Passively Driven Integrated Microfluidic System for Separation of

Han-Sheng Chuang , Stephen C. Jacobson , Steven T. Wereley. Microfluidics and .... C. Lopez-Garcia , R. L. Monson , K. Haubert , M. B. Wheeler , D. J...
0 downloads 0 Views 185KB Size
Anal. Chem. 2003, 75, 1671-1675

Passively Driven Integrated Microfluidic System for Separation of Motile Sperm Brenda S. Cho,† Timothy G. Schuster,‡ Xiaoyue Zhu,† David Chang,§ Gary D. Smith,*,‡,|,⊥,# and Shuichi Takayama*,†,I

Department of Biomedical Engineering, Department of Urology, Department of Chemical Engineering, Department of Obstetrics & Gynecology, Department of Physiology, Reproductive Sciences Program, and Macromolecular Science and Engineering, University of Michigan, Ann Arbor, Michigan 48109

This paper describes a self-contained integrated microfluidic system that can separate motile sperm from small samples that are difficult to handle using conventional sperm-sorting techniques. The device isolates motile sperm from nonmotile sperm and other cellular debris, based on the ability of motile sperm to cross streamlines in a laminar fluid stream. The device is small, simple, and disposable yet is an integrated system complete with sample inlets, outlets, sorting channel, and a novel passively driven pumping system that provides a steady flow of liquid; it requires no external power source or controls. The device fulfills a need in clinical settings where small amounts of sperm need to be sorted. It also opens the way for convenient bioassays based on sperm motility including at-home motile sperm tests. A major focus of efforts in device miniaturization is to develop integrated microfluidic devices, so-called labs-on-a-chip.1-5 There are two main goals to such research: (i) to reduce time, equipment costs, and reagent consumption of existing macroscopic protocols and (ii) to perform procedures that are difficult or impossible to perform with macroscopic devices. Recent advances in microfabrication (e.g., soft lithography) have made it possible to easily construct almost any desired structure.6,7 There is still a challenge, however, in making functional integrated * Corresponding authors. E-mail: [email protected]. [email protected]. † Department of Biomedical Engineering. ‡ Department of Urology. § Department of Chemical Engineering. | Department of Obstetrics & Gynecology. ⊥ Department of Physiology. # Reproductive Sciences Program. I Macromolecular Science and Engineering. (1) Harrison, D. J.; Fluri, K.; Seiler, K.; Fan, Z. H.; Effenhauser, C. S.; Manz, A. Science 1993, 261, 895-897. (2) Burns, M. A.; Johnson, B. N.; Brahmasandra, S. N.; Handique, K.; Webster, J. R.; Krishnan, M.; Sammarco, T. S.; Man, P. M.; Jones, D.; Heldsinger, D.; Mastrangelo, C. H.; Burke, D. T. Science 1998, 282, 484-487. (3) Fu, A. Y.; Spence, C.; Scherer, A.; Arnold, F. H.; Quake, S. R. Nat. Biotechnol. 1999, 17, 1109-1111. (4) McClain, M. A.; Culbertson, C. T.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 2001, 73, 5334-5328. (5) Lagally, E. T.; Emrich, C. A.; Mathies, R. A. Lab Chip 2001, 1, 102-107. (6) Anderson, J. R.; Chiu, D. T.; Jackman, R. J.; Cherniavskaya, O.; McDonald, J. C.; Wu, H. K.; Whitesides, S. H.; Whitesides, G. M. Anal. Chem. 2000, 72, 3158-3164. (7) Unger, M. A.; Chou, H. P.; Thorsen, T.; Scherer, A.; Quake, S. R. Science 2000, 288, 113-116. 10.1021/ac020579e CCC: $25.00 Published on Web 02/22/2003

© 2003 American Chemical Society

systems that have a self-contained power source and are also practical to manufacture and use. A successful strategy pioneered by Weigl et al. for chemical analysis has been to integrate passively driven pumping systems and microfluidic channels onto simple, disposable chips.8 In this paper, we describe the use of a related concept for the development of a microscale integrated system to sort motile sperm from nonmotile cells. The device, which we call a microscale integrated sperm sorter (MISS), incorporates all functions necessary for sperm sorting onto a simple disposable polymeric microchannel device. A key component of the MISS is a novel gravity-driven pumping system that can maintain a steady flow rate over time regardless of the volumes of fluid in the reservoirs. The MISS can isolate motile sperm from regular semen samples as well as from samples that are difficult or not possible to process using traditional sperm-sorting methods (Figure 1). The MISS is inexpensive, small (just slightly larger than a U.S. penny), and easy to use. It fills a clinical need to select the most viable sperm for in vitro fertilization procedures and opens new possibilities in disposable, at-home diagnostic tests for male infertility. Approximately 10% of couples have infertility problems, of which roughly half involve a lack of or abnormal sperm.9 Currently, the most advanced treatment for such male-related infertility is an in vitro fertilization technique called intracytoplasmic sperm injection (ICSI),10 where an oocyte is fertilized by the direct injection of a sperm. ICSI has drastically reduced the number of viable sperm required for fertilization, giving hope of conception to extremely severe cases of sperm-based infertility. Because ICSI bypasses all normal sperm selection processes, however, it is important to have an efficient artificial selection process to maximize the probability of successful pregnancy, birth, and healthy offspring.11 While theoretically only a single sperm is needed per harvested ooctye, the practicality of processing, isolating, and locating the most viable sperm is challenging, or in some cases impossible, with current techniques such as centrifugation and swim-up processes.12 Thus, doctors frequently resort to hand sorting through dead sperm and debris to find a “good” (8) Weigl, B. H.; Bardell, R.; Schulte, T.; Williams, C. Proceedings of MicroTAS 2000, Enschede, The Netherlands, 2000; pp 299-302. (9) Mosher, W. D.; Pratt, W. F. Fertil. Steril. 1991, 56, 192-193. (10) Palermo, G.; Joris, H.; Devroey, P.; Van Steirteghem, A. C. Lancet 1992, 340, 17-18. (11) Schultz, R. M.; Williams, C. J. Science 2002, 296, 2188-2190. (12) Smith, S.; Hosid, S.; Scott, L. Fertil. Steril. 1995, 63, 591-597.

Analytical Chemistry, Vol. 75, No. 7, April 1, 2003 1671

Figure 1. Microscale integrated sperm sorter. (a) A photograph of the device. (b) Three-dimensional geometrical depiction of the twoinlet, two-outlet microfluidic channel designs with horizontally oriented fluid reservoirs that serve as passively driven, steady-flow rate pumps. A sperm sample is placed in one inlet reservoir and medium is placed in another inlet reservoir. The liquids and sperm flow in the x-y plane, and gravity is in the z direction. Upon sorting, one outlet reservoir will contain mostly motile sperm and the other outlet reservoir will contain motile and nonmotile sperm. The heights of the inlet reservoirs are 3.0 mm, and the outlet reservoirs 2.0 mm. The capillary force difference and height difference generates a steady hydraulic pressure which is the main driving force for the fluid pumping.

sperm (e.g., most motile and of a distinct morphology), a procedure that can take hours in some cases. Isolation of viable sperm from samples with severely low sperm counts (oligozoospermia) or from small amounts of cryopreserved spermsfor example, from patients that have preserved sperm prior to undergoing chemotherapysrequires new efficient technologies. EXPERIMENTAL SECTION Materials and Reagents. Poly(dimethylsiloxane) (PDMS) was obtained from Dow Corning (Sylgard 184), phosphatebuffered saline (PBS) from Invitrogen Corp., bovine serum albumin (BSA) fraction V from Sigma, HEPES buffered human tubal fluid with 0.2% BSA (Processing medium; PM) from Irvine Scientific, and propidium iodide (PI) from Molecular Probes. Preparation of Microfluidic Devices. Microfluidic spermsorting channels were made using soft lithographic methods.13 Briefly, PDMS was cast onto a master mold with the desired reservoir and channel features and cured. Resulting PDMS stamps were plasma oxidized to seal channels onto a glass cover slide. (13) Duffy, D. C.; McDonald, J. C.; Schueller, O. J.; Whitesides, G. M. Anal. Chem. 1990, 62, 4974-4984.

1672 Analytical Chemistry, Vol. 75, No. 7, April 1, 2003

Channels and reservoirs were further coated with a 1% BSA solution (dissolved in PBS) to improve liquid flow inside channels and to minimize nonspecific adsorption of cells to channel walls. The sperm inlet and nonmotile sperm outlet channel widths were 100 µm and their lengths 5000 µm. The sperm-free media inlet and motile sperm outlet channel widths were 300 µm and their lengths 5000 µm. The main channel or midchannel width was 500 µm and its length was 5000 µm. All channel heights were 50 µm. Procedure for Sorting and Analysis of Motile Sperm. Semen samples were obtained with Institution Review Board approval from men undergoing infertility evaluation. Sorting tests were performed using washed semen samples. In the order listed, 60 µL of processing medium was added to the media inlet reservoir, 50 µL of a washed semen sample to the sample inlet reservoir, and 2 µL of medium to each of the outlet reservoirs. Sperm-sorting yields were calculated taking these dilution factors into account. The numbers of motile sperm were determined by a Makler counting chamber (Sefi-Medical Instruments, Haifi, Israel). For visualization of membrane-compromised sperm, which generally corresponds to nonmotile sperm, 3 µL of PI (Molecular Probes, www.probes.com, 60 mM dissolved in processing medium) was added to sperm samples prior to sorting. A Texas Red filter set (577-nm excitation, 620-nm emission) was used to view red fluorescence from stained cells. An inverted microscope (Nikon TE 300, www.nikon-usa.com) with a CCD camera (Hamamatsu ORCA-100, www.hamamatsu.com) was used to capture images and record movies. RESULTS AND DISCUSSION Mechanism of Sorting. The MISS uses a sorting system where nonmotile sperm flow along their initial streamlines and exit one outlet whereas motile sperm can deviate from their initial streamlines and exit through a different outlet (Figure 2). This sorting mechanism is related to the “filtering” mechanism used in the H-filter where rapidly diffusing small molecules exit through a different outlet from larger molecules and particles that diffuse more slowly.14 The difference between the two devices is that the MISS takes advantage of active movement of cells whereas the H-filter takes advantage of passive diffusion of particles. These types of sorting are possible because, in small channels, multiple laminar streams can flow parallel to each other with no turbulent mixing at the interface between the streams (typical Reynolds numbers for the flow of sperm sample and media inside the MISS were on the order of 0.1).14-17 Nonmotile human sperm, ∼60 µm in length, and nonmotile particles on the same order of magnitude in size diffused slowly (D ) 1.5 × 10-13 m2/s; 690 s to diffuse 10 µm) and remained within their initial streamlines. In contrast, motile human sperm swim at velocities greater than 20 µm/s at 25 °C.18 This rapid mobility allows motile sperm, but not the nonmotile sperm, to distribute themselves randomly within the width of a 500-µm channel within seconds. The MISS was (14) Brody, J. P.; Yager, P. Sens. Actuators, A 1997, 58, 13-18. (15) Weigl, B. H.; Yager, P. Science 1999, 283, 346-347. (16) Kenis, P. J. A.; Ismagilov, R. F.; Whitesides, G. M. Science 1999, 285, 8385. (17) Takayama, S.; Ostuni, E.; LeDuc, P.; Naruse, K.; Ingber, D. E.; Whitesides, G. M. Nature 2001, 411, 1016. (18) World Health Organization. WHO laboratory manual for the examination of human semen and sperm-cervical mucus interaction, 4th ed.; Cambridge University Press: Cambridge, U.K., 1999.

Figure 2. Video images and schematic figure of sperm sorting. (a) Phase contrast images of sperm sample entering channel at the inlet junction, motile sperm swimming out of their initial streamline and spreading throughout the width of the channel, and motile sperm being sorted at the outlet junction. (b) Cartoon illustration of the video images shown in (a). The dashed line represents the interface between the parallel laminar streams. At the outlet junction, the motile sperm are evenly distributed throughout the width of the channel. The majority of the nonmotile sperm, however, are positioned in the initial streamline, which corresponds to the upper stream in this image. The relative flow rates of the inlet streams and outlet streams (upper stream/lower stream) are ∼1:3 (see Supporting Information for a movie of the process).

designed specifically to give sperm a residence time of 20 s in the main separation channel. A bifurcation placed at the end of this separation channel allows efficient collection of only the motile sperm that deviated from its initial inlet stream (Figure 2 and Supporting Information). Component Integration. The MISS integrates all functions necessary for sperm sortingssuch as inlet/outlet ports, fluid reservoirs, pumps, power source, and separation columnsonto a simple chip design that is practical to manufacture and use (Figure 1). A key design feature is the set of four horizontally oriented fluid reservoirs that also function as sample inlet/outlet ports and a fluid pumping system. The orientation, geometry, and size of these reservoirs are designed to balance gravitational forces and surface tension forces and provide a pumping system that generates a steady flow rate over extended periods of time regardless of the volume of fluid in the reservoirs. This contrasts with conventional gravity-driven pumping systems whose flow rates decrease over time as the volume of fluid in the inlet reservoir decreases. The diameters of the reservoirs were selected to be small enough that surface tension prevents liquid from spilling out of the horizontally oriented reservoirs but large enough to hold sufficient amounts of sample (tens to hundreds of microliters) and allow convenient sample introduction and recovery. This balance of forces allows the reservoirs to be arranged horizontally without the liquid inside spilling out. The horizontal reservoir arrangement, in turn, holds the height difference between the fluid in the inlet and outlet reservoirs the same (1.0mm height difference between inlet and outlet reservoir ceilings, Figure 1) regardless of the volume of fluid present in the reservoirs and maintains a constant hydraulic pressure even as the amount of fluid in the reservoirs changes.19

Passive Pumping Mechanism. The passively driven pumping system described here is unique in that it uses horizontally oriented reservoirs to overcome the problem of traditional gravitydriven pumping, where the pressure decreases as the amount of liquid in the reservoir decreases. Furthermore, the structure of the pump is greatly simplified compared to other mechanical or nonmechanical pumping systems, allowing easy manufacture and integration of the pump into a small, integrated device. Finally, the use of gravity and surface tension as the driving-force contributes to the overall small size of the MISS by eliminating the need for power supplies, such as batteries. Taking gravity, surface tension, and channel resistance into consideration, the MISS was designed to give a steady flow rate of sperm with a residence time of ∼20 s inside the main sorting channel. More specifically, the MISS is designed so that the flow resistance of the fluid reservoirs is more than 106 times less than that of the microfluidic channels and, therefore, negligible. Thus, the resistance of the channels, calculated to be 2.8 × 1012 kg/(s/m4), approximates the total resistance of the system. Since a bulk flow rate of 0.008 µL/s is required to achieve the desired residence time of 20 s and the total resistance is 2.8 × 1012 kg/(s/m4), the net pressure drop required to drive the fluid is 23 N/m2. To achieve this desired pressure drop, we designed the dimensions of the reservoirs such that capillary forces (3.0-mm-diameter inlet reservoir versus 2.0-mm-diameter outlet reservoir) would be 13 N/m2 and the pressure drop across the microfluidic channel of the MISS due to hydrostatic forces (1.0-mm height difference) would be 9.8 N/m2. For calculation of the capillary force, we approximated the contact angle to be 0° (the contact angle of water (19) Zhu, X.; Phadke, N.; Chang, J.; Cho, B.; Huh, D.; Takayama, S. Proceedings of MicroTAS 2002, Nara, Japan, 2002; pp 151-153.

Analytical Chemistry, Vol. 75, No. 7, April 1, 2003

1673

Figure 4. Purity of sorted sperm samples. The purity ratio, shown in the bar graphs, is the number of motile sperm to the total number of sperm in a given reservoir. The blue bars represent the motile sperm purity of the sperm sample inlet, and the purple bars represent the motile sperm purity of the sorted motile sperm outlet.

Figure 3. Video images and schematic of the sorting of sperm sample treated with PI. (a) Fluorescent image of PI-stained sperm. The stained fraction, which represents membrane-compromised sperm, remains in the initial sperm sample streamline and exits out the nonmotile sperm outlet. (b) Phase contrast image of the experiment shown in (a). The motile, nonmembrane-compromised sperm are seen exiting the motile sperm outlet. (c) Schematic illustration of the sperm sorting process observed in the fluorescent and phase contrast images shown in (a) amd (b) (see Supporting Information for movies on the process).

on BSA-coated PDMS is very small), the surface tension of the washed semen sample to be ∼0.040 N/m (less than that of water due to “impurities” such as proteins), and the viscosity of the washed semen sample to be similar to that of water. The observed bulk flow rate of 0.008 µL/s for a dilute particle suspension in 1% BSA solution was approximately equal to that of the calculated flow rate. Actual sperm samples sometimes had lower flow rates due to larger apparent viscosity. Smaller flow rates for the sperm sample stream would result in slightly lower yields but do not affect the purity of the sperm recovered at the sorted sperm outlet. Sperm Sorting. Sperm-sorting efficiencies of the MISS were evaluated by three methods: (i) tracking the movement of motile sperm in the channel by phase contrast microscopy (Figures 2 and 3), (ii) tracking movement of PI-stained cells in the channel by fluorescence microscopy (Figure 3), and (iii) using a Makler counting chamber, a grid-based sperm-counting device, to determine numbers of motile sperm and nonmotile sperm in the inlet 1674 Analytical Chemistry, Vol. 75, No. 7, April 1, 2003

and outlet reservoirs (Figure 4). The sperm-tracking experiments shows the process of how motile sperm can swim out of its initial streamline (Figure 2 and Supporting Information). PI stains membrane-compromised cells such as dead cells and thus allows the nonmotile sperm to be highlighted and visualized with red fluorescence while the motile sperm remain unstained (Figure 3 and Supporting Information). The bar graphs in Figure 4 compare percentage of sperm that are motile before and after sorting. The purity of motile sperm after sorting was nearly 100% regardless of motile sperm purity before sorting (Figure 3). The yields (39, 42, and 43%), defined as the ratio of the number of motile sperm in the motile sperm outlet reservoir to the total number of motile sperm in the sperm sample inlet reservoir, were comparable to or greater than the recovery rates (0.8-50%) of sperm processed using conventional sorting methods such as direct swim-up, swimup from a pellet of centrifuged sperm, or density gradient separation.12,20 We also observed that sperm morphology, another important trait that correlates with successful pregnancies, also improved after sorting with the MISS (strict sperm morphology: 9.5 ( 1.1% normal before sorting to 22.4 ( 3.3% normal after sorting). Kruger Strict sperm morphology is a set of criteria or standards whereby sperm must fit within specific measurements (head width and length, tail length, acrosome making up a certain percentage of the sperm head) and lack abnormalities (e.g., pin head, round head, crimped tail).21-23 CONCLUSION New single-cell-based treatments such as ICSI require new technologies for cellular manipulations in order to make procedures more efficient and safe.11 The MISS increases the efficiency of sperm isolation from samples with few motile sperm that are difficult or impossible to process using conventional methods. The MISS may also decrease risks24 associated with ICSI by isolating viable sperm using a mild, biomimetic sorting mechanism based on sperm motility; it avoids centrifugation and sample compaction, (20) Florence, L. H. N.; Liu, D. Y.; Baker, H. W. G. Hum. Reprod. 1992, 7, 261266. (21) Katz, D. F.; Diel, L.; Overstreet, J. W. Biol. Reprod. 1982, 26, 566-570. (22) Grow, D. R.; Oehninger, S.; Seltman, H. J.; Toner, J. P.; Swanson, R. J.; Kruger, T. F.; Muasher, S. J. Fertil. Steril. 1994, 62, 559-567. (23) Kruger, T. F.; Coetzee, K. Hum. Reprod. Update 1999, 5, 172-178. (24) Hansen, M.; Kurinczuk, J. J.; Bower, C.; Webb, S. N. Engl. J. Med. 2002, 346, 725-730.

which have been reported to cause sublethal damage to sperm.12,25,26 The simple and user-friendly design of the MISS allows straightforward system transfer from the microfabrication laboratory to the clinical setting where sperm isolation is needed (Schuster et al.; unpublished). In addition to clinical use, the combination of MISS with a colorimetric readout of sperm number (www.embryotech.com, for example) is envisioned to be useful as a self-contained at-home test for preliminary screening of male infertility and vasectomy, vasectomy reversal success, or both. Since sperm motility is a sensitive indicator of toxicity,27 the MISS may also be useful as a toxicology test. Integration of additional function into MISS is envisioned soon. For example, serial connections of multiple MISS systems should improve motile sperm yield; parallel operation of multiple MISS systems would greatly increase the volume of sample that can be processed. Changes in the geometries and dimensions of the channels and reservoirs would alter flow rates and vary motile sperm yield and purity. Additionally, the use of multiple outlet channels may allow sorting of sperm based on more subtle differences in motility. Although we have focused on sperm, the MISS should be applicable to the sorting of any microorganisms that can swim efficiently. Finally, the design concept of MISSsto developing self(25) Alvarez, J. G.; Lasso, J. L.; Blasco, L.; Nunez, R. C.; Heyner, S.; Caballero, P. P.; Storey, B. T. Hum. Reprod. 1993, 8, 1087-1092. (26) Aitken, R. J.; Clarkson, J. S. J. Androl. 1998, 9, 367-376. (27) Bavister, B.; Andrews, J. C. J. In Vitro Fertil. Em. 1988, 5, 67-75. (28) Beebe, D. J.; Walker, G. M. Lab Chip 2002, 2, 57-61. (29) Handique, K.; Burke, D. T.; Mastrangelo, C. H.; Burns, M. A. Anal. Chem. 2000, 72, 4100-4109. (30) Puntambekar, A.; Choi, J. W.; Ahn, C. H.; Kim, S.; Makhijani, V. Lab Chip 2002, 2, 213-218. (31) Beebe, D. J.; Moore, J. S.; Bauer, J. M.; Yu, Q.; Liu, R. H.; Devadoss, C.; Jo, B. H. Nature 2000, 404, 588-590. (32) Stroock, A. D.; Dertinger, S. K. W.; Ajdari, A.; Mezic, I.; Stone, H. A.; Whitesides, G. M. Science 2002, 295, 647-651.

contained, readily fabricated, functional microdevices that can manipulate living cells without the need for electronics or external power sourcessis of interest for a broad range of biomedical applications. The increasing number of microfluidic components that function without external power or control such as the steady flow rate micropump system of the MISS, gravity-,8,19 absorbent,8 or evaporation-driven28 pumps, chemically29 or structurally programmed30 valves, stimuli-responsive hydrogel valves,31 passive fluid mixers,32 or diffusion-based filters,14 provides many opportunities for expanding the possibilities of self-contained functional microfluidic systems for cellular manipulations. ACKNOWLEDGMENT We thank Brian Johnson and Mark Burns for assistance in channel fabrication and use of their clean room facilities, Joseph Bull and James Grotberg for helpful discussions, and Wei Gu for assistance in fabrication of devices. We thank the Whitaker Foundation, NICHD, and the Engineering Technology Development Fund for financial support. SUPPORTING INFORMATION AVAILABLE Three movies (one showing separation of motile from nonmotile sperm in the microfluidic channel, one showing flow of propidium iodide-stained sperm viewed under green excitation light, and another showing flow of propidium iodide-stained sperm along with nonstained motile sperm viewed under phase contrast illumination). This material is available free of charge via the Internet at http://pubs.acs.org.

Received for review September 17, 2002. Accepted January 16, 2003. AC020579E

Analytical Chemistry, Vol. 75, No. 7, April 1, 2003

1675