Introduction

Assisted reproductive technologies (ART), such as in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI), are cornerstone practices in fertility clinics, farm animal breeding programs, and developmental biology research, despite their still low and variable efficiency1,2. A major contributing factor to ART success is the quality of the sperm used. However, traditional sperm selection techniques, focused on sorting by motility and viability, are proving insufficient3. Multiple other sperm characteristics including epigenetic modifications, chromatin packaging, and sperm mRNA cargo, are determinants of the activation and progression of embryonic development4,5,6. Similarly, improved methods are needed to evaluate sperm quality for artificial insemination. In the context of human fertility, there is a great overlap in the semen parameters between fertile and infertile men when using traditional methods7,8. In farm animals, the selection of males for breeding purposes based on semen characteristics has a major economic impact as a single semen collection may be used to inseminate thousands of females9,10. An optimal sperm quality assay needs to evaluate both the ability to reach the site of fertilization and to activate embryonic development11.

The latest advances in sperm selection techniques have attempted to simulate the physiological transport of sperm through the female reproductive tract12,13,14. Most notably, the oviduct not only provides a means for gametes to meet, but also actively regulates sperm viability, maturation, and function15. One key interaction between mammalian sperm and the oviduct occurs in the isthmus, where sperm survive for hours, days or even months (in bats) after binding to oviduct epithelial cells (OECs)16. This sperm reservoir is believed to compensate for asynchrony between mating and ovulation and has also been described in the human Fallopian tube17. Sperm heads bind to specific oligosaccharides or glycans that are part of the OECs glycocalyx18,19. In pigs, two glycan motifs have been identified in the isthmus; these motifs are mainly attached to N-linked glycoproteins, sulfated Lewis X trisaccharide (suLeX) and a biantennary 6-sialylated oligosaccharide (bi-SiaLN)20,21,22. Similar oviduct glycans also bind bovine and human sperm23,24,25. In addition, sperm bind to non-glycosylated proteins present in the oviduct, such as fibronectin26 and annexins27. Intriguingly, the recognition of oviduct glycans by sperm membrane receptors is sufficient to trigger changes in sperm function. Thus, porcine sperm bound to suLex or bi-SiaLN coupled to inert microbeads survived for a longer incubation period than free-swimming sperm or sperm bound to fibronectin-coated beads28. This enhanced viability could be partly explained by a decreased activity of the electron transport chain leading to lower intracellular reactive oxygen species29, which are involved both in sperm capacitation and death30.

Compelling evidence suggests that the sperm reservoir in the oviduct also acts as a sperm selection mechanism. First, a correlation between male reproductive performance and the ability of an ejaculate to bind OECs has been noted for both bulls and boars31,32,33,34,35,36,37. Second, several in vitro studies have characterized the population of sperm that adhere to OECs after co-culture. Overall, bound sperm are mostly non-capacitated38,39,40,41 and have higher DNA integrity31,42 than sperm that remain free after culture. Third, bovine IVF efficiency (quantified as the frequency of development to blastocysts) was increased by using sperm that adhered to OECs during a pre-incubation step43. Considering that fertilization is skewed toward sperm that survive longer in the oviduct reservoir, the preferential binding of sperm with higher embryonic potential would increase reproductive success. While the benefits of using OECs cultures for IVF have been well-established38,43,44, the complexity of maintaining a live OEC culture (the cells rapidly lose their epithelial differentiation), the lack of standardization, and the risk of biocontamination have limited their application. An engineered inert technology that replicates oviduct function would therefore be highly advantageous for use in both IVF and sperm quality assays.

The aim of the present study was to develop an in vitro model of the sperm reservoir by using immobilized oviduct glycans on functionalized glass surfaces (i.e., coverslips and slides coated with specific functional groups of the molecules). After validating the system using porcine sperm and determining the optimal culture conditions, its potential for porcine IVF was tested. Although there are many reports of oviduct cells influencing fertilization in vitro, including some arranged in three dimensions45, this is the first report of in vitro fertilization employing a specific component of the sperm reservoir in the oviduct.

Results

Porcine sperm bind with high specificity to functionalized glass surfaces coated with suLeX

Two types of functionalized glass surfaces were tested to immobilize oviduct glycans and model the oviduct reservoir (Fig. 1): 3D-functionalized streptavidin (Fig. 1A) and functionalized polyethyleneglycol (PEG)-biotin slides with attached streptavidin (Fig. 1B). The streptavidin coating of both surfaces enabled the adhesion of biotinylated glycans via a non-covalent bond (Fig. 1C). On the coverslips (Fig. 1A), sperm were incubated in medium droplets placed on the surface and covered with mineral oil. In contrast, the slides (Fig. 1B) were covered with a hybridization chamber, creating a space for introducing the medium and incubating the sperm. To attach a glycan, medium containing 0.02 mg/mL of biotinylated suLeX was placed in droplets onto the coverslips or inside the chamber in the slides. The systems were incubated at ~ 22⁰C for 30 min. After uncoupled suLeX was removed by refreshing the medium, sperm were added to each droplet or chamber and incubated at 39⁰C. Free sperm were removed by replacing the medium, and the density of sperm bound to the coverslip or slide (number of sperm per mm²) was evaluated. In addition to suLeX, two molecules were used as negative controls in some of the experiments to confirm that sperm adhere to suLeX and not to other components of the system: LacNAc, a disaccharide that does not bind uncapacitated porcine sperm28, and suLeA, a trisaccharide closely related to suLeX that binds bovine but not porcine sperm22,23,24. Fibronectin was used as a non-glycan positive control in some experiments because previous studies with fibronectin-coated beads demonstrated its ability to bind porcine sperm, although it did not extend sperm lifespan as effectively as glycan-bound beads28.

Fig. 1
figure 1

Schematic representation of the two in vitro systems developed to model the sperm-oviduct interaction. (a) 3-D functionalized streptavidin coverslip: the coverslips are placed on a petri dish, medium droplets containing sulfated LeX (suLeX) are placed on the streptavidin regions of the coverslip that are surrounded by hydrophobic surface, and the whole system is covered with mineral oil. (b) Polyethyleneglycol (PEG)-biotin slides: the slides are coated with streptavidin, a hybridization chamber is placed on the slide, and the chambers are filled with medium containing suLeX. (c) The streptavidin coating in both glass surfaces allows biotinylated suLeX to be tightly immobilized to the surface. Once the sperm are added and incubated in the system, membrane receptors on the head recognize and attach to suLeX bound to the glass.

Full size image

A series of experiments were performed to validate the use of functionalized coverslips coated with suLeX to bind porcine sperm and determine the ideal culture conditions for the following experiments (Fig. 2). The first experiment showed that porcine sperm were bound to coverslips coated with suLeX with high specificity (Fig. 2A and D). After incubation with 1 × 106 sperm/mL for 30 min, there was minimal binding to the coverslips with no added biotinylated molecules (control droplets with only medium; 76.1 ± 22.9 sperm/mm2) and to coverslips with immobilized LacNAc (56.3 ± 12.0 sperm/mm2). In contrast, many more sperm were bound to immobilized suLeX (5468.6 ± 897.1 sperm/mm², p < 0.0001) as well as to the positive control fibronectin (4205.5 ± 623.5 sperm/mm², p < 0.0001). The concentration of sperm in the medium recovered from washing the suLeX droplets was assessed with a hemocytometer. Of the sperm added to the suLeX droplets, 90.0 ± 3.1% were retained on the coverslip. In the second experiment, different concentrations of sperm (0.25, 1, and 5 × 106 sperm/mL) were tested (Fig. 2B and E). The binding density on suLeX-coverslips was proportional to the concentration of sperm incubated in each droplet, although only 5 × 106 sperm/mL was significantly different (p < 0.05). A 10 × 106 sperm/mL incubation concentration was also tested, but the number of sperm bound to the coverslip made counting the sperm density impossible and a very large number of sperm remained free-swimming or clumped after washing (Fig. 2E). In the third experiment, different culture times (30, 60 and 90 min) were tested to determine the length of time required for sperm to reach maximal binding (Fig. 2C). Moreover, there was no significant difference in the binding density between all the time points to suLeX-coated coverslips and therefore, a culture time of 30 min was used for all following experiments. Finally, a very low binding density was observed in the suLeA droplets during culture-time (92.0 ± 25.1 sperm/mm²) and sperm-concentration (95.8 ± 35.5 sperm/mm²) experiments (Fig. 2B and C), comparable to LacNAc results (Fig. 2A), confirming the validity of both as negative controls.

Fig. 2
figure 2

Porcine sperm bind to suLeX immobilized on a coverslip with high affinity and specificity. (a) Binding density (per mm2) to suLeX-coverslips, compared to that of fibronectin, LacNAc, or control (no added biotinylated glycan) after incubation with 1 × 106 sperm/mL for 30 min. (b) Binding density to suLeX- and suLeA-coverslips after incubation with increasing sperm concentrations for 30 min. (c) Binding density to suLeX- and suLeA-coverslips after incubation with 1 × 106 sperm/mL for increasing incubation times. The error bars indicate the mean ± SEM. Different letters indicate p < 0.05. (d) Representative microscopic images of sperm immobilized on coverslips pre-coated with suLeX, LacNAc, fibronectin, and control. Scale bar = 20 μm. Arrowheads show clumped sperm that were not bound to the coverslips. (e) Representative microscopic image of sperm immobilized on coverslips pre-coated with suLeX after incubation of 10 × 106 sperm/mL during 30 min. Scale bar = 20 μm.

Full size image

A similar set of experiments was performed to test the ability of a second system, which consisted of hybridization chambers on functionalized slides, to bind porcine sperm (Fig. 3). The results were equivalent to those of the coverslip experiments. Porcine sperm were bound with high specificity to chambers coated with suLeX (2741.2 ± 149.6 sperm/mm²) and fibronectin (2839.5 ± 244.2 sperm/mm²) compared to those coated with LacNAc (60.9 ± 12.2 sperm/mm², p < 0.0001) or control medium only (66.6 ± 16.4 sperm/mm², p < 0.0001; Fig. 3A and D). Furthermore, a proportional increase in the binding density according to the sperm concentration was detected, with only 5 × 106 sperm/mL being statistically significant (p < 0.0001, Fig. 3B). No differences in the binding density were observed with increasing incubation time (30, 60, or 90 min; Fig. 3C).

Fig. 3
figure 3

Porcine sperm bind to suLeX immobilized on a slide with high affinity and specificity. (a) Binding density (per mm2) to suLeX-slides, compared to that of fibronectin, LacNAc and the control after incubation with 1 × 106 sperm/mL for 30 min. (b) Binding density to suLeX- and suLeA-slides after incubation with increasing sperm concentrations for 30 min. (c) Binding density to suLeX- and suLeA-slides after incubation with 1 × 106 sperm/mL for increasing incubation times. The error bars indicate mean ± SEM. Different letters indicate p < 0.05. (d) Representative phase contrast microscopic images of sperm immobilized on coverslips pre-coated with different glycans. Scale bar = 50 μm. Scale bar in inset image = 20 μm. Arrowheads show clumped sperm that were not bound to the coverslips.

Full size image

In the coverslip system, when sperm were incubated with control medium and LacNAc, free-swimming sperm remained after washing within the droplet as well as regions where sperm had clumped together (Fig. 2D). On the contrary, practically no free-swimming sperm remained after washing the coverslips with immobilized suLeX and fibronectin. In the slide system, the removal of free sperm was less efficient, as more free-swimming sperm were observed even with suLeX and fibronectin (Fig. 3D). A greater binding density was observed on the coverslip than on the slide, when the same sperm concentrations in the binding step were compared. The lower binding observed on the slides is likely attributed to the medium replacement method. In specific, the suction force created by aspirating the contents of the hybridization chamber with a pipette through the fill-up port may have removed attached sperm. However, sperm binding on slides were more uniformly distributed throughout the chamber compared to the coverslips, where a higher density was seen on the edges than the center of the droplets.

In a preliminary experiment, different concentrations of suLeX and LacNAc were also tested for their ability to be immobilized on coverslips and bind sperm. No significant differences in sperm binding were observed (Supplementary Fig. S1) and, hence, the lowest tested concentration of suLeX (0.02 mg/mL) was used for all subsequent experiments.

The coverslip system was chosen for the next set of experiments, as it enabled the incubation of COCs in a manner similar to conventional IVF methods, using droplets of medium covered with mineral oil.

Progesterone and mature cumulus oocyte complexes induce sperm release from suLeX-coated coverslips

A second set of experiments was performed to determine if sperm can be released from glycans coupled to coverslips (Fig. 4), allowing them to fertilize oocytes.

Fig. 4
figure 4

Progesterone and cumulus-oocyte complexes (COCs) induce sperm release from suLeX immobilized on coverslips. (a) Binding density (per mm2) to suLeX-coverslips after incubation of 1 × 106 sperm/mL for 30 min, followed by incubation for 1 h with control Modified Tris-buffered medium (mTBM), capacitating mTBM (containing 2.5 mM caffeine), capacitating mTBM supplemented with 800 nm progesterone (P4), and capacitating mTBM supplemented with 4 mM 4-aminopyridine (4-AP). (b) Binding density to suLeX-coverslips after incubation of 0.25 × 106 sperm/mL for 30 min, followed by 5 h with live COCs or denatured COCs. The error bars indicate mean ± SEM. Different letters indicate p < 0.05. (c) Representative microscopic images of sperm immobilized on suLeX-coverslips and cultured with COCs. Scale bar = 50 μm. Arrowheads show sperm bound to cumulus cells.

Full size image

In the first experiment, 1 × 106 sperm/mL were incubated for 30 min in the suLeX coverslip. After washing the droplets to remove free sperm, 800 nM of progesterone or 4 mM of 4-aminopyridine (4-AP, an inducer of hyperactivation that was used as positive control) were added to the droplets and incubated for 1 additional hour (Fig. 4A). The concentration of P4 used was chosen based on a previous study that tested the release of porcine sperm from suLeX-coated beads46 and is 2–20 times higher than the concentration reported in post-ovulatory oviduct fluid of different species47,48,49,50. A capacitating Modified Tris-buffered medium (mTBM) with 2.5 mM caffeine (a milder hyperactivating agent) was used to represent fertilization conditions. In addition, a mTBM without caffeine was used as control medium. No significant differences were observed in the binding density between the use of capacitating or control medium. Compared with the control medium, progesterone reduced the number of bound sperm by 37.2% (p < 0.0001) and by 27.4% compared with the capacitating medium (p < 0.01). Similarly, compared to control medium, 4-AP reduced the number of bound sperm by 33.7% (p < 0.0001) and by 23.5% compared with capacitating medium (p < 0.01).

In the second experiment, 0.25 × 106 sperm/mL were incubated for 30 min in the suLeX coverslip. After washing the droplets to remove free sperm, mature COCs were transferred to the droplets (after 42 h of IVM) and incubated for 5 h under fertilizing conditions (Fig. 4B and C). Compared with capacitating mTBM, mature COCs reduced the number of bound sperm by 55.2% (p < 0.0001). However, COCs denatured by heating (previously exposed to 60oC for 5 min in a warm bath) lead to a similar reduction of bound sperm (56.9%, p < 0.0001). In both groups, some sperm were observed bound to the remaining cumulus cells (Fig. 4C). These were not included in the bound-sperm count. For each replicate, the concentration of sperm in the medium that was recovered after the initial sperm incubation was assessed with a hemocytometer. Of the sperm added to the droplets, 78.3 ± 1.9% were retained on the coverslips after removal of unbound sperm. Assuming that at least 55.2% of bound sperm would be released by COCs (compared to the control group), it was estimated that the concentration of free-swimming sperm after the addition of COCs would be only 43.2% of the sperm added to the droplet.

Porcine sperm bound to suLeX-coated coverslips pre-fertilization can successfully fertilize COCs

A protocol was developed to perform IVF using the glycan-coated coverslip, named glycan-IVF (Fig. 5). Briefly, 0.25 × 106 sperm/mL were first incubated for 30 min in the coverslips and allowed to bind to the suLeX. After removing free sperm, COCs were added to the system and incubated for 5 h to allow release of bound sperm and fertilization. The glycan-IVF was validated by comparing with a conventional IVF protocol, which was performed concomitantly as a control (Fig. 6). For the conventional IVF a concentration of 0.1 × 106 sperm/mL was used. Based on the previous experiment, approximately 43% of the sperm initially added to the coverslips would be released by COCs and available for IVF. Therefore, 0.25 × 106 sperm/mL were used for glycan-IVF to yield approximately 0.1 × 106 sperm/mL of free-swimming sperm. After IVF, presumptive zygotes incubated for 7 additional hours and the formation of pronuclei was evaluated with Hoechst 33342 to determine the fertilization frequency. Sperm bound to suLeX pre-fertilization achieved similar fertilization frequencies (total number of zygotes/total oocytes, Figs. 6A and 75.1 ± 2.2%) and IVF efficiency (monospermic zygotes/total oocytes, Figs. 6C and 39.7 ± 4.9%) compared to sperm without prior binding (72.0 ± 3.2% and 33.3 ± 3.5%, respectively). Additionally, the analysis of unfertilized eggs confirmed that 90% of the oocytes achieved maturation, since they were either fertilized or at the metaphase II stage.

Fig. 5
figure 5

Schematic representation of IVF using sperm bound to glycans pre-fertilization (glycan-IVF). Functionalized streptavidin-coverslips were coated with suLeX in droplets of medium. Sperm were added to the droplets and incubated to allow them to bind suLeX. The medium in the droplet was replaced to remove free sperm. Mature cumulus-oocyte complexes (COCs) were added to the droplet and co-cultured with bound sperm. COCs induce the release of some of the sperm enabling them to fertilize the oocytes.

Full size image
Fig. 6
figure 6

Porcine sperm bound to suLeX coverslips pre-fertilization successfully fertilize oocytes in vitro. IVF was performed under conventional conditions (Control; 5 h with 0.1 × 106 sperm/mL) and on suLeX coverslips with sperm bound pre-fertilization (glycan-IVF; 5 h with 0.25 × 106 sperm/mL). Pronuclear formation was evaluated 12 h after the beginning of IVF by Hoechst staining. (a) Frequency of fertilized oocytes from total cultured oocytes, (b) Frequency of polyspermic zygotes from total fertilized oocytes, and (c) IVF efficiency or frequency of monospermic zygotes from total cultured oocytes. The error bars indicate mean ± SEM.

Full size image

The fertilizing ability of sperm on glycan-coverslips over time was further investigated (Fig. 7). Sperm (0.1 × 106 sperm/mL) were incubated on coverslips coated with suLeX, LacNAc, or fibronectin, or with no molecule (control) for 30 min to allow sperm binding, followed by additional 0, 6, 12, or 24 h. Without removing free sperm, mature COCs were transferred to the system and cultured for 5 h to allow fertilization. There was a reduction in fertilization frequency for longer sperm pre-incubation time in all groups, but suLeX maintained sperm fertilizing ability for a longer time (Fig. 7A). Thus, when sperm were pre-incubated for 24 h, almost no oocytes were fertilized in the control group (1.0 ± 1.0%), whereas 12.0 ± 3.6% were fertilized in the suLeX group (p < 0.05). Regarding polyspermy (Fig. 7B), a lower percentage of polyspermic zygotes (polyspermic zygotes/total zygotes) was observed in the suLeX group after 0 h of preincubation (17.7 ± 3.3%) compared to the control group (41.3 ± 6.9%, p < 0.05). The polyspermic frequencies were not statistically analyzed for the 12 and 24 h groups due to the absence of polyspermic zygotes in most groups and replicates at these time points. Finally, the IVF efficiency (monospermic zygotes/total oocytes; Fig. 7C) was greater for sperm pre-incubated with suLeX for 0 h (53.1 ± 2.9%) compared to the control (35.8 ± 3.0%, p < 0.01), suLeA (40.5 ± 1.6%, p < 0.05) and fibronectin (40.0 ± 2.0%, p < 0.05). Although IVF efficiency was reduced at longer sperm pre-incubation times in all the treatments, it was still 10.9 ± 2.7% at 24 h, which was greater than 1.0 ± 1.0% observed in the control group (p < 0.05).

Fig. 7
figure 7

Porcine sperm bound to suLeX coverslips pre-fertilization maintain their fertilizing ability over longer culture time and produce a lower frequency of polyspermic zygotes. IVF was performed on coverslips coated with suLeX, suLeA, fibronectin or no coupled molecule (control). Sperm were cultured for 30 min to allow sperm binding to coverslips followed by an additional incubation of 0, 6, 12, or 24 h before the addition of cumulus-oocyte complexes. Pronuclear formation was evaluated 12 h after the beginning of IVF with Hoechst staining. (a) Frequency of fertilized oocytes from total cultured oocytes, (b) Frequency of polyspermic zygotes from total fertilized oocytes, and (c) IVF efficiency or frequency of monospermic zygotes from total cultured oocytes. The error bars indicate mean ± SEM. Different letters indicate significant differences (p < 0.05) within the culture time.

Full size image

Discussion

After mating, mammalian sperm form a reservoir in the isthmus by adhering to specific glycans on the oviduct epithelium51. This functional storage is believed to play an active role in maintaining fertility by promoting the survival of superior sperm14. In the present study, an in vitro model of the oviduct sperm reservoir was developed using immobilized suLeX on functionalized glass surfaces (Fig. 1). The binding specificity of porcine sperm to suLeX was validated with two systems based on either coverslips or slides. After establishing suitable culture conditions, the ability of the suLeX-coverslip system to sustain sperm viability and fertilization potential was evaluated.

The specific binding of porcine and bovine sperm to oviduct glycans in vitro has been previously reported22,24. In these studies, biotinylated glycans (such as suLeX, suLeA, and bi-SiaLN) were coupled to microscopic streptavidin-beads to mimic the insolubility of cell-bound glycans. This approach has proven to be useful for studying sperm retention and storage in the oviduct. For example, binding to immobilized glycans extends the lifespan of porcine and bovine sperm24,28. In the present study, a novel method for sperm retention by glycan-coated glass surfaces was evaluated. This method has substantial advantages over glycan-coated beads; mainly uniformity of the sperm adhesion, easier sperm quantification, and easier sperm recovery. Under the microscope, sperm appeared to be attached through their heads maintaining a free and beating flagellum (Supplementary Video S2). The ability of non-capacitated porcine sperm to bind suLeX but not the positional isomer suLeAwas confirmed, as previously reported22. Interestingly, a higher sperm binding density was observed than previously reported for glycan-beads, which may be due to a higher concentration of biotinylated suLeX adhered to the surface. Moreover, binding occurred rapidly (within 30 min after adding sperm) and the system also saturated promptly (i.e., no additional increase in binding density was observed after 30 min).

After preliminary experiments, the ability of sperm to detach from the suLeX coverslip was assessed. The presence of progesterone reduced the number of sperm adhering to the suLeX-coverslip, as previously reported for suLeX-coated beads46,52 and OECs in vitro46,53. The mechanism of action by which progesterone triggers sperm detachment from glycans involves hyperactivation through CatSper-mediated Ca2+ increase46,52. Specifically, the progesterone-induced change in sperm tail velocity causes a rotation energy that is enough to detach the sperm from an adherent surface, a mechanism described as hammer-throw escape54. In the present study, 4-AP, a known inducer of sperm hyperactivation52,55, also decreased the number of attached sperm similar to progesterone, whereas caffeine alone, which increases the beat frequency of porcine sperm56, had no significant effect. It is important to note that, compared to what was reported for glycan-coated beads, progesterone released fewer sperm from coverslips46 (approximately 68% and 37%, respectively), most likely due to the nature of the coverslip system that allows for gravity to maintain contact between sperm and the glycans at the bottom of the culture wells.

A decrease in the proportion of coverslip-bound sperm was also observed after incubation with mature COCs, an effect that may be mediated by the release of progesterone by cumulus cells57. However, COCs denatured by heating induced a similar response. This unexpected effect may be attributed to the residual presence of COCs-secreted factors in the medium in which COCs were transferred, even though, as a precaution, the culture medium was replaced repeatedly before the transfer of COCs to the droplet with sperm. Alternatively, glycosaminoglycans present in the cumulus cell matrix such as hyaluronic acid58, may remove sperm from suLeX surface by providing other molecules to which sperm adhere59. Both progesterone and glycosaminoglycans, though heat sensitive, are expected to resist incubation at 60oC60,61. It is generally accepted that the hormonal milieu around ovulation promotes the detachment of sperm from the isthmus, ensuring they reach the ampulla at the appropriate time to fertilize the oocyte62, however the role of COCs is not well defined. Although both in vitro and in vivo trials have reported that the presence of COCs in the upper oviduct is sufficient to trigger the release of sperm from the isthmus62,63, a recent study suggested that COCs-derived factors indirectly regulate sperm transport in the oviduct by changing the OECs’ gene expression64. Moreover, others have suggested that sperm release from the isthmus is dependent on binding competition with oviduct fluid glycosaminoglycans65. Although the proposed suLeX-coverslip system might not accurately mimic physiological processes due to the uncharacteristic proximity of COCs to oviduct glycan-bound sperm, it can nonetheless serve as a research tool to study the functional changes in sperm that mediate their attachment and detachment from the oviduct epithelium.

The next set of experiments demonstrated that suLeX-coverslips can sustain IVF. In this glycan-IVF system, porcine sperm were bound to suLeX-coverslips before the addition of mature COCs and incubation for 5 h. To guarantee that only sperm adhered to glycans were used, the culture medium within the droplets was repeatedly replaced. As a result, no or very few free-swimming sperm were observed before the transfer of COCs. Compared to conventional IVF, glycan-IVF produced a comparable percentage of fertilized oocytes, indicating that sperm maintained their fertilizing ability after being attached and detached from the coverslip. The culture of sperm with OECs before or during IVF is known to improve IVF success in a variety of animals38,43,44,66,67,68,69, at least partly through the attachment of sperm to the OECs. Additionally, if bovine sperm were allowed to attach to an OEC-monolayer and later recovered after progesterone-induced release, these sperm yielded more developing embryos than the yield from control sperm43. Correspondingly, when porcine sperm that did not attach to OECs were recovered and used for IVF, they produced lower IVF yields compared to control sperm38. It is plausible that selective attachment to oviduct glycans could replicate these outcomes. In fact, different characteristics of the sperm that are able or not able to adhere to oviduct glycans have been reported. Porcine sperm that attach to soluble or immobilized suLeXare mostly non-capacitated20,22, as is also observed for sperm attached to OECs38,39,40,41, and exhibit specific intracellular Zn2+ distribution21. However, it is important to note that sperm binding in vivo is a complex process, involving not only glycans but also other factors that may cooperate to retain sperm, such as proteins (e.g., annexin 1, 2, 4, and 5) and lipids (e.g., choline phospholipids)70. Therefore, the exclusive use of suLeX to simulate sperm storage and selection in the oviduct may not fully replicate the benefits of an OEC culture. Further studies are needed to validate whether glycan-IVF can indeed improve IVF success, including embryo yield and quality, through the selection of superior sperm.

IVF using porcine cells is particularly inefficient due to the high incidence of polyspermic fertilization, which can be partly explained by the high number of capacitated sperm in contact with the oocyte71,72. Reducing the sperm concentration or IVF incubation time can limit polyspermy but often decreases embryo yield. Instead, glycan-IVF may prevent polyspermy without affecting fertility, by reducing the number of free-swimming sperm while allowing only suitable sperm to fertilize. In the present study, the frequency of polyspermy was reduced when COCs were transferred after sperm were allowed to bind immobilized suLeX, but not other glycans. This result is consistent with the role of the oviduct in suppressing polyspermy73. Furthermore, culture on coverslips coated with fibronectin, an adhesive protein that binds sperm in a glycan-independent manner, did not decrease the polyspermy frequency nor improved the IVF efficiency. This finding supports the hypothesis that a positive selection of sperm may occur through attachment-detachment to suLeX, whereas binding and release from fibronectin may involve a different process or changes in sperm function. Moreover, attachment to suLeX-coverslips extended sperm fertilizing ability. When COCs were transferred 24 h after sperm were allowed to bind to immobilized suLeX, the fertilization frequency and IVF efficiency were greater than those of sperm that were cultured on a coverslip without glycans. In vitro studies with soluble suLeX or suLeX coupled to beads demonstrated that attachment to glycans reduces intracellular ROS levels29 and extends sperm lifespan28. Other mechanisms of sperm survival in the oviduct include the activity of antioxidant-related proteins, such as superoxide dismutase and redoxins, the expression of which is increases after mating74. Although the cryopreservation of porcine sperm has improved in recent years75, extending the lifespan of fresh sperm for use in IVF could prove to be of significant value.

In summary, an in vitro model of the oviduct reservoir was developed by immobilizing oviduct glycans on glass surfaces. This system holds significant potential as a sperm quality assay and IVF technique because it provides a simple method that avoids the biological variability and complexities of OEC cultures. While the suLeX-coverslip is easily adapted to the culture medium droplets commonly used for IVF, the slides may be more indicated for research purposes involving cell staining and microscopic evaluation. Additional studies are needed to determine whether the glycan-IVF system enhances IVF success by promoting fertilization by glycan-bound sperm that detach in the presence of COCs. The consequences of limited sperm selection in vitro are uncertain but may hinder embryonic development by allowing sperm with functional and genetic defects to fertilize oocytes. As more sperm-binding oviduct glycans are identified, they could be incorporated into this functionalized glass system to further represent the complexity of the oviduct epithelium.

Materials and methods

Two functionalized glass surfaces were used to immobilize oviduct glycans and model the oviduct reservoir (Fig. 1): 3D-functionalized streptavidin coverslips (PolyAn Scientific, Berlin, Germany) and functionalized PEG-biotin slides (Microsurfaces Inc., Englewood, NJ). The coverslip customized design (Fig. 1A) included eight circular areas (5 mm diameter) with covalently linked streptavidin that were surrounded by hydrophobic regions to contain medium droplets. The coverslips were maintained at 4ºC in an argon or nitrogen atmosphere until use. The PEG/biotin-slides (Fig. 1B) were coated with 15 µg/mL streptavidin (Sigma-Aldrich, Burlington, MA) diluted with 10% glycerol in 1X PBS, following manufacturer’s instructions, and maintained at −80ºC with a nitrogen atmosphere until use. A Secure-Seal™ hybridization chamber (Grace Bio-Labs, Bend, OR) consisting of 8 chambers (8–9 mm diameter and 1 mm depth) was adhered to the top of the slides to prevent evaporation while culturing sperm. Glycans (Fig. 1C) were covalently coupled to a biotinylated poly[N-(2-hydroxyethyl)acrylamide] core of approximately 10–30 kDa that contained by molarity 20% glycan and 5% biotin76.

Culture media were prepared in the lab following standard protocols. Media components were purchased from Sigma-Aldrich, unless otherwise indicated. The media composition was as follows: phosphate buffered saline (PBS; 136.9 mM NaCl, 2.7 mM KCl, 10.1 mM Na₂HPO₄, 1.8 mM KH₂PO₄), Hepes buffered saline (31.5 mM NaCl, 5.0 mM KCl, 18.2 mM Hepes, pH 7.2), Hank’s buffered salt solution (HBS; 1.3 M NaCl, 40.0 mM KCl, 10.0 mM CaCl₂, 5.0 mM MgCl₂, 140.0 mM fructose, 5% bovine serum albumin, BSA), non-capacitating mTBM (113.1 mM NaCl, 3.0 mM KCl, 7.5 mM CaCl₂, 20.0 mM Tris base, 11.0 mM glucose, 5.0 mM Na-pyruvate, 0.2% BSA), differential motility Tyrode’s albumin lactate pyruvate (dmTALP; 100.0 mM NaCl, 3.1 mM KCl, 2.1 mM CaCl₂, 1.5 mM MgCl₂, 0.3 mM KH₂PO4, 10.0 mM NaHCO3, 1.0 mM Na-pyruvate, 25.0 mM Hepes, 0.4% lactic acid, 0.6% BSA, pH 7.3), oocyte handling medium (TCM-199 buffered with 25 mM HEPES and supplemented with 0.05 mg/mL gentamicin), IVM medium (TCM-199 supplemented with 3.05 mM glucose, 0.91 mM Na-pyruvate, 0.50 mM cysteine, 1% penicillin-streptomycin-amphotericin, 1% polyvinyl alcohol, 0.01 U/mL luteinizing hormone from human pituitary, 0.01 U/mL follicle stimulating hormone from porcine pituitary (MP Biomedicals™, Irvine, CA), 10 ng/mL epidermal growth factor, embryo culture medium (PZM-5 from Cosmo Bio USA Inc., Carlsbad, CA, supplemented with 0.4% BSA).

Unless otherwise stated, the rest of the materials were obtained from Thermo Fisher Scientific Inc. (Waltham, MA).

Sperm preparation

Ejaculated porcine semen was obtained from fertile boars from a commercial source (PIC, Hendersonville, TN). Fresh semen was kept at 15–18ºC for no more than 3 days and prewarmed to 37ºC before use. Only samples with more than 80% motility (assessed before each experiment) were used.

For the experiments with coverslips, 500 µL of pooled semen (from two boars) was layered on top of a discontinuous Percoll gradient consisting of 1.5 mL of 45% Percoll and 1.5 mL of 90% Percoll. To make the 90% Percoll, Percoll® (Cytiva, Marlborough, MA) was diluted with 10x HBS. The 45% Percoll was prepared by diluting 90% Percoll with Hepes buffered saline. The Percoll gradient was centrifuged at 600 X g for 20 min. The supernatant was discarded, and the sperm pellet was resuspended with 2 mL Hepes buffered saline and centrifuged again at 800 X g for 5 min. The supernatant was discarded again, and the pellet was resuspended with 100 µL mTBM.

For the experiments with slides, 2 mL of pooled semen was washed through a Percoll cushion containing 4 mL of dmTALP, 0.6 mL of HBS and 5.4 mL of Percoll®, and centrifuged for 10 min at 800 X g. The supernatant was discarded, and the resulting pellet was resuspended in 10 mL of dmTALP and centrifuged for 5 min at 600 X g. The supernatant was again discarded, and the pellet was resuspended in 1 mL of dmTALP.

After washing, the sperm concentration was assessed by counting with a hemocytometer and adjusted by diluting with the specific medium for each experiment. Different media were used based on the specific culture conditions employed for coverslips and slides (mTBM and dmTALP, respectively). Coverslips, which were used for IVF, were incubated under a controlled 5% CO₂ atmosphere, where mTBM maintains an optimal pH of 7.2–7.4 for IVF. In contrast, slides were cultured under atmospheric CO₂ conditions, where dmTALP, adjusted to a pH of 7.2, performs well in supporting sperm viability and capacitation.

Sperm binding to glycans on streptavidin-modified coverslips

Functionalized coverslips containing 8-regions coated with streptavidin were placed in culture dishes (100 mm diameter). Droplets (25 µL) of mTBM containing 0.02 mg/mL of biotinylated suLeX or corresponding controls (see experimental design) were placed on the streptavidin regions, immediately covered with mineral oil and incubated at room temperature (~ 22⁰C) for 30 min. The droplets were then washed five times to remove uncoupled glycans by doubling the volume of the droplet with mTBM and removing the volume added. Washed sperm were added to each droplet up to a volume of 50 µL (see experimental design for Sperm Binding with Increasing Sperm Concentration) and incubated at 39⁰C in humidified air with 5% CO2 for 30–90 min (see experimental design for Sperm Binding During The Incubation Time) to enable sperm binding. Free sperm were removed by aspirating and replacing half of the medium five times.

Sperm binding to glycans on streptavidin-modified slides

Streptavidin-coated slides were mounted with SecureSeal™ Hybridization Chambers to form eight independent culture sites with a volume of approximately 65 µL. Biotinylated suLeX or corresponding controls (see Experimental Design) diluted in dmTALP (0.02 mg/mL) was added to each chamber. The mixture was incubated at ~ 22⁰C for 30 min. The medium from each chamber was removed and refilled twice with mTBM to remove uncoupled glycans. Pre-washed sperm were added to each chamber (see experimental design for specific sperm concentrations) and incubated at 39⁰C in humidified air for 30–90 min (see experimental design for Sperm Binding With Increasing Sperm Concentration) to enable sperm binding.

Assessment of sperm binding density

To assess the number of bound sperm, the coverslips and slides were observed under an inverted Zeiss HBO50 A/C microscope (Zeiss, Oberkochen, Germany) with 200X or 400X magnification. Five images were taken within each droplet and chamber using an Axiocam 503 mono and ZEN 2 Lite software. The sperm in each image were counted using ImageJ (NIH, Bethesda, MD) and divided by the average size (mm2) of the images. Only sperm bound to the bottom with visible heads were counted. Agglutinated sperm were not included in the final count. The average of the 5 images was calculated to determine the number of sperm bound per mm2 in each droplet/chamber (binding density).

Oocyte collection and in vitro maturation

Ovaries from adult sows were obtained from an abattoir located 80 miles from the laboratory (Calihan Pork Processors, Peoria, IL), maintained in warm 0.9% saline, and processed within three h of collection. Oocytes were aspirated from follicles between three and eight mm in diameter using an 18-gauge needle attached to a 10 mL syringe. The collected fluid was placed in a conical tube after aspiration allowing oocytes and debris to form a pellet at the bottom of the tube. The pellet was resuspended in oocyte handling medium. Using a dissecting microscope, oocytes with at least two layers of compact cumulus cells and homogenous cytoplasm were selected. Cumulus-oocyte complexes (COCs) were transferred in groups of approximately 40 COCs to Nunc™ Cell-Culture dishes containing 500 µL IVM medium per well and covered with 200 µL of mineral oil. COCs were incubated for 42 h at 39⁰C in humidified air with 5% CO₂77.

In vitro fertilization and embryo culture

After 42 h of IVM, the COCs were co-cultured with sperm following either conventional IVF procedures or with a modified protocol on the glycan-coated coverslip system (glycan-IVF, Fig. 5) for 5 h at 38.5⁰C in humidified air with 5% CO2. For conventional IVF, mature COCs were first transferred to 25 µL droplets on a culture dish (30 mm diameter) and 25 µL of pre-washed sperm were added up to a final concentration of 0.1 × 106 sperm/mL. For glycan-IVF, sperm were bound to glycan-coated coverslips before the transfer of mature COCs. Briefly, coverslips were coated with suLeX as described for the glycan binding assay, sperm (0.25 × 106 sperm/mL) were incubated at 39⁰C in humidified air with 5% CO2, and after 30 min free sperm were removed. In both systems, sperm-COC culture was performed in 50 µL droplets of mTBM supplemented with 2.5 mM caffeine (capacitating mTBM) and approximately 20 COCs were cultured per droplet. After 5 h, presumptive zygotes were denuded of cumulus cells by pipetting with 1% hyaluronidase. Zygotes were then transferred to 50 µL droplets of embryo culture medium (~ 20 COCs/droplet) and further incubated for 7 h under the same culture conditions.

Assessment of pronuclear formation

After 12 h from the beginning of IVF, presumptive zygotes were fixed in 100% ethanol containing 1 µL/mL Hoechst33342 and kept at 4⁰C overnight. They were mounted on slides with one drop of Vectashield Antifade Mounting Medium (Vector Labs, Burlingame, CA) and observed under a Zeiss Axiovert 200 M (λex 365 nm, λem 445 nm). Zygotes were classified according to the presence of pronuclei (PN) as follows: fertilized (at least two visible PN), non-fertilized (no visible PN), asynchronous (presence of 1 PN and one condensed sperm head), monospermic (presence of two PN), or polyspermic (presence of three or more PN). In addition, unfertilized oocytes were classified as follows to evaluate the state of maturation: metaphase II (chromatin condensed in two sets of chromosomes), metaphase I (chromatin condensed in one set of chromosomes), or germinal vesicle (uncondensed chromatin inside spherical nucleus).

Experimental design

  1. 1.

    Validation of the glycan-coverslip system for binding porcine sperm.

1.1. Sperm binding to different concentrations of suLeX.

The streptavidin circular areas on the functionalized coverslips were coated with different concentrations (0.02, 0.2 and 1 mg/mL) of biotinylated suLeX and biotinylated N-acetyllactosamine (LacNAc) in a 25-µL droplet, following the previously described protocol. Sperm were added to each droplet increasing the volume to 50 µL (1 × 106 sperm/mL) and incubated for 30 min. After washing to remove free sperm, the binding density was determined.

1.2. Specificity of sperm binding to suLeX.

The streptavidin circular areas on the functionalized coverslips were coated with 0.02 mg of biotinylated suLeX, biotinylated LacNAc, biotin-labeled fibronectin purified from bovine plasma (Cytoskeleton, Denver, CO), or with no molecule (control) in a 25-µL droplet. Sperm (1 × 106 sperm/mL) were incubated for 30 min in each droplet (50 µL) before the binding density was determined.

1.3. Sperm binding during the incubation time.

The streptavidin circular areas on the functionalized coverslips were coated with 0.02 mg/mL of either biotinylated suLeX or biotinylated sulfated Lewis A (suLeA) in a 25-µL droplet. Sperm (1 × 106 sperm/mL) were added to each droplet (50 µL) and incubated for a total of 90 min. The binding density was evaluated after 30, 60, and 90 min in the suLeX droplets and after 90 min in the suLeA droplets.

1.4. Sperm binding with increasing sperm concentration.

The streptavidin circular areas on the functionalized coverslip were coated with 0.02 mg/mL of either biotinylated suLeX or biotinylated suLeA in a 25-µL droplet. Sperm at different concentrations (0.25, 1, and 5 × 106 sperm/mL) were added to the suLeX droplets (50 µL) and 5 × 106 sperm/mL were added to the suLeA droplets (50 µL). The coverslip was incubated for 30 min before the binding density was assessed.

  1. 2.

    Validation of the glycan-slide system for binding of porcine sperm.

2.1. Specificity of sperm binding to suLeX.

The chambers on the streptavidin-slides were coated with 0.02 mg of either biotinylated suLeX, biotinylated LacNAc, or biotin-labeled fibronectin (65 µL), or just 65 µL of mTBM medium (control) following the previously described protocol. Sperm (1 × 106 sperm/mL) were added to each chamber and incubated for 30 min. After removing free sperm, the binding density was determined.

2.2. Sperm binding during the incubation time.

The chambers on the streptavidin-slides were coated with 0.02 mg of either biotinylated suLeX (six chambers) or biotinylated suLeA (two chambers) (65 µL). Sperm (1 × 106 sperm/mL) were added to each chamber and incubated for a total of 90 min. The binding density was evaluated after 30, 60, and 90 min in the suLeX chambers and after 90 min in the suLeA chambers.

2.3. Sperm binding with increasing sperm concentration.

The chambers on the streptavidin-slides were coated with 0.02 mg of either biotinylated suLeX (six chambers) or biotinylated suLeA (two chambers) (65 µL). Sperm at different concentrations (0.25, 1, and 5 × 106 sperm/mL) were added to the suLeX chambers, and sperm at 5 × 106 sperm/mL were added to the suLeA chambers. The system was incubated for 30 min before assessing the binding density.

  1. 3.

    Assessment of sperm release from glycans immobilized on glass coverslips.

3.1. Sperm release from glycan-coverslips induced by progesterone and 4-aminopyridine.

Following the previously described protocol, the functionalized coverslips were coated with 0.02 mg of biotinylated suLeX in a 25-µL droplet. Sperm were added to each droplet increasing the volume to 50 µL (1 × 106 sperm/mL) and were incubated for 30 min. After removing free sperm, the following releasing treatments were added: control mTBM, capacitating mTBM (supplemented with 2.5 mM of caffeine), 800 nM of progesterone diluted in capacitating mTBM, 4 mM of 4-AP (Tocris Bioscience, Bristol, UK) diluted in capacitating mTBM. The coverslips were incubated for one hour at 39ºC in humidified air with 5% CO2 to enable sperm release. After removing free sperm, the binding density was evaluated.

3.2. Sperm release from glycan-coverslips induced by mature COCs.

Following the previously described protocol, the functionalized coverslips were coated with 0.02 mg of biotinylated suLeX in a 25-µL droplet. Sperm were added to each droplet increasing the volume to 50 µL (0.25 × 106 sperm/mL) and were incubated for 30 min. After removing free sperm, the media on the droplets were replaced with capacitating mTBM and the following releasing treatments were added (two droplets per treatment): 20 mature COCs (42 h of IVM), 20 denatured COCs (induced by heating in a warm bath at 60⁰C for 5 min), and only capacitating mTBM (control). The coverslips were incubated for 5 h at 39⁰C in humidified air with 5% CO2 to enable sperm release. After removing COCs and free sperm, the binding density was evaluated.

  1. 4.

    Validation of the sperm fertilizing ability on a glycan-coated coverslip system (glycan-IVF).

4.1. Fertilizing capacity of sperm bound to glycans pre-fertilization compared to conventional IVF.

IVF was performed following conventional and glycan-IVF protocols, as described above. After 12 h from the beginning of sperm-COC culture, the PN formation was assessed to determine the percentage of fertilized oocytes (total number of zygotes/total oocytes), the rate of polyspermic zygotes (polyspermic zygotes/total zygotes) and the IVF efficiency (monospermic zygotes/total oocytes). A total of 95 oocytes were analyzed per system.

4.2. Fertilizing capacity of sperm bound to glycans pre-fertilization over time.

IVF was performed following the glycan-IVF protocol with some modifications. The streptavidin circular areas on the functionalized coverslips were coated with 0.02 mg of either biotinylated suLeX, biotinylated suLeA, biotin-labeled fibronectin, or mTBM alone. Sperm (0.1 × 106 sperm/mL) were incubated for 30 min to allow sperm binding followed by an additional incubation of 0, 6, 12, or 24 h. Mature COCs were then directly transferred to the droplets without previously removing free sperm and incubated at 39⁰C in humidified air with 5% CO2. After 5 h, COCs were denuded, transferred to embryo culture medium, and incubated for an additional 7 more h. After 12 h from the beginning of the sperm-COC culture, PN formation was assessed to determine the frequency of fertilized oocytes, the frequency of polyspermic zygotes, and the IVF efficiency. Between 75 and 103 oocytes were assessed per treatment and culture time.

Statistical analysis

For all the experiments, three replicates were performed with two droplets/chambers (i.e., experimental unit) per treatment group (n = 6/group). For Experiment 4.2., each of the four incubation times was analyzed as a separate experiment, as they were performed independently. The data were analyzed using RStudio 2022.07.2 + 576 for Windows. Differences among means were determined using two-way analysis of variance followed by Tukey’s multiple-comparison test. Treatment was set as the fixed factor and replication as the random variable. Before analysis, the Shapiro-Wilk test was used to determine data normality. Data from binding density data followed a normal distribution. Data from pronuclear formation that did not present a normal distribution were square root arcsine transformed before analysis, and the normality of variance was validated after transformation. The results are shown as means ± SEM and differences were considered to be statistically significant when the p-value was less than 0.05.