Fertilization in mammals requires sperm to bind to the zona pellucida (ZP) that surrounds the egg. Galactose (Gal) or N-acetylglucosamine (GlcNAc) residues on the glycans of ZP protein 3 (ZP3) have been implicated as mouse sperm receptors. However, Mgat1–/– eggs with modified N-glycans lacking terminal Gal and GlcNAc residues are fertilized. To determine if Gal and GlcNAc on O-glycans of the ZP are required for fertilization, a conditional allele of the T-synthase gene (T-synF) was generated. T-syn encodes core 1 β1,3-galactosyltransferase 1 (T-synthase), which initiates the synthesis of core-1-derived O-glycans, the only O-glycans on mouse ZP3. T-synF/F:ZP3Cre females in which T-synF was deleted at the beginning of oogenesis generated eggs lacking core-1-derived O-glycans. Nevertheless, T-synF/F:ZP3Cre females were fertile and their eggs bound sperm similarly to controls. In addition, T-syn–/– embryos generated from T-syn null eggs developed until ∼E12.5. Thus, core-1-derived O-glycans are not required for blastogenesis, implantation, or development prior to midgestation. Moreover, T-syn–/–Mgat1–/– eggs lacking complex and hybrid N-glycans as well as core-1-derived O-glycans were fertilized. The combined data show that mouse ZP3 does not require terminal Gal or GlcNAc on either N- or O-glycans for fertilization.

The specificity of sperm-egg recognition in mammals is mediated primarily by the zona pellucida surrounding ovulated eggs. Mouse sperm are quite promiscuous and bind to human eggs, but human spermatozoa will not bind to mouse eggs. The mouse zona pellucida contains three glycoproteins, ZP1, ZP2, and ZP3, which are conserved in rat and human. The recent observation that human zonae pellucidae contain a fourth protein raises the possibility that the presence of four zona proteins will support human sperm binding. Using mass spectrometry, four proteins that are similar in size and share 62-70% amino acid identity with human ZP1, ZP2, ZP3, and ZP4/ZPB were detected in rat zonae pellucidae. However, although mouse and rat spermatozoa bind to eggs from each rodent, human sperm bind to neither, and the presence of human follicular fluid did not alter the specificity of sperm binding. In addition, mutant mouse eggs lacking hybrid/complex N-glycans or deficient in Core 2 O-glycans were no more able to support human sperm binding than normal mouse eggs. These data suggest that the presence of four zona proteins are not sufficient to support human sperm binding to rodent eggs and that additional determinants must be responsible for taxon-specific fertilization among mammals.After passage through the lower female reproductive tract, mammalian spermatozoa fertilize ovulated eggs in the ampulla of the oviduct. A key event in successful fertilization is sperm binding to the surface of the extracellular zona pellucida that surrounds the egg. Following zona penetration and fusion with the egg plasma membrane, peripherally located cortical granules within the egg exocytose their contents, which modify the zona matrix such that sperm no longer bind. These events are carefully orchestrated to ensure that a single sperm fertilizes a single egg (1).Despite decades of investigation, the molecular basis of mammalian sperm-egg recognition remains controversial. Human sperm are particularly fastidious and bind to old world primate eggs but not to eggs of other species. In contrast, mouse sperm are quite promiscuous, binding with near universality to eggs from other mammalian orders (2). The mouse zona pellucida is composed of three major glycoproteins, ZP1, ZP2, and ZP3, one of which, ZP2, is proteolytically cleaved following fertilization (3). Mouse lines have been established that lack each of the zona proteins as well as lines in which human ZP2 and/or human ZP3 replace endogenous mouse proteins (4). Mice without ZP1 form a zona pellucida matrix to which mouse sperm bind and Zp1 null females are fertile, albeit with decreased fecundity (5). Mice in which endogenous proteins are replaced with human ZP2, human ZP3, or both are also fertile but do not support human sperm binding (6, 7). Thus, mouse ZP1 is not required for sperm-egg recognition, and human ZP2 and ZP3 are not sufficient to support human sperm binding.These results suggest that "humanized" zona matrices either lack a factor or contain a factor that prevents the binding of h...

To identify roles in spermatogenesis for major subclasses of N- and O-glycans and Notch signaling, male mice carrying floxed C1galt1, Pofut1, Notch1 or Mgat1 alleles and a testis-specific Cre recombinase transgene were generated. T-synthase (C1GALT1) transfers Gal to generate core 1 and core 2 mucin O-glycans; POFUT1 transfers O-fucose to particular epidermal growth factor-like repeats and is essential for canonical Notch signaling; and MGAT1 (GlcNAcT-I) transfers GlcNAc to initiate hybrid and complex N-glycan synthesis. Cre recombinase transgenes driven by various promoters were investigated, including Stra8-iCre expressed in spermatogonia, Sycp1-Cre expressed in spermatocytes, Prm1-Cre expressed in spermatids, and AMH-Cre expressed in Sertoli cells. All Cre transgenes deleted floxed alleles, but efficiencies varied widely. Stra8-iCre was the most effective, deleting floxed Notch1 and Mgat1 alleles with 100% efficiency and floxed C1galt1 and Pofut1 alleles with ~80% efficiency, based on transmission of deleted alleles. Removal of C1galt1, Pofut1, or Notch1 in spermatogonia had no effect on testicular weight, histology, or fertility. However, males in which the synthesis of complex N-glycans was blocked by deletion of Mgat1 in spermatogonia did not produce sperm. Spermatogonia, spermatocytes, and spermatids were generated, but most spermatids formed giant multinucleated cells or symplasts, and apoptosis was increased. Therefore, although core 1 and 2 mucin O-glycans, NOTCH1, POFUT1, O-fucose glycans, and Notch signaling are dispensable, MGAT1 and complex N-glycans are essential for spermatogenesis.

SummarySister chromatid cohesion mediated by the cohesin complex is essential for chromosome segregation in mitosis and meiosis [1]. Rec8-containing cohesin, bound to Smc3/Smc1α or Smc3/Smc1β, maintains bivalent cohesion in mammalian meiosis [2, 3, 4, 5, 6]. In females, meiotic DNA replication and recombination occur in fetal oocytes. After birth, oocytes arrest at the prolonged dictyate stage until recruited to grow into mature oocytes that divide at ovulation. How cohesion is maintained in arrested oocytes remains a pivotal question relevant to maternal age-related aneuploidy. Hypothetically, cohesin turnover regenerates cohesion in oocytes. Evidence for post-replicative cohesion establishment mechanism exists, in yeast and invertebrates [7, 8]. In mouse fetal oocytes, cohesin loading factor Nipbl/Scc2 localizes to chromosome axes during recombination [9, 10]. Alternatively, cohesion is maintained without turnover. Consistent with this, cohesion maintenance does not require Smc1β transcription, but unlike Rec8, Smc1β is not required for establishing bivalent cohesion [11, 12]. Rec8 maintains cohesion without turnover during weeks of oocyte growth [3]. Whether the same applies to months or decades of arrest is unknown. Here, we test whether Rec8 activated in arrested mouse oocytes builds cohesion revealed by TEV cleavage and live-cell imaging. Rec8 establishes cohesion when activated during DNA replication in fetal oocytes using tamoxifen-inducible Cre. In contrast, no new cohesion is detected when Rec8 is activated in arrested oocytes by tamoxifen despite cohesin synthesis. We conclude that cohesion established in fetal oocytes is maintained for months without detectable turnover in dictyate-arrested oocytes. This implies that women’s fertility depends on the longevity of cohesin proteins that established cohesion in utero.