香京julia种子在线播放

Centrioles are essential for animal development and physiology, as demonstrated by a variety of experiments that have tested the centriole’s role directly (Bettencourt-Dias et al., 2011). Most of these experiments have been performed in vitro, in human immortalized cells, or in animal models, limiting our knowledge of the centriole’s role in human reproduction. In general, the centriole’s role is expected to be similarly essential in humans because of its conservation throughout animal evolution (Carvalho-Santos et al., 2011). However, the centrioles of sperm and the early embryo of murine animals are exceptions to this evolutionary conservation (Figure 1). While humans and many other mammals have centrioles in their spermatozoa and early embryos, mice, rats, and hamsters (the most common experimental mammals) do not have recognizable centrioles in their spermatozoa and early embryos (Schatten et al., 1986; Sathananthan et al., 1996; Phillips et al., 2014). These major differences in centriole appearance raise the question: What exactly is the role of the centriole in human fertilization and early embryonic development?

One extreme and unlikely idea is that sperm centrioles are not needed in humans because they are undetectable in mice sperm and appear to be expendable for murine live birth. However, an alternative hypothesis is that mice have evolved a novel biology of sperm and early embryonic centrioles, and, therefore, studying their role in reproduction is not applicable to their role in human reproduction and their clinical implication. Indeed, all studied non-murine mammals do have centrioles in their sperm and early embryo, and, in certain animals, some evidence indicates that these centrioles are necessary. Here, we will explore this evidence and the fact that the sperm of humans and non-murine mammals have two centrioles (one with typical structure and one atypical), not one, as was concluded in the past (Avidor-Reiss and Fishman, 2018). We start with a brief general introduction to centrioles, then progress to discussion of sperm and early embryonic centrioles. Then, we describe clinical studies that implicate the centriole in human reproduction. Finally, we propose future directions to resolve the question of the role of the sperm centriole role in the embryo. This review focuses on the role of the mature sperm (spermatozoon) centrioles in the zygote, and it does not address the role of centrioles in germline (spermatogonia, spermatocyte, and meiosis) development [reviewed in Riparbelli and Callaini (2011)].

Centrioles are evolutionarily conserved cellular components essential for fertilization, cell development, and animal physiology through their function in the cell [reviewed in Nigg and Raff (2009)]. The centriole is a cylinder built from nine triplet microtubules arranged in a ring-like formation to form a barrel structure. In the centrosome, the centrioles are surrounded by pericentriolar material (PCM) (Vorobjev and Chentsov Yu, 1982) [reviewed in Lange and Gull (1996), Mennella et al. (2012)]. In cilia and flagella, the centriole triplets are continuous with the doublet microtubules of the axoneme. Centrioles are composed of more than one hundred proteins (Preble et al., 2000) [reviewed in Winey and O’Toole (2014), Nigg and Holland (2018)], and, while new components of centriolar protein structure continue to be discovered, determining their functionality can be a challenge, both because they often have essential roles in early development [reviewed in Schatten and Sun (2010)], and because their function may vary from one cell type to another [previously reviewed in Loncarek and Bettencourt-Dias (2018)].

Most cells have two centrioles during early interphase. Most centrioles form by “duplication,” where each of the two-preexisting centrioles direct the formation of one new procentriole, providing a mechanism to control the number of centrioles formed. Centrioles may infrequently form de novo, in the absence of preexisting centrioles or in specialized cell types, but often times, this mechanism results in too many centrioles being formed (Loncarek and Khodjakov, 2009; Shahid and Singh, 2018). Progression from a procentriole to a mature centriole correlates with cell-cycle progression (Lange and Gull, 1996). A pair of centrioles is made of the cell’s older mother centriole and a younger daughter centriole. As a result, the number of centrioles in the cell oscillates between two centrioles during early interphase and four centrioles during late interphase and mitosis (Fukasawa, 2007).

Centrioles are integral to the cilia, which confer motility and extra-cellular communication potential to the cell. In most cell types, the cilium is formed by the mature centriole within the cell, which is known as the basal body [reviewed in Avidor-Reiss and Gopalakrishnan (2013)]. However, it was recently proposed that in sperm, the flagellum is formed by the younger daughter centriole, but this proposal requires further support (Garanina et al., 2019). Centrioles are also involved in the cell division process, but their contributions vary in distinct cell types. In general, centrioles recruit PCM to form a centrosome. This centrosome associates with microtubules to assemble an aster that localizes to the spindle poles during mitosis. The asters’ main function is to determine the axis of cell division and the number of spindle poles (Bobinnec et al., 1998). Interestingly, some animal cells [oocytes, importantly (Calarco et al., 1972)] and many plant cells (Joshi and Palevitz, 1996) lack centrioles. This implies that centrioles are not necessary in all cell types and that some cells can successfully complete mitosis and cellular organization without centriolar influence. Therefore, the significance of centrioles may vary in the several cell types that normally contain them.

Oocytes contain most of the elements necessary for zygotic development (i.e., Golgi, ER, and proteasomes), including centriolar and PCM proteins, but they lack assembled centrioles. Oocyte centrioles are eliminated, either by extrusion or inactivation, during oogenesis in humans and most studied animals (Hertig and Adams, 1967; Sathananthan et al., 1985; Schatten et al., 1986; Pickering et al., 1988; Crozet, 1990; Navara et al., 1994; Crozet et al., 2000) [reviewed in Sathananthan et al. (2006)]. Limited studies have explored the mechanism of oocyte centriole elimination, but they propose a mechanism reliant on some of the same proteins that are involved in centrosome biology, such as PLK1 (Pimenta-Marques et al., 2016). Centriole elimination by the oocyte necessitates contribution of both spermatozoan centrioles to the zygote to ensure an appropriate number of centrioles in the embryo cell (Connolly et al., 1986; Schatten, 1994; Manandhar et al., 2005; Pimenta-Marques et al., 2016). More information can be found in a recent review on how animal oocytes assemble spindles in the absence of the centriole at Severson et al. (2016).

In mice, centriole inheritance to the embryo diverges from other mammalian models. In mice, centrioles appear to completely degenerate during spermiogenesis (Manandhar et al., 1998), and, as a result, sperm centrioles may not be present within the zygote (Schatten, 1994; Hewitson et al., 1997) (Figure 1C). It has been proposed that zygotic centrioles are inherited maternally or form de novo (Schatten et al., 1985, 1986, 1991; Gueth-Hallonet et al., 1993; Hewitson et al., 1997). The support for non-paternal inheritance is based on several observations, including: (1) zygotes do not appear to have a dominant sperm aster and, instead, have maternal mini-asters, (2) the sperm axoneme is present in the oocyte but lacks microtubules and shows no association with mitotic poles, (3) centrioles are observed only starting in the 32/64 cell stage of early embryos (Gueth-Hallonet et al., 1993), and (4) intracytoplasmic sperm injection (ICSI) with disassociated sperm nuclei is sufficient for embryo development (Kuretake et al., 1996; Yan et al., 2008). The question of why the centrioles of the murine zygote are not inherited from the sperm, as is seen with other mammals, remains unanswered.

The ability to form a parthenogen (an embryo that is developed from an unfertilized egg that lacks centrioles) and parthenogenic cell lines is often referenced to suggest that sperm centrioles are not essential in mammals. However, it is important to note that mammalian parthenogenesis does not lead to viable offspring (Wininger, 2004). It has been proposed that imprinting defects are the main barrier to viability in mammalian parthenogenesis (Kono, 2006; Miller et al., 2019), but concomitant centriole abnormalities present during mammalian parthenogenesis may also contribute to this barrier.

The importance of sperm centrioles in mammalian parthenogenic embryos is apparent, in that parthenogenic cells have an abnormal number of centrioles, as expected from a de novo mechanism of centriole formation (Brevini et al., 2012). In the absence of preexisting centrioles (as would be expected in parthenogenesis), an unregulated number of centrioles is expected to form de novo in the activated oocytes or subsequent cell types of the embryo (Hinchcliffe, 2011). The presence of too many or too few centrioles, as is observed in mammalian parthenogenic cell lines, can lead to chromosomal instability and, ultimately, cell death (Sir et al., 2013; Godinho and Pellman, 2014). These centriole abnormalities can partially explain the presence of high levels of chromosomal abnormalities in parthenogenic embryos and their ultimate inability to develop viable offspring (Bhak et al., 2006). Therefore, it is possible that non-murine zygotes need two centrioles to produce a viable embryo, and further study of parthenogenic centrioles is necessary, particularly to determine at what stage they form, how many are formed, and how they function.

Though studies in humans are lacking, two sperm centrioles are likely present in the early zygotes and four during mitosis. The human zygote inherits the PC from sperm, which is well established (Sathananthan et al., 1991). Also, the human zygote is likely to inherit the DC because (1) the spermatozoa have a remodeled DC that is attached to the axoneme (Fishman et al., 2018), and (2) the base of the axoneme is located at one of the spindle poles (Asch et al., 1995; Van Blerkom, 1996; Simerly et al., 1999; Kovacic and Vlaisavljevic, 2000).

Kai et al. (2015) showed the presence of two centrosomes in zygotes with two pronuclei (presumably fertilized by a single sperm) by immunofluorescence staining for centrosomes. Sathananthan et al. (1991) identified centrioles within the centrosomes of zygotes with two pronuclei by electron microscopy following in vitro fertilization; two centrosomes with one or two centrioles were observed. Both Kai et al. (2015) and Sathananthan et al. (1991) have observed four centrosomes in early zygotes with three pronuclei, suggesting that dispermic embryos provide extra centrioles that form extra centrosomes. Interestingly, only three and not four spindle poles were observed in these zygotes. The presence of only three poles in dispermic embryos, despite the presence of four centrosomes, raises the question of whether the number of poles is determined by the sperm centrioles or pronuclei. However, these studies are consistent with the idea that human sperm normally contribute two centrioles to the zygote.

After membrane fusion, the centrosome forms one sperm aster near the head, which then enlarges throughout most of the zygote cytoplasm (Van Blerkom, 1996; Simerly et al., 1999). During pronuclear migration, this aster splits into two. Later, each aster localizes to one pole of the forming bipolar mitotic spindle. Interestingly, after initial formation, the asters collapse and are either small or unrecognizable during metaphase. This collapse is transient, as the asters reappear in anaphase. Why asters collapse during metaphase is unknown and needs further investigation.

It is commonly thought that human sperm centrioles are essential for pronuclear migration, based on studies on abnormal sperm asters in the zygote. Immunofluorescence analysis of fertilized zygotes with arrested development has shown disorganized and diminished sperm aster microtubule arrays, along with the lack of pronuclear formation and/or migration (Asch et al., 1995; Van Blerkom, 1996). These concurrent findings suggest that the sperm aster may be responsible for normal pronuclear development. These studies are limited by the fact that microtubules are also nucleated by non-centriolar microtubule-organizing centers, and that, in many of these studies, a defect in the centriole was not found or studied, making the specific role of the centriole and sperm aster in the zygote unclear.

After entering the egg, the PC is thought to be released from the sperm neck structures before recruiting PCM and forming astral microtubules, a process proposed to be mediated by proteasomes (Wojcik et al., 2000; Rawe et al., 2008; Sutovsky, 2018), but it is unknown if the DC is detached from the neck structures. Spermatozoan proteasomes have been localized to the neck region/midpiece in close association with the centrioles; however, relatively little is known about their function. Spermatozoa from humans and bovine, preloaded with function-blocking anti-proteasome antibody, resulted in disrupted sperm aster formation and pronuclear development/apposition of human oocytes, despite the lack of observable centriole structural deficits (Rawe et al., 2008). This suggests that spermatozoan proteasomes may play an important role in centriole contribution to zygote development. More pharmacological and genetic studies should be done to investigate the mechanism and the precise role of the proteasome in sperm and zygote centrioles.

Sperm with impaired centrioles were long thought to result in embryos with abnormal cleavage and infertility (Sathananthan, 1994; Schatten, 1994). This idea is now widely accepted and is supported by some epidemiological studies using artificial insemination to overcome barriers to fertilization in defective sperm samples. However, this idea is based on few studies, which analyzed 1–10 patients and few fertile donors as a control (Table 1). The phenotypes observed in these studies are diverse, providing opportunity for many inferences but making direct correlation with centriole phenotype difficult. Also, many studies are complicated by the presence of other defects in the sperm (in addition to the centriole), which may have a significant impact on the embryo phenotype. In particular, one can expect that some forms of centriolar defect may originate early in germline development and will result in abnormal DNA content due to abnormities in mitosis or meiosis in addition to the centriole defect, as was found for the mutation in the centriolar protein Centrobin (Liška et al., 2009). Unfortunately, these drawbacks negatively impact the confidence that we have in inferring the role of the sperm centriole in embryo development. However, the totality of these studies suggests that sperm centrioles are likely essential for normal embryo development.

Intracytoplasmic sperm injection is a useful treatment for male infertility due to interference with sperm translocation to the egg and fusion with it, but not for infertility due to dysfunction of sperm components that are needed after fusion (Palermo et al., 1992). Indeed, ICSI allows successful fertilization for many couples, but this may require multiple attempts, and, in many other couples, this fails [reviewed in Neri et al. (2014)]. This suggests that in these failed cases, infertility may be caused by processes downstream of gamete fusion that depend on the male contribution. However, the multifactorial nature of reduced semen quality has prevented many of the underlying contributory molecular mechanisms from being fully elucidated.

The only routine clinical assay to assess sperm phenotype is semen analysis, and this assay by itself cannot implicate the sperm centriole in infertility. Semen analysis compares the quality of the patient’s semen and sperm against several standardized parameters and helps identify specific defects in the male which are seen in up to one-third of infertile couples (Isidori et al., 2006) (Table 1). One group of sperm defects with several phenotypes that could originate from centriolar defect is idiopathic male factors (male infertility due to a sperm defect that has no explanation, i.e., when the male is infertile despite having normal semen analysis, history, and physical examination, and when female factor infertility has been ruled out) (Chemes and Rawe, 2010). For example, it was suggested that abnormal centrioles will form an abnormal sperm tail (which is one phenotype of teratozoospermia, defined as presence of sperm with abnormal morphology), a defective sperm axoneme resulting in poor motility (asthenozoospermia), or a combination of the two (asthenoteratozoospermia). Some of these cases of male factor infertility may be due to a subtle sperm centriolar defect–a hypothesis that will be examined in the following sections.

Several assays for studying typical sperm centrioles are available. However, the technology is either insufficiently specific or inappropriately laborious. Sperm cell centrioles has been assessed by electron microscopy, a laborious technique that is inadequate for large-scale studies and is inaccessible in most clinical settings (Chemes et al., 1987b; Moretti et al., 2016). Sperm centriolar protein content has been assessed by Western blot, which studied total protein and is not specific to the centriole and cannot conclusively implicate the centriole in infertility (Hinduja et al., 2010). Sperm centriole function has been assessed by microinjection of human sperm into bovine, rabbit, or hamster oocytes followed by immunofluorescence staining for aster formation, which has been useful in studying abnormal sperm centrioles in some infertility cases (Rawe et al., 2002; Terada et al., 2004; Yoshimoto-Kakoi et al., 2008). However, this method is now illegal according to the Dickey-Wicker amendment (the U.S. federal bill that prohibits funding for the creation of human embryos for research purposes). Because of these limitations, sperm centriole defect in infertile men has only been demonstrated in small case studies (Nanassy and Carrell, 2008; Chemes, 2012; Moretti et al., 2016; Sha et al., 2017). Therefore, there is a need for a specific and high-throughput method for assessing sperm centrioles and determining their association with certain embryonic phenotypes and outcomes.

Previous animal and epidemiological studies of sperm centrioles provide a basis for a correlation between centriole abnormality and embryo development pathology; however, further investigation is needed for conclusiveness on this subject. This investigation should include new model mammals and more conclusive clinical research.

Since mice do not have detectable centrioles in the sperm and early embryo, other mammalian models with these centrioles should be developed. For instance, centriole inheritance in rabbits resembles that of humans, and they are suitable models because they are the smallest non-murine mammal with established genetic engineering (Liu et al., 2004; Terada, 2007). The generation of a model rabbit for studying centriole-based infertility would be very useful; however, to do so, the tools necessary for specifically interfering with the sperm centriole must first be identified. Characterizing the mechanisms of sperm centriole formation and function as well as their unique properties will provide the insights necessary to develop these tools. CRISPR/Cas9 editing can be used to alter sperm-specific sequences, which will enable the development of a specific model animal.

Despite the correlation between centriole dysfunction and infertility that has been proposed in the epidemiological literature, the extent and significance of this relationship is still not known. The main reason is that most human studies to date have not demonstrated a clear case of specific centriole dysfunction. Therefore, a diagnostic test should be developed to better identify sperm with specific centriolar defect. Also, because such defect is likely to affect only a small fraction of all infertility cases, the diagnostic test needs to be easy and rapid. One additional approach for improving epidemiological studies is to include advanced semen analysis that examines DNA and egg activation factors, followed by testing sperm that have only centriolar defect. Ultimately, these efforts should lead to more conclusive clinical research, such as large prospective cohort studies and randomized controlled trials.

Sperm centriole defect may be mediated by genetic, environmental, or infectious factors; however, no directed effort has been made to identify these factors. Therefore, there is a need to identify sperm centriole defect in human, to find their cause, and to characterize its impact on male infertility.

TA-R conceived, supervised, and wrote the manuscript. MM wrote the manuscript and searched the literature. EF prepared the figures. PS performed the clinical perspective.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer GC declared a past co-authorship with one of the authors TA-R to the handling Editor.