미세유체기술을기반으로연동자극을모사한소수정란체외배양시스템

Transcription

1 석사학위논문 Master s Thesis 미세유체기술을기반으로연동자극을모사한소수정란체외배양시스템 Microfluidic in vitro culture system mimicking peristaltic stimulation in bovine embryos 배채윤 ( 裵彩允 Bae, Chae Yun) 바이오및뇌공학과 Department of Bio and Brain Engineering KAIST 2009

2 미세유체기술을기반으로연동자극을모사한소수정란체외배양시스템 Microfluidic in vitro culture system mimicking peristaltic stimulation in bovine embryos

3 Microfluidic in vitro culture system mimicking peristaltic stimulation in bovine embryos Advisor: Professor Je-Kyun Park by Chae Yun Bae Department of Bio and Brain Engineering KAIST A thesis submitted to the faculty of the KAIST in partial fulfillment of the requirements for the degree of Master of Science in Engineering in the Department of Bio and Brain Engineering. Daejeon, Korea Approved by Professor Je-Kyun Park (Major Advisor)

4 미세유체기술을기반으로연동자극을 모사한소수정란체외배양시스템 배채윤 위논문은한국과학기술원석사학위논문으로학위논문심사 위원회에서심사통과하였음. 2009년 12월 22일 심사위원장 박제균 ( 인 ) 심사위원 최철희 ( 인 ) 심사위원 정기훈 ( 인 )

5 MBiS 배채윤. Bae, Chae Yun. Microfluidic in vitro culture system mimicking peristaltic stimulation in bovine embryos 미세유체기술을기반으로연동자극을모사한소수정란체외배양시스템 Department of Bio and Brain Engineering p. Advisor Prof. Park, Je-Kyun. Text in English Abstract This work demonstrates a novel microfluidic in vitro cultivation system for embryos that improves their development using a partially constricted channel that mimics peristaltic muscle contraction (the gravity-driven stimulation system). To investigate the compressive effects of constriction geometry on embryonic development, different constriction widths of the channel were designed. Bovine embryos were loaded on the channel and the whole system was incubated on a tilting machine to provide embryo movement via gravity. The fertilized embryos were cultivated on the gravity-driven stimulation system until the blastocyst, hatching, or hatched blastocyst stages. The proportion of eight-cell development among total embryos in the constricted channel (56.77±13.7%; mean±sd) was superior to that in the straight channel (23.97±11.0%). Moreover, the membrane based mechanical stimulation was also designed because of needs for independent control of flow and compressive force (the pressure-driven stimulation system). Therefore, the possibility to improve developmental ratio for bovine embryos was suggested as supporting more precise in vivo like environment. i

6 Contents Abstract i Table of contents ii Nomenclature iv List of Tables vii List of Figures viii 1. Introduction Assisted reproductive technology, in vitro culture Microfluidic in vitro culture system Major issues Research objectives 5 2. Materials and Methods Gravity-driven stimulation Channel design Fabrication Preparation of bovine embryos Operation of the gravity-driven stimulation system Blastocyst double staining Pressure-driven stimulation Channel design Fabrication Operation of the pressure-driven stimulation Characterization of the pressure-driven stimulation 19 ii

7 3. Results and Discussion Gravity-driven stimulation Microfluidic embryo manipulation Microfluidic in vitro culture Effect on the early development Pressure-driven stimulation Characterization of the membrane Compressive force control Microfluidic in vitro culture Flow control Conclusions 39 References 42 Summary in Korean 48 iii

8 Nomenclature Alphabetic Letters g V r K F n acceleration of gravity volume of embryos Stokes radius of embryo stiffness of embryo normal force iv

9 Greek Letters ρ embryo ρ medium θ μ η δ density of the embryo density of the medium plane angle of the tilting machine coefficient of friction fluid viscosity embryo deformation v

10 Abbreviations IVM IVF IVC CWs FBS IVP in vitro maturation in vitro fertilization in vitro culture constriction widths fetal bovine serum in vitro production vi

11 List of Tables Table 1. Review of microfluidic in vitro culture system. vii

12 List of Figures Figure 1. Schematic view and design of a channel mimicking segmental contractions in an oviduct. Figure 2. Relationship between the applied force and embryo deformation determined by embryo diameter and CW. Figure 3. Design of the gravity-driven stimulation system. Figure 4. Configuration and picture of the gravity-driven stimulation system. Figure 5. Design of the pressure-driven stimulation system. Figure 6. Schematic and picture of the pressure-driven stimulation system. Figure 7. Plot of embryo velocity and fluid flow rate depending on tilting angle. Figure 8. Bovine embryo development in the gravity-driven stimulation system. Figure 9. Double-stained blastocysts. Figure 10. The percent of bovine embryos at early developmental stages in microchannels with 150 or 160 mm CW compared with a straight microchannel as a control. viii

13 Figure 11. Plot of pressure measurement depending on syringe displacement. Figure 12. Plot and pictures of the culture channel height depending on pressure loading and unloading. Figure 13. Plot and pictures of the embryo deformation depending on pressure loading and unloading. Figure 14. Morphology of bovine embryos as the result of in vitro culture on the pressure-driven stimulation system. Figure 15. Plot of the flow rate depending on the syringe displacement ix

14 Chapter 1 Introduction 1.1 Assisted reproductive technology, IVC Emerging assisted reproductive technologies have been developed in in vitro production (IVP) for several decades and applied to the treatment of human infertility [1], embryo manipulation [2], and nuclear transfer [3]. IVP techniques are also important in livestock breeding which is directly related to economic aspects of the agricultural industry [4]. Continuous refinement of IVP [5] has been achieved through biochemical changes such as modification of medium components [6], their concentrations and supplementations of new compounds at the specific point as well as environmental changes in physiological conditions, such as temperature [7], ph, shear stress [8], and osmolality [9]. Although these efforts have revealed many factors that are critical to developing embryos, the developmental ratio, particularly that of mammalian embryos such as porcine and bovine species, still remains low [10]. In a single IVP run, through the complicated four steps such as in vitro maturation(ivm), sperm selection, in vitro fertilization(ivf) and in vitro culture(ivc), oocytes and embryos are moved as many as twenty times between media, requiring removal from their incubator environment. Therefore, the most important step for pregnancy ratio, IVC would be - 1 -

15 improved by automation and improved handling to reduce the environmental stress. Moreover, relative inefficiency of mammalian IVP is also believed to be due to the different environments of in vitro and in vivo conditions. Therefore, trials to identify key factors that positively affect embryo development in an in vivo environment are extremely important, and several studies to mimic in vivo conditions have been proposed [11] 1.2 Microfluidic IVC system Microfluidic technology is considered one of the methods to create in vivo like biological microenvironment and to investigate microscale physical behavior through the manipulation of micro and nanoliter volumes of fluid [12, 13]. Therefore, the physical control of biological sample is relatively precise, simple and easy to support in vivo like environment. Moreover, the automation of external equipment is easily applied to the microfluidics with improved handling technology. Microfluidic based IVP has been applied in researches on several processes, such as spermatozoa sorting [14], the removal of cumulus cells [15], IVF [16] and IVC [17, 18]. For IVC in particular (Table 1), however, static culture [17] and volume related study [19] were briefly considered in microfluidics

16 Moreover, microfluidic research for in vivo like physiological microenvironment based on embryo culture system represent a more dynamic flow culture; thus, novel culture systems using Braille-display [20] have been suggested. However, although these studies were innovative trials over the concept of chemical modification, few reports have considered in vivo dynamic physical stimulation in an oviduct, such as peristaltic constriction, which moves embryos from the fallopian tube to the uterus. Therefore, this study demonstrated microfluidic IVC system based on mechanical stimulation mimicking peristaltic stimulation for bovine embryos [21]. Microdrop Static microfluidic Braille microfluidic Stimulation microfluidic Embryos θ Tilting machine [5] First, 1986 [17] Beebe, 2004 [20] Takayama, 2007 [21] Park, 2009 Conventional method First culture Dynamic (braille) Peristaltic stimulation PROS Simple manipulation PROS Improved development PROS Moderate mass transfer PROS Peristaltic stimulation Bovine embryo CONS Static culture condition CONS Mouse embryo Static culture CONS Mouse embryo CONS Relatively fast flow Table 1. Review of microfluidic in vitro culture system

17 1.3 Major issues Muscular activity of the oviduct has been recognized as one of factors for ovum transport, which is an initial step in normal in vivo reproductive events [22]. In the female oviduct, fertilization occurs in the ampulla-isthmus junction and the fertilized zygotes are transported into the isthmus, where they are delayed (Fig 1a). And then they slowly enter the uterus. The reason they are delayed at the entrance of uterus is oviductal peristalsis due to the circular muscle contraction which encloses isthmus region of oviduct. When the circular muscle was contracted, intraluminal pressure cycle and electrical pulses have beens recorded in the rabbit oviduct [23, 24] and the inner diameter of oviduct was reduced when the muscle contraction occurred. Embryo itself is given to compressive force during development because of muscle contraction. Therefore, this study demonstrated microfluidic IVC system with two different methods based on mechanical stimulation mimicking peristaltic stimulation for bovine embryos. Another major issue in this study is to improve low efficiency for mammalian embryos. Previous researchers applied their microfluidic system only to mouse embryos which is relatively easy to culture in vitro [17, 19, 20]. Almost 80% of mouse embryos were developed until blastocysts in the - 4 -

18 biochemical modified medium. In the bovine embryos, on the other hand, less than 30% of embryos were only developed until blastocysts in the same media condition [25]. To improve IVP of mammalian embryos, relatively more attention than mouse embryos needs to focus on the mechanical platform on which IVP occurs. Therefore, bovine embryos were applied to the microfluidic IVC system with mechanical stimulation. 1.4 Research Objectives This work aims to propose a novel microfluidic IVC system based on the mechanical stimulation derived from peristalsis using bovine embryos. To investigate the effects of mechanical stimulation on the development of bovine embryos in a microfluidic channel, gravity driven stimulation system via constriction was introduced for initial work. A micro-modulated tilting machine was adopted to realize a simple dynamic flow culture system using gravity due to the IVC is a long-term culture process. In addition, pressure driven stimulation system was also characterized because of needs for independent control of flow rate and compressive force. The membrane based pressure driven stimulation was manipulated by syringe pump and quantitatively characterized by membrane deflection and embryo deformation

19 Chapter 2 Materials and methods 2.1 Gravity-driven stimulation Channel design During transport, embryos are under the influence of two different mechanics, ciliary motility and muscle contractility. Muscular contraction is closely related to the transport, stimulation, and denudation of embryos due to deformation and squeezing of the tubal wall [26]. There are three parts in the oviduct such as ampulla, ampulla-isthmus junction and isthmus. As the oviduct is close to the uterus, inner diameter of the oviduct is narrower. There are three different muscle layers in the isthmus, which is the narrowest part of the oviduct: the outer longitudinal or spiral-shaped uterine layer, the intermediate circular layer, and the innermost longitudinal layer. Segmental contractions in the isthmus are generated by intermediate circular muscle (Fig 1a). The embryo is pressed against mucosal folds through the contracted oviductal segment [27]. The effect of the oviductal intermediate circular muscle was mimicked by designing microchannels with constriction areas and constriction width (CW) (Fig 1b)

20 (a) A B C A Ampulla 2 cell Fertilized egg Ampulla-isthmus junction B Peristalsis via muscle contraction and relaxation 4 cell C Isthmus Blastocyst Intermediate circular muscle layer (b) Embryo Constriction area Direction of embryo movement Constriction width Figure 1. Schematic view and design of a channel mimicking segmental contractions in an oviduct. (a) Segmental contractions with cross-section view and longitudinal sectional view (b) Peristalsis-mimicked schematic with constriction area and constriction width (CW) To determine appropriate CWs, a range of applied forces that positively affected embryos was considered relative to embryo deformation (Fig 2)

21 Based on a previous study on human embryos [28], the maximal force to maintain elastic deformation was set as a critical threshold value for mechanical stimulation. In general, the average diameter of bovine embryos in the morula stage was ± 4.04 μm and ranged from 148 to 170 μm. Based on the results shown in Fig 2, the CWs used for mechanical stimulation were 150 and 160 μm (Fig 3a). Figure 2. Relationship between the applied force and embryo deformation determined by embryo diameter and CW. The critical force applied to embryos (0.2 μn) and stiffness of embryos was determined according to [28]

22 Each microchannel was 200 μm 50 mm 200 μm (w l h) and contained (except for the control) ten constriction areas. The channels also included two side reservoirs, two channel necks, and an in/outlet port (Fig 3b). Channel necks were 50 μm in width to prevent embryos from leaking. The in/outlet port was positioned at the middle of each channel for loading embryos. (a) a No constriction channel ba ca 200 μm 200 μm 200 μm 150 μm 200 μm 160 μm ab ac (b) Channel neck 50μm 150μm 200μm Side Reservoir In/outlet port Side Reservoir Figure 3. Design of the gravity-driven stimulation system. (a) Channel design of the microfluidic IVC device. (b) Dimensions of CW for the microfluidic IVC device

23 2.1.2 Fabrication The gravity-driven stimulation device was prepared by conventional photolithography, poly(dimethylsiloxane) (PDMS) (sylgard 184; Dow corning, Midland, MI) replica molding, and aligning processes. The features were fabricated using an SU (MicroChem, Newton, MA) photoresist. Briefly, the spin speeds were set at 500 rpm for 10 s and 1450 rpm for 30 s to make a fluidic channel 200 μm in height. Then samples were soft baked at 65 for 25 min and 95 for 30 min, exposed to ultraviolet light for 36.1 s at 650 mj/cm 2, and then hard baked at 65 for 1 min and 95 for 30 min. Finally, the resultant wafer was immersed into SU-8 developer for 30 min and rinsed with isopropanol for 15 s. The patterned silicon wafer was sufficiently silanized and then submersed in PDMS and cured. After punching side reservoirs and the in/outlet port, the device was bonded with a glass slide by the O 2 plasma asher (270 W, 30 s). As shown in Fig 4a, a mineral oil reservoir made of PDMS was attached to both side reservoirs using the O2 plasma asher (270 W, 30 s). Then PDMS was pasted outside of the reservoirs and allowed to cure. The final device is shown in Fig 4b

24 2.1.3 Preparation of bovine embryos All materials were purchased from Sigma-Aldrich (St. Louis, MO) unless stated otherwise. Bovine oocyte maturation medium (IVM medium) [29], Percoll [30], SP-TALP [30], Fert-TALP [29], CR1aa medium supplemented with 0.3% BSA (IVC-BSA medium) [31], and CR1aa medium supplemented with 10% fetal bovine serum (FBS) (IVC-FBS medium) [31] were used as reported previously. Collected bovine ovaries were transported to the laboratory in 0.9% normal saline in a thermo flask containing 150 μg/ml penicillin and 100 μg/ml streptomycin. The selected oocytes were incubated in IVM medium at 38.5 for 22 h. After incubation, the oocytes were washed to remove cumulus cells that were growing around the oocytes. To prepare IVF, frozen bovine semen was included in a two-phase Percoll tube to separate motile sperm cells from others and centrifuged at 2000 rpm for 20 min. Motile sperm cells from the bottom of the tube were sprinkled on SP-TALP and centrifuged at 1000 rpm for 10 min. Prepared oocytes were loaded on Fert- TALP with 2 μl heparin and 2 μl sperm cells. The IVF process was continued with incubation at 38.5 for 22 h. Fertilized oocytes were prepared for both conventional microdrop [31] and the microfluidic IVC device procedures. All

25 procedures were performed according to the ethical guidelines of the Institutional Review Board at KAIST Operation of the gravity-driven stimulation system Before the bovine embryos were loaded on the device, each channel was filled with 50 μl IVC-BSA media via the in/outlet port and 200 μl mineral oil was poured into the oil reservoirs to prevent media evaporation. The fertilized oocytes were collected and twenty embryos were loaded in a microchannel through the in/outlet port, which was blocked by the inlet cap during incubation (Fig 4a). The microfluidic IVC device with the IVF oocytes was incubated at 38.5 for 44 h on the tilting machine, which made a cyclic ± 10 tilt over 1 min (Fig 4c). After oocytes had completed early development, a new device filled with 50 μl IVC-FBS medium was prepared for later-stage development. At this point, cell number was recorded for each stage and both four-cells and eight-cells were loaded rapidly. The device was also incubated at 38.5 for 1 wk to get bovine blastocysts on the tilting machine (Fig 4d)

26 (a) Mineral oil Inlet cap Oil reservoir (b) medium Slide glass Embryos (c) Embryos (d) θ Figure 4. Configuration and picture of the gravity-driven stimulation system. (a) Configuration of the gravity-driven stimulation system (b) Picture of the fabricated the device. Scale bar is 1 cm. (c) Configuration of tilting machine and the system. (d) Picture of a micro-modulated tilting machine Blastocyst double staining Blastocyst quality was examined by double staining. The collected blastocysts were treated by pronase to remove zona pellucid, and then incubated in rabbit anti-pig whole serum for 1 h. Both propidium iodide (Invitrogen, Carlsbad, CA) and Hoechst (Invitrogen, Carlsbad, CA) were applied for 1 h and then stained blastocysts were fixed on glass slides. A fluorescent microscope

27 (Olympus 1 51, Hahioji, Japan) was used to detect the number of cells and the color of stained cells. 2.2 Pressure-driven stimulation Channel design The pressure-driven stimulation device was considered to control flow rate and compressive force independently. A channel schematic of pressure-driven stimulation system is described in Fig 5. There were two separated layers such as a culture channel layer and a stimulation control channel layer. The stimulation control channel layer was a bottom layer with 30 μm PDMS membrane and the culture channel layer was an attaching top layer. A pair of culture channels in the culture channel layer was separated at the side of media inlet and intersected on the side of media outlet. Each channel had 180 μm in widths and 200 μm in height. Before both channel intersected on the media outlet side, channel necks were prevented embryos from leaking in 50 μm widths. The embryo inlet/outlet ports were on the between mechanical or non-mechanical stimulation area and media inlet ports. One of the culture channels was aligned on the stimulation channel (3 mm 4 mm (w

28 l)) for mechanical stimulation area with 4 mm in length. Both culture channels were crossed on the blocking channel to confine the embryos to mechanical or non-mechanical stimulation area. Both the stimulation control channel and the blocking control channel had 60 μm in height on the same layer. The stimulation control channel size was and the blocking control channel size was 200 μm 4 mm (w l). Culture channel layer Stimulation control channel layer Stimulation control port 200μm 180μm Embryo in/outlet port Mechanical stimulation Media outlet port Media inlet port Channel neck 50μm Non-mechanical stimulation Blocking control port Figure 5. Design of the pressure-driven stimulation system

29 2.2.2 Fabrication Two different layers of the pressure-driven stimulation device were separately fabricated by conventional photolithography, poly(dimethylsiloxane) (PDMS) (sylgard 184; Dow corning, Midland, MI) replica molding, and aligning processes (Fig 6a). The culture channel was fabricated using an SU (MicroChem, Newton, MA) photoresist. Briefly, the spin speeds were set at 500 rpm for 10 s and 3000 rpm for 30 s to make a fluidic channel 200 μm in height. Then samples were soft baked at 65 for 6 min and 95 for 40 min, exposed to ultraviolet light for 32 s at 650 mj/cm 2, and then hard baked at 65 for 5 min and 95 for 15 min. Finally, the resultant wafer was immersed into SU-8 developer for 15 min and rinsed with isopropanol for 15 s. The stimulation/blocking control channel was fabricated using an SU (MicroChem, Newton, MA) photoresist. The spin speeds were set at 500 rpm for 10 s and 1700 rpm for 30 s to make a fluidic channel 200 μm in height. Then samples were soft baked at 65 for 3 min and 95 for 6 min, exposed to ultraviolet light for 25 s at 650 mj/cm 2, and then hard baked at 65 for 1 min and 95 for 4 min. Finally, the resultant wafer was immersed into SU-8 developer for 3 min and rinsed with isopropanol for 15 s

30 (a) Culture channel (b) 170~200 μm 90 μm 60 μm Stimulation/Blocking control channel Figure 6. Schematic and picture of the pressure-driven stimulation system. (a) Schematic of the pressure-driven stimulation device: culture channel layer (upper layer; height 200 μm) and stimulation/blocking control channel layer (bottom layer; height 60 μm). Membrane height was about 30 μm. (b) Picture of the pressure-driven stimulation device. Both the patterned silicon wafers were sufficiently silanized. PDMS was spin coated on the stimulation/blocking control channel wafer. The spin speeds were set at 500 rpm for 10 s and 1400 rpm for 30 s to make 90 μm height membrane on the patterned wafer. The membrane was cured on the flat surface. Then, the culture channel wafer was submersed in PDMS and cured. After punching the media in/outlet port and the embryo in/outlet port in the culture channel, it was bonded with PDMS membrane on the stimulation/blocking channel by the O 2 plasma asher (270 W, 1 min). Then the integrated layers were detached from the stimulation/blocking control channel wafer and each

31 inlet port of the stimulation/blocking control channel was punched. Finally, the device was bonded with a glass slide by the O 2 plasma asher (270 W, 30 s). The final device is shown in Fig 6b Operation of the pressure-driven stimulation One day before the loading embryos, the culture channel and the stimulation/blocking control channel were filled with IVC-BSA media. The blocking control channel and the stimulation control channel were connected with 50 μl and 100 μl gas tight syringe. The culture channel was connected with 2.5 ml gas tight syringe (Fig 6b). The IVC-BSA media flowed through the culture channel with 1 μl/h flow rate. The media outlet of culture channel was linked to the 1 ml EP-tube for a media reservoir. The whole device with connected syringes and the pump were maintained in the incubator at 38.5 for 22 h. The embryo was also prepared with same direction of the section The fertilized oocytes were collected and twenty embryos were loaded with 1 μl/h flow rate through the embryo in/outlet ports of each channel, which was blocked by the inlet cap during incubation. Then, the media was injected into the blocking control channel to confine the embryos on the mechanical

32 stimulation or non-mechanical stimulation area. Both the compressive force and the flow rate were controlled by the 100 μl and 2.5 ml syringe fixed on the micro-modulated pump. Each amplitude, duration and frequency was controlled by the micro-modulated pump. The syringe connected device with the bovine embryos was maintained in the incubator at 38.5 for 22 h Characterization of the pressure-driven stimulation Each syringe was fixed on the micro-modulated pump. To estimate how much pressure on the syringe, digital pressure switch (SMC digital pressure switch ZSE40F) was filled with water and connected to the water filled syringe. The membrane deflection and the embryo deformation were measured by microscopic (Olympus 1 51, Hachioji, Japan) images. To observe the side view of the channel, a piece of silicon mirror was fixed on the right next to the culture channel and PDMS was cured. Then the silicon mirror and patterned wafer were removed. The flat side of surface was put on the microscopy. Then, the culture channel, membrane deflection and embryo deformation were examined

33 Chapter 3 Results and Discussion 3.1 Gravity-driven stimulation Microfluidic embryo manipulation Fertilized embryos are sensitive to temperature [32], and thus the cell manipulation time for IVC outside of CO 2 incubators should be minimized. Hence, a simple cell loading technique using micropipettes was adopted. Direct connecting external equipment to a microfluidic device may change the temperature of the CO 2 incubators, even if the embryo injection time outside the CO 2 incubator is minimized. Therefore, a micro-modulated tilting machine as shown in Fig 4d was used to create a driving force to move embryos. By simply inserting the embryos through the in/outlet port of the microfluidic device and using a tilting machine, preparation for IVC was completed simply and rapidly within 1 min. To apply a peristalsis effect to the embryos, they must penetrate the constricted areas. The penetration of embryos occurs when the moving force of embryos (F 1 ) is larger than the frictional force (F 2 ):

34 (1) (2) where and are the densities of the embryo and medium, respectively, is the volume of an embryo, is the acceleration of gravity, is the plane angle of the tilting machine, is the fluid viscosity, is the Stokes radius of the embryo, is the terminal velocity of the embryo, is the coefficient of friction, is the normal force exerted on embryo deformation, is the stiffness of an embryo, and is embryo deformation. Thus, the angle of the tilting machine and the difference between the CW and diameter of an embryo are the principal parameters for mechanical stimulation. A tilt of 10 achieved over 1 min was sufficient to penetrate the constriction area of embryos even though they were of various sizes. Therefore, this angle was applied and maintained for 5 min, and then reversed 10 over 1 min, to resemble a seesaw

35 The plane angle of tilting machine affected the medium flow rate and embryo velocity. As the tilted angle was increased, both embryo velocity and fluid flow rate were dramatically increased (Fig 7). The actual embryo velocity was almost 2300 μm/s at 10 tilting condition. The medium flow rate also was more than 7000 nl/min at the critical angle, 10 tilting condition. Both embryo velocity and fluid flow rate was relatively higher than the profiles from in vivo condition. If the total length of ovum transport is assumed about 3 cm, and the duration is also assumed 5 days, then average velocity for ovum transport would be calculated 0.07 μm/s. The fluid flow rate was also considered as less than 100 nl/min [24]. Hence, both values should be minimized. Otherwise, high shear stress would be negatively affected during the embryo development. Because the fertilized embryos were sticky due to the presence of a zona pellucida [33], surface modification was considered to maintain the movement of embryos in the microfluidic channels. However, BSA was presented in the IVC medium so that the bovine embryos moved well along the straight channel without interacting with the PDMS walls, even though they were attached and stuck to each other. Evaporation may change the osmolality of a medium, and prevent embryos from normal development. Thus to minimize evaporation, water was filled in the petridish containing a device and mineral oil was poured

36 into the side media reservoirs. In addition, the hole of the in/outlet port was capped with a PDMS fragment. Embryo velocity (μm/s) Embryo movement Fluid diffusion 0 degree 5 degree 10 degree Tilting condition (angle) Fluid Flow flow rate rate (nl/min) Figure 7. Plot of embryo velocity and fluid flow rate depending on tilting angle Microfluidic in vitro culture Microfluidic channels were filled with IVC media, and fertilized bovine embryos were loaded. The medium volume was 50 μl and the volume inside the channel was 20 μl, which was the same as the amount used in the conventional microdrop method

37 The embryos loaded into the IVC device were cultivated in IVC BSA media for 44 h. Cleaved embryos at the four and eight-cell stages were collected from the in/outlet port using a micropipette and transferred to another device filled with IVC FBS media. The transferred embryos were cultivated for 5~7 days. Figure 8 shows bovine embryos cultivated in a microfluidic channel for 7 days. Each embryo was theoretically stimulated 64,800 times by 160 μm CW (Tilting angle: 10, Frequency: 0.15 Hz). The black arrow indicates an embryo in the blastocyst stage of up to 180 μm in size. Because this embryo was too big to penetrate the constriction areas, it moved partially between adjacent constriction areas. When cultivated for a further 1~2 days, such embryos continued development until hatching, or the hatched blastocyst stage, as indicated by the dotted arrow of Fig 8d. Although a hatched zona pellucida > 200 μm in size sometimes blocked the movement of embryos (as shown in Fig 8d), this did not cause any major problems because this only happened at the end of the IVC. On the contrary, because the size of fertilized embryos does not differ significantly until the early morula stage, we simply observed the penetration of embryos in the microfluidic channels by comparing their initial and final positions

38 (a) (b) (c) Blastocyst (d) Hatched Blastocyst Blastocyst Figure 8. Bovine embryo development in the gravity-driven stimulation system. (a) Picture of an embryo before penetrating a constriction area. (b) Picture of an embryo after penetrating the constriction area. (c) Picture of embryos developed for 1wk. Black solid arrow indicates an embryo in the blastocyst stage. (d) Picture of embryos developed for 10 days. Dotted arrow and solid arrow indicate hatched blastocyst and inner cell mass of the blastocyst, respectively. All scale bars are 100 μm. To verify the quality of embryos cultivated in the microfluidic channel with gravity-driven stimulation, they were stained with Hoechst (blue, inner cell

39 mass) and propidium iodide (red, trophoblast) (Fig 9). Figure 9a shows an embryo in the blastocyst stage cultivated using the conventional microdrop method. The high proportion of inner cell mass over trophoblast and more staining reflects the higher quality of the blastocysts. Actually, it was doubted whether a blastocyst cultivated in the microfluidic channel was normally developed and comparable with that from the conventional culture method in quality. However, the double-staining method revealed that embryos cultivated in the microfluidic IVC device were comparable with those cultivated using the conventional method. (a) X 200 (b) X 200 Conventional microdrop Microfluidic IVC system Figure 9. Double-stained blastocysts. (a) Double-stained blastocyst cultivated using the conventional microdrop method. (b) Double-stained blastocyst cultivated in the gravity-driven stimulation system. Blue and red colors represent inner cell mass and trophoblast, respectively

40 3.1.3 Effect on the early development Conventional embryology has focused on the chemical composition of the culture environment, and modifications thereof, to enhance embryo development [7, 34]. Recent findings of the key factors that positively affect embryo development in the in vivo environment have been critical and efforts to mimic in vivo conditions have been progressed. However, few studies have considered dynamic physical stimulation, such as peristaltic constriction, in an oviduct. By comparing the development ratio of embryos cultivated in a straight channel to those cultivated in channels with constricted areas, the effects of mechanical stimulation on the early development of bovine embryos were examined. Figure 10 shows the percent of embryos at each early developmental stage relative to total embryo numbers, such that the summed percent of each stage (the two-, four- and eight cell stages) should be 100% (n56). Therefore, a larger percent in the eight-cell stage represents a better environment for embryo development. The percent of embryos in the straight channel and constricted channels with 150 and 160 μm CWs in the eight-cell stage were ± 11.0 (mean ± SD), ± 11.0, and ± 13.7%, respectively. As seen in Fig 10, the microchannels with constricted areas were superior to the

41 microchannel without mechanical stimulation. We also examined embryonic development in a microchannel with a 140 μm CW; however, a preliminary experiment indicated that only about 10% of the embryos developed until the eight-cell stage and many of them did not exhibit healthy morphology. In addition, when more than 0.2 μn force was applied to human embryos, they but did not return to their original state after deformation, i.e. they were severely damaged [26]. As shown in Fig 2, embryos greater than 150 μm were exposed to harsh external forces in the microchannel with a 140 μm CW. Assuming that bovine embryos are similar to human embryos, it can be inferred that this stress negatively affected embryonic development. This also corresponds to the results shown in Fig 10. Although the microchannel with a 150 μm CW improved embryo development more than did the control channel with no mechanical stimulation, the condition was not an optimal range of stimulation, representing that the force caused by the dimension was relatively big condition to embryos when compared with the result of 160 μm width. Many factors affect embryo cultivation, such as the angle of the tilting machine, number of constricted areas, CW, time required to achieve the angle, and time held at the angle. Because IVC requires long culture times, we could not examine embryonic development to optimize each of these factors. Therefore,

42 more detailed studies are needed to realize more effective conditions for embryonic development. However, an advanced concept to eliminate the transfer process of embryos and the need to remove the device from the CO 2 incubator would be desirable to develop an improved IVC system. Cleavage ratio (%) Straight microchannel Microchannel with 150 mm CW Microchannel with 160 mm CW 2 cell 4 cell 8 cell Developmental stages ** * Figure 10. The percent of bovine embryos at early developmental stages in microchannels with 150 or 160 μm CW compared with a straight microchannel as a control. Asterisks denote statistically significant differences ( * P < 0.05, ** P = 0.001) between the indicated pairs

43 3.2 Pressure-driven stimulation Characterization of the membrane The gravity-driven stimulation device had two major problems such as relatively uncontrollable high value of flow rate and irregular stimulation in the partially blocked channel by variable sized embryos. Although the early development ratio in the mechanical stimulation channel was remarkably increased compared with the straight channel, device itself did not show significantly improved result compared to the conventional microdrop because of the high shear stress. Therefore, pressure-driven stimulation design was considered with certain points: independent controlling of flow rate and finecontrolling of compressive force. To apply finely deflected membrane, the miro-modulated pump was used to investigate down to the minutest details. To identify how much pressure could be applied to the membrane, the pressure on the 100 μl syringe was measured as its displacement increased. Therefore, linear relationship (, R 2 = ) between syringe displacement and measured pressure (Fig 11). Although we used only small scale of syringe displacement such as from 0 to 1000 μm, we calibrated syringe displacement into the pressure with linear fit relationship

44 Pressure (kpa) Syringe displacement (μm) Figure 11. Plot of pressure measurement depending on syringe displacement Compressive force control It was applied for the pressure-driven stimulation system to give compressive force on embryos directly by controlling membrane deflection. To find out how much deflect the membrane depending on the pressure loading, the side view of culture channel was observed and the channel height was measured (Fig 12). As the pressure was loaded on the stimulation control channel, the membrane was deflecting and the height of culture channel was decreasing. To find out hysteresis of membrane deflection, the culture channel

45 height was measured at the loaded and unloaded pressure on the membrane. Both pressure increasing and decreasing condition, the degree of membrane deflection was comparable. Channel height (μm) Pa Pressure increasing Pressure decreasing Pressure (Pa) 130 Pa 230 Pa Figure 12. Plot and pictures of the culture channel height depending on pressure loading and unloading

46 Although the plot of the culture channel height depending on the pressure was shown as linear, the variation of measured error was relatively high. With consideration of clear relationship between PDMS membrane deflection and loading pressure, it would be possible to measure more precise membrane deflection with different methods. Even though the membrane deflection was measured relatively rough, it is easy to compare with embryo deformation. It would be suggested to press the side of embryos into the other side in order to observe embryo deformation easily. Otherwise, the side view detection was inevitable to verify a deformed sphere shape because the pressure was loaded below the embryos. Embryos were also observed and the height of deformed shape was measured to verify their deformation in the pressure-driven stimulation system (Fig 13). As the loaded pressure was increased, the membrane was deflecting and the deformed height of embryos was decreasing. The embryo size has become decrease at the point of 100 Pa because the actual size of embryo was smaller than the culture channel height. Although the embryo size was decreasing with linear relationship during loading pressure from 100 Pa to 250 Pa, on the other hand it was increasing with different slope of linear relationship during unloading pressure from 250 Pa to 0 Pa. Nonelastic

47 deformation of embryos was supposed to have such hysteresis due to the excessive loading pressure. Therefore, when the embryos were cultured on the device, total loading pressure was controlled at the range of less than 250 Pa for elastic deformation of bovine embryos. Embryo height (μm) Pressure increasing Pressure decreasing Pa Pressure (Pa) 130 Pa 230 Pa Figure 13. Plot and pictures of the embryo deformation depending on pressure loading and unloading

48 3.2.3 Microfluidic in vitro culture Bovine embryos have been cultured on the pressure-driven stimulation system for 2 days in order to find out the possibility of normal cleavage during the early development (Fig 14). The medium flow on the both mechanical and non-mechanical stimulation channel was the continuous with 10 nl/min flow rate for 44 h. The mechanical stimulation channel was only induced compressive force (rate of rise: 15 Pa/s, duration: 10 sec, frequency: Hz) for 44 h. Although the morphology of cleaved embryos was only observed on the microscopy, embryos from mechanical stimulation were clearly developed into several numbers of cells. They also showed compacted cleaved cells because of sine waved compressive force. According to the previous work, compaction is one of the most distinct in morphology between in vivo and in vitro produced embryos [35]. If the cleaved embryos are checked by molecular biology work such as gene expression or ECM networking on the embryos, the biological effect between two conditions could be discussed further. Moreover, the condition of flow and compressive force should be optimized to find out in vivo like condition representing on the pressure-driven stimulation system

49 (a) (b) Figure 14. Morphology of bovine embryos as the result of in vitro culture on the pressure-driven stimulation system. (a) 2 days in vitro culture from non-mechanical stimulation. (b) 2 days in vitro culture from mechanical stimulation. Scale bar is 100 μm. Although the controllable range became wider in this system compared to the gravity-driven stimulation system, it would be complicated to apply every condition in a practical manner. Therefore, in vivo like condition would be suggested to the system to find out optimal conditions of flow and compressive force for bovine embryos. According to the previous biology work, intraluminal pressure and fluid flow rate of the in vivo and in vitro oviduct was recorded. Although the loading pressure influenced not only embryo deformation but also PDMS membrane absorption, the stimulation range and pattern would be deliberated. The perfusion catheter recorded intraluminal pressure in the oviduct was recorded as pulsatile signal [23] (rate of rise:

50 Pa/s, duration: 0.57 sec, frequency: 0.17 Hz). The flow control in a microfluidic device would be suggested fluid pulses of media at physiological frequencies (0.135 Hz in rabbit oviduct [24]). Therefore, the range of flow rate and compressive force would be suggested from in vivo reference for optimized condition in the pressure-driven stimulation system Flow control To slow down the default flow rate on the culture channel, we used syringe pump. When the 2.5 ml syringe was controlled with certain amount of displacement, flow velocity was measured by tracing streamlines of 1 μm beads (Fig 15). In this graph, syringe displacement was programmed as loading 3 μm/s for 5 s, unloading 3 μm/s for 5 s and resting for 15 s. Then the velocity profile was showed as pulsatile flow (frequency about 0.05 Hz). If the syringe displacement was programmed as loading 1 μm/s for 2 s and resting for 29 s, the continuous flow (flow rate about 10 nl/min) would be observed (result was not shown). In this result, the precise flow control in the pressure-driven stimulation system was successfully verified compared with the gravity-driven stimulation system. When the flow on the 10 controlled tilting machine occurred, diffusion time was about 10.6 s (continuous flow rate about

51 nl/min) and the velocity of embryo movement was about 2000 μm/s. However, medium flow in the pressure-driven stimulation system was controlled on the range of 10 nl/min and even pulsatile flow could be generated. And the embryo movement was relatively minimized. Therefore, the flow induced physical stress on the embryos was reduced. Flow rate (nl/min) Flow rate Syringe (2.5 ml) displacement Time (sec) Syringe displacement (um) Syringe displacement (μm) Figure 15. Plot of the flow rate depending on the syringe displacement

52 Chapter 4 Conclusions Microfluidics is very useful for studying sophisticated biological in vivo environments and for customizing in vitro experiments. We demonstrated that mechanical stimulation via constriction in the gravity-driven stimulation system could affect and encourage the early development of bovine embryos. In vitro fertilized bovine embryos were successfully cultured in microfluidic systems. To the best of our knowledge, this is the first trial to examine the effects of mechanical stimulation on embryonic growth to mimic peristaltic constriction in vivo. Moreover, controllable flow and compressive force were implemented on the embryos using pressure-driven stimulation system. Although much supplementary work should follow, this work is expected to provide new insight into the development of microfluidic platforms for assisted-reproductive technology research. There are several points and challenges to improve this work. First of all, the physical stimulation condition for bovine embryos should be optimized for the pressure-driven stimulation system. Although we investigated overall range of the pressure and flow rate in this system, appropriate physical condition for

53 improvement in development of bovine embryos would be examined. In this system, we would like to separate the flow control from the pressure generating because of uncontrollable high flow rate in the gravity-driven stimulation system. Nevertheless, two different factors would be influenced each other. When the pressure was loaded in a pulsatile form, the flow was generated in a certain pattern even though any flow has been not given. Consequently, the integrated effect from pressure-driven flow should be considered. To find out appropriate physical condition in this system, embryo condition should be also guaranteed with a technical improvement. According to the preexperiment, early developmental ratio of bovine embryos was significantly decreasing as the manipulation time outside of incubator was increasing. As we considered, gravity-driven stimulation system had effects on short and easy manipulation process. However, pressure-driven stimulation system with improved flow control had relatively complicated connection between external equipment and the device. Therefore, the manipulation time was increased. Technical improvement of an experimenter would be one of the solutions. Moreover, reformed external equipment or redesigned to compressive method could be considered. Various systems would be also given consideration to improve user interface. The tilting was one of the simple

54 methods however it was hard to control many physical factors precisely and independently. Therefore the other methods such as centrifugal force, magnetic force or hydrostatic force would be considered as alternative system for giving mechanical stimulation in microfluidic IVC system for bovine embryos. On the process to verify the embryos, multifarious evidence should be confirmed. We did morphology observation and double staining. According to previous biological researches, molecular biology work such as gene expression and ECM networking would be one of the methods to confirm the embryo quality. The result from embryo transfer was also necessary to verify the pregnancy rate comparison between in vivo produced embryos and the microfluidic IVC produced embryos. Finally, several improved design would be suggested to embody in vivo peristalsis more precisely. Other type of physical forces would be considered such as hydrostatic pressure. There are several other factors to influence in vitro culture for embryos: the number of loading embryos, osmolality change and the volume of media. Therefore, the channel design would be considered with such different factors. Single embryo would be also pressed in a cube frame. Separate control with two grating shaped membrane would express more in vivo like peristalsis on the channel

55 References [1] M. Hagmann, "Fertility therapy may aid gene transfer," Science, 284, pp , [2] I.K. Glasgow, H.C. Zeringue, D.J. Beebe, S.J. Choi, J.T. Lyman, N.G. Chan, M.B. Wheeler, "Handling individual mammalian embryos using microfluidics," IEEE Trans. Biomed. Eng., 48, pp , [3] I.A. Polejaeva, S.H. Chen, T.D. Vaught, R.L. Page, J. Mullins, S. Ball, Y. Dai, J. Boone, S. Walker, D.L. Ayares, A. Colman, K.H. Campbell, "Cloned pigs produced by nuclear transfer from adult somatic cells," Nature, 407, pp , [4] W.H. Velander, H. Lubon, W.N. Drohan, "Transgenic livestock as drug factories," Sci. Am., 276, pp , [5] J.J. Parrish, Susko-Parish, J. L., Jeibfried-Rutledge, M. L., Critser, E. S., Eyestone, W. H., First, N. L., "Bovine in vitro fertilization with frozenthawed semen," Theriogenology, 25, pp , [6] D.K. Gardner, M. Lane, "Culture of viable human blastocysts in defined sequential serum-free media," Hum. Reprod., 13, pp ; discussion 160., [7] B.M. Ravindranatha, S. Nandi, H.M. Raghu, S.M. Reddy, "In vitro

56 maturation and fertilization of buffalo oocytes: effects of storage of ovaries, IVM temperatures, storage of processed sperm and fertilization media," Reprod. Domest. Anim., 38, pp , [8] Y. Xie, F. Wang, W. Zhong, E. Puscheck, H. Shen, D.A. Rappolee, "Shear stress induces preimplantation embryo death that is delayed by the zona pellucida and associated with stress-activated protein kinasemediated apoptosis," Biol. Reprod., 75, pp , [9] R. Li, K. Whitworth, L. Lai, D. Wax, L. Spate, C.N. Murphy, A. Rieke, C. Isom, Y. Hao, Z. Zhong, M. Katayama, H. Schatten, R.S. Prather, "Concentration and composition of free amino acids and osmolalities of porcine oviductal and uterine fluid and their effects on development of porcine IVF embryos," Mol. Reprod. Dev., 74, pp , [10] A.T. Oliveira, R.F. Lopes, J.L. Rodrigues, "Gene expression and developmental competence of bovine embryos produced in vitro with different serum concentrations," Reprod. Domest. Anim., 41, pp , [11] G.M. Walker, M.S. Ozers, D.J. Beebe, "Insect cell culture in microfluidic channels," Biomed. Microdevices, 4, pp , [12] D.J. Beebe, G.A. Mensing, G.M. Walker, "Physics and applications of