MOLECULAR REPRODUCTION A N D DEVELOPMENT 33:81-88 (1992)
Subzonal Insemination of a Single Mouse Spermatozoon With a Personal Computer-Controlled Micromanipulation System KAZUHIKO KOBAYASHI,' KEISUKE KAT0,2 MASAHIKO SAGA,3 MASAMI YAMANE,2 CAPPY M. ROTHMAN,4 AND SHYOSO OGAWA' 'Laboratory of Animal Reproduction, Ikuta-Campus, Me@ University, Kawasaki, Japan; 2Laboratory of Machine Design, Waseda University, Tokyo, Japan; 3Department of Obstetrics and Gynecology, St. Marianna University School of Medicine, Kawasaki, Japan; 4Center for Reproductive Medicine, Century City Hospital, Los Angeles, California ABSTRACT A personal computer-controlled micromanipulation system was developed for automatic injection of spermatozoa into the perivitelline space of mouse ova. A pair of three-dimensional hydraulic micromanipulators driven by pulse motors was used for this automatic system. The pulse signals that regulate the motors are initiated by the computer program, and these signals cause the micromanipulator to move the microtool precisely.The computer program was designed to perform the most effective movements of the sperm injection needle used during manual micromanipulation.Prior to the manipulation, the computer locates the tip of the injection needle and the end of the egg-holding pipette in the microscope field using image processing.The trajectory of the injection needle is determined according to these initial positions. Using this robotic system, subzonal insemination with a single mouse spermatozoon was attempted in a total of 143 ova. The sperm insertion was successfully completed in all cases without damaging any of the ova. Spermatozoa treated with ionophore A23187 and those without the treatment were used. The fertilization rate (68.8%)of the ova inseminated with treated sperm was significantly higher than that (37.5%)obtained with the nontreated sperm (P < 0.05).These findings suggest the feasibility and potential for further applications of a robotic microinsemination system and, in addition, that a higher fertility rate in the subzonal insemination of mouse ova can be achieved with the ionophore treatment of spermatozoa. 0 1992 Wiley-Liss, Inc.
Key Words: Sperm injection, lonophore treatment, Computer-controlled microsurgery
INTRODUCTION The successful injection of spermatozoa underneath the zona pellucida of oocytes has been achieved, resulting in a human birth (Ng et al., 1990) and in live offspring in mice (Mann, 1988).In a previous paper (Kobayashi e t al., 1992), we described a micromanipulation technique that allows easy insertion of sperm into the perivitelline space of mouse oocytes.
0 1992 WILEY-LISS, INC.
This paper reports the development of a robotic computer-driven micromanipulation system for subzonal injection of live sperm. This system applies micromanipulation methods reported previously (Ogawa et al., 1986; Yamakawa et al., 1988; Ogawa and Yamakawa, 1988) for use with a n automated system. Our aims in automating the micromanipulation system were to improve its repeatability and thereby to prevent oocyte damage due to human error and to adapt it for use by inexperienced operators. While we demonstrate the efficiency of this automated system using mouse gametes, its features would be particularly advantageous in future application for assisted reproduction in humans. MATERIALS AND METHODS Preparation of Mouse Oocytes and Sperm Mature female mice (ICR strain) were superovulated with the injection of 5 IU pregnant mare serum gonadotropin (PMSG) and 5 IU human chorionic gonadotropin (hCG) at 48 h r intervals. The oocytes with cumulus cells were collected from their oviducts about 14 h r after the hCG injection. The cumulus cells were removed from the oocytes by treating them with hyaluronidase (150 IUlml). Sperm cells were obtained from the caudae epididymides of adult males of the same strain. The sperm were cultured in TYH medium (Toyoda et al., 1971) for about 1 h r at 37°C in a humidified atmosphere of 5% C 0 2 in air to complete capacitation. Some of the sperm samples were further treated with 1 p M of ionophore A23187 to induce the acrosome reaction [Lassalle and Testart, 19881. Prior to microinjection, the preincubated and the ionophore-treated sperm suspensions were diluted (1:4) in 2% methyl cellulose solution to slow sperm velocity and facilitate the capture of individual sperm into the injec-
Received January 27, 1992; accepted March 30,1992. Address reprint requests t o Kazuhiko Kobayashi, c/o Cappy M. Rothman, MD, Century City Medical Plaza, 2080 Century Park East, Suite 907, Los Angeles, CA 90067.
82
K. KOBAYASHI ET AL.
lon tubing filled with silicone oil. The tank moves up and down on rails through the power of the pulse motor. This system generates pressure by utilizing the difference in height between the surface of the silicone oil in the tank and the oil covering the medium droplet containing gametes on the microscope stage. When the oil tank is moved up so that the oil surface in the tank is higher than that on the microscope stage, the pressure within the pipette becomes positive. On the other hand, reversing these positions through lowering the tank produces negative pressure, and the oocyte is captured and held at the end of the pipette. The internal pressure of the sperm injection needle was regulated by a syringe system equipped with a pulse motor. The injection needle is linked to a glass syringe (3 ml in volume) fitted to the manipulator via Fig. 1. Micromanipulator for pressure control within the egg-holdpolyethylene tubing. They are all filled with air. The ing pipette. The egg-holding pipette (HP) is linked to the oil tank (OT) piston of the syringe is pushed or pulled by driving the of the manipulator by Teflon tubing (TT). All of them are filled with silicone oil (SO). The tank moves up and down on the rails (R) by the motor, and positive or negative pressure is generated driving of the pulse motor (PMI, which is connected via the belt iB). inside the injection needle. Difference of level between the surface of the oil in the tank and that of These four micromanipulators were arranged as folthe paraffin oil (PO) covering the medium drop (MD) on the microlows. A pair of three-dimensional hydraulic micromascope (MS) stage generates positive or negative pressure within the nipulators was installed on both sides of a n inverted holding pipette. When the oil tank is moved up so that the oil level in the tank ia) is higher than that on the microscope stage (b), internal microscope (IMT-2; Olympus Optical Co. Ltd., Tokyo, pressure of the holding pipette is positive. When a is lower than b, the Japan) equipped with differential interference contrast pressure is negative, which captures the ovum a t the pipette end. (Nomarski, IMT2-NIC; Olympus). A sperm injection When a and b are on the same plane, pressure within the pipette is needle and a n egg-holding pipette, which were conneutral nected to the corresponding pressure regulators, were attached to the right and left three-dimensional microtion needle (Mann, 1988). A small droplet of the sperm manipulators, respectively (Fig. 2a). suspension and one of Dulbecco’s phosphate-buffered saline supplemented with 18% calf serum (D-PBS) were Computer-Controlled Micromanipulation System The computer-controlled micromanipulation system placed within a well formed on a coverslip (43 x 50 mm) and covered with light-weight paraffin oil. After the used for the present study was made by improving the transfer of three or four oocytes into the D-PBS droplet, software of our previous model (Ogawa and Yathe coverslip was placed on the warmed microscope makawa, 1988). The control system is illustrated by a stage. These procedures for the preparation of mouse diagram in Figure 3. The automatic micromanipulagametes have been described in detail elsewhere tion is achieved by the following process. The microscope image is stored in a memory unit (frame memory, (Kobayashi et al., 1992). IMM-256V8-41; Keio Electric Co. Ltd., Tokyo, Japan) Preparation of Microtools as visual signals through a high-resolution video camThe microtools used here were prepared according to era ((22400; Hamamatsu Photonics Co. Ltd., Hamathe method presented in our previous report (Koba- matsu, Japan) attached to the microscope. A personal yashi et al., 1992). The sperm injection needle was bev- computer (PC-9801VX2; NEC, Tokyo, Japan) processes eled so that outer diameter a t the opening of its tip the visual data and locates the sperm injection needle measured -10 km. Outer and inner diameters a t the and the egg-holding pipette on the microscope field. end of the egg-holding pipette were -35 and -15 pm, Utilizing this information, the computer determines respectively. the movement of the injection needle for completion of the microsurgery and calculates the optimum quantity Micromanipulator of control signals (pulse signals) for each of the pulse A pair of three-dimensional hydraulic micromanipu- motors attached to the micromanipulator. Then, the lators (MO-102M; Narishige Scientific Inst. Lab., To- pulse signals are sent to the motor drive unit, and in kyo, Japan) was used for this system. A total of three turn the microtool attached to the micromanipulator is pulse motors were fitted to each micromanipulator, operated. providing movement in the X, Y, and Z axes. The following process describes the automatic detecTo hold the oocyte a t the end of the egg-holding pi- tion of the microtools on the microscope field by the pette, negative pressure was produced within the pi- image processing (Fig. 4a-d). Noise is removed by pette by a micromanipulator. Figure 1shows the struc- “smoothing processing” from the microscope image, ture of the manipulator. The holding pipette is which is fixed in the frame memory. The margins of the connected to the oil tank of the manipulator with Tef- objects on the image are enhanced by “differential pro-
AUTOMATIC SUBZONAL MICROINSEMINATION
Fig. 2. Computer-controlled micromanipulation system. a: MS, inverted microscope; TMP and TMN, three-dimensional hydraulic micromanipulators for the egg-holding pipette and the sperm injection needle; PX, PY, and PZ, pulse motors providing X, Y, and Z axis movements of the corresponding three-dimensional micromanipulators; PRP and PRN, pressure regulators for the egg-holding pipette and for the injection needle; VC, video camera. b VB, video booster; FM, frame memory unit; PC, personal computer; MD, motor drive unit; HD, hard disc drive; KB, computer keyboard; M, mouse; MO, MP, and MC, monitors for the original and the processed microscope image and for the computer.
Fig. 3. Diagram of the micromanipulation system controlled by a personal computer.
cessing,” and then only the margins are extracted by “binary processing” (Fig. 4b). The computer searches for the margin having the same shape as the pattern of the microtool (Fig. 4d), which was previously inputted. With this method, the computer recognizes the microtools on the microscope image (Fig. 4c). The movement of the sperm injection needle for the insertion of its tip into the perivitelline space (PVS)
83
was developed on the basis of the manual micromanipulation procedure for subzonal insemination as our previous report (Kobayashi et al., 1992) illustrated. For the computer to reproduce the needle movement easily, the path of the needle was slightly modified so that it became a sequence of segments approximating the original track. In this way, a sequence of needle movements consisting of eight steps was developed (Fig. 5). For each microinjection, the trajectory of the sperm injection needle was adjusted by the computer program according to the initial positions of the needle tip and the end of the egg-holding pipette. The micromanipulator was operated so that the attached microtool moved precisely along the determined trajectory. The micromanipulator that generates the internal pressure of the egg-holding pipette was also controlled automatically by the computer program during the microinjection. The pressure was applied in the same manner as in the manual sperm injection (see Fig. 5). Each pulse motor attached to micromanipulators can also be controlled through the program for computer keyboard operation. With this program, all the micromanipulators can be used as manual manipulators by pressing the designated key. Aspiration of a sperm into the injection needle and capture of the ovum to be injected at the end of the egg-holding pipette were performed by keyboard operation.
Sperm Injection Into the Perivitelline Space of the Oocyte A normal-appearing sperm with good motility was chosen from the droplet of sperm suspension and gently aspirated tail first into the injection needle. The needle was moved into the droplet containing oocytes. A mature oocyte, which had the first polar body, was selected and held at the end of the egg-holding pipette by suction. Then, the positions of the microtools were adjusted so that their tips appear within the monitor screen displaying the microscope image. Following these manipulations, using computer keyboard operation, automatic subzonal sperm insertion was carried out (see Fig. 5). In some cases, a special manipulation of the sperm injection needle was employed within the PVS to promote the adhesion of the sperm head onto the ooplasmic membrane. The oblique opening of the injection needle is held on the ooplasmic membrane by keyboard operation (Fig. 6a). The sperm is then pushed forward with positive pressure so that the sperm head is forced to adhere to the membrane (Fig. 6b). A narrow space between the needle opening and the ooplasmic membrane is made, releasing the sperm tail (Fig. 6c). Then, the injection needle is gently withdrawn from the PVS (Fig. 6d).
Assay for Fertilization of the Inseminated Ova The ova injected with spermatozoa were washed in fresh TYH medium and incubated for 6 h r in the same
84
K. KOBAYASHI ET AL.
Fig. 4. Automatic detection of the microtool tips by image processing. a: Prior to image processing, the microtools are positioned so that their tips appear within the respective squares superimposed on the microscope image, which is displayed on the monitor screen. b The image is processed by the computer and only the margins ofthe objects are extracted (HP, egg-holding pipette; IN, sperm injection needle). c:
The computer locates the margin that has the same contour as each microtool within the square and marks the positions of their tips with a cross on the displayed image (arrowheads). d The computer recognizes the microtools on the image using the patterns of the egg-holding pipette (HP) and the injection needle (IN) that were previously inputted. X 300.
medium. After culturing, the ova were observed under the inverted microscope to assess fertilization. Ova with two pronuclei and the second polar body were deemed to be fertilized.
Statistical Analysis
Experimental Design Automated subzonal insemination was attempted using the computer-controlled micromanipulation system. The incidence of successful sperm insertion in the automatic system was examined. For these microinseminations, a single sperm obtained from either the preincubated or the ionophoretreated population was used. In combination with some of the injections, the sperm-adhesion treatment was employed. Thus four types of insemination classified according to the application of the ionophore and adhesion treatment were tested. From the results of these insemination groups, the effects of the ionophore treatment and that of the sperm-adhesion treatment on the fertilization rate were also evaluated.
The significance of difference between experimental groups was determined by x2 test.
Fig. 5. Stepwise illustration of the coniputer-controlled subzonal insemination procedure. a: An injection needle containing a spermatozoon (arrowhead) and an ovum to be injected. b: The sperm injection needle is moved so as to press the “12 o’clock” side of the zona pellucida. As a result, only the zona is held between the needle tip and the end of the holding pipette. c: The zona pellucida is pinned against the pipette end by the needle. Then the negative pressure within the holding pipette is reduced to prevent the pipette from pulling the sperm from the needle and/or the ooplasm through the punctured site in the zona pellucida during the subsequent needle insertion into the lumen of the pipette through the zona. d: The injection needle is further advanced, sticking the zona pellucida without touching the ooplasmic membrane. e: The injection needle is drawn back while still piercing the zona pellucida. f: The ovum is shifted so that another portion of the zona can be captured by the holding pipette. g:The ovum is sucked and held by the pipette again. h The injection needle is slightly withdrawn, introducing the t i p into the perivitelline space. The state of the internal pressure of the holding pipette at each step is indicated by the direction of arrows as follows: t, negative pressure; +,positive pressure; -, neutral. x300.
AUTOMATIC SUBZONAL MICROINSEMINATION
Fig. 5.
85
86
K. KOBAYASHI ET AL.
Fig. 6 . Manipulation assisting adhesion of the sperm head onto the ooplasmic membrane. a: After insertion of the sperm injection needle into the perivitelline space, the oblique orifice of the needle is held on the ooplasmic membrane. b: The sperm head is pressed against the
membrane, with positive pressure within the needle. c: The sperm tail is then released by making a small space between the orifice and the ooplasmic membrane. d The sperm adhering to the ooplasmic membrane with motility. Arrowheads, sperm head. ~ 3 0 0 .
TABLE 1. Results of the Subzonal Insemination With the Computer-ControlledMicromanipulation System
TABLE 2. Fertilization by Microinsemination Underneath the Zona Pellucida Using the Computer-ControlledMicromanipulation System
No. of ova
No. of o v a with operation
tested
comoleted (%)
143
143 (100)
No. of ova damaged (%b) 0 (0)
Sperm injected Ionophoretreated
Preincubated
RESULTS AND DISCUSSION In the present study, all the sperm microinjections using the computer-controlled system were carried out by students with no previous experience in micromanipulation. Subzonal insemination controlled by a personal computer was attempted in a total of 143 ova. The insertion of a sperm under the zona pellucida (ZP) was successfully completed in all cases. No sign of cell damage was observed in any of these ova after the micromanipulation (Table 1). It took -2 min to perform the subzonal insemination using the robotic system (from the detection of the microtools to the release of sperm under the ZP). Of the 143 ova used for this experiment, 72 were injected with the preincubated sperm and 71 with those treated with ionophore. As is shown in Table 2, of the ova inseminated with the preincubated sperm, 40 re-
No. of ova
Manipulation
assisted s p e r m adhesiona
No. of ova tested
+ +-
32 39 40 32
fertilized (70)
22 (68.8Ib 14 (35.9)" 15 137.5)' 9 (28.1)"
"Insemination was performed i n combination w i t h (+) o r w i t h o u t (-) the sperm-adhesion t r e a t m e n t . bJValues w i t h different superscripts are significantly differ-
ent (P< 0.05).
ceived the sperm-adhesion treatment and 32 did not. Fertilization was found in 15 (37.5%)and nine (28.18) of these ova, respectively. These two fertility rates did not differ significantly. In the insemination with the ionophore-treated sperm, 32 ova received the sperm-adhesion treatment and the remaining 39 did not, of which 22 (68.8%)and 14 (35.981, respectively, were judged to be fertilized. There was a significant difference between these fertility rates ( P < 0.05). The rate of fertilization using the
AUTOMATIC SUBZONAL MICROINSEMINATION TABLE 3. Adherence of the Spermatozoa1 Head Onto the Ooplasmic Membrane and Fertilization of the Ova After the Sperm-Adhesion Treatment
Sperm injected Ionophoretreated Preincubated
No. of ova tested (A) 32
40
No. of ova showing sperm adhesion (+) and no adhesion (-1 (B) (B/A%)
+ +
-
26 (81.3)" 6 (18.7)
-
18 (4E1.0)~ 22 (55.0)
No. of ova fertilized (C) (C/B%) 20 (76.9)" 2 (33.3)d 10 (55.6)" 5 (22.7)'
a-fSignificant differences were found between a and b (P < 0.005),c and d (P< 0.051, and e and f (P < 0.05).
ionophore-treated sperm in combination with the adhesion treatment was significantly higher ( P < 0.05) than either rate obtained by insemination with only preincubated sperm. Thus the insemination with the ionophore-treated sperm using the adhesion treatment yielded a distinctly higher rate of fertilization than any of the other groups. These results indicate that, in the subzonal insemination of mouse ova, a n increase in the fertility rate can be achieved not only with the ionophore treatment of sperm but also with the application of the manipulation, which assists sperm adhesion to the ooplasmic membrane. It can also be noted that the data on the fertilization events obtained with the robotic sperm insertion are in agreement with the results of our previous report on the manual subzonal insemination of mouse ova (Kobayashi e t al., 1992). From these findings, it is suggested that the computer-controlled system enables the operator inexperienced in micromanipulation to perform the subzonal insemination with skill comparable to that of the expert, without the fatigue encountered when manually injecting sperm under a microscope, and t h a t a fertilization rate equal to that obtained with manual insemination can be achieved with the robotic micromanipulation. Detailed data on the adherence of the sperm heads to the ooplasmic membrane after the sperm-adhesion treatment and fertilization of the inseminated ova are summarized in Table 3. Sperm adherence to the ooplasmic membrane was observed in 26 (81.3%) of 32 ova injected with the ionophore-treated sperm and 18 (45.0%)of 40 ova with the preincubated sperm, of which 20 (76.9%) and 10 (55.6%), respectively, were found to be fertilized. In cases when there was no sperm adhesion, two (33.3%)of six ova injected with ionophoretreated sperm and five (22.7%) of 22 ova with those preincubated were recorded as being fertilized after culturing for 6 hr. Irrespective of the ionophore treatment, sperm that successfully adhered to the ooplasmic membrane fertilized ova a t a significantly higher rate (P < 0.05) than those that did not adhere. A significant increase in sperm adherence was obtained when the ionophore treatment was used (P < 0.005), resulting in a n apparently higher fertilization rate compared with insemina-
87
tion without the treatment. These results suggest that the adherence of the sperm head onto the ooplasmic membrane after the adhesion treatment might indicate the potential for fertilization and, in addition, that sperm fertility with the subzonal insemination is enhanced by the ionophore treatment. The above data on the relationship between sperm adhesion and fertilization of inseminated ova are consistent with data obtained in our previous experiment on the microinsemination in mice ova (Kobayashi et al., 1992). In this series of experiments, we noted some difficulty during aspiration of mouse sperm into the injection needle. When a high negative pressure was rapidly applied within the injection needle, it temporarily caught the sperm head at the opening of the needle, and the sperm became immotile immediately. Fertilization was not induced by injection of such sperm under the Z P of the ovum (data not shown). For these reasons, the application of negative pressure to the sperm injection needle was carefully controlled to allow smooth passage of the sperm head through the needle opening: Rapid suction is applied only until the tail is captured, and then the sperm head is gradually pulled into the needle by slight suction. One reason for the success of this computer-controlled system is the application of a simple and effective micromanipulation procedure that we previously developed for manual use (Kobayashi e t al., 1992). Due to the high elasticity of the mouse ZP, the tip of the sperm injection needle cannot easily pierce the ZP even in the manual micromanipulation unless our method is employed. The Z P is apt to be depressed and distorted by the advance of the injection needle instead of being pierced. It has been frequently experienced in the computer-controlled subzonal insemination without our manipulation method that the ovum is shifted as a consequence of the ZP distortion during the micromanipulation. In those cases, penetration by the sperm injection needle was not achieved because the needle precisely followed the trajectory determined according to the initial position of the ovum. In our microinsemination procedure, the sperm injection needle is first moved so a s to pin the Z P against the end of the egg-holding pipette (Fig. 5a-c) and then pierces with the advance into the lumen of the eggholding pipette (Fig. 5c,d). With this method, the injection needle movement determined by the computer is always effective because it is not hindered by the elasticity of the ZP. As a result, the insertion of the needle tip into the PVS of the mouse ova is performed accurately with the computer-controlled micromanipulation system. This automatic manipulation system is advantageous in that i t allows accurate detection of the positions of the sperm injection needle and the egg-holding pipette by image processing. Hence the computer can determine the optimum movement of the microtool for successful injection. On the other hand, disadvantages lie in the fact that it requires the adjustment of the computer program that controls the micromanipulator
88
K. KOBAYASHI ET AL.
€or use in species other than mice, because the determination of the microtool movement by image processing is based on the assumption t h a t the size of the ovum is fixed and additionally in that image processing for the location of the microtools takes a considerable time (-20 sec). The results of this study indicate the feasibility of and potential for further application of robotic microinsemination. The development of software that serves various micromanipulations and more sophisticated methods of image processing will facilitate wide applicability of the complicated technique of micromanipulation.
ACKNOWLEDGMENTS A part of this work was supported by Research Funds from IST, Meiji University. REFERENCES Kobayashi K, Okuyama M, Fujimoto G , Rothman CM, Hill DL, Ogawa S (1992):Subzonal insemination with a single spermatozoon
using manipulation assisted sperm adhesion onto the ooplasmic membrane in mouse ova. Mol Reprod Dev 31:223-229. Lassalle B, Testart J (1988):Human sperm injection into the perivitelline space (SI-PVS) of hamster oocytes: Effect of sperm pretreatment by calcium-ionophore A23187 and freezing-thawing on the penetration rate and polyspermy. Gamete Res 20:301-311. Mann JR (1988):Full term development of mouse eggs fertilized by a spermatozoon microinjected under the zona pellucida. Biol Reprod 38:1077-1083. Ng SC, Bongso TA, Ratnam SS, Sathananthan H (1990):M.1.S.T.-A technique of sub-zonal sperm transfer. In: “Proceedings of VII World Congress on Human Reproduction, Helsinki.” Abstract 256. Ogawa S, Takahashi H, Mizuno J , Kashiwazaki N, Yamane M, Narishige E (1986): Personal computer-controlled microsurgery of fertilized eggs and early embryos. Theriogenology 25335-345. Ogawa S, Yamakawa H (1988):Micromanipulation system controlled by personal computer for microsurgery of ova and early embryos. In: “Synip Biotech Anim Breed Jpn-Ger Center.”Berlin, pp 67-83. Toyoda Y, Yokoyama M, Hoshi T (1971):Studies on the fertilization of mouse eggs in vitro. I. In-vitro fertilization of eggs by fresh epididyma1 sperm. Jpn J Anim Reprod 16:147-151. Yamakawa H, Yamanoi J, Sunouchi H, Yamane M, Narishige E, Ogawa S (1988):An improved micromanipulation system controlled by personal computer for early embryos. Jpn J Anim Reprod 34%19.
Subzonal Insemination of a Single Mouse Spermatozoon With a Personal Computer-Controlled Micromanipulation System KAZUHIKO KOBAYASHI,' KEISUKE KAT0,2 MASAHIKO SAGA,3 MASAMI YAMANE,2 CAPPY M. ROTHMAN,4 AND SHYOSO OGAWA' 'Laboratory of Animal Reproduction, Ikuta-Campus, Me@ University, Kawasaki, Japan; 2Laboratory of Machine Design, Waseda University, Tokyo, Japan; 3Department of Obstetrics and Gynecology, St. Marianna University School of Medicine, Kawasaki, Japan; 4Center for Reproductive Medicine, Century City Hospital, Los Angeles, California ABSTRACT A personal computer-controlled micromanipulation system was developed for automatic injection of spermatozoa into the perivitelline space of mouse ova. A pair of three-dimensional hydraulic micromanipulators driven by pulse motors was used for this automatic system. The pulse signals that regulate the motors are initiated by the computer program, and these signals cause the micromanipulator to move the microtool precisely.The computer program was designed to perform the most effective movements of the sperm injection needle used during manual micromanipulation.Prior to the manipulation, the computer locates the tip of the injection needle and the end of the egg-holding pipette in the microscope field using image processing.The trajectory of the injection needle is determined according to these initial positions. Using this robotic system, subzonal insemination with a single mouse spermatozoon was attempted in a total of 143 ova. The sperm insertion was successfully completed in all cases without damaging any of the ova. Spermatozoa treated with ionophore A23187 and those without the treatment were used. The fertilization rate (68.8%)of the ova inseminated with treated sperm was significantly higher than that (37.5%)obtained with the nontreated sperm (P < 0.05).These findings suggest the feasibility and potential for further applications of a robotic microinsemination system and, in addition, that a higher fertility rate in the subzonal insemination of mouse ova can be achieved with the ionophore treatment of spermatozoa. 0 1992 Wiley-Liss, Inc.
Key Words: Sperm injection, lonophore treatment, Computer-controlled microsurgery
INTRODUCTION The successful injection of spermatozoa underneath the zona pellucida of oocytes has been achieved, resulting in a human birth (Ng et al., 1990) and in live offspring in mice (Mann, 1988).In a previous paper (Kobayashi e t al., 1992), we described a micromanipulation technique that allows easy insertion of sperm into the perivitelline space of mouse oocytes.
0 1992 WILEY-LISS, INC.
This paper reports the development of a robotic computer-driven micromanipulation system for subzonal injection of live sperm. This system applies micromanipulation methods reported previously (Ogawa et al., 1986; Yamakawa et al., 1988; Ogawa and Yamakawa, 1988) for use with a n automated system. Our aims in automating the micromanipulation system were to improve its repeatability and thereby to prevent oocyte damage due to human error and to adapt it for use by inexperienced operators. While we demonstrate the efficiency of this automated system using mouse gametes, its features would be particularly advantageous in future application for assisted reproduction in humans. MATERIALS AND METHODS Preparation of Mouse Oocytes and Sperm Mature female mice (ICR strain) were superovulated with the injection of 5 IU pregnant mare serum gonadotropin (PMSG) and 5 IU human chorionic gonadotropin (hCG) at 48 h r intervals. The oocytes with cumulus cells were collected from their oviducts about 14 h r after the hCG injection. The cumulus cells were removed from the oocytes by treating them with hyaluronidase (150 IUlml). Sperm cells were obtained from the caudae epididymides of adult males of the same strain. The sperm were cultured in TYH medium (Toyoda et al., 1971) for about 1 h r at 37°C in a humidified atmosphere of 5% C 0 2 in air to complete capacitation. Some of the sperm samples were further treated with 1 p M of ionophore A23187 to induce the acrosome reaction [Lassalle and Testart, 19881. Prior to microinjection, the preincubated and the ionophore-treated sperm suspensions were diluted (1:4) in 2% methyl cellulose solution to slow sperm velocity and facilitate the capture of individual sperm into the injec-
Received January 27, 1992; accepted March 30,1992. Address reprint requests t o Kazuhiko Kobayashi, c/o Cappy M. Rothman, MD, Century City Medical Plaza, 2080 Century Park East, Suite 907, Los Angeles, CA 90067.
82
K. KOBAYASHI ET AL.
lon tubing filled with silicone oil. The tank moves up and down on rails through the power of the pulse motor. This system generates pressure by utilizing the difference in height between the surface of the silicone oil in the tank and the oil covering the medium droplet containing gametes on the microscope stage. When the oil tank is moved up so that the oil surface in the tank is higher than that on the microscope stage, the pressure within the pipette becomes positive. On the other hand, reversing these positions through lowering the tank produces negative pressure, and the oocyte is captured and held at the end of the pipette. The internal pressure of the sperm injection needle was regulated by a syringe system equipped with a pulse motor. The injection needle is linked to a glass syringe (3 ml in volume) fitted to the manipulator via Fig. 1. Micromanipulator for pressure control within the egg-holdpolyethylene tubing. They are all filled with air. The ing pipette. The egg-holding pipette (HP) is linked to the oil tank (OT) piston of the syringe is pushed or pulled by driving the of the manipulator by Teflon tubing (TT). All of them are filled with silicone oil (SO). The tank moves up and down on the rails (R) by the motor, and positive or negative pressure is generated driving of the pulse motor (PMI, which is connected via the belt iB). inside the injection needle. Difference of level between the surface of the oil in the tank and that of These four micromanipulators were arranged as folthe paraffin oil (PO) covering the medium drop (MD) on the microlows. A pair of three-dimensional hydraulic micromascope (MS) stage generates positive or negative pressure within the nipulators was installed on both sides of a n inverted holding pipette. When the oil tank is moved up so that the oil level in the tank ia) is higher than that on the microscope stage (b), internal microscope (IMT-2; Olympus Optical Co. Ltd., Tokyo, pressure of the holding pipette is positive. When a is lower than b, the Japan) equipped with differential interference contrast pressure is negative, which captures the ovum a t the pipette end. (Nomarski, IMT2-NIC; Olympus). A sperm injection When a and b are on the same plane, pressure within the pipette is needle and a n egg-holding pipette, which were conneutral nected to the corresponding pressure regulators, were attached to the right and left three-dimensional microtion needle (Mann, 1988). A small droplet of the sperm manipulators, respectively (Fig. 2a). suspension and one of Dulbecco’s phosphate-buffered saline supplemented with 18% calf serum (D-PBS) were Computer-Controlled Micromanipulation System The computer-controlled micromanipulation system placed within a well formed on a coverslip (43 x 50 mm) and covered with light-weight paraffin oil. After the used for the present study was made by improving the transfer of three or four oocytes into the D-PBS droplet, software of our previous model (Ogawa and Yathe coverslip was placed on the warmed microscope makawa, 1988). The control system is illustrated by a stage. These procedures for the preparation of mouse diagram in Figure 3. The automatic micromanipulagametes have been described in detail elsewhere tion is achieved by the following process. The microscope image is stored in a memory unit (frame memory, (Kobayashi et al., 1992). IMM-256V8-41; Keio Electric Co. Ltd., Tokyo, Japan) Preparation of Microtools as visual signals through a high-resolution video camThe microtools used here were prepared according to era ((22400; Hamamatsu Photonics Co. Ltd., Hamathe method presented in our previous report (Koba- matsu, Japan) attached to the microscope. A personal yashi et al., 1992). The sperm injection needle was bev- computer (PC-9801VX2; NEC, Tokyo, Japan) processes eled so that outer diameter a t the opening of its tip the visual data and locates the sperm injection needle measured -10 km. Outer and inner diameters a t the and the egg-holding pipette on the microscope field. end of the egg-holding pipette were -35 and -15 pm, Utilizing this information, the computer determines respectively. the movement of the injection needle for completion of the microsurgery and calculates the optimum quantity Micromanipulator of control signals (pulse signals) for each of the pulse A pair of three-dimensional hydraulic micromanipu- motors attached to the micromanipulator. Then, the lators (MO-102M; Narishige Scientific Inst. Lab., To- pulse signals are sent to the motor drive unit, and in kyo, Japan) was used for this system. A total of three turn the microtool attached to the micromanipulator is pulse motors were fitted to each micromanipulator, operated. providing movement in the X, Y, and Z axes. The following process describes the automatic detecTo hold the oocyte a t the end of the egg-holding pi- tion of the microtools on the microscope field by the pette, negative pressure was produced within the pi- image processing (Fig. 4a-d). Noise is removed by pette by a micromanipulator. Figure 1shows the struc- “smoothing processing” from the microscope image, ture of the manipulator. The holding pipette is which is fixed in the frame memory. The margins of the connected to the oil tank of the manipulator with Tef- objects on the image are enhanced by “differential pro-
AUTOMATIC SUBZONAL MICROINSEMINATION
Fig. 2. Computer-controlled micromanipulation system. a: MS, inverted microscope; TMP and TMN, three-dimensional hydraulic micromanipulators for the egg-holding pipette and the sperm injection needle; PX, PY, and PZ, pulse motors providing X, Y, and Z axis movements of the corresponding three-dimensional micromanipulators; PRP and PRN, pressure regulators for the egg-holding pipette and for the injection needle; VC, video camera. b VB, video booster; FM, frame memory unit; PC, personal computer; MD, motor drive unit; HD, hard disc drive; KB, computer keyboard; M, mouse; MO, MP, and MC, monitors for the original and the processed microscope image and for the computer.
Fig. 3. Diagram of the micromanipulation system controlled by a personal computer.
cessing,” and then only the margins are extracted by “binary processing” (Fig. 4b). The computer searches for the margin having the same shape as the pattern of the microtool (Fig. 4d), which was previously inputted. With this method, the computer recognizes the microtools on the microscope image (Fig. 4c). The movement of the sperm injection needle for the insertion of its tip into the perivitelline space (PVS)
83
was developed on the basis of the manual micromanipulation procedure for subzonal insemination as our previous report (Kobayashi et al., 1992) illustrated. For the computer to reproduce the needle movement easily, the path of the needle was slightly modified so that it became a sequence of segments approximating the original track. In this way, a sequence of needle movements consisting of eight steps was developed (Fig. 5). For each microinjection, the trajectory of the sperm injection needle was adjusted by the computer program according to the initial positions of the needle tip and the end of the egg-holding pipette. The micromanipulator was operated so that the attached microtool moved precisely along the determined trajectory. The micromanipulator that generates the internal pressure of the egg-holding pipette was also controlled automatically by the computer program during the microinjection. The pressure was applied in the same manner as in the manual sperm injection (see Fig. 5). Each pulse motor attached to micromanipulators can also be controlled through the program for computer keyboard operation. With this program, all the micromanipulators can be used as manual manipulators by pressing the designated key. Aspiration of a sperm into the injection needle and capture of the ovum to be injected at the end of the egg-holding pipette were performed by keyboard operation.
Sperm Injection Into the Perivitelline Space of the Oocyte A normal-appearing sperm with good motility was chosen from the droplet of sperm suspension and gently aspirated tail first into the injection needle. The needle was moved into the droplet containing oocytes. A mature oocyte, which had the first polar body, was selected and held at the end of the egg-holding pipette by suction. Then, the positions of the microtools were adjusted so that their tips appear within the monitor screen displaying the microscope image. Following these manipulations, using computer keyboard operation, automatic subzonal sperm insertion was carried out (see Fig. 5). In some cases, a special manipulation of the sperm injection needle was employed within the PVS to promote the adhesion of the sperm head onto the ooplasmic membrane. The oblique opening of the injection needle is held on the ooplasmic membrane by keyboard operation (Fig. 6a). The sperm is then pushed forward with positive pressure so that the sperm head is forced to adhere to the membrane (Fig. 6b). A narrow space between the needle opening and the ooplasmic membrane is made, releasing the sperm tail (Fig. 6c). Then, the injection needle is gently withdrawn from the PVS (Fig. 6d).
Assay for Fertilization of the Inseminated Ova The ova injected with spermatozoa were washed in fresh TYH medium and incubated for 6 h r in the same
84
K. KOBAYASHI ET AL.
Fig. 4. Automatic detection of the microtool tips by image processing. a: Prior to image processing, the microtools are positioned so that their tips appear within the respective squares superimposed on the microscope image, which is displayed on the monitor screen. b The image is processed by the computer and only the margins ofthe objects are extracted (HP, egg-holding pipette; IN, sperm injection needle). c:
The computer locates the margin that has the same contour as each microtool within the square and marks the positions of their tips with a cross on the displayed image (arrowheads). d The computer recognizes the microtools on the image using the patterns of the egg-holding pipette (HP) and the injection needle (IN) that were previously inputted. X 300.
medium. After culturing, the ova were observed under the inverted microscope to assess fertilization. Ova with two pronuclei and the second polar body were deemed to be fertilized.
Statistical Analysis
Experimental Design Automated subzonal insemination was attempted using the computer-controlled micromanipulation system. The incidence of successful sperm insertion in the automatic system was examined. For these microinseminations, a single sperm obtained from either the preincubated or the ionophoretreated population was used. In combination with some of the injections, the sperm-adhesion treatment was employed. Thus four types of insemination classified according to the application of the ionophore and adhesion treatment were tested. From the results of these insemination groups, the effects of the ionophore treatment and that of the sperm-adhesion treatment on the fertilization rate were also evaluated.
The significance of difference between experimental groups was determined by x2 test.
Fig. 5. Stepwise illustration of the coniputer-controlled subzonal insemination procedure. a: An injection needle containing a spermatozoon (arrowhead) and an ovum to be injected. b: The sperm injection needle is moved so as to press the “12 o’clock” side of the zona pellucida. As a result, only the zona is held between the needle tip and the end of the holding pipette. c: The zona pellucida is pinned against the pipette end by the needle. Then the negative pressure within the holding pipette is reduced to prevent the pipette from pulling the sperm from the needle and/or the ooplasm through the punctured site in the zona pellucida during the subsequent needle insertion into the lumen of the pipette through the zona. d: The injection needle is further advanced, sticking the zona pellucida without touching the ooplasmic membrane. e: The injection needle is drawn back while still piercing the zona pellucida. f: The ovum is shifted so that another portion of the zona can be captured by the holding pipette. g:The ovum is sucked and held by the pipette again. h The injection needle is slightly withdrawn, introducing the t i p into the perivitelline space. The state of the internal pressure of the holding pipette at each step is indicated by the direction of arrows as follows: t, negative pressure; +,positive pressure; -, neutral. x300.
AUTOMATIC SUBZONAL MICROINSEMINATION
Fig. 5.
85
86
K. KOBAYASHI ET AL.
Fig. 6 . Manipulation assisting adhesion of the sperm head onto the ooplasmic membrane. a: After insertion of the sperm injection needle into the perivitelline space, the oblique orifice of the needle is held on the ooplasmic membrane. b: The sperm head is pressed against the
membrane, with positive pressure within the needle. c: The sperm tail is then released by making a small space between the orifice and the ooplasmic membrane. d The sperm adhering to the ooplasmic membrane with motility. Arrowheads, sperm head. ~ 3 0 0 .
TABLE 1. Results of the Subzonal Insemination With the Computer-ControlledMicromanipulation System
TABLE 2. Fertilization by Microinsemination Underneath the Zona Pellucida Using the Computer-ControlledMicromanipulation System
No. of ova
No. of o v a with operation
tested
comoleted (%)
143
143 (100)
No. of ova damaged (%b) 0 (0)
Sperm injected Ionophoretreated
Preincubated
RESULTS AND DISCUSSION In the present study, all the sperm microinjections using the computer-controlled system were carried out by students with no previous experience in micromanipulation. Subzonal insemination controlled by a personal computer was attempted in a total of 143 ova. The insertion of a sperm under the zona pellucida (ZP) was successfully completed in all cases. No sign of cell damage was observed in any of these ova after the micromanipulation (Table 1). It took -2 min to perform the subzonal insemination using the robotic system (from the detection of the microtools to the release of sperm under the ZP). Of the 143 ova used for this experiment, 72 were injected with the preincubated sperm and 71 with those treated with ionophore. As is shown in Table 2, of the ova inseminated with the preincubated sperm, 40 re-
No. of ova
Manipulation
assisted s p e r m adhesiona
No. of ova tested
+ +-
32 39 40 32
fertilized (70)
22 (68.8Ib 14 (35.9)" 15 137.5)' 9 (28.1)"
"Insemination was performed i n combination w i t h (+) o r w i t h o u t (-) the sperm-adhesion t r e a t m e n t . bJValues w i t h different superscripts are significantly differ-
ent (P< 0.05).
ceived the sperm-adhesion treatment and 32 did not. Fertilization was found in 15 (37.5%)and nine (28.18) of these ova, respectively. These two fertility rates did not differ significantly. In the insemination with the ionophore-treated sperm, 32 ova received the sperm-adhesion treatment and the remaining 39 did not, of which 22 (68.8%)and 14 (35.981, respectively, were judged to be fertilized. There was a significant difference between these fertility rates ( P < 0.05). The rate of fertilization using the
AUTOMATIC SUBZONAL MICROINSEMINATION TABLE 3. Adherence of the Spermatozoa1 Head Onto the Ooplasmic Membrane and Fertilization of the Ova After the Sperm-Adhesion Treatment
Sperm injected Ionophoretreated Preincubated
No. of ova tested (A) 32
40
No. of ova showing sperm adhesion (+) and no adhesion (-1 (B) (B/A%)
+ +
-
26 (81.3)" 6 (18.7)
-
18 (4E1.0)~ 22 (55.0)
No. of ova fertilized (C) (C/B%) 20 (76.9)" 2 (33.3)d 10 (55.6)" 5 (22.7)'
a-fSignificant differences were found between a and b (P < 0.005),c and d (P< 0.051, and e and f (P < 0.05).
ionophore-treated sperm in combination with the adhesion treatment was significantly higher ( P < 0.05) than either rate obtained by insemination with only preincubated sperm. Thus the insemination with the ionophore-treated sperm using the adhesion treatment yielded a distinctly higher rate of fertilization than any of the other groups. These results indicate that, in the subzonal insemination of mouse ova, a n increase in the fertility rate can be achieved not only with the ionophore treatment of sperm but also with the application of the manipulation, which assists sperm adhesion to the ooplasmic membrane. It can also be noted that the data on the fertilization events obtained with the robotic sperm insertion are in agreement with the results of our previous report on the manual subzonal insemination of mouse ova (Kobayashi e t al., 1992). From these findings, it is suggested that the computer-controlled system enables the operator inexperienced in micromanipulation to perform the subzonal insemination with skill comparable to that of the expert, without the fatigue encountered when manually injecting sperm under a microscope, and t h a t a fertilization rate equal to that obtained with manual insemination can be achieved with the robotic micromanipulation. Detailed data on the adherence of the sperm heads to the ooplasmic membrane after the sperm-adhesion treatment and fertilization of the inseminated ova are summarized in Table 3. Sperm adherence to the ooplasmic membrane was observed in 26 (81.3%) of 32 ova injected with the ionophore-treated sperm and 18 (45.0%)of 40 ova with the preincubated sperm, of which 20 (76.9%) and 10 (55.6%), respectively, were found to be fertilized. In cases when there was no sperm adhesion, two (33.3%)of six ova injected with ionophoretreated sperm and five (22.7%) of 22 ova with those preincubated were recorded as being fertilized after culturing for 6 hr. Irrespective of the ionophore treatment, sperm that successfully adhered to the ooplasmic membrane fertilized ova a t a significantly higher rate (P < 0.05) than those that did not adhere. A significant increase in sperm adherence was obtained when the ionophore treatment was used (P < 0.005), resulting in a n apparently higher fertilization rate compared with insemina-
87
tion without the treatment. These results suggest that the adherence of the sperm head onto the ooplasmic membrane after the adhesion treatment might indicate the potential for fertilization and, in addition, that sperm fertility with the subzonal insemination is enhanced by the ionophore treatment. The above data on the relationship between sperm adhesion and fertilization of inseminated ova are consistent with data obtained in our previous experiment on the microinsemination in mice ova (Kobayashi et al., 1992). In this series of experiments, we noted some difficulty during aspiration of mouse sperm into the injection needle. When a high negative pressure was rapidly applied within the injection needle, it temporarily caught the sperm head at the opening of the needle, and the sperm became immotile immediately. Fertilization was not induced by injection of such sperm under the Z P of the ovum (data not shown). For these reasons, the application of negative pressure to the sperm injection needle was carefully controlled to allow smooth passage of the sperm head through the needle opening: Rapid suction is applied only until the tail is captured, and then the sperm head is gradually pulled into the needle by slight suction. One reason for the success of this computer-controlled system is the application of a simple and effective micromanipulation procedure that we previously developed for manual use (Kobayashi e t al., 1992). Due to the high elasticity of the mouse ZP, the tip of the sperm injection needle cannot easily pierce the ZP even in the manual micromanipulation unless our method is employed. The Z P is apt to be depressed and distorted by the advance of the injection needle instead of being pierced. It has been frequently experienced in the computer-controlled subzonal insemination without our manipulation method that the ovum is shifted as a consequence of the ZP distortion during the micromanipulation. In those cases, penetration by the sperm injection needle was not achieved because the needle precisely followed the trajectory determined according to the initial position of the ovum. In our microinsemination procedure, the sperm injection needle is first moved so a s to pin the Z P against the end of the egg-holding pipette (Fig. 5a-c) and then pierces with the advance into the lumen of the eggholding pipette (Fig. 5c,d). With this method, the injection needle movement determined by the computer is always effective because it is not hindered by the elasticity of the ZP. As a result, the insertion of the needle tip into the PVS of the mouse ova is performed accurately with the computer-controlled micromanipulation system. This automatic manipulation system is advantageous in that i t allows accurate detection of the positions of the sperm injection needle and the egg-holding pipette by image processing. Hence the computer can determine the optimum movement of the microtool for successful injection. On the other hand, disadvantages lie in the fact that it requires the adjustment of the computer program that controls the micromanipulator
88
K. KOBAYASHI ET AL.
€or use in species other than mice, because the determination of the microtool movement by image processing is based on the assumption t h a t the size of the ovum is fixed and additionally in that image processing for the location of the microtools takes a considerable time (-20 sec). The results of this study indicate the feasibility of and potential for further application of robotic microinsemination. The development of software that serves various micromanipulations and more sophisticated methods of image processing will facilitate wide applicability of the complicated technique of micromanipulation.
ACKNOWLEDGMENTS A part of this work was supported by Research Funds from IST, Meiji University. REFERENCES Kobayashi K, Okuyama M, Fujimoto G , Rothman CM, Hill DL, Ogawa S (1992):Subzonal insemination with a single spermatozoon
using manipulation assisted sperm adhesion onto the ooplasmic membrane in mouse ova. Mol Reprod Dev 31:223-229. Lassalle B, Testart J (1988):Human sperm injection into the perivitelline space (SI-PVS) of hamster oocytes: Effect of sperm pretreatment by calcium-ionophore A23187 and freezing-thawing on the penetration rate and polyspermy. Gamete Res 20:301-311. Mann JR (1988):Full term development of mouse eggs fertilized by a spermatozoon microinjected under the zona pellucida. Biol Reprod 38:1077-1083. Ng SC, Bongso TA, Ratnam SS, Sathananthan H (1990):M.1.S.T.-A technique of sub-zonal sperm transfer. In: “Proceedings of VII World Congress on Human Reproduction, Helsinki.” Abstract 256. Ogawa S, Takahashi H, Mizuno J , Kashiwazaki N, Yamane M, Narishige E (1986): Personal computer-controlled microsurgery of fertilized eggs and early embryos. Theriogenology 25335-345. Ogawa S, Yamakawa H (1988):Micromanipulation system controlled by personal computer for microsurgery of ova and early embryos. In: “Synip Biotech Anim Breed Jpn-Ger Center.”Berlin, pp 67-83. Toyoda Y, Yokoyama M, Hoshi T (1971):Studies on the fertilization of mouse eggs in vitro. I. In-vitro fertilization of eggs by fresh epididyma1 sperm. Jpn J Anim Reprod 16:147-151. Yamakawa H, Yamanoi J, Sunouchi H, Yamane M, Narishige E, Ogawa S (1988):An improved micromanipulation system controlled by personal computer for early embryos. Jpn J Anim Reprod 34%19.
