A Multiwell Microfluidic Device for Analyzing and Screening Non-Hormonal Contraceptive Agents

Birth control and family planning play pivotal roles in the economic growth and reduction of maternal, infant, and child mortality. Current contraceptives, such as hormonal agents and intrauterine devices, target only a small subset of reproductive processes and can have serious side effects on the health of women. To develop novel contraceptive agents, we established a scalable microfluidic device for analyzing and screening the effects of potential contraceptive agents on the maturation of the cumulus-oocyte complex. The microfluidic device performs on-chip incubation for studying oocyte maturation and cumulus expansion and isolates the microwells by oil-water interfaces to avoid crosstalk between the wells. A filter membrane is incorporated in the device to simplify incubation, medium exchange, washing, and fluorescence staining of oocytes. Cumulus expansion can be monitored directly in the device and oocyte maturation can be examined after enzymatic removal of cumulus cells and on-chip fluorescence staining. We evaluate the performance of the device by studying the influence of three drugs known to block oocyte maturation and/or cumulus expansion.

Contraceptives are essential in birth control, family planning, and economic growth [1]. Contraception prevents unintended pregnancy, reduces the need for abortion, and improves the health and well-being of women. Current contraceptive options include hormonal agents administrated orally, transdermally, intravaginally, and intrauterinely, barrier methods (condoms and diaphragm), and spermicides. Existing hormonal agents target only a small subset of reproductive processes, such as ovulation, fertilization, and implantation, and can have serious side effects on the health of women [2, 3]. As an example, estrogen-containing contraceptives (e.g., oral pills) can cause vascular disease and increase the risk of certain hormone-responsive cancers, while progesterone-only contraceptives can lead to irregular or heavy bleeding during menstruation [4–6]. Non-hormonal agents that target other aspects of the reproductive process represent new opportunities in developing novel contraceptives. For instance, cumulus expansion and oocyte maturation occur during ovulation and are essential reductive processes of female fertility that have not received much attention in the development of contraceptive agents. Cumulus expansion is a highly regulated process where the somatic cell surrounding the oocyte generate a protein matrix that effectively increases the size of the cumulus-oocyte complex (COC) and greatly facilitates transfer of the egg to the oviduct. Oocyte maturation, which is an essential process for normal ovulation and fertilization, involves both completion of meiosis and asymmetrical cell division. Targeting both cumulus expansion and oocyte maturation may enable novel contraceptive agents and combination regimens that are potent, safe, and easy to use.

The development of novel contraceptives can often be limited by the paucity of an automated platform for studying oocyte development, which involves complex and labor-intensive manual procedures. Cell-based drug screening platforms that expose target cells with individual drugs include microreactors [7–17], droplets [18–20], and hydrogels [21, 22]. To maintain the cells during biochemical procedures (e.g., washing), engineering techniques, including geometrical design [23–28], microarrays [29–32], membrane [33–37], and dielectrophoresis [38–44], have been incorporated into drug screening systems. However, these techniques do not address the requirements for the analysis of COC development that requires multiple enzyme loading, cumulus cell removal, and straining procedures [45–50]. A multiplex platform for analyzing COC development has not been developed, which limits the progress of novel non-hormonal contraceptive development.

In this study, we report a multiwell microfluidic device for analyzing and screening non-hormonal contraceptive agents. The microfluidic device allows the study of multiple drug conditions and performs the multistep biochemical assay procedures, such as incubation, enzymatic removal of cumulus cells, washing, and staining in situ. We show the scalability of the design and optimize the device for studying cumulus cell expansion and oocyte maturation. In particular, the COC is incubated in the multiwell array that is separated by oil-water interfaces for evaluating the influence of contraceptive agents. A filter membrane is integrated into the multilayer fluidic chamber to facilitate COC incubation, medium exchange, enzyme loading, and fluorescence staining in the same device. Cumulus cell expansion is measured by monitoring the COC before and after incubation for 16 hours in the presence of epidermal growth factor. To observe oocyte maturation, the cumulus cells are removed enzymatically and the oocytes are stained in situ. We demonstrate the performance of the device by investigating the effects of three contraceptive agents targeting various aspects of COC development.

The microfluidic design consists of a top polydimethylsiloxane (PDMS) layer, a filter membrane, and a bottom PDMS layer (Figure 1A). In this design, the COCs were loaded into the device from the medium inlet and retained on the filter membrane in the microwells. Drugs were applied in the microwells from the drug inlets individually (Figure 1B, step 1). To avoid chemical diffusion between adjacent wells, two oil channels were incorporated in the design. A fluorocarbon (FC-40) was applied in the oil inlets to create oil-water interfaces for isolating the microwells (Figure 1B, step 2). The COCs were incubated in the presence of drugs in the microwell for 16 hours. The inlets were sealed with tape or PDMS. The fertility of COCs was evaluated by examining cumulus cell expansion and oocyte maturation after the incubation. Cumulus expansion was estimated by the occupying region of the COCs before and after incubation (Figure 1C, steps 1, 2). To facilitate the observation of oocyte development, cumulus cells were dissociated from the oocytes with hyaluronidase from the medium inlet (Figure 1B, step 3 and Figure 1C, step 3). The filter membrane was chosen to retain the oocytes while removing the cumulus cells in the washing step. The parallel processing nature of microfluidics allows enzyme loading and cumulus cell removal in all microwells simultaneously, which simplifies the assay protocol for analyzing and screening of contraceptives. The filter membrane also facilitates the implementation of multistep biochemical assays. The device performed fluorescence staining of the nucleus and cytoskeleton (i.e., DAPI for nucleus staining and Sir-actin for cytoskeleton staining), for evaluating oocyte development, such as polar body formation during metaphase II (Figure 1B, step 4 and Figure 1C, step 4).

Figure 1. A multiwell microfluidic device for screening of novel contraceptive agents.

(A) The scalable microfluidic system consists of two channel layers and a filter membrane sandwiched in between the layers. The multilayer design allows exposure of enzyme, washing, staining, and other biochemical assays directly performed in the microfluidic device. The wells are isolated by oil interfaces during incubation for studying the influences of drugs on cumulus-oocyte complex (COC) maturation. (B) The drug screening process includes (step 1) COC and drug loading, (step 2) oil isolation, (step 3) enzyme loading, and (step 4) fluorescence staining. Black dots represent the open ports and the arrows indicate the flow direction in each step. (C) Cumulus cell expansion and polar body formation are directly measured in the multiwell microfluidic device by (steps 1–2) on-chip incubation, (step 3) cumulus cells removal and (step 4) analysis of oocyte maturation.

The microfluidic device was fabricated by molding two PDMS layers and sealing a filter membrane between the channel layers (Figure 2A). Briefly, two PDMS layers were fabricated using soft lithography. The top layer mold consisted of microwells and microchannels at a height of 1 mm and the connection channels at the height of 200 μm. The multi-height design was achieved by bonding a laser machined polyester layer and an acrylic layer on a glass slide, where the pattern of the acrylic layer did not include the connection channel. The height of the connection channel was reduced to avoid crosstalk between microwells. The mold of the bottom layer was fabricated by polyester laser machining. The thickness of this layer was minimized to allow observation under an inverted microscope. The top layer served as the working layer (e.g., the COC culture) and the bottom layer functioned as the systematic outlet in the drug screening experiments. A filter membrane was sandwiched between the two PDMS layers to culture COCs and facilitate biochemical assays (Supplementary Information Figure S1). The membrane binding process was achieved by contacting the PDMS layers with a thin liquid PDMS layer spin-coated on a silicon wafer, bonding the filter membrane on one PDMS layer, aligning the other PDMS layer with the first layer on the membrane, and curing the devices to solidify the bonding among the three layers [51].

Figure 2. Characterization of the microfluidic system.

(A) Fabrication of the scalable microfluidic device. (B) Independent drug conditions (represented using the color dyes) are generated by incorporating a water-oil interface (blue dotted lines) to isolate the wells. A 4-plex device is shown as an example. Inserts: the microwell, water-oil interface, and channel connecting the wells were examined microscopically. This image was taken after 24 hr incubation. Scale bar, 2 mm. (C) The ability of the device for maintaining independent drug conditions is characterized quantitatively by examining fluorescent dye diffusion between microwells. (D) The ability of the microfluidic device for performing medium exchange and washing is characterized by sequentially and alternatively loading fluorescent dye (Rho110) and PBS.

In the multiwell device, the diffusion of contraceptive agents between adjacent wells should be avoided in order to analyze multiple drugs simultaneously. For a short incubation time, long diffusion channels can be created to minimize the crosstalk between the microwells (Supplementary Information Figure S2). For COC development, which requires 16 hours, an oil-water interface was created to isolate the microwells robustly. To evaluate the stability of the oil-water interface, the device was loaded with phosphate buffered saline (PBS) and food dyes (Figure 2B). Oil was applied at the oil inlets to create the oil-water interfaces. The oil-water interface was observed at the entrances (inset: b); whereas the channel was filled with oil (inset: a) and the microwells were filled PBS (inset: c). Due to the immiscible oil-water interface and the low diffusivity of chemicals across the interface, the contraceptives were confined within the wells and the crosstalk among microwells was avoided. To analyze the isolation efficiency, a fluorescent dye, Rhodamine 110, was applied in the microwells and the fluorescence intensity was monitored in the microwells (point c), the interface (point b), and the middle of the channel between adjacent wells (point a). The fluorescence intensity remained unchanged at these positions over 24 hr at 37°C, which was similar to the COC incubation condition (Figure 2C).

To perform biochemical procedures in the drug screening process, we evaluated the capability of the device for performing medium exchange. In particular, PBS and Rhodamine 110 were sequentially and alternatively loaded into the microwell microfluidic device. Each of the solutions was loaded at 10 μL/s for 5 min. The loading speed was chosen to minimize the shear stress on COCs. The fluorescence intensity in the microwells was measured after each medium exchange (Figure 2D). The fluorescence intensity was at a high and stable level when the fluorescent dye was applied. The fluorescence intensity returned to the background level when PBS was applied, suggesting effective washing of the dye to an undetectable level. To illustrate the scalability of the approach, a 32-well design was fabricated and demonstrated for medium exchange in all wells simultaneously (Supplementary Information Movie S1). We also performed the medium exchange procedure with spheroids formed by human umbilical cord vein endothelial cells (HUVEC) to evaluate the biocompatibility (Supplementary Information Figure S3). The HUVEC spheroids, which have a similar size to the oocytes, were retained and preserved intact during this process.

To determine that oocyte maturation occurred normally in the device, oocytes were incubated in normal maturation medium for 16 hours followed by removal of cumulus cells and staining of the spindle as described previously [52, 53]. The majority of oocytes (52 out of 62; 84%) showed a normal barrel-shaped spindle, small polar body and chromosomes aligned on the spindle in a straight line (Supplementary Information Figure S4). To evaluate the feasibility of the device for analyzing contraceptive agents, the influences of three contraceptive agents, including SB431542 (SB), N,N,N’,N’-tetrakis(2-pyridylmethyl)-1,2-ethylenediamine (TPEN), and pyrithione (PYR), on COC development were examined. The COCs were incubated in the presence of the contraceptives in the device. Cumulus cell expansion was monitored at 16 hr and oocyte maturation was examined by fluorescence staining after cumulus cell removal (Figure 3A). Cumulus cell expansion was quantitatively analyzed by the occupying area of COCs before and after incubation (Figure 3B) and oocyte maturation was evaluated by the portion of oocytes with a polar body and the morphology of the polar body (Figure 3C).

Figure 3. Screening for non-hormonal contraceptive agents on the microfluidic system.

(A) The contraceptive agents affect the cumulus expansion and oocyte maturation in the microfluidic device. The cytoskeleton and nucleus of the oocyte is stained to evaluate the maturation. Scale bar: 50 μm. (B) The cumulus expansion is inhibited with contraceptive agents. (C) The oocyte maturation is evaluated with ratio of oocytes with a polar body. The data (i.e., mean ± SEM) is representative for two independent experiments and at least 10 COCs were analyzed for each case. One-way ANOVA followed by Dunnett’s post hoc test was performed in Graphpad Prism 8: ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.

In the control group, the COCs expanded during the incubation period and asymmetric division during oocyte maturation created polar bodies that were small and spherical. Most of the oocytes (63%) matured with polar bodies. In the SB group, the treatment damaged the cumulus structure surrounding the oocyte and inhibited the cumulus expansion whereas SB has no observable effect on oocyte maturation. The results suggest SB potentially functions by interfering the cumulus structure, which may block the implantation of COCs on the oviduct. This observation is consistent with a previous study using SB [54]. In contrast, TPEN reduced cumulus cells expansion significantly and reduced the number of oocytes with polar bodies. Furthermore, the polar bodies displayed abnormal morphology, which might be attributed to spindle defects, failure of cytokinesis, and chromosome segregation defects [52]. This abnormality can block cell development after fertilization [53]. We also investigated the effects of a zinc ionophore, pyrithione (PYR) on COC development, which is not fully understood. Similar to other drugs, the expansion of COC was attenuated significantly. The formation of polar bodies was also completely inhibited in the presence of PYR. Therefore, our results suggest PYR functions by interrupting the formation of polar bodies as well as inhibiting COC expansion. These results were also confirmed by performing the experiments manually in 96-well plates (Supplementary Information Figure S5), supporting the applicability of the multiwell microfluidic device for analyzing and screening contraceptive reagents.

In this study, we demonstrate a multiwell microfluidic device for analyzing and screening non-hormonal contraceptive agents. This microfluidic device is specifically designed for targeting COC development. The device is capable of evaluating multiple drugs in parallel by isolating the microwells with oil-water interfaces. The multilayer microfluidic device also incorporates a filter membrane to facilitate COC incubation, medium exchange, enzyme loading, removal of cumulus cells, and fluorescence staining in each microwell. The procedures were optimized for measuring cumulus cell expansion and evaluating oocyte maturation with cumulus cell removal and fluorescence staining. The feasibility of the device for analyzing and screening non-hormonal contraceptives was demonstrated by studying the effects of three agents on COC development.

Despite its importance, screening of non-hormonal contraceptives is an understudied area and an automated platform for analyzing COC development is not available. Existing drug screening platforms are often designed to study drug influence on immobilized targets/cells (e.g., mix-and-measure assay) or require complicated engineering designs or manual procedures to perform biochemical procedures. Our device directly addresses these unmet needs in contraceptive development. The design of the device is highly scalable, the fabrication is easy to perform, and the implementation of the experiment is straightforward.

With the scalability and simplicity, this device provides a promising platform for medium to high throughput screening of contraceptive agents. In the future, multiplex devices with oocyte loading mechanisms should be designed. Automated imaging and data analysis algorithms can be incorporated to fully automate the screening and analysis process. Other reproductive processes, such as on-chip fertilization, should also be integrated into the platform for analyzing and screening novel contraceptives and combination regimens.

Experimental Section

Materials

Polydimethylsiloxane (PDMS) was obtained from Dow Corning (Sylgard 184). The polyester (PETE) membrane with 10 μm pore size was purchased from SterliTech Inc. Acrylic (0.762 mm) for fabricating the top mold was from ePlastics. A transparent polyester membrane (PP2500, 3M) was employed to fabricate the top layer and bottom layer molds (nominal thickness 100 μm). Rhodamine 110 (Rho110) and FC-40 oil were purchased from Sigma-Aldrich. Rhodamine was dissolved in PBS.

Wash media was MEM-α (gibco, cat # 12000-022 lot # 1779618) supplemented with 2.2 g/L sodium bicarbonate (Sigma, cat # S6014-500G, lot # SLBV7249), 0.23 mM sodium pyruvate (Sigma, cat # P4562-25G), 10 μg/ml streptomycin sulfate, 10 IU/ml penicillin G (Corning, cat # 30-002-CI, lot # 30002253), 3 mg/mL bovine serum albumin (Sigma, cat # A1470-25G, lot # SLBN1886V), and 25 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) (Sigma, cat # H6147-26G, lot # 016K54321) in addition to 10 μM milrinone (CalBioChem, cat # 475840, lot # 000080279), a PDE3A-specific inhibitor to prevent premature resumption of meiosis. Maturation medium was MEM-α supplemented with 10% fetal bovine serum (Atlanta Biologicals, cat # S11550, lot # C0048), 10 ng/mL epidermal growth factor (Corning cat# 354001), a cumulus cell expansion stimulator, and 25 mM HEPES. The oocyte response was studied for three drugs, including SB (Sigma cat# S4317), TPEN (Sigma, cat # P4413-100mg, lot # 120M1573V), and PYR (Sigma cat# H3261). The concentration of the drugs was 100 μM, which was 10 times as the final concentration. Wash media supplied with hyaluronidase (2 mg/mL) was utilized to detach the oocyte with the cumulus cells. Paraformaldehyde (PFA) and Triton X-100 (Sigma-Aldrich) were applied for cell fixation and permeabilization. The concentration of PFA and Triton X-100 was 4% and 0.25% in PBS. Fluorescent dyes, 4′,6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich) and Sir-Actin (Cytoskeleton Inc.), were applied for nucleus and cytoskeleton staining to determine the cell structure at the single cell level. The concentration of DAPI and Sir-Actin was 300 nM and 1000 nM, respectively.

Microfluidic device

The microfluidic device consisted of three layers and was fabricated using a soft lithography process (Figure 2A). The top layer mold consisted of a polyester membrane layer and an acrylic layer. The layers were bonded with double-sided tape (3M) and subsequently laser machined (Universal Laser Systems). The height of the polyester layer and the acrylic layer were 200 μm and 0.862 mm, respectively. The pattern of the polyester layer included the microwells, the microchannels, and the connecting channels between the wells and channels. The acrylic layer contained identical microwells and microchannels but not the connection channels. The dimension of the microwell was 6 mm by 4 mm with well-rounded edges. The width of the channel was 1.5 mm. The acrylic layer was aligned and bonded with the polyester layer to create the mold, which consisted of microwells and microchannels at the height of approximately 1 mm and connection channels at the height of 200 μm. This mold was bonded on a glass slide. PDMS (at a ratio of 10:1 between pre-polymer and cross-linker) was poured on the mold and cured for 1 hr at 80°C. The PDMS layer was peeled off and the ports were punched using a puncher of 1.5 mm in diameter (Miltex Biopsy Punch). This layer served as the top layer. The bottom layer mold was laser machined using the polyester membrane bonded with double-sided tape. The thickness is 200 μm. PDMS (at a ratio of 10:1 between pre-polymer and cross-linker) was poured on the mold and cured for 1 hr at 80°C. This PDMS layer were peeled off and utilized as the bottom layer. Both PDMS layers were attached with a thin layer of liquid PDMS, which was spin-coated on a silicon wafer at 3000 rpm for 1 min (KW-4A, Chemat Technology Inc.). The filter membrane was sandwiched between these two layers with the assistance of this liquid PDMS layer. The device was cured for 2 hr at 70°C to solidify the liquid PDMS and bond the layers.

Device characterization

We developed a method to avoid drug diffusion between adjacent wells. To characterize the method, the device was first preloaded with PBS at 200 μL/min for 5 min using a syringe pump (PHD ULTRA, Harvard Apparatus). Rho110 (100 μM, 3.6 μL) was pipetted into each well through the inlets connecting with the wells. FC-40 was loaded in the oil channels to separate microwells during incubation. The fluorescence signals at the microwells, the oil/water interface above the microwells, and the middle of channel connecting microwells were monitored over time using a fluorescence microscope (DMI4000B, Leica) equipped with a CCD camera (SensiCam QE, PCO). The microscope objective was 20X and the exposure time was 75 ms. The fluorescence intensity was measured in ImageJ to determine the spatial and temporal distribution of the fluorescent dye and the diffusion among microwells.

The medium exchange process was characterized. The device was alternatively loaded was PBS and Rho110 (10 μM) at 10 μL/s for 5 min. Five rounds were performed. The fluorescence signal at the microwells was monitored. The exposure time was 15 ms. The fluorescence intensity was measured in ImageJ.

Animal use and oocyte collection

5 IU pregnant mare serum gonadotropin (PMSG) (ProSpec, cat # hor-272, lot # 918PGSMP) in sterile PBS (100 μL total volume) was injected intraperitoneally into 17–19 day old CD1 mice (Mus musculus) to induce ovulation. Ovaries were collected approximately 48 h after PMSG injection. COCs containing fully-grown oocytes were released from PMSG-primed mouse ovaries aseptically via gentle puncture with a syringe and needle then placed in wash media. Euthanasia was performed via CO₂ inhalation followed by cervical dislocation. Animals were used according to the Guide for the Care and Use of Laboratory Animals (Institute for Learning and Animal Research) in addition to review and approval by the Institutional Animal Care and Use Committee at The Pennsylvania State University.

Contraceptive analysis

The device was preloaded with culture medium. COCs were transferred to the device and the medium was replaced with maturation media at 200 μL/min. Drugs, including SB, TPEN, and PYR, were individually pipetted into the microwells through the connecting inlets (i.e., 2.4 μL in each well). The final concentration of each drug was 10 μM in the maturation media. FC-40 was loaded in the oil channels at 200 μL/min for 3 min. The oocytes were incubated at 37°C with 5% CO₂ and 5% O₂ for 16 hr. After the incubation process, the oil was removed by applying medium through the oil channels at 200 μL/min for 3 min. Hyaluronidase was loaded in the microwells at 200 μL/min for 5 min and the device was incubated at 37°C for 20 min to separate the oocytes from the cumulus cells. PBS was loaded at 500 μL/min for 5 min. During this process, the small cumulus cells were washed away whereas the oocytes were retained on the filter membrane. For staining, PFA was loaded at 200 μL/min for 5 min and the oocytes were fixed for 30 min. Triton X-100 was loaded at 200 μL/min for 5 min and the oocytes were permeabilized for 10 min. PBS was loaded at 500 μL/min for 5 min to wash the cells. The Sir-Actin was then loaded in the device at 200 μL/min for 5 min and the cytoskeleton was stained for 30 min. DAPI was loaded at 200 μL/min for 5 min and the cell nucleus was stained for 5 min. PBS was loaded at 500 μL/min for 5 min to reduce the background fluorescence signal. The oocytes were monitored using the same imaging system mentioned above. The exposure time was set as 300 ms. The images were analyzed using ImageJ. To analyze the data, one-way ANOVA followed by Dunnett’s post hoc test was performed using Graphpad Prism 8.

Data and materials availability:

All data generated or analyzed during this study are included in this published article and its supplementary information files.