Sperm selection by the oviduct: perspectives for male fertility and assisted reproductive technologies

Abstract

The contribution of sperm to embryogenesis is gaining attention with up to 50% of infertility cases being attributed to a paternal factor. The traditional methods used in assisted reproductive technologies for selecting and assessing sperm quality are mainly based on motility and viability parameters. However, other sperm characteristics, including deoxyribonucleic acid integrity, have major consequences for successful live birth. In natural reproduction, sperm navigate the male and female reproductive tract to reach and fertilize the egg. During transport, sperm encounter many obstacles that dramatically reduce the number arriving at the fertilization site. In humans, the number of sperm is reduced from tens of millions in the ejaculate to hundreds in the Fallopian tube (oviduct). Whether this sperm population has higher fertilization potential is not fully understood, but several studies in animals indicate that many defective sperm do not advance to the site of fertilization. Moreover, the oviduct plays a key role in fertility by modulating sperm transport, viability, and maturation, providing sperm that are ready to fertilize at the appropriate time. Here we present evidence of sperm selection by the oviduct with emphasis on the mechanisms of selection and the sperm characteristics selected. Considering the sperm parameters that are essential for healthy embryonic development, we discuss the use of novel in vitro sperm selection methods that mimic physiological conditions. We propose that insight gained from understanding how the oviduct selects sperm can be translated to assisted reproductive technologies to yield high fertilization, embryonic development, and pregnancy rates.

Understanding how sperm are selected during their transport through the oviduct provides the underpinning for rational development of new technologies for treating infertility.

Introduction

Infertility in humans, defined as the inability to become pregnant after 1 year of unprotected sexual intercourse, has a paternal origin in up to 50% of cases (see review [1]). Of these infertile men, it is estimated that around 20% have undetermined infertility, either idiopathic (abnormal semen analysis of unknown cause) or unexplained (infertile but with normal results on conventional semen analysis) [1]. The overlap in the parameters of semen quality between fertile and infertile men implies that the traditional methods used for assessing sperm quality, mainly based on concentration, motility, morphology, and viability, are insufficient [1, 2]. Regarding farm animals, male infertility can have a major economic impact since ejaculates from the same male are used to inseminate thousands of females. Although we lack large-scale studies, a great proportion of livestock reproductive failure is also attributed to male factor, with some estimates as high as 40–50% (see reviews [3, 4]). Thus, there is a growing interest in developing new sperm quality assays that better predict fertility outcomes by assessing sperm ability not only to reach the site of fertilization but also to fertilize and activate successful embryonic development (see review [5]). Indeed, increasing numbers of studies are revealing the hidden contribution of sperm to healthy embryonic development, which includes multiple genetic and non-genetic characteristics such as sperm deoxyribonucleic acid (DNA) integrity, sperm centrioles, and different components of the sperm epigenome (see reviews [6–8]). It is therefore becoming apparent that an abnormal paternal component can lead to early embryonic loss and infertility, even when fertilization is successful.

Even more concerning is the possibility of having a successful pregnancy that becomes an individual with health problems. Some studies in humans have linked advanced paternal age with both a higher prevalence of psychiatric disorders and adverse events in pregnancy [9, 10]. This could be related to epigenetic errors in the sperm, as differentially methylated DNA regions (DMR) involving neurological signaling pathways have been identified between sperm from 50- to 35-year-old men, with corresponding DMR in blastocysts after intracytoplasmic sperm injection (ICSI) [11]. Furthermore, animal studies have shown transgenerational epigenetic abnormalities transmitted from the sperm to the embryo that lead to developmental defects in the offspring [12–14]. Another clear example is sperm DNA damage, which has been reported to induce fragmented paternal chromosomes after in vitro fertilization (IVF) leading to random distribution of the chromosomal fragments in the first cell division and blastocysts with chaotic mosaicism [15]. Sperm DNA integrity testing is now recommended as part of the workup for couples experiencing recurrent pregnancy loss. A high percentage of sperm with DNA damage in the ejaculate is indicative of lower chances of pregnancy in couples attempting natural conception or insemination [16]. Recent data also support a link between oxidative stress, DNA damage, the mutational load transmitted to the embryo, and long-term health issues of the progeny, including imprinting disorders (see review [17]).

On the other hand, it is important to consider how sperm factors affect the success of assisted reproductive technologies (ART). ART, including IVF and ICSI, are common practices in fertility clinics. In 2019, ⁓2.1% of children born in the US were conceived through ART, numbers that are increasing every year [18]. However, despite major improvements in ART over the past 40 years, ART success rate still requires improvement and greater emphasis on the role of the paternal component would be a natural goal. For men presenting with low sperm numbers, poor morphology, and/or motility, the reflex treatment is to perform ICSI. However, other characteristics that cannot be compensated for, such as DNA damage, messenger ribonucleic acid (mRNA), and chromatin packaging, can have greater consequences for ART success, as they affect the activation and progress of embryonic development [6–8]. Furthermore, the most prevalent sperm selection techniques used for ART (swim-up and density-gradient centrifugation) are focused on sorting by motility and viability [19], which may still allow defective sperm to fertilize.

The transfer of in vitro-produced (IVP) embryos in farm animals has also drastically increased over the past 20 years and is now a major industry in cattle. IVF interest relies on its potential to accelerate genetic progress compared with natural mating and artificial insemination (AI) [20] and to allow the transfer of genetically interesting traits worldwide while avoiding biosecurity risks. According to data collected by the International Embryo Technology Society in 2020, a total of 878 181 IVP embryos were transferred globally, compared with 294 670 in vivo-derived embryos [21]. In addition, IVF is the cornerstone for technologies such as transgenesis and stem cell reprograming. However, as in humans, the success of IVF in animals is not optimal. For instance, in cattle, it was estimated that only 30% of immature oocytes reach the blastocyst stage and 12% lead to a live birth [22]. The idiosyncrasies of some species limit even more the success of this technology. In pigs, IVF is especially inefficient due to a high incidence of polyspermic fertilization, which is partly explained by the high number of capacitated sperm in contact with the oocyte in IVF (see reviews [23, 24]).

Overall, this information has resulted in growing research interest to understand natural sperm selection for improving fertility outcomes (see reviews [25, 26]). Sperm transport through the female reproductive tract entails a selection process that dramatically reduces the numbers arriving at the fertilization site (see review [27]). As we will discuss in the following sections, there is also an abundance of evidence that the last portion of the female reproductive tract where fertilization occurs, the oviduct, may play a major role in sperm selection. Once considered just a “pipeline” that connected the ovary with the uterus, the oviduct is now known to modulate sperm viability, capacitation, and function (see reviews [28, 29]).

This paper aims to review natural sperm selection by the oviduct and its importance for fertility. Understanding the sperm-oviduct crosstalk may aid in developing new technologies for treating infertility, including sperm quality assays and selection protocols for ART that better resemble physiological conditions.

Sperm selection in the female reproductive tract

During the journey of sperm through the female reproductive tract toward the egg, sperm encounter many obstacles that regulate their movement and survival (see reviews [29–33]. These include the acidic pH of the vagina, the viscous mucus and folds in the cervix, and an inflammatory response in the uterus. Overall, these barriers favor viable sperm with an intact plasma membrane, normal morphology, and progressive motility. In the present review, we will focus though on the last part of the female reproductive tract, the oviduct. Once sperm are at the anterior portion (fundus) of the uterine horn, they go through the utero-tubal junction (UTJ) to enter the oviduct and move to the ampulla, where they will meet and fertilize the oocyte. Some sperm arrive in the oviduct within minutes in what is referred to as the rapid transport phase, first described in the rabbit [34], which includes uncapacitated and often dead cells. These are distinguished from sperm in the sustained transport phase that arrive in the oviduct later and are usually the fertilizing sperm. The oviduct is a complex organ with distinct morphological and functional regions: the UTJ, the isthmus, the isthmus-ampullary junction, the ampulla, and the infundibulum. Different factors affect sperm migration in the oviduct, mainly anatomical (the intricate folding of the mucosa that limits the luminal space), rheological (the viscosity and flow of the oviductal fluid), and chemical (chemoattractants and chemorepellents present in the oviductal fluid). In addition to these regions, there is also temporal regulation related to the timing of the estrous cycle.

Of the millions of sperm that are ejaculated, only a very small number (10–1000 sperm) reach the ampulla (see reviews [26, 27]). The limited number of sperm reaching the oocyte allows the oocyte to minimize polyspermy. Whether this population of sperm that reach the site of fertilization is also superior in terms of fertilizing competence is not fully understood. Strong evolutionary evidence from multiple species, including insects, birds, reptiles, and mammals, makes it reasonable to believe that the female reproductive tract would select fit sperm for ensuring fertilization and healthy embryonic development (see review [27]). Holt and Fazeli [27] proposed a “molecular sperm passport hypothesis” in which females exert a cryptic choice based on a molecular dialog between the sperm and the female reproductive tract. This hypothesis is supported by studies in domestic animals in which heterospermic inseminations (inseminations with two ejaculates at the same time) have shown skewed fertilization success toward one of the males, even if both have similar sperm quality parameters (see review [35]). An example is a study in pigs in which 11 females were inseminated with semen from two boars that resulted in a 7:3 ratio of embryo paternity, which was correlated with higher sperm binding to zona pellucida (ZP) when the same ejaculates were used for homospermic inseminations [36]. Similarly, in a larger study, the zona-binding ability was a significant component of an algorithm that estimated fertility with 70% success in heterospermic inseminations [37]. In another study of cattle, the biased fertilization success toward one of the males was positively correlated with sperm progressive motility and membrane integrity, and negatively correlated with DNA fragmentation [38]. Others point out that sperm selection in the female reproductive tract may involve compatible genetic characteristics between male and female or even the sex chromosome cargo of the sperm (see review [39]). Specifically, in a study in humans, the female reproductive tract produced antibodies against N-glycolylneuraminic acid (Neu5Gc), which removed only the sperm with that specific sialic acid in their membrane [40]. In pigs, X- and Y-chromosome-bearing sperm elicited a distinct transcriptomic response in contralateral oviducts, with immune-related genes being altered most [41].

Regarding the oviduct, its role in sperm selection was elegantly shown in the early 1970s in an in vivo study in rabbits. Cohen and McNaughton showed that sperm recovered high in the oviduct were the most potent sperm to conceive when re-inseminated into rabbits [42]. We will present evidence below of selection of sperm for characteristics mediated by a variety of mechanisms, including the interaction with oviductal epithelial cells (OECs) and cumulus-oocyte complexes (COCs) (Figure 1). However, these results should be taken with caution as they are limited to in vitro studies due to the difficulty of studying gametes in vivo. Moreover, they do not account for the selection that takes place in the lower female reproductive tract. In other words, sperm added to oviduct epithelial cells in vitro may not represent the population entering the oviduct. Only very recently have some technologies allowed the characterization of sperm in situ at specific locations of the mouse oviduct. For instance, a recent study applied tissue clearing and a three-dimensional confocal imaging approach to characterize the main sperm calcium channel, CatSper [43]. They observed a lower percentage of sperm with intact CATSPER1 close to the UTJ compared with the ampulla and hypothesized that CATSPER1 acts as a marker of sperm death and elimination via phagocytosis in the oviduct, triggered in a time and space-dependent manner [43]. As more of these technologies are applied, we will be able to validate some of the current information about the interactions between gametes and the oviduct.

Figure 1.

Locations and mechanisms of sperm selection in the oviduct.

Sperm migration through the utero-tubal junction

Studies in pigs [44] and sheep [45] have revealed that only around 1 out of 10 million of the inseminated sperm reach the oviduct. In humans, only around 1000 sperm are found in the Fallopian tubes 8–15 hour after intercourse [46]. The UTJ presents mucosal folds, a constricted lumen, and a viscous mucus, acting as a physical barrier that dramatically reduces the number of sperm passing deeper into the oviduct. For instance, in sows, 24 hour after intrauterine insemination, of all the sperm recovered, 42% were found in the UTJ and only 0.5% in the isthmus [47]. Moreover, the UTJ acts as a selective barrier, with studies in pigs, mice, and hamsters, showing that all sperm arriving at the isthmus are acrosome intact [48], and a higher proportion relative to the uterus are alive [49], uncapacitated [50], and morphologically normal [51]. More intriguingly, a study in mice revealed that the UTJ may also select sperm according to their DNA integrity [52]. After mating with males pre-treated with γ-radiation and scrotal heat stress to induce DNA damage, sperm recovered from the oviduct present fewer DNA breaks than sperm recovered from the uterus [52]. Alternately, in this experiment, the UTJ might be selecting against sperm that have been damaged in other ways in addition to DNA strand breaks.

Other sperm functional characteristics are also decisive to render entrance to the oviduct. At least 18 sperm proteins play a role in sperm migration through the UTJ, including proteins from the A Disintegrin and Metalloprotease domain (ADAM) family (see reviews [53, 54]). Mouse knock-out (KO) models for these proteins are infertile due to the inability of their sperm to migrate through the UTJ after mating, although presenting normal sperm counts and morphology. They also show impaired ZP binding during fertilization of cumulus-free oocytes. The mechanistic cause of this phenotype is still unresolved, but it may be related to impaired motility and hyperactivation, although this is not clearly observed for all the proteins required for UTJ transport. Interestingly, a KO mouse for testis-expressed gene 101 (Tex101), which codes for a glycosylphosphatidylinositol-anchored protein that interacts with ADAM3, has revealed that sperm cooperative behavior may be involved in the entrance to the oviduct [55]. In this study, wild-type sperm formed unidirectional clusters on the UTJ, whereas Tex101-null sperm only formed small groups with disordered orientation [55]. The ability to form the cluster seems related to the swimming pattern, which in Tex101-null sperm is more asymmetrical, preventing sperm from attaching. Their hypothesis is that sperm tails beating in a synchronized manner in the clusters generate a strong tail pendulum force transiently pushing the UTJ open. This hypothesis needs further investigation, but it is consistent with similar observations of sperm deficient in other proteins. It is also important to consider that all these studies come from mouse models and the importance of these proteins has not been identified in other animals. Nevertheless, the UTJ seems to select a subpopulation of sperm that can bind the ZP and fertilize oocytes.

Sperm storage in the isthmus

After passing the UTJ, the mammalian sperm forms a reservoir by binding the epithelium of the isthmus (see review [56]). This increases the odds of fertilization when there is asynchrony between mating and ovulation, that is when ovulation is delayed after mating. In pigs and cattle, sperm are stored for 24–48 hours, but storage can last for several days, months, and even years in other vertebrates and invertebrates with internal fertilization [56]. Functional sperm storage is also observed in the human Fallopian tube, although the mechanisms have been less investigated [57]. Once the sperm are released, they can ascend to the ampulla to fertilize the oocyte. Some studies suggest that this release is mediated by the hormonal milieu around ovulation [58–60] and by capacitation [61–63], ensuring that sperm would be near the oocyte at the appropriate time.

There are several studies showing that sperm incubation with OECs can improve IVF success, perhaps due to effects on sperm adhesion to the OECs [64–66]. Sperm heads bind specific oligosaccharides containing glycan motifs that are part of the OECs glycocalyx (see reviews [67, 68]). For instance, in pigs, sulfated Lewis X trisaccharide (suLe^X) and a biantennary 6-sialylated oligosaccharide (bi-SiaLN), motifs found in larger glycans, are responsible for binding and retaining sperm [61, 63]. Both motifs are attached mostly to N-linked glycoproteins localized on the apical surface of the porcine isthmus. Blocking the sperm membrane proteins with soluble suLe^X and bi-SiaLN reduced sperm binding to OECs by 60% [61, 63]. Similarly, bovine sperm bind sulfated Lewis A (suLe^A) trisaccharide closely related to suLe^X [69, 70]. In humans, an oviduct glycan related to the glycan on asialofetuin could be responsible for retaining sperm [71]. The sperm membrane proteins responsible for binding to the oviduct epithelium have also been investigated (see review [33]). The accessory gland protein AQN1 binds mannose and galactose residues of oviduct cells, although it does not bind Le^X and bi-SiaLN motifs [72, 73]. Similarly, specific bovine seminal plasma proteins (BSPs) mediate sperm binding to fucose residues on oviduct epithelial glycans [74, 75]. Interestingly, capacitation induced the loss of these proteins, pointing to a possible mechanism of sperm release from the oviduct [72, 75]. In humans, the carbohydrate-binding protein fucosyltransferase-5 has been identified as a candidate mediating sperm-oviduct binding [71]. However, sperm from the cauda epididymis that have not been exposed to seminal plasma proteins can still bind OECs [76], which suggests that other sperm receptors may be involved. Consistent with this, lactadherin, a suLe^X-binding protein on porcine epididymal sperm, has been implicated in binding to the oviduct [77].

Several articles have observed a correlation between the ability of sperm to bind OECs and fertility outcomes, supporting the importance of the sperm reservoir for fertility (see Table 1). Some sperm characteristics that are routinely assessed for measuring the quality of an ejaculate have also been correlated to the percentage of sperm that bind OECs, for instance morphology [78], and plasma membrane integrity [79, 80] (Table 1). Other studies highlight the relationship between fertility, the ability to bind the oviduct, and β-defensins, which are immune-related peptides present on the sperm membrane that are involved in their survival in the female reproductive tract [81]. Thus, a haplotype that includes various β-defensins genes was related to bull fertility and the ability of sperm to bind the oviduct epithelium [82]. Yet, the formation of sperm storage may not be necessary for fertility, particularly if semen deposition is timed closely to ovulation. As previously discussed by Hunter [83], sperm were able to fertilize after surgical insemination directly into the ampulla in rabbits, sheep, and pigs. Moreover, in a recent study in cattle, the authors only observed a small, although significant, difference in sperm binding to OECs between high- and low-fertility males, concluding that the sperm’s ability to bind the oviduct cannot alone explain the large fertility variation among bulls [84].

Table 1.

Correlation between sperm binding to the oviduct and fertility

Species	Oviduct Model	Results	References
Bull	OEC aggregates	Sperm-binding capacity from frozen–thawed samples is positively correlated with non-return rates after AI (only when membrane integrity is >60%).	[194]
Bull	OEC aggregates	Sperm-binding capacity from frozen–thawed samples is positively correlated with the conception rate.	[103]
Boar	Ex vivo oviduct explant	Ejaculates from subfertile males have a lower binding index compared with the mean of the population. The rate of cytoplasmic droplets in a sperm sample is negatively correlated with the binding index.	[78]
Bull	OEC aggregates	Ejaculates from males with higher non-return rate after AI have higher binding index compared with males with lower non-return rate. The sperm-binding capacity is positively correlated with the relative volume shift in response to hypo-osmotic stress (indicative of membrane integrity).	[79]
Bull	OEC aggregates	Ejaculates from males with high-field fertility have a higher binding index compared with low-fertility males.	[84]
Buffalo Bull	OEC aggregates	Sperm-binding index of frozen–thawed sperm ejaculates is positively correlated with the conception rate after AI. The binding index is negatively correlated with the rate of sperm with compromised membranes.	[80]
Boar	OEC aggregates	A multiple linear regression model that includes sperm-binding capacity, among other quality parameters (morphology, motility, and boar stud entry date), is predictive of pregnancy rate after AI.	[193]

Binding Capacity or Binding Index = sperm bound per unit area of oviduct explant, OEC monolayers, or OEC aggregates.

In domestic animals, it is also important to consider that frozen–thawed semen is commonly used for AI. The cryopreservation and thawing processes, as well as the use of protein extenders, induce biochemical and physiological changes in the sperm, including alterations of plasma membrane proteins that may affect the interaction with the oviduct (see review [85]). In fact, studies of bovine and equine sperm have shown a decrease of their ability to bind OECs following cryopreservation [86, 87]. And in porcine sperm, the use of Beltsville Thawing Solution extender for following storage at 17°C reduced the sperm binding to oviduct explants compared with Androstar® [88].

Many reports indicate that sperm in close contact with the oviduct may improve fertility by regulating sperm function (see reviews [28–30]). Interestingly, some of the effects on sperm physiology are mediated by direct membrane interaction, such as the maintenance of sperm viability and the prevention of capacitation [89–93]. For instance, sperm attached to OECs maintained reduced intracellular Ca²⁺, a mechanism that may be involved in the regulation of capacitation [91, 93]. Similarly, the binding of porcine sperm to the oviduct glycan suLe^X attached to beads maintains their viability and the binding to insoluble suLe^X lowers intracellular Ca²⁺ concentrations, compared with free-swimming sperm [94]. This indicates that recognition and binding to oviduct glycans are enough to trigger effects on sperm function. However, the specific mechanisms by which sperm–glycan interaction lengthens sperm viability and possibly delays capacitation are still unknown. Some potential intermediaries are a decrease in reactive oxygen species (known to be involved in sperm death and capacitation as reviewed by Aitken et al. [95]), adenosine triphosphate (ATP) consumption (related to a decrease in motility), and the plasma membrane remodeling that normally occurs during capacitation.

At the same time, sperm interaction modifies the function of the oviduct. Bovine OECs co-incubated with sperm responded by the de novo synthesis of proteins, an effect that was partly prevented when a porous membrane blocked the direct contact between sperm and OECs [96]. Porcine sperm also induced an upregulation of heat-shock proteins in OECs only when directly in contact with the cells [97]. Moreover, the live imaging of an ex vivo bovine oviduct explant revealed an increase in the ciliary beat frequency of OECs after 15 minutes of the addition of sperm [98]. In the same study, transmission electron microscopy of the OECs after sperm incubation revealed an active endoplasmic reticulum in ciliated and secretory oviduct cells, which suggests an increase in protein synthesis and trafficking [98]. As discussed by Kölle [99], the oviduct may also respond to sperm interaction with the secretion of nutrients that would support sperm viability.

A more intriguing hypothesis is that the interaction with the oviduct not only modifies sperm function but also selects a sperm population of superior characteristics with greater embryonic developmental potential. In other words, only sperm with specific characteristics would bind the isthmus, survive, and continue traveling to the ampulla to fertilize the oocyte. A plethora of in vitro studies supports this narrative (see Table 2). Overall, they show that OECs preferentially bind uncapacitated sperm [64, 93, 100, 101] with higher DNA integrity [102, 103]. Some of the differences observed between bound and free-swimming sperm were detected only after 3 minute of incubation [93]. This supports a selective process based on sperm’s ability to bind the oviduct epithelium, rather than just be in close contact with the OECs. Some IVF studies achieved higher fertilization success by selecting sperm that adhered to OECs after a co-incubation period [64–66, 104], denoting that the selected sperm characteristics are also important for embryonic development. Two studies have demonstrated that both porcine and bovine sperm that were damaged by sex sorting using flow cytometry had reduced binding to oviduct cells (see reviews [98, 105]). Even though sex-sorted sperm had a decreased binding to the oviduct epithelium, the motility characteristics of the bound sperm were similar between normal and sex-sorted sperm [98]. Although most of these studies come from animal models, a study in humans revealed a similar selective binding to OECs [102]. Thus, a higher proportion of bound sperm presented normal motility, a functional membrane, an intact acrosome, and higher chromatin integrity compared with free-swimming sperm [102]. Several studies using porcine cells indicate that oviduct glycans are responsible for selective sperm binding to the isthmus. First, a lower proportion of capacitated compared with uncapacitated sperm bind soluble or immobilized suLe^X [61, 63]. Second, only sperm presenting specific Zn²⁺ signatures, according to its intracellular distribution, bind oviduct glycan-coated beads [62]. Immobilized oviduct glycans also model other oviduct functions. Binding to OEC and immobilized glycans extends the lifespan of porcine sperm [94]. Thus, the selectivity of sperm binding to oviduct cells also seems to extend to adhesion to oviduct cell glycans.

Table 2.

Evidence of sperm selection by binding to OECs

Parameter	Species	Oviduct Model	Results	References
DNA Damage	Bull	OEC aggregates	Sperm-binding index from frozen–thawed samples is negatively correlated to sperm DNA fragmentation index (sperm chromatin structure assay).	[103]
	Man	Bovine OEC monolayer	Bound sperm have lower abnormal chromatin and mean ratio of denatured DNA (sperm chromatin structure assay) compared with free sperm.	[102]
Capacitation	Boar	OEC monolayer and ex vivo oviduct explant	Bound sperm show lower levels of tyrosine phosphorylation and different phosphotyrosine distribution patterns (immunofluorescence) compared with free sperm.	[101]
	Boar	OEC suspension	A higher proportion of bound sperm are uncapacitated (fluorescent antibiotic chlortetracycline staining) compared with control samples.	[100]
	Boar	OEC monolayer	Free sperm have higher levels of tyrosine phosphorylation (immunofluorescence) and higher rates of phosphatidylserine translocation (Annexin V-Cy3™ Apoptosis Detection Kit) compared with bound and unselected sperm.	[64]
	Boar	OEC monolayer	Bound sperm show lower cytosolic calcium concentration (fluo-3-AM staining) and tyrosine phosphorylation of membrane proteins (distribution of Cy-3 staining), and higher membrane integrity (propidium iodide) and motility, compared with free sperm.	[93]
Fertilization Competence	Bull	OEC monolayer	Bound sperm, collected after heparin-induced release, show higher ZP binding and cleavage rate after IVF, compared with free sperm and sperm cultured with a porous insert that prevented direct contact with OECs.	[65]
	Bull	OEC suspension	Free sperm show lower fertilization rate after IVF and mean number of sperm bound to the ZP than control sperm. Free sperm have lower membrane integrity (hypo-osmotic-swelling-test).	[104]
	Bull	OEC monolayer	Bound sperm, collected after progesterone-induced release, produce higher cleavage and blastocyst rates after IVF, compared with control sperm.	[66]
	Boar	OEC monolayer	Free sperm produce lower penetration and fertilization rates compared with bound and control sperm.	[64]

Binding Index = sperm bound per unit area of oviduct explant, OEC monolayers, or OEC aggregates. Control = sperm that was not cultured with OECs.

Sperm transport through the oviduct and guidance to the oocyte

Using the mouse, a species in which sperm can be observed in an intact oviduct, it was demonstrated that peristalsis moves free-swimming sperm through the oviduct, particularly the isthmus [106]. In addition, there are other mechanisms that regulate mammalian sperm guidance to the ampulla and the oocyte: rheotaxis, thermotaxis, and chemotaxis (see review [31]).

Rheotaxis refers to the motility oriented against a fluid flow. In vitro studies using microfluidic chambers in different species, including humans, have revealed that sperm present rheotactic behavior [31]. Furthermore, rheotaxis appears to guide sperm within the oviduct, as the blockage of the fluid flow by clamping the UTJ in an ex vivo culture reduced the sperm arriving at the ampulla [107]. Moreover, sperm that swim against the flow gradient have better motility parameters independent of the capacitation state [107]. This indicates that only more fit sperm show a rheotactic response. However, there is controversy about whether rheotaxis is a passive process or whether the fluid flow triggers a signal in sperm to change their motility behavior. Rheotaxis has been associated with an asymmetrical movement of the flagellum midpiece that directs the position of the sperm [108], and an increase of intracellular Ca²⁺mediated by CatSper [107], whereas others have not observed any of these changes during rheotaxis [109].

There is also evidence that thermotaxis, or the migration toward higher temperature, guides sperm within the oviduct toward the ampulla (see reviews [31, 110, 111]). In rabbits, the ampulla is 2°C warmer than the isthmus [112] and increases 0.8°C after ovulation [113]. Some studies also indicate that thermotaxis acts as a sperm selective mechanism by guiding only capacitated and “fittest” sperm. First, only capacitated sperm present thermotaxis [113]. Second, an in vitro assay, in which mouse and human sperm were exposed to a gradient of temperature (35–38°C), selected sperm with higher DNA integrity and lower chromatin compaction [114]. Mouse sperm selected by this method produced higher blastocyst, implantation, and live-birth rates after ICSI, compared with sperm selected by swim-up [114]. Last, bovine ejaculates with higher conception rates after AI present a higher rate of sperm with a thermotactic response in vitro [115].

Regarding chemotaxis, the response to a substance concentration gradient, many molecules are known to guide sperm motility, such as atrial natriuretic peptide, progesterone, nitric oxide, odorant substances, and chemokines (see review [31]). Moreover, follicular fluid (FF) [116] and COCs [117] attract sperm. This has led to the hypothesis that progesterone, both present in FF and produced by COCs [118], is a major chemoattractant guiding sperm toward the oocyte after ovulation. In vitro studies in different species, including human, cow, rabbit, horse, and mouse, have shown sperm migration toward gradients of picomolar concentration of progesterone [31]. Sperm chemotaxis toward progesterone is regulated by changes in the flagellar activity [119–121]. But questions arise about the progesterone concentration and conditions used in these in vitro studies that may not represent physiological chemotaxis in the oviduct. On the other hand, recent studies have identified specific chemokines produced by COCs that guide sperm in the oviduct. In cattle, the stromal cell-derived factor 1 (SDF1) provides direction to sperm movement, whereas the inhibition of SDF1 synthesis blocks sperm taxis toward COCs [122]. In humans, the chemokine ligand 20 (CCL20) chemo-attracts sperm through the CCR6 receptor, and the neutralization of CCL20 blocks chemotaxis toward the FF [123].

Similar to thermotaxis, only a subpopulation of sperm shows a chemotactic response, which would have a greater chance to find and fertilize the egg. For instance, only sperm with an intact acrosome are guided by progesterone [124] and FF [125]. Moreover, a higher percentage of human sperm that respond to FF chemotaxis are capacitated compared with non-chemo-responsive sperm [126]. A possible explanation is that the thermo- and chemoreceptors achieve the proper location and configuration after the capacitation-induced reorganization of the plasma membrane (as suggested [111]). In fact, CCR6 exposure on the human sperm surface is more abundant after capacitation [127]. Moreover, ˂30% of human sperm express CXCR4 [128], which denotes that only some sperm would respond to the SDF1 chemokine. Further evidence of sperm selection based on their ability to respond to chemotaxis comes from IVF studies that have tested different chemotaxis chambers for sperm selection. Overall, they showed that sperm that are chemo-attracted to either progesterone or FF have higher DNA integrity and achieved higher fertilization rates compared with unselected sperm [129–132]. Yet, these studies do not represent physiological conditions and do not compare results with non-chemo-responsive sperm after being exposed to the same chambers.

It is generally accepted that thermotaxis and rheotaxis first guide sperm in the oviduct, and chemotaxis takes an increasing role as sperm move closer to the oocyte. However, the combinatory effect of these processes together with the anatomy of the oviduct should be considered for understanding sperm migration. Some authors have studied sperm motility behavior along the oviduct by representing its morphology with microfluidic systems. For instance, sperm present optimal upstream swimming and swim faster near surfaces [133, 134], and present progressive surface-aligned motility when the surface has low curvature [135]. At the same time, sperm hyperactivation regulates the swimming behavior along physical boundaries, which reduces their guiding effect on sperm movement [136]. In this study, hyperactivation induced a “pseudo-chemotaxis” effect, as sperm stayed longer in chambers with hyperactivation agonists [136]. Since sperm are hyperactivated in the lower oviduct, these results suggest that chemotaxis plays a role in sperm guidance far before being close to the oocyte. All these sperm migration properties can be applied to generate more sophisticated microfluidic devices that better mimic the oviduct. For instance, a microfluidic device for human sperm recreates the microgrooves in the Fallopian tube with oval-shaped micro-pockets in which only viable and highly motile sperm accumulate [137].

Sperm interaction with oviductal fluid

As others have extensively reviewed, oviductal fluid regulates sperm function through different components, including proteins, mRNA, and extracellular vesicles [28–30, 111, 138, 139]. Oviductal fluid modulates sperm molecular composition, motility, viability, capacitation, and acrosome reaction. Effects vary depending on the estrous cycle and are progressively coordinated during transport in the oviduct. For instance, studies of sperm migration in the oviduct in situ show a higher rate of acrosome reaction among sperm in the ampulla compared with the lower isthmus [43, 140]. But years of in vitro studies have revealed that sperm do not capacitate synchronously when physiological conditions are replicated [141]. All this information taken together supports the model that some sperm subpopulations are selected by the oviduct during capacitation, which could allow the fittest sperm to fertilize once they reach the oocyte. It is also noteworthy that studies in taxonomically diverse species show a bias in fertilization toward specific males that is regulated by female reproductive fluids (see review [142]).

Some oviductal fluid physical characteristics, including osmolarity and pH, may pose a selective pressure on sperm. Oviductal fluid is hyperosmotic (350–355 mOsm/kg in bovine) compared with seminal plasma and has a higher pH (7.4) compared with the uterus (6.8) [143]. Indeed, the sperm’s ability to adapt to osmotic and pH changes in vitro (maintaining their viability, motility, and morphology) has been related to fertility outcomes [143–145]. Some osmo-responsive genes are upregulated on good osmo-adapter sperm ejaculates [143]. This has prompted the use of hypo-osmotic swelling tests and other tests based on the assessment of sperm volume in response to hypo-osmotic stress to determine sperm quality [146].

Sperm interaction with the cumulus-oocyte complex before fertilization

Right before fertilization, sperm encounter some last selective barriers as they interact with the COC, including the extracellular matrix formed by the cumulus cells and the ZP. The importance of the cumulus oophorus for successful fertilization is well established (see reviews [147–149]). Upon maturation, the expanded cumulus forms a matrix mainly with hyaluronic acid, an extracellular glycosaminoglycan, that sperm must traverse to reach the oocyte [147]. At the same time, the cumulus oophorus entraps and guides sperm toward the oocyte, facilitating fertilization [149]. Multiple IVF studies in domestic animals have shown that the fertilization and blastocyst formation rates are higher when cumulus cells are maintained in comparison to when they are removed before IVF (denuded oocytes) [150–153]. Moreover, the addition of cumulus cells during IVF culture with denuded oocytes can recover the rates of sperm penetration [153]. More intriguing is the observation that comes from some mouse KO models for ADAM3 and other sperm proteins, which are essential for transport through the UTJ [148]. Sperm from these mice cannot fertilize denuded zona-intact oocytes in vitro, but they do fertilize intact COCs, indicating that cumulus cells may somehow compensate for the lack of these sperm proteins.

The cumulus oophorus limits and potentially selects the sperm that reach the oocyte. Hyperactivation plays a role in the ability of sperm to traverse the cumulus matrix [148]. But hyaluronidase, contained in the plasma membrane and inner acrosomal membrane of the sperm head [154], appears to be the key factor for sperm to disperse and penetrate the extracellular matrix. This was shown using a mouse with sperm deficient in both hyaluronidase 5 and 7, which had a decreased ability to fertilize in vitro and produce offspring in vivo [154]. Evidence of the role of hyaluronic acid and the extracellular matrix in sperm selection comes from an evolutionary study in mice [155]. The authors identified different proteins in COCs that stabilize hyaluronic acid, three of which have been positively selected rapidly through evolution, which implies an important role in fertilization [155]. They also observed that mouse females that were raised with many males, in comparison to just one male, produced COCs more resistant to hyaluronidase, denoting a sperm selective mechanism conditioned by the environment [155]. Furthermore, a system consisting of a column filled with cumulus cells has been developed to select human sperm for IVF/ICSI. Sperm that go through the column show higher rates of normal morphology and good motility parameters, compared with sperm that do not pass the column [156], and induce higher fertilization and pregnancy rates after ICSI, compared with conventional sperm selection with a density gradient [157]. A similar study, which tested the ability of sperm to go through a group of COCs in a microfluidic device, found not only better motility parameters, but also lower DNA fragmentation compared with sperm that did not traverse the COCs [158].

On the other hand, the interaction of sperm with hyaluronic acid should be considered as a potential selection mechanism throughout much of the female reproductive tract, as hyaluronic acid is found extensively not only in the cumulus matrix but also in the cervical mucus and oviductal fluid [27]. Hyaluronic acid-binding proteins have been identified in sperm of various species [159]. And in some, such as in buffalo, sperm binding to hyaluronic acid has been correlated with fertility after AI [160]. A plethora of studies in humans has shown that sperm that bind hyaluronic acid have better quality parameters. For instance, all sperm bound to hyaluronic acid were viable and had a higher frequency of intact acrosomes compared with free sperm [161]. The ability of sperm to bind hyaluronic acid has been also positively correlated to motility and morphology parameters [162]. Even more interestingly, sperm that bind hyaluronic acid showed reduced DNA damage and chromosomal disomies [37, 163–166]. These results have led to the development of a sperm selection system for human ICSI that consists of a hyaluronic acid-coated surface, named Physiological ICSI (PICSI). PICSI leads to higher production of embryos, higher embryo quality, and lower embryo disomies [164, 167–169]. Although some meta-analyses of the clinical use of PICSI have not identified a clear improvement in overall fertility outcomes (see review [25]), PICSI has been consistently correlated with reduced miscarriage rates [170, 171]. A recent study [166] that further analyzed the data from the previous clinical trial [170] corroborated the mitigation of miscarriages, especially in older women. The authors of this study hypothesized that the sperm abnormalities (DNA damage) that were avoided by selecting with PICSI did not affect the establishment of clinical pregnancy, but rather the capacity to maintain it [166].

Once sperm move through the cumulus cells, they bind to the ZP. The binding between sperm membrane proteins and ZP glycoproteins is an essential mechanism for sperm–oocyte recognition and fertilization (see reviews by authors [172–174]). Three to four zona glycoproteins (ZP1, ZP2, ZP3, and ZP4) compose the ZP in mammals. There is controversy about the zona components that bind sperm. Genetic experiments in mice indicate that a peptide within ZP2 binds mouse sperm but biochemical experiments using human and mouse zona proteins suggest glycan structures in zona proteins bind sperm [172]. Interestingly, the glycan structures reported to bind human sperm (glycans containing sialylated Le^X) [175, 176] are similar but not identical to oviduct glycans that bind porcine sperm [63].

Many older studies concluded that the adhesion to the ZP triggers the sperm acrosome reaction, and therefore the release of enzymes needed for zona penetration and fusion with the oocyte [173]. After this reaction, ZP-binding proteins would be released from the sperm surface. However, recent studies have put this paradigm into question. In vitro experiments have shown that sperm that begin the acrosome reaction before reaching the ZP are primarily the sperm that fertilize oocytes [177] and that release of acrosomal proteins is a gradual process in which acrosomal proteins are not synchronously released [178]. Moreover, a study in situ of mouse sperm in the oviduct has revealed that most sperm in the ampulla region have already undergone the acrosome reaction [43, 140]. This implies that the sperm–ZP interaction is more complex than what was believed 20 years ago. Nevertheless, the number of sperm bound to ZP of cumulus-free oocytes is conventionally used as a research tool estimating IVF success in experimental settings. In human, the sperm-binding ratio in a hemizona assay (a comparative assay in which the ZP from an oocyte is halved and sperm from a treatment and control group are allowed to bind each half) has been correlated with the fertilization rate after IVF [179]. Moreover, the ability of both bovine and porcine sperm to bind the ZP in vitro has been correlated to the fertility of the male after AI, although these assays are more accurate when other sperm characteristics are included in the predictive models [37, 180, 181].

Considering that only mature sperm expressing certain proteins recognize the ZP, this mechanism may act as a last selective barrier for sperm to successfully fertilize. Some reports on humans indicate that sperm that bind to the ZP have better quality. Hence, sperm from fertile men showed higher rates of ZP binding compared with infertile men [182], and sperm morphology and hyperactivated motility were correlated with sperm-binding ability to the ZP [183]. Moreover, the zona pellucida-binding assay (ZPBA) has been tested as a selection method for ICSI in humans, which involves incubating sperm with extra oocytes and aspirating only the ones that are bound to the ZP for ICSI (see review [184]). Some have obtained a higher rate of good-quality embryos with this method [185], although an overall increase in fertility success has not been achieved [186]. It is also worth noting that sperm–ZP interaction may not be a relevant sperm selection mechanism in vivo, as unfit sperm would have been eliminated before. In fact, indirect selection for ZP-binding may occur in the UTJ, considering that the knockout of genes that affect movement into the oviduct also affects ZP binding [54]. Interestingly, a study in pigs found a robust correlation (R = 0.94) between the sperm-binding ability to ZP and OECs aggregates suggesting the sperm requirements for adhesion to both matrices are very similar [181].

Concluding remarks

The research studies presented herein show that sperm interaction with the oviduct plays an important role in fertility. During transport through the oviduct, sperm are exposed to various discriminating pressures that favor the selection of a sperm subpopulation that is more fit. Sperm are selected not only based on their motility and viability, but also due to other predetermined characteristics that lead to healthy embryonic development, including DNA integrity. However, we still lack a full understanding of the mechanisms of selection, the specific parameters selected, and how they affect embryonic development. Moreover, in vivo and in situ studies are needed to validate the in vitro results. Nevertheless, the above research indicates that the current sperm quality analysis and selection techniques used in the context of ART are severely deficient. A shift in the focus of these techniques to sophisticated functional assays that include several sperm properties that the female tract appears to select for would potentially improve the success of fertility treatments. To that end, microfluidic devices are exceptional tools to simulate the oviduct anatomy and rheological properties, while allowing the incorporation of other elements that test the sperm performance, for instance, the ability of sperm to bind to oligosaccharides or respond to chemoattractants. Others have previously reviewed currently used and prospective techniques for selecting sperm for ART in humans [19, 146, 184, 187–190]. In Table 3, we present a summary of those that are based on physiological selection in the oviduct. Note that we have not included sperm incubation with OECs. Even though selection of sperm through this technique could improve IVF success [64–66], some major disadvantages discourage their clinical application (i.e., the risk of bacterial contamination, technical difficulties such as establishing and maintaining a differentiated OEC culture, and recovering the sperm that bind OECs). As for the assessment of sperm quality, a combination of traditional and novel assays is needed because the evaluation of multiple sperm parameters better predicts the fertility potential of a specific male [2, 37, 181, 191–193].

Table 3.

Summary of in vitro sperm selection methods based on physiological selection in the oviduct

Technique	Rationale	Mechanism	References
Microfluidic based systems	Functional sperm present chemotactic, thermotactic, and rheotactic behaviors.	Sperm are exposed in a microfluidic device to a chemical gradient, thermal gradient, or different fluid flows. Sperm that accumulate on specific parts of the device are selected.	[195–200]
Hyaluronic acid-based selection	Functional sperm present specific receptors that allow them to bind to hyaluronic acid.	PICSI: Sperm are cultured on a Petri dish coated with hyaluronic acid. Sperm bound to the bottom of the dish are selected.	[160–163,167,201]
		Hyaluronan-rich medium: A droplet of medium with sperm is put in contact with a droplet of medium with hyaluronic acid. Sperm in the interface of the droplets are selected.	[165, 201, 202]
Cumulus oophorous column	Functional sperm go through the cumulus oophorous matrix.	Sperm that traverse a capillary filled with cumulus cells are selected.	[156–158]
ZPBA	Functional sperm bind the oocyte ZP during fertilization.	Sperm are cultured with oocytes and those attached to ZP are aspirated and used for ICSI.	[184–186]

It is worth noting that our current understanding of sperm transport and interaction with the oviduct comes mostly from animal studies. Still, the plethora of studies that show common mechanisms across genetically distant species and the promising results seen in ART outcome after applying a “natural” sperm selection approach highlight the value of these animal studies for all animals. The development of broadly applicable laboratory selection methods that mimic selection by the female reproductive tract is expected to improve ART success and reduce abnormal embryogenesis due to paternal factor.

Conflict of interest

The authors have declared that no conflict of interest exists.

Authors contributions

S.S-H. prepared the manuscript. D.S. and D.M. reviewed and edited the manuscript.