 | Mol. Pharmaceutics 2014, 11, 2009−2015

Molecular Pharmaceutics

Article

Figure 1. Schematic of the experimental process. (a) HUVECs were cultured in a microchannel. The dimensions of the straight channel were 4 cm long, 2 mm wide, and 100 μm high. (b) The syringe pump and manometer were connected to the system, and medium (with different concentrations of drug) was injected into the channel. The components of the system are shown in Figure S1 (see Supporting Information). (c) The media in the channel and tubes were collected for the ET-1 assay. (d) After 16 h, the media in the channel and tubes were collected for the ET-1 assay. (e) The cells were directly stained in the channel. The objective magnifications of the cell images are (a) 10× and (e) 40×.



MATERIALS AND METHODS Materials. Dulbecco’s phosphate-buffered saline (DPBS) was purchased from Gibco (US). Human fibronectin was purchased from Millipore (US). HUVECs, endothelial cell medium (ECM), trypsin/EDTA solution, and trypsin neutralization solution were purchased from ScienCell Research Laboratories (US). Hydralazine hydrochloride (CAS: 304-201), 3.7% paraformaldehyde solution, Triton X-100, and bovine serum albumin (BSA) were purchased from Sigma-Aldrich (US). Anti-CD62P (AK4) antibody was purchased from Abcam (US). Alexa Fluor 546 goat antimouse IgG antibody, Alexa Fluor 488 phalloidin, and 4′,6-diamidino-2-phenylindole (DAPI) were purchased from Invitrogen (US). A human ET-1 immunoassay kit (QET00B) was purchased from R&D Systems (US). A photoresist (SU-8 2075) and developer were purchased from MicroChem (US). Poly(dimethylsiloxane) (PDMS) was purchased as a Sylgard 184 kit from Dow Corning (US). Fabrication of the Microfluidic Chip. A straight channel with a length of 4 cm, a width of 2 mm, and a height of 100 μm was used in this work. The microchannel structure was fabricated in PDMS using a standard soft photolithography technique.35−39 Briefly, the desired channel pattern was drawn using L-edit software (Tanner Research, Inc.) and printed on a

antihypertensive drugs, an elastic material-based platform would be preferable. In this study, we used an easily fabricated and assembled microfluidic system to study the response of endothelial cells to different concentrations of an antihypertensive drug while under pressure. The system includes a syringe pump, a microfluidic chip, a digital manometer, and several joints and tubes. Human umbilical vein endothelial cells (HUVECs) were first cultured on the chip. Hydralazine hydrochloride was then added to the medium at one of four different concentrations (0, 50, 250, or 500 μmol/L) and infused into the channel using the syringe pump. Next, atmospheric pressure (control), 12 kPa (medium pressure), or 18 kPa (high pressure) was applied to the cells in the channel for 6 h. Hydralazine hydrochloride, which was used as a model drug, is a smooth-muscle relaxant used for the treatment of hypertension.31 Of note, hydralazine hydrochloride has been reported to have significant effects on preventing endothelial dysfunction (a common feature of essential hypertension).32−34 After the pressure and drugloading experiments, the ECs were stained for P-selectin and filamentous (F-)actin. To further evaluate the injury of the ECs in the channel under different experimental conditions, the cells were assayed for the release of endothelin-1 (ET-1). 2010

dx.doi.org/10.1021/mp5000532 | Mol. Pharmaceutics 2014, 11, 2009−2015

Molecular Pharmaceutics

Article

collected and adjusted to 5 mL. The media were centrifuged, and the supernatants were dispensed into Eppendorf tubes and stored at −20 °C until the ET-1 assay (Figure 1c). The cells were cultured for an additional 16 h in medium with the same concentration of hydralazine hydrochloride, and the medium in the channels was then collected to measure ET-1 release after the pressure experiments (Figure 1d). The cells were subsequently used for staining (Figure 1e). Control chips (no pressure) were also placed in the incubator in culture medium containing hydralazine hydrochloride (0, 50, 250, or 500 μmol/ L). We also performed a parallel experiment. Four chips were connected in series in normal ECM and maintained under 12 or 18 kPa of pressure for 6 h. After the pressure experiments, the ECM was replaced with medium containing one of four concentrations of hydralazine hydrochloride (0, 50, 250, or 500 μmol/L). The chips were then placed in an incubator for 16 h, after which the medium in the channels was collected for the ET-1 assay. Cell Staining. The HUVECs in the channels were stained for P-selectin, actin filaments, and nuclei using anti-CD62P antibody (primary antibody) and Alexa Fluor 546 goat antimouse IgG antibody (secondary antibody), Alexa Fluor 488 phalloidin, and DAPI, respectively. The anti-CD62P antibody, Alexa Fluor 546 goat antimouse IgG antibody, and Alexa Fluor 488 phalloidin were prepared at a concentration of 10 μg/mL in phosphate-buffered saline (PBS) solution. The DAPI was diluted to a concentration of 0.2 μg/mL in PBS solution. All of the solutions were infused into the channels using 1 mL syringes. The cells were then gently washed twice with PBS, fixed with 3.7% paraformaldehyde solution in PBS for 10 min at RT, and washed with PBS. After the cells were permeabilized with 0.1% Triton X-100 in PBS solution, 100 μL of anti-CD62P antibody solution containing 1% BSA was added to the fixed cells and incubated at 37 °C for 1 h. After washing with PBS three times, 100 μL of Alexa Fluor 546 goat antimouse IgG antibody solution containing 1% BSA was added to the fixed cells and incubated at RT for 1 h, followed by washing with PBS three times. The cells were also stained for F-actin with 100 μL of Alexa Fluor 488 phalloidin solution for 20 min at RT, after which 100 μL of DAPI solution was infused into the channels for 2 min at RT. The fixed cells in the channels were washed three times. Fluorescence images were recorded using a microscope (Ti-s, Nikon, Japan) with a CCD camera (Ds-Ri1, Nikon, Japan). ET-1 Assay. The ET-1 level in each sample was measured using a commercial enzyme immunoassay technique-based kit. The intensity of light emitted by the 96 wells was measured by a microplate reader (GENios Plus, Tecan) and expressed in relative light units (RLUs). The ET-1 concentration was determined from a standard curve that was generated using a logistic fit of the concentrations and RLUs of the standards. Statistical Analysis. The data are expressed as means ± standard deviations (more than three replicates were conducted in each experiment). Student’s t tests and analysis of variance (ANOVA) followed by Tukey’s multiple comparison tests were performed to analyze the differences between two experimental groups and among more than two experimental groups, respectively.

transparency. The transparency then served as the photomask in the contact photolithography. A three-inch silicon wafer was used as the substrate and was spin-coated with the negative photoresist SU-8. After prebaking, exposure to UV light, postbaking, and developing, the negative replicas of the designed channel structure were created on the silicon wafer. The silicone elastomer PDMS (Sylgard 184) contains a polymer base and curing agent that were then mixed at a ratio of 10:1 (weight) and poured onto the wafer and cured in a drying oven (80 °C) for 60 min. This PDMS layer was separated from the wafer and cut into small pieces encompassing the entire channel structure. Holes were drilled at the inlet and outlet locations with a blunted and beveled syringe needle. Finally, oxygen plasma treatment was performed using a plasma cleaner (PDC-32G, Harrick Scientific Products, Inc.) to bond the PDMS layer to a clean glass slide, resulting in a complete microfluidic chip. Cell Culture. Cells were cultured on the microfluidic chip. Prior to cell seeding, the chips, tubes, and joints were autoclaved for 20 min at 121 °C. Additionally, a fibronectin solution at a concentration of 50 μg/mL was injected into the channels to promote cell adhesion. The chips were then incubated for at least 1 h in a 37 °C incubator, after which we removed the fibronectin solution and dried the chips in a clean hood at room temperature (RT). Next, HUVECs cultured in flasks (Corning) were removed with trypsin/EDTA solution when approximately 80% were confluent and then resuspended in medium to a density of 2.0 × 106 cells/mL. Thereafter, the cell suspension (10 μL) was carefully injected into the channels, and the chips were incubated at 37 °C with 5% CO2. After 3−4 h, most of the cells had adhered to the glass surface. The medium was then replaced every 8−10 h until the pressure experiments. The cells were used between passages 5 and 7. Cell numbers in the channels were calculated from the number of cell nuclei in the nuclear staining images. Five sets of randomly captured images in each experimental condition were used. Pressure Experiment Procedure. The system included a syringe pump (KDS100, KD Scientific), a syringe, a microfluidic chip, a digital manometer (M382, As-one), and joints and tubes (1 × 2 mm and 3 × 5 mm). A photograph of these components is shown in Figure S1 (see Supporting Information). The syringe pump was used for medium perfusion and pressure generation, and the digital manometer was used as a pressure monitor. A schematic of the pressure experiments is shown in Figure 1. The cells in the channels were ready for the pressure experiments when they attained 70−80% confluence (Figure 1a). First, the medium in the channels was replaced. Then, 6 mL of medium containing one of four different concentrations of hydralazine hydrochloride (0, 50, 250, or 500 μmol/L) was loaded into a 10 mL syringe. Next, the syringe was connected to the inlet of the microfluidic chip with the 1 × 2 mm tube, and the probe of the manometer was connected to the outlet of the chip with the 3 × 5 mm tube. The syringe was then mounted on the syringe pump. The microfluidic chip was placed in an incubator at 37 °C with 5% CO2, and the pump and the manometer were left outside the incubator. We turned on the pump and set a high flow rate (at approximately 30−50 μL/min) to let the pressure of the system reach 12 or 18 kPa in 20−30 min. The pump was then adjusted to a lower flow rate (5−10 μL/min) to stabilize the pressure at 12 ± 0.5 or 18 ± 0.5 kPa for 6 h. After 6 h, the pump was stopped, and the media in the channels and the tubes were



RESULTS AND DISCUSSION Construction of the Device. We used an easily assembled microfluidic system to study the responses of HUVECs to the 2011

dx.doi.org/10.1021/mp5000532 | Mol. Pharmaceutics 2014, 11, 2009−2015

Molecular Pharmaceutics

Article

Figure 2. F-actin and nuclear staining of the HUVECs in the channels: (a−d) 0 kPa; (e−h) 12 kPa; (i−l) 18 kPa; (a,e,i) 0 μmol/L; (b,f,j) 50 μmol/ L; (c,g,k) 250 μmol/L; (d,h,l) 500 μmol/L; (m,n) cell number and cell area of HUVECs in the channels. The objective magnification was 40×. * indicates statistical significance when compared with cells treated with 50, 250, and 500 μmol/L hydralazine hydrochloride at 95% confidence level.

Effects of Pressure and the Drug on the Cytoskeleton and the Expression of P-Selectin. We studied the effects of pressure and different drug concentrations on the F-actin cytoskeleton and on the expression of P-selectin by comparing the fluorescence images of the cells in the microchannels. After the pressure experiments, the cells were cultured for an additional 16 h, and the F-actin, P-selectin, and nuclei of the HUVECs were then stained and imaged using fluorescence microscopy. The F-actin and nuclear fluorescence images of the cells under each experimental condition, obtained at 40× magnification, were overlaid, and these images are shown in Figure 2. Under the condition of 0 kPa (atmospheric pressure) and no drug, the cells were randomly oriented, and several actin filaments appeared within the cytoplasm (Figures 2a and S2, Supporting Information). When the pressure was increased to 12 kPa (medium pressure), the actin filaments became apparent, and the number of actin filaments significantly increased (Figure 2e). When the pressure was further increased to 18 kPa (high pressure), most of the cytoskeletons collapsed (Figure 2i). When an increased concentration of hydralazine hydrochloride was added to the medium, a more prominent network of actin filaments occupied the cytoplasm. Generally, endothelial cells are structurally adhered together by occludin, adherens junctions, and F-actin.42 The F-actin cytoskeleton provides and maintains an integral cellular framework and is involved in changes in cellular shape.43 In this study, pressure had a strong influence on the structure of F-actin in the cell. In particular, after hydralazine hydrochloride was added, the Factin in the cells that experienced the high pressure rearranged into an organized network. Cell numbers in the channels were also calculated from the number of cell nuclei in the nuclear staining images. The results showed that cell number decreased while drug concentrations of the medium increased at the same

antihypertensive drug hydralazine hydrochloride. This system can be considered as a biomimetic model of blood vessel blockage because we used the probe of the manometer to clog the outlet of the channel. Two typical pressures (12 and 18 kPa) were applied in this study because hypertension was defined as a blood pressure persistently at or above 140/90 mmHg (18.6/12.0 kPa).40 The hydraulic diameter, d, of the microchannel was approximately 190 μm (d = 4A/L, where A is the cross-sectional area and L is the perimeter), which is within the size range of human blood vessels. Moreover, it is easy to fabricate microchannels of different dimensions to mimic different vessel sizes, which vary from a diameter of approximately 25 mm for the aorta to only 8 μm for capillaries.33 During the 6 h pressure experiment, the cells cultured in the channel were exposed to 12 or 18 kPa of pressure, with low shear stress caused by the medium flowing from the syringe to the manometer. The average shear stress, τ, was estimated by the following formula: τ = 6 μQ/wh2 ≈ 0.556 dyn/cm2, where μ is the fluid viscosity (approximately 10−7 N·s/cm2), Q is the flow rate (approximately 4 mL/6 h), w is the channel width (0.2 cm), and h is the channel height (0.01 cm). Low shear stress (