Introduction

Extensive studies have been done in animals regarding polyspermy. However, some areas are unknown in human polyspermy due to the ethical issues concerning studying human fertilization. Application of in vitro fertilization (IVF) allows us to observe fertilization during the interval of the procedure. Routinely, fertilization status is checked 16-20 hours after insemination in vitro. Approximately 7 % of fertilized human eggs are polyspermic, which is defined by appearance of three pronuclei, so-called triploid [19].

Theoretically, the triple pronuclei routinely observed in the IVF laboratories could be a result of two occurrences: 1) triple nuclei derive from two sperm and oocyte nucleus; 2) triple nuclei derive from one sperm, oocyte nucleus, and failure of second polar body extrusion. Therefore, the clinically observed polyspermic eggs are not always truly polyspermic. Regardless, occurrence of multipronuclear formation in the fertilized oocytes reflects poor oocyte quality. With that being said, do embryos have the ability to self-correct during early development? In this review, the following areas will be discussed.

  • Polyspermy block

  • Polyspermy and oocyte quality

  • Perivitelline space (PVS)/cortical granules (CG)

  • Multiploidy - failure of the second polar body extrusion

  • Multiploidy - ability of embryo’s self-correction

  • Multiploidy and mosaicism

Polyspermy Block

Polyspermic block in mammalian eggs has been well reviewed [7, 24, 39, 47]. Commonly accepted theories are a temporary fast block followed by permanent block, which is mediated by exocytosis of cortical granules to prevent additional sperm penetration. However, Mio et al. recently [29] described dynamic changes of how human polyspermy occurs during real-time development by using time-lapse cinematography. Their discoveries challenged traditional concepts that two types of polyspermy block were thought to exist in marine animals and mammals, including human-“oolemma block” and “zona reaction.” The authors reported that once the leading sperm penetrated the zona pellucida (ZP) and attached to the oocyte membrane, any following sperm within the ZP immediately would stop penetrating within 10 seconds. Then, the oocyte membrane block happens.

More recently, Zumoffen et al. [51] reported that a protein called lactoferrin, detected in human oviductal secretion, is involved in gamete interaction, which may be associated with polyspermy block. In vivo, the oviductal modification of ZP resistance to proteolytic digestion has been demonstrated to influence fertilization and this prefertilization mechanism is considered to contribute to the control of polyspermy. ZP resistance to proteolysis, induced by oviductal fluid, was observed in mouse, rat, hamster, rabbit, sheep, goat, pig, and cow, but not in human [30]. In the circumstances of IVF, the oocytes are aspirated directly from the ovaries. They are not exposed to the oviductal proteins. Obviously, proteolytic digestion is not a prerequisite for human oocytes to be fertilized normally. However, the singletons derived from IVF/ICSI have poorer perinatal outcomes than those naturally born ([35], systematic review and meta-analysis). In pigs, zona hardening is regulated by the oviductal proteins. Such interaction is associated with frequencies of monospermy and polyspermy [31]. Despite differences of ZP resistance among species and human [30], the oviductal contribution to egg quality and occurrence of polyspermy cannot be ignored. Further studies are needed to address the difference of oocytes/sperm exposed with or without oviductal environment. Perhaps the short-term exposure to the oviductal proteins/fluid would enable the oocytes to reach full cytoplasmic maturation, which could reduce the occurrence of polyspermy.

Polyspermy and Oocyte Quality

Although polyspermy was considered abnormal fertilization, in some species, physiological polyspermy does occur ([23•] review). Dale and DeFelice [7] opened a debate about whether mammalian oocytes actively repel supernumerary sperm. Considering the fact that the ratio of sperm:oocyte at final destiny in nature is so low, these authors indicated that polyspermy preventing mechanisms may not be necessary.

Aged eggs appear to have reduced ability to prevent polyspermy at the level of the oolemma, i.e., to establish a membrane block to polyspermy [8]. However, women aged 45 years or older had the same fertilization rate via IVF and ICSI [5••]. These authors suggested that zona hardening does not appear to be a consequence of reproductive aging, which would challenge the application of assisted hatching and also the theory of zona hardening in the prevention of polyspermy. The experiment using pig as a model also suggested that the ZP and oolemma are not competent factors for the prevention of polyspermy [43].

Mammalian oocyte maturation includes two independent processes-nuclear maturation and cytoplasmic maturation [11]. In the controlled ovarian stimulation, the oocytes are retrieved from follicles in different sizes. Although most oocytes completed first meiosis (metaphase II stage), with released first polar body, their cytoplasmic maturation varies. More recently, it was reported that mouse oocyte growth and differentiation are genetically dissociable from the chromosomal events of the first meiosis [10••], which further proved that completion of oocyte nuclear maturation reaching metaphase II does not mean the oocytes are ready to accept the sperm. The oocyte cytoplasm could be at different status: immature, mature, or overmature. Even with natural conditions, the egg is ovulated with different status of cytoplasmic maturation. Most likely, polyspermy happens to those eggs with immature cytoplasm or overmature cytoplasm (aging). Therefore, theoretically, for the clinical application of conventional IVF, the insemination time should be based on their status of cytoplasmic maturity, meaning that oocytes from the same patient should be inseminated at different times.

Progesterone levels are closely related to the oocyte maturation status, especially the ratio of estrogen/progesterone. Increased level of progesterone decreases ZP hardening, resulting in a high rate of polyspermy in pigs [30, 31]. In humans, serum premature progesterone rise on the day of human chorionic gonadotropin (hCG) administration has been reported to be correlated negatively with live birth rate [20, 34]. These authors concluded the progesterone premature rise did not affect pregnancy rates when transferring the embryos in frozen cycles, indicating endometrium damage occurrence. However, the results are controversial. Check et al. [6] published results of a retrospective cohort analysis that egg donor progesterone level (>3.4 ng/ml) may have lower pregnancy rate in recipients. Studies regarding premature progesterone rise and its association with polyspermy may shed light on improving egg quality by using different ovarian stimulation protocols.

Application of in vitro maturation (IVM) revolutionized the field of assisted reproductive technologies. Many publications have reported successful outcomes using different approaches of IVM. Midkine was reported to facilitate mammalian oocyte cytoplasmic maturation [22]. To date, it is still unknown whether IVM oocytes have lower polyspermic rate because of application of ICSI in most cases [12, 38]. However, Walls et al. [46] demonstrated IVM oocytes (150 oocytes of 8 cycles) achieved the same fertilization rate and clinical pregnancy rate comparing two insemination approaches: conventional IVF and ICSI. As application of conventional IVF to in vitro matured oocytes, more information about polyspermic occurrence of IVM would allow us to understand the relationship of oocyte quality and polyspermy.

Perivitelline Space and Cortical Granules

Perivitelline space is the space between oolemma and zona pellucida of the oocyte. This space does not develop in oocyte at germinal vesicle stage until oocyte starts meiotic maturation. Glycoproteins secreted by cumulus cells are involved in making the space more obvious [42]. In mice, polyspermy occurs at a higher rate in the oocytes with smaller PVS treated with Tunicamycin, an inhibitor of glycoprotein synthesis [25]. In the physiological condition, the size of PVS is not a reason for causing polyspermy. It is a result of oocyte maturity, especially for patients treated with the controlled ovarian stimulation. Ovarian apoptosis happens at every stage of follicular development. The oocytes retrieved from these follicles could come from normal-size follicles but already at atretic stages. The oocytes from the atretic follicles could be at immature or overmature, thus resulting oocytes with bigger or small PVS. The human oocyte grading [49] based on the size of PVS and status of polar body (fragmented vs. nonfragmented) specified the best synchronized oocyte morphology-normal PVS and nonfragmented first polar body. The results of this publication were from patients who underwent intracytoplasmic sperm injection (ICSI) treatment involving a single sperm injection. To date, no studies have been performed concerning whether this group of oocytes (with normal PVS and intact first polar body) have lower polyspermic rate using conventional IVF. More criteria are needed regarding the association of oocyte morphology and polyspermy. In another words, what type of morphological features of oocytes has higher incidence of polyspermy in humans?

Cortical granules (CG) in mammalian oocytes have been well studied over the past decades (reviewed by [27]). The content of CGs is released into the PVS immediately after fertilization (exocytosis) for the establishment of polyspermy block. Distribution of CGs in immature oocytes and in vitro matured oocytes varies under different culture conditions comparing with the CG distribution in vivo matured oocytes in mice [28]. The total number of GCs in unfertilized oocytes was increased during 3- to 6-hour culture in vitro. Increasing number of GCs also was associated with fertilization and embryo development [37]. These authors indicated that proliferation of GCs could be used as criteria for nuclear maturation. However, if CGs developed from Golgi complexes during oocyte development [15, 27], their total number could be more related to the cytoplasmic maturation rather than nuclear maturation. Using a mouse model, Liu et al. [28] also suggested that CGs were associated with oocyte cytoplasmic maturation. Future studies are needed regarding the mechanisms of CGs maturation and oocyte cytoplasmic maturation, thus the polyspermy rate would be reduced via improving cytoplasmic maturity.

Moreover, cortical granules can be released before fertilization, after fertilization and even up to the first cleavage stage [27]. Observing more than 3,000 embryos cultured in time-lapse EmbryoScope, we found that CGs do not exist in PVS of the oocyte at germinal vesicle stage but exist in the PVS of the oocytes at metaphase II stage. They do not seem to have much change morphologically until 2-cell stage. Then, after 4-cell stage, the total number of granules in PVS is reduced and disappeared by Day 3 of the culture (unpublished data). Apparently, the cortical granules may continue their contribution to the early development during cleavage stage, perhaps up to blastocyst stage to facilitate embryo hatching. Mouse ovastacin, a cortical granule protease, cleaves zona pellucida protein 2 [4]. Swine CGs contain hydrolytic enzymes, proteases, and peroxidases [1]. Hopefully, future studies would prove the concept of contribution of CGs to blastocyst hatching, which would shed light on understanding the controversial results about application of the assisted hatching in human IVF field.

Multiploidy - Failure of the Second Polar Body Extrusion

Mammalian oocytes undergo two rounds of meiotic divisions. The first round results oocytes at Metaphase II stage with the first polar body. The second meiotic division happens after fertilization by releasing the second polar body (PB2). The polar body extruding processes are driven by the spindle migration which is controlled by cytoplasmic actin filaments [2, 26].

Failure of the second polar body extrusion is related to the oocyte polarization during which a series of signaling cascade is associated with the emission of PB2. These signal pathways are not well studied in humans. In mice, it was recently reported that Cdc42 activation (Cdc42-GTP) is related to the mammalian oocyte polarization [9], together with downstream of cytoplasmic Ran-GTP gradient and upstream of N-WASP activation at the oocyte cortex. These authors also reported that Cdc42-GTP accumulates, in a polarized fashion, in the cortex overlying the meiotic spindle during both meiosis I and II in mice.

The second polar body contains one set of segregated chromatids. If not released, it would form a nucleus-like structure, which looks like the third pronucleus in the fertilized egg. Clinically, it also is called triploid. Such triploids also occur in the event of application of intracytoplasmic sperm injection (ICSI), which only involves injection of a single sperm. From our observation by time-lapse EmbryoScope, 77 of 3,213 (2.39 %) eggs injected were triploid. These embryos do become 8-cell embryo and even form blastocysts. Under such circumstances, triploid embryos generated by ICSI are more prone to generating genetically abnormal babies. A severe case was reported that a triploid infant, 69, XXX karyotype, resulted from nondisjunction at maternal second meiosis [3••, 18]. However, a small-scale study that included 32 triploid embryos by ICSI and 18 by IVF has shown that 50 % of these embryos could become normal blastocysts as detected by fluorescence in situ hybridization (FISH) and parental inheritance analysis by PCR [14••]. These authors suggested the ability of the self-correction depends on the parental origin of the extra pronucleus, which indicated that the triploid embryos via ICSI have higher chance of becoming normal embryos, whereas the embryos derived by IVF could be at a risk of genetic abnormalities. Regardless, the triploid embryos derived from ICSI should be discarded in the IVF laboratories.

Multiploidy - Ability of Embryo’s Self-Correction

As mentioned earlier, triploid embryos via ICSI could become normal embryos, but to date it is still unknown whether human embryos derived from polyspermic fertilization possess the ability of self-correction.

Studies in pigs have confirmed that the polyspermic eggs were able to develop to term, resulting piglets with normal chromosome numbers [17••]. The same group also reported that the polyspermic eggs developed to the blastocyst stage at a developmental rate similar to that of normal 2PN eggs [16]. In mice, the embryo at 2-cell stage is able to engulf sperm present in the perivitelline space [44]. Ultrastructural studies showed that approximately 50 % of blastomeres at the 3- and 4-cell stage contain accessory sperm heads and additional sperm head was seen inside the lysosomal body [50].

Lysosomal proteolysis is triggered by posttranslational protein modification by ubiquitination in mammalian oocytes. The deubiquitinating enzymes (DUBs) could reverse ubiquitination process. Two members of Ubiquitin C-terminal hydrolase family of DUBs, UCHL1 and UCHL3, are associated with murine oocyte cortex (UCHL1) and meiotic spindle (UCHL3). Their mRNAs are highly expressed in murine oocytes at germinal vesicle stage and metaphase II stage [33]. Antibodies against UCHL3 into mature metaphase II oocytes blocked fertilization by reducing sperm penetration of the zona pellucida and sperm incorporation into the ooplasm [32]. Sutovsky suggested that polyspermy can be ameliorated by modulating sperm-associated deubiquitinating enzymes ([41•], review). Perhaps the balancing of ubiquitination and deubiquitination processes controls the proteolysis process in oocytes and sperm, which may be associated with removing extra sperm inside the oocyte cytoplasm and blastomeres of embryos at the cleavage stages. Future studies are needed in these areas.

Multiploidy and Mosaicism

Multiploidy resulting from polyspermy or retention of secondary polar body, or in rare cases binuclear sperm, is the major cause of embryo mosaicism. Patients with diploid/triploid chromosome mosaicism present significant clinical abnormalities, including similar genetic syndromes to aberrant genomic imprinting [36], dysmorphic formation [45], cutaneous pigmentary dysplasia [48], and hydatidiform mosaic mole [40]. More recently, Huisman et al. [21] reported that patients have high rate of somatic mosaicism (10/44; 23 %) for an NIPBL (encodes Nipped-B-like protein) gene mutation, which results in Cornelia de Lange syndrome.

Clinically, many mosaicism cases were underdiagnosed or neglected. The patients could have normal karyotype with abnormal clinical phenotype [3••]. The amniocentesis is not sufficient to make diagnosis of diploidy-triploid mosaicism [13]. Analyzing cleaved embryos by preimplantation genetic diagnosis (PGD) also has the risk of missing the mosaicism. The application of trophectoderm PGD at blastocyst stage also would be subjected to the underdiagnosis of mosaicism due to the fact that inner cell mass eventually differentiates into the human body. Therefore, genetic prenatal testing for the babies via application of assisted reproductive technologies should be provided routinely to all patients.

Conclusions

Collectively, polyspermy is not a normal phenomenon. To avoid it from happening requires improving the quality of the oocytes. If additional sperm inside the oocyte did not affect the embryonic genome, the embryos may have some ability of removing the additional sperm. Under such conditions, if the quality of oocyte is not favorable, the polyspermic embryos would not survive, not because of the polyspermy but because of poor oocyte quality. Therefore, improving oocyte quality by variety of ovarian stimulation protocols in the ART field is a key issue.