Anal. Chem. 2006, 78, 3193-3197
Fluorescence Monitoring of ATP-Stimulated, Endothelium-Derived Nitric Oxide Production in Channels of a Poly(dimethylsiloxane)-Based Microfluidic Device Teresa D’Amico Oblak, Paul Root, and Dana M. Spence*
Department of Chemistry, Wayne State University, 5101 Cass Avenue, Detroit, Michigan 48202
Intracellular nitric oxide (NO) production in a microfluidic endothelium is detected using fluorescence microscopy. Bovine pulmonary artery endothelial cells (bPAECs) were loaded with the fluorescence probe diaminodifluorofluorescein diacetate (DAF-FM DA), and the subsequent fluorescent DAF-FM DA/NO adduct was measured. Solutions of bradykinin, a well-known stimulus of endothelium-derived NO, activated nitric oxide synthase (NOS) in the immobilized bPAECs. This activation was inhibited using L-nitro arginine methyl ester (L-NAME), a competitive inhibitor of NOS. Importantly, the NO production was also stimulated with adenosine triphosphate (ATP) using concentrations as low as 1 µM. Previous reports on stimulating NO production using an immobilized endothelium in microfluidic channels were limited by the requirement of ATP concentrations of at least 100 µM, a value that is not physiologically relevant. The ability to monitor NO production with ATP concentrations that are similar to in vivo levels of ATP in the microcirculation represents a major advance in the use of microfluidic technology as an in vitro model of the microcirculation. Recently, in an attempt to define the role of adenosine triphosphate (ATP) as a stimulus for endothelium-derived nitric oxide (NO) in the microcirculation, our group used a microfluidic device to create an in vitro mimic of an important in vivo process. Specifically, we were able to immobilize bovine pulmonary artery endothelial cells (bPAECs) in a fibronectin-coated channel patterned on a microfluidic device derived from poly(dimethylsiloxane) (PDMS).1 Using this system, it was demonstrated that NO could be measured at a micromolded electrode, patterned from carbon ink,2 that was integrated into the channel of the microfluidic device. Moreover, the NO that was detected at the electrode was produced and secreted from the endothelial cells that were immobilized to the channel surface. In this sense, the microfluidic channels represent a model of an important in vivo process; namely, the ability of endothelial cells to produce NO (a potent vasodilator) in the presence of an ATP stimulus. (1) Spence, D. M.; Torrence, N. J.; Kovarik, M. L.; Martin, R. S. Analyst 2004, 129, 995-1000. (2) Kovarik, M. L.; Torrence, N. J.; Spence, D. M.; Martin, R. S. Analyst (Cambridge, U.K.) 2004, 129, 400-405. 10.1021/ac052066o CCC: $33.50 Published on Web 04/04/2006
© 2006 American Chemical Society
Although the creation of the in vitro model of the endothelium on the microfluidic device possessed many advantages over our previous report that employed microbore tubing as the vessel mimic,3 there were still facets of this work that needed to be improved. For example, NO could generally not be detected amperometrically at the carbon ink electrode unless high micromolar amounts of ATP were introduced across the chip-based endothelium. In fact, 100 µM ATP was required to measure NO release on a regular basis, a concentration that is roughly 2 orders of magnitude higher than those levels available to stimulate NO production in vivo.4 Although these previous studies suggested that ATP is, indeed, a stimulus for NO in vivo, such concentrations of extracellular ATP in the microcirculation probably do not exist. This assumption is based on the fact that typically, studies involving the measurement of deformation-induced ATP release from erythrocytes (red blood cells, RBCs) have reported ATP release values that were in the low micromolar to high nanomolar range.5-9 It has been reported that RBCs are the source of the ATP that is stimulating the endothelium-derived NO in the microcirculation, specifically the lung.5 Therefore, it is imperative that a detection scheme be developed that is capable of measuring NO in the presence of micromolar to submicromolar levels of an ATP stimulus. Here, we demonstrate the ability to monitor NO derived from bPAECs in a microfluidic device using the fluorescent probe DAFFM DA.10,11 DAF-FM DA is an intracellular fluorescein based dye that permeates and is retained inside the cell wall once esterases deacylate the dye forming DAF-FM, a slightly fluorescent compound. In the presence of nitric oxide and oxygen, DAF-FM forms (3) Kotsis, D. H.; Spence, D. M. Anal. Chem. 2003, 75, 145-151. (4) Sprague, R. S.; Olearczyk, J. J.; Spence, D. M.; Stephenson, A. H.; Sprung, R. W.; Lonigro, A. J. Am. J. Physiol. 2003, 285, H693-H700. (5) Sprague, R. S.; Ellsworth, M. L.; Stephenson, A. H.; Lonigro, A. J. Am. J. Physiol. 1996, 271, H2717-H2722. (6) Sprague, R. S.; Ellsworth, M. L.; Stephenson, A. H.; Kleinhenz, M. E.; Lonigro, A. J. Am. J. Physiol. 1998, 275, H1726-H1732. (7) Sprung, R. J.; Sprague, R. S.; Spence, D. M. Anal. Chem. 2002, 74, 22742278. (8) Fischer, D. J.; Torrence, N. J.; Sprung, R. J.; Spence, D. M. Analyst 2003, 128, 1163-1168. (9) Price, A. K.; Fischer, D. J.; Martin, R. S.; Spence, D. M. Anal. Chem. 2004, 76, 4849-4855. (10) Kojima, H.; Sakurai, K.; Kikuchi, K.; Kawahara, S.; Kiino, Y.; Nagoshi, H.; Hirata, Y.; Nagano, T. Chem. Pharm. Bull. 1998, 46, 373-375. (11) Kojima, H.; Urano, Y.; Kikuchi, K.; Higuchi, T.; Hirata, Y.; Nagano, T. Angew. Chem., Int. Ed. 1999, 38, 3209-3212.
Analytical Chemistry, Vol. 78, No. 9, May 1, 2006 3193
Figure 1. A PDMS chip was fabricated to contain 100 µm × 100 µm × 2 cm channels. bPAECs were cultured into the microfluidic channels, and the chip was reversibly sealed to a glass substrate, making sure to align the channel and the fluid access port. Solutions were perfused across the cells (2 µL min-1) using 75-µm microbore tubing to connect the syringe pump to the microfluidic adaptor on the fluid access port. Fluorescence measurements were acquired using an Olympus IX71M microscope equipped with an electrothermally cooled CCD camera and MicroSuite software.
a fluorescent benzotriazole that has an excitation at 495 nm and emits at 515 nm. Without the production of nitric oxide, DAF triazoles are not formed in cells, indicating the appropriateness for utilizing DAF dyes in the detection of nitric oxide. Key to the NO measurements reported here is that the endothelial cells were stimulated with concentrations of ATP that are similar to those levels believed to be released from RBCs in vivo. Moreover, the levels of ATP used to stimulate NO in this study are closer to the submicromolar levels of ATP required to stimulate vasorelaxation in isolated rabbit lungs.4 EXPERIMENTAL SECTION Preparation of the Microfluidic Device. PDMS channel structures were produced on the basis of previously published methods.12,13 Briefly, masters for the production of PDMS microchannels were made by coating a 4-in. silicon wafer (Silicon, Inc., Boise, ID) with SU-8 10 negative photoresist (MicroChem Corp., Newton, MA) using a spin coater (Brewer Science, Rolla, MO) operating with a spin program of 2000 rpm for 20 s. The photoresist was prebaked at 95 °C for 5 min prior to UV exposure with a near-UV flood source (Autoflood 1000, Optical Associates, Milpitas, CA) through a negative film (2400 dpi, Jostens, Topeka, KS), which contained the desired channel structures. All channel structures were drawn in Freehand (PC version 10.0, Macromedia, Inc. San Francisco, CA). Following this exposure, the wafer was postbaked at 95 °C for 5 min and developed in Nano SU-8 developer (Microchem Corp.). The thickness of the photoresist was measured with a profilometer (Alpha Step-200, Tencor Instruments, Mountain View, CA), which corresponded to the channel depth of the PDMS structures. (12) McDonald, J. C.; Duffy, D. C.; Anderson, J. R.; Chiu, D. T.; Wu, H.; Schueller, O. J. A.; Whitesides, G. M. Electrophoresis 2000, 21, 27-40. (13) McDonald, J. C.; Whitesides, G. M. Acc. Chem. Res. 2002, 35, 491-499.
3194 Analytical Chemistry, Vol. 78, No. 9, May 1, 2006
A 20:1 mixture of Sylgard 184 elastomer and curing agent (Ellsworth Adhesives, Germantown, WI) was used (in comparison to the normally used 10:1 mixture) to increase the adhesiveness of the PDMS to aid in the reversible bonding procedure. This mixture was poured onto the silicon wafer and cured at 70 °C for ∼2 h. After this time, the PDMS layer was then removed from the master. A chip containing channels of 100 µm depth × 100 µm width × 2 cm length was used for all studies reported here. The channel depth corresponds to the height of the master, which was measured with the aforementioned profilometer. Cell Immobilization. The fabricated polymer microchip was reversibly sealed to a glass substrate, making sure to align the channel and the fluid access port. Glass plates (7.33 cm wide, 10.0 cm length, and 1 mm thick) were purchased from Bio-Rad (Hercules, CA) and drilled in-house for fluid access. Drilling was accomplished using a 1.5-mm diamond bit and a Dremel rotary tool. A microfluidic adaptor (Nanoport, Upchurch Scientific, Oak Harbor, WA) assembly was used to connect the fluid access hole to microbore tubing. Once the fluid access port was placed in line with the channel, the chip was reversibly sealed to the plate, taking care to eliminate any air bubbles trapped between the glass and the microfluidic chip. A solution of 100 µg/mL fibronectin (Sigma Chemical Co., St. Louis, MO), was pumped through the channel at 1 µL min-1 for 1 h at room temperature. The fibronectin solution was then removed by vacuum aspiration. The chip was removed from the glass substrate, and a suspension of bPAECs was then pipetted into the channel and incubated at 37 °C and 5% CO2 for 2 h or until the cells had attached to the fibronectin layer. Once the endothelial cells had adhered to the microfluidic channel, the PDMS chip was sealed to the glass substrate as described above, and the complete experimental setup shown in Figure 1 was used to monitor the NO production in the layered cells. This
Figure 2. Bradykinin was introduced at a flow rate of 2 µL min-1 to bPAECs immobilized in the channels of a microfluidic device. The stimulated NO production was monitored via the formation of the fluorescent DAF-FM DA/NO adduct. Figure 2a was obtained immediately after 1 mM bradykinin was introduced to the channel containing the bPAECs, whereas Figure 2b was obtained in the same location after 30 min had elapsed. The intracellular fluorescence intensities are summarized in part c. L-Nitro arginine methyl ester , was used to demonstrate that the increase in DAF-FM DA/NO fluorescence upon stimulation with 10 µM and 1 mM bradykinin, respectively, was the result of endothelial NO production. Cells incubated with 1 mM L-NAME prior to the addition of bradykinin resulted in no significant increase in fluorescence. The asterisks indicate a significant increase in fluorescence intensity (relative to the addition of buffer to the cells) upon addition of bradykinin to the bPAECs (p < 0.001).
method is similar to previous work involving the immobilization of PC 12 cells in channels patterned in PDMS.14 Measurement of Nitric Oxide. Hank’s Balanced Salt Solution (HBSS, Sigma) was pumped through the channel at 2.0 µL min-1 for 10 min to remove excess media in the channel. A solution of HBSS supplemented with 5 mM L-arginine (Sigma) and 5 µM DAF-FM DA (Molecular Probes, Eugene, OR), which lies in the manufacturer’s suggested range for this probe, was pipetted into the channels. The DAF-FM DA solution was allowed to incubate with the cells for 30 min in the cell incubator to ensure a sufficient amount of dye had accumulated in the endothelial cells. Excess DAF-FM DA was removed by pumping HBSS over the channel at 2.0 µL min-1 for 10 min, and a fluorescence image was obtained to establish the background fluorescence. To measure NO production in the cells, solutions of bradykinin (Sigma), adenosine triphosphate (Sigma), or L-nitro arginine methyl ester (L-NAME, Sigma), a competitive inhibitor of nitric oxide synthase, were prepared in HBSS pumped over the channel until the channel was completely filled. Fluorescence images were taken every 5 min using an Olympus IX71M microscope (Olympus America, Melville, NY) with an electrothermally cooled CCD (Orca, Hamamatsu) and MicroSuite software (Olympus America). The microscope incorporated an FITC filter cube (Chroma Technology Corp.) containing the excitation (460-500 nm) and emission (505-560 nm) filters. Fluorescence intensities were averaged from 1 × 103 pixels2 centered inside an immobilized cell. RESULTS AND DISCUSSION Measurement of Bradykinin-Stimulated NO. To demonstrate that the bPAECs immobilized in the channels of the (14) Li, M.; Spence, D. M.; Martin, R. S. Electroanalysis 2005, 17, 1171-1180.
microfluidic device were indeed capable of producing NO, bradykinin, a well-known stimulus of NO production,15 was introduced to the cells at a flow rate of 2 µL min-1. The fluorescence signal, resulting from the DAF-FM DA/NO adduct, was then monitored every 5 min up to 30 min. As shown in Figure 2, the bPAECderived NO production increases upon introduction of 1 mM bradykinin to the immobilized cells in the microfluidic channel. Figure 2a is a fluorescence image obtained immediately after 1 mM bradykinin was introduced to the channel containing the bPAECs, whereas Figure 2b is an image obtained in the same location after 30 min had elapsed. To demonstrate that bradykinin was stimulating NO production, and that the increase in the fluorescence intensity was due to increments in intracellular NO, the bPAECs were incubated with L-nitro arginine methyl ester prior to stimulation with bradykinin. L-NAME is a well-established inhibitor of nitric oxide synthase (NOS).16 The data in Figure 2c verify that the increases in fluorescence seen in Figure 2a and b were the result of NOS activation upon addition of 10 µM and 1 mM bradykinin solutions, respectively. Furthermore, there is no increase in fluorescence upon addition of bradykinin to the bPAECs that had been incubated with L-NAME while immobilized in the channels of the microfluidic device. Note that some fluorescence exists prior to addition of bradykinin to the DAF-FM DA-loaded cells in Figure 2a and b. This is to be expected because, although bradykinin stimulates the activation of NOS, there will exist some basal levels of NO in the bPAECs prior to activation of NOS by bradykinin; (15) Bogle, R. G.; Coade, S. B.; Moncada, S.; Pearson, J. D.; Mann, G. E. Biochem. Biophys. Res. Commun. 1991, 80, 926-932. (16) Sprague, R. S.; Stephenson, A. H.; Dimmit, R. A.; Weintraub, N. A.; Branch, C. A.; McMurdo, L.; Lonigro, A. J. Am. J. Physiol. 1995, 296, H1941-H1948.
Analytical Chemistry, Vol. 78, No. 9, May 1, 2006
3195
Figure 3. Response of immobilized bPAECs to increasing concentrations of ATP. The fluorescent DAF-FM DA/NO adduct was monitored at 0 min (a) and 30 min (b) upon addition of 10 µM ATP. Intracellular fluorescence intensities are summarized in part c after the addition of solutions containing ATP at concentrations of 1, 10, and 100 µM. The asterisks indicate a statistically significant difference between the fluorescence resulting from the addition of buffer and the fluorescence resulting from the addition of ATP (p < 0.001).
however, the intracellular level of NO does not increase upon bradykinin stimulation when the cells are incubated with L-NAME. Moreover, these basal levels are not believed to be due to shear stress. Prior to addition of bradykinin, HBSS buffer was pumped through the endothelium-coated channel of the microfluidic device without any measurable increase in fluorescence. Therefore, although it is well-established that shear-stress-induced NO exists in endothelial cells, we were unable to measure shear-stressinduced NO production with the flow rates employed in this study. The data in Figure 2c represent the average fluorescence intensity of the DAF-FM DA-loaded bPAECs (shown in the images in Figure 2a and b) and the studies involving L-NAME. The data in Figure 2c provide evidence that the immobilized bPAECs have the ability to respond to a known stimulus of NO production and, moreover, that this NO production can be altered in the presence of a known inhibitor (L-NAME). Measurement of ATP-Stimulated NO. The studies involving endothelium-derived NO production using bradykinin as the stimulus were performed because it is well-established that bradykinin is an excellent stimulus of NO production;15 however, it is also known that ATP, when released from RBCs flowing through the intact circulation in vivo, also stimulates NO production.4,5 Specifically, ATP binds to a P2y purinergic receptor found on the endothelial cells that line circulatory vessels.17 Previously, our group was able to demonstrate that ATP-induced NO production could be determined amperometrically in a channel fabricated in a PDMS-based microfluidic device.1 Unfortunately, the concentration of ATP employed in that study was 100 µM, a value that is ∼2 orders of magnitude greater than values typically released from the RBCs upon mechanical deformation.4,6-9 Here, we report the ability to monitor endothelium NO production in bPAECs im(17) Motte, S.; Perotton, S.; Boeynaems, J. M. Circ. Res. 1993, 72, 504-510.
3196 Analytical Chemistry, Vol. 78, No. 9, May 1, 2006
mobilized in the channel of a microfluidic device using ATP concentrations as low as 1 µM. The images in Figure 3 show the resultant fluorescence from the DAF-FM DA probe reacting with NO in the presence of varying amounts of ATP added to the bPAECs. Figure 3a is a fluorescence image of cells at 0 min stimulated with 10 µM ATP, whereas Figure 3b is a fluorescence image of the same cells taken 30 min after the addition of the 10 µM ATP solution. Figure 3c summarizes the average fluorescence intensities of a predefined area (1 × 103 pixels2) immediately after the ATP was added to the cells (t ) 0 min) and after an elapsed period of time (t ) 30 min). The data in Figure 3c, which also shows data from cells incubated with ATP concentrations of 1 and 100 µM, respectively, further suggest that the ATP is stimulating NO production and that the NO production is ATP-concentration-dependent. More importantly, the data in Figure 3c show statistically significant differences in fluorescence intensity upon addition of ATP concentrations as low as 1 µM (when compared to a buffer containing no ATP). This value of 1 µM is important because it has been demonstrated that, in the isolated rabbit lung, a minimum of 0.3 µM ATP was a required component of the perfusate to demonstrate a significant change in mean arterial pressure.4 The data reported here suggest that measuring endotheliumderived NO via dynamic fluorescence microscopy may be an improvement over previous reports of measuring NO derived from a chip-based endothelium using amperometry as the detection scheme.1 However, it is important to note that the amperometric technique was measuring NO that had been released into the extracellular milieu. In this sense, the amperometric measurement may not be measuring the total ATP-stimulated NO production (because not all NO is released from a cell upon synthesis).
However, amperometry has the advantage of being a noninvasive mode of measuring endothelium-derived NO. Therefore, multiple stimulations or perturbations to the cell can be carried out and monitored. The work described here will advance attempts to mimic in vivo environments (such as the microcirculation) on an in vitro platform. This is especially true when considering that, in vivo, RBCs traveling through the microcirculation release high nanomolar to low micromolar amounts of ATP. Therefore, using a fluorescence detection platform to monitor the NO production may enable the incorporation of RBCs across the chip-based endothelium to more completely investigate the role of RBCs in the control of vascular tone in the microcirculation. The ability to monitor the interaction between RBCs flowing across an endothelium provides a valuable tool for researchers investigating biological events in the circulation. Finally, in vivo, cells often
interact with other cells (and often do so as tissues). Therefore, the ability to monitor cellular signaling events with different detection schemes may prove to be a useful tool as researchers continue to build more elaborate and complex cellular scaffolds using microfluidic technology. ACKNOWLEDGMENT The authors thank Michael Owczarek for assisting with the setup of the facility used to culture the cells employed in this work. This work was funded by the National Institutes of Health (HL073942).
Received for review November 22, 2005. Accepted March 10, 2006. AC052066O
Analytical Chemistry, Vol. 78, No. 9, May 1, 2006
3197
