| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Animal Science,3
Obstetrics and Gynecology,4
Veterinary Pathobiology,5 University of Missouri– Columbia, Missouri 65211
 |
ABSTRACT |
|---|
fertilization, gametogenesis, meiosis, ovum, sperm
|
INTRODUCTION |
|---|
|
STRUCTURAL AND FUNCTIONAL DEFINITIONS OF THE CENTROSOME |
|---|
The centriolar cylinder and the fibrous pericentriolar material comprise more than 100 different types of proteins [7–9]. Among them, the 
-tubulin ring complexes (TuRC), comprising -tubulin and accessory proteins, are directly involved in microtubule nucleation [10, 11]. TuRC are embedded in the pericentriolar matrix [12], possibly anchored by the centrosome protein pericentrin [reviewed in 13]. Several cytoplasmic and nuclear proteins associate with centrosomes in a microtubule-dependent or independent manner. Dynein and proteins coupled with dynein move to the centrosome through microtubular tracks by minus end–directed motor activity [14]. Some proteins localize to centrosomes only during dividing stages [8]. Nuclear-mitotic apparatus protein (NuMA) is a nuclear protein during interphase, but it associates with centrosomes during spindle assembly after nuclear envelope breakdown [15]. Some newly discovered centriolar/centrosomal proteins have been described in recent reviews [16, 17].
Unlike other cellular organelles, centrosomes do not have a definite shape, size, or limiting boundary. Their shapes and activities change dynamically during different stages of the cell cycle. The centrosome structure is highly variable in different cell types and organisms. Homologous organelles are spindle pole bodies (SPBs) of yeast [18], basal bodies of flagellates [19], deuterosomes of ciliated epithelial cells [20], blepharoplasts of lower plants [21], etc. Due to extremely diverse forms in different types of cells, it has been problematic to propose a universally applicable definition of the centrosome.
As with structural plasticity, centrosomes perform varied functions. They are the major microtubule organizing centers (MTOCs) of cells. Centrosomes emanate microtubule asters during interphase and spindle microtubules during dividing stages. In cycling cells, the centrosomes duplicate into two before entry into mitosis [22], split, and form the poles of bipolar spindles during prometaphase. This ensures inheritance of a complete functional centrosome to each daughter cell. Ultrastructurally centrosome duplication is reflected by the centriole replication [5, 6, 23]. At the G1/S transition, the mother and daughter centrioles form globular fibrous material, called the procentriole, associated with the proximal region [6, 22]. They grow into daughter centrioles during S and G2 periods, finally forming two pairs of centrioles that are indicative of duplicated centrosomes. The microtubular triplets of the original centrioles do not act as a template but serve as a site where the procentriole is laid down. Centriolar microtubules grow within the fibrillar matrix of the procentriole, seemingly around a cartwheel structure.
The centrosomes also play an important role in cytokinesis [24, 25]. During spermiogenesis, the distal centrioles form microtubular axonemes of the sperm tails. The basal bodies/centrioles of ciliated epithelia generate motile cilia. It has been argued that the formation of the motile apparatus in male gametes is the universally conserved and indispensable function of the centriole [26]. During fertilization, except in rodents, the sperm centrosome assembles a microtubular aster. The female pronucleus moves toward the sperm pronucleus along the astral microtubules [27, 28] using a dynein motor system [29, 30]. Centrosomes house many key molecules of cell cycle regulation [31–37]; research has led to speculation that centrosomes may trigger a phosphorylation cascade resulting in entry into mitosis. Cells in which centrosomes are depleted by microsurgery or laser ablation are able to complete mitosis and G1 stage, but become arrested at the G1/S transition [24, 38], suggesting that the centrosomal integrity is crucial for entry into S phase.
Centrosomes may be involved in several other cellular functions in addition to those described above [39]. Centrosome functions are crucial for eukaryotic cells. Abnormal centrosomal activity leads to genomic instability [40, 41]. Normality of centrosomes is strictly maintained in the cycling cells by intimate coordination between the nuclear and centrosomal cycles, ensured by coordinated checkpoint control mechanisms at the G1/S [24, 42], G2/M, and M/G1 transitions [43]. When cells exit the mitotic cycle and undergo differentiation, the centrosomes undergo profound changes to accommodate the specialized morphological and functional requirements.
Male and female gametes are highly specialized cells that are produced after a prolonged dictyate stage, meiotic cell divisions, and intricate morphogenesis. Centrosomes undergo degeneration and profound modification during the final stages of gametogenesis to meet the specific needs of gametic functions and fertilization. In the following sections, centrosome reduction in male and female gametes and their significance will be described in detail.
|
MALE GAMETOGENESIS |
|---|
-tubulin, and pericentrin (G. Manandhar and G. Schatten, unpublished observations). Male meiotic spindles of most animal species display two centrioles in each pole [45]. Centrosomes are structurally and functionally intact until the end of meiosis. The haploid cells formed after meiotic divisions are round spermatids. Morphogenesis of fully differentiated spermatozoa from the round spermatids (spermiogenesis) takes place in the seminiferous tubules in close association with Sertoli cells. During this process, the spermatids shed the excess cytoplasm as residual bodies. Morphologically fully formed spermatozoa are released into the seminiferous tubule lumen in a process called spermiation. Further maturation takes place in the epididymis. Mammalian spermiogenesis has been reviewed elsewhere [46–48].
|
CENTROSOME REDUCTION DURING SPERMIOGENESIS |
|---|
-Tubulin is discarded in the residual bodies during the spermiation stage (Fig. 1B) [50]. Centriolar disintegration begins in the testis and continues through the epididymis. Mouse spermatozoa completely lose -tubulin, centrin, and centrioles [44, 50]. Complete centriolar degeneration has also been reported in rat spermatozoa (Fig. 1C) [51].
|
Nonrodent mammalian spermatozoa degenerate centrosomes partially. Mature spermatozoa of rhesus monkeys lose -tubulin but retain centrin [52]. The proximal centrioles remain intact in sheep [53], bull [54], rhesus monkeys [55], and humans [56, 57], but the distal centrioles undergo various degrees of degeneration in several species (Fig. 1D) [58–61]. Rhesus and human spermatozoa show highly degenerated distal centrioles with 50% of their microtubular triplets lost [62]. Among the remaining, the majority is collapsed, and their A-tubules are filled with dense material. The central lumen of the residual distal centriole is occupied by a proximally extended microtubular duplex of the axoneme.
The mammalian model of sperm centrosome reduction also holds true in lower animals. Xenopus spermatozoa do not display -tubulin [63, 64] but possess some other centrosomal proteins such as pericentrin [63], CTR2611 [64], and Spc98p [65]. Drosophila spermatocytes display -tubulin, centrosomin, and centrin that are discarded from the mature spermatozoa [66]. Centrioles degenerate to various extents in different lower animals [45]. Snail (Lymnaea stagnalis) spermatozoa lose both centrioles at maturity [67]. The proximal centriole disappears during prometaphase II, so that the round spermatids possess only one centriole. During late spermiogenesis, the microtubular cylinder of the centriole is replaced by nine amorphous columns [67]. In insects, centrioles either disappear during spermiogenesis [68] or become modified into various "acentriolar" structures [69]. Centrioles have been shown in Xenopus [64, 70] and Drosophila spermatozoa (Fig. 1E) [71].
Why do centrosomes degenerate during spermiogenesis? The answer to this question is largely speculative. Centrosome degeneration has been studied in some experimental models. In cultured cells, centrosomes degenerate when they are physically damaged by X-rays [72], laser irradiation [73], or treated with antimitotic drugs [74]. Centrosome degeneration in cultured cells exposed to antiglutamylated tubulin antibody closely mimics spermatogenic centrosome reduction. The antibody binds to the glutamylated sites of tubulin, making them inaccessible to the centriolar organizer proteins resulting in disintegration of centrioles. Centriole loss is accompanied by the disjunction of centrosomal material from the pericentriolar region and scattering into the cytoplasm. Analogous centrosome degeneration in gametes may be related to depletion of cytoplasmic reserves of centrosomal constituents, since the synthetic activities of the nuclei are totally shut down during the late spermiogenesis stages. In Chlamydomonas and Paramecium, microtubular triplets of the basal bodies/centrioles degenerate as the result of a deficiency of 
-tubulin [75, 76]; 
-tubulin [77, 78]; Bld10-p [79]; and Vfl1 [80], etc. [17]. It is very likely that centriole degeneration during animal spermiogeneses could be related to deficiency of homologous molecules. The role of a ubiquitin-proteasome system in centriole degeneration is also an interesting field to be investigated [81]. Pericentrin and Spd-2 are involved in recruiting the centrosomal materials around the pericentriolar lattice in frog egg extract and nematode embryonic cells [82, 83]. Conversely, a lack of these proteins could be implicated in disjunction and/or loss of pericentriolar material from the degenerating centrioles of gametogenic cells. The existence of divergent molecular pathways of pericentriolar material disjunction and centriolar disintegration would explain why these two events are temporally separated during spermiogenesis. Centrioles of various differentiating somatic cells lose pericentriolar material and cease to function as MTOC before degeneration [49].
There is no proven explanation why spermatozoa of some animal species lose both centrioles while other species only partially degenerate the distal one. Perhaps the distal centrioles are derived from the mother centrioles, which are one cell cycle generation older than the proximal centrioles [6], and hence are more vulnerable to degeneration. Molecular markers of the old (mother) and new (daughter) centrioles have been recently described [84–87] that might help to track the ‘aging’ or ‘senescence’ of the centrioles during spermiogenesis.
|
SIGNIFICANCE OF SPERM CENTROSOME REDUCTION |
|---|
The finding that centrioles are not present in all animal spermatozoa is intriguing. Because many invertebrate species and rodents eliminate both centrioles during spermiogenesis, centrioles must originate de novo in embryos of those animals. In nonrodent mammals including monkeys and humans, the distal centrioles exist in highly degenerated and modified form. A question was raised whether such residual distal centrioles would regenerate and/or replicate in the zygotes [49, 62].
Centriole biogenesis in embryonic cells is yet an unresolved enigma in cell biology. As described above, new centrioles originate from procentrioles that are associated with the proximal region of the mature centrioles but not from the microtubular triplets of the centrioles. New centrioles also can arise de novo without being associated with a pre-existing centriole. In ciliated epithelia, the basal bodies originate from deuterosomes [92, 93] that are dense aggregates of fibrous substances surrounded by granular material in which several developing procentrioles are embedded [94]. In fertilized mouse embryos and parthenogenetic rabbit embryos, centrioles arise during the blastocyst stage by de novo synthesis [95, 96]. The precursor bodies of new centrioles of those embryonic cells are similar to procentrioles or deuterosomes [96, 97]. Evidently the procentriole-like bodies of the embryonic cells form solitary centrioles that replicate in an orthogonal manner afterward [97]. These observations gave rise to a question: If centrioles can originate de novo, why has the geometrically intricate orthogonal mode of replication been evolutionarily conserved? Perhaps it serves a very important purpose by strictly regulating centriole replication in a 2 
4 manner, equivalent to 1 2 centrosomes in each cell cycle that is crucial for bipolar spindle formation. Parthenogenetically activated sea urchin and insect eggs form multiple centrioles [98–100], possibly as the result of the lack of control on the initiation site. Cells can switch to a default pathway of de novo synthesis when their existing centrioles are experimentally destroyed, but in this mode cells cannot be restrained over the number of the resurrected centrioles [101]. Genes that control the fidelity of centriole replication are beginning to be discovered [102–104].
Preservation of proximal centrioles and degeneration of distal centrioles in nonrodent mammalian and some invertebrate spermatozoa suggests that centriolar pairs are asymmetrically replicated during fertilization. For example, in sea urchins the proximal centriole replicates in a normal fashion. The fibrous, globule-like residual distal centriole also assembles a new centriole without resurrecting its own microtubular triplets [89]. Perhaps a similar mode of centriole replication is valid in nonrodent mammals during fertilization. This could be the reason why one pole of the first cleavage spindle of sheep zygotes possesses two centrioles, whereas at the other pole only one exists [91]. A similar distribution of centrioles was also observed in monospermic androgenetic eggs [91]. An extensive electron microscopic study of human zygotes has failed to find centriolar duplexes in all spindle poles during the first cleavage [56, 57]. Nevertheless, centriolar duplexes occur invariably in all late-stage embryonic cells [95, 105].
These studies indicate that the loss of one or both spermatozoan centrioles does not impede in the regeneration of new centrioles in embryos. As discussed above, new centrioles originate from the procentrioles associated with the mature centrioles, not from the centrioles themselves. Hence degeneration of sperm centriolar microtubules might not affect their ability to organize procentrioles and support the generation of new centrioles. The hypothetical ‘polar organizers’ of starfish eggs/embryos described by Sluder and colleagues [106–108] may be equivalent to the procentrioles of the cycling cells. They colocalize with centrioles or exist independently when centrioles are absent. A typical example of de novo regeneration of centrioles is seen in Lymnaea stagnalis zygotes. The spermatozoa introduce basal bodies that are devoid of microtubular triplets [68]. New centrioles arise from deuterosome/procentriole-like structures independent from the sperm basal bodies [109]. Despite the oocytes' ability to generate centrioles without male contribution, replication mediated by a preexisting mature or residual centriole is seemingly more efficient.
|
FEMALE GAMETOGENESIS |
|---|
In mammals, the primary oocytes resume meiosis I division in response to luteinizing hormone (LH) secreted by the pituitary. Among the stockpile of arrested oocytes, only fully grown ones are competent to respond to LH at a given time to the developmental stimulus. The stimulated oocytes complete meiosis I division by forming secondary oocytes and first polar bodies. The secondary oocytes enter meiosis II but become arrested again at metaphase II stage. In most mammals, the metaphase II oocytes are released from the follicles and become ready for fertilization.
|
CENTROSOME REDUCTION DURING OOGENESIS |
|---|
A remarkable event of centrosome reduction during animal oogenesis is the loss of centrioles (Fig. 2). Premeiotic and meiotic stages of mouse oocytes have been thoroughly studied using electron microscopy. Oogonia and fetal oocytes display normal centrioles until pachytene stage, but lack them in the subsequent stages (Fig. 2, A, B, and E) [95]. During the germinal vesicle breakdown (GVBD) stage in mouse oocytes, multiple perinuclear MTOCs appear [95, 110], which gradually coalesce to form the poles of metaphase spindles. Centrioles are not observed in the perinuclear MTOCs or spindle poles of the meiotic oocytes [95]. Besides mice, lack of centrioles in the metaphase II spindle poles has been documented in rabbits [111], cows [105], sea urchins [89], Xenopus [112], humans [113], and several other species (Fig. 2E) [reviewed in 95].
|
Centrosomes of Drosophila oocytes help to transport nutrients from the surrounding nurse cells before being degenerated. In the beginning they nucleate microtubules, along which the inactive centrioles of the surrounding nurse cells migrate into the oocyte [114] and organize a large MTOC [115]. The microtubules emanating from the aggregated centrioles grow into nurse cells through the ring canals [114], directing the nutrients and mRNA flow into the oocyte from the nurse cells [114, 115]. In the course of degeneration, the centriolar aggregate eventually loses pericentriolar material and ceases microtubule-nucleating function [116]. Finally in the GV stage, MTOCs disperse into the cytoplasm and disappear beyond detection. The final fate of centrioles has not been investigated at the ultrastructural level, but they are presumed to disintegrate completely before the oocytes enter meiosis [116, 117].
Whereas mammalian and insect oocytes lose centrioles during pachytene arrest, echinoderm and mollusk oocytes degrade them during meiotic divisions (Fig. 2, A, D, and G). In starfish oocytes, the meiosis I spindle poles contain two centrioles, but the meiotic II spindle poles have only one [118]. Hence the two centrioles received by the secondary oocytes do not duplicate before meiosis II division probably because of the lack of S period. However, during that period, the unduplicated centriolar pair separates, each localizing to the opposite poles of the meiosis II spindle. Consequently, the fully formed eggs receive one centriole (Fig. 2G) [118]. What happens to the centrioles retained by the eggs is yet to be investigated. Most probably they rapidly degenerate as evidenced by the fact that the asters formed in parthenogenetically activated sea urchin eggs do not encompass centrioles during the early stages [119, 120]. A similar mode of centriole degeneration has been reported during oogenesis of Mytilus edulus [121] and crayfish [122].
The mode of centriole loss during oogenesis of pulmonary snail Lymnaea stagnalis is midway between the Drosophila and sea urchin types (Fig. 2, A, C, and F). Lymnaea oocytes possess centrioles [109] but do not duplicate before entering the dividing phase of meiosis. Hence the meiosis I spindle poles contain one centriole. The secondary oocyte inherits one centriole that is distributed to the meiosis II outer spindle pole and is extruded with the meiosis II polar body. Finally the mature eggs are devoid of centrioles (Fig. 2F) [109]. Interestingly, the inner pole of metaphase II spindle engulfs the sperm basal body, but its amorphous cylindrical structure does not resemble a centriole [67].
The molecular mechanism of oocyte centriole degeneration is totally unknown as in the case of spermatid centriole degeneration; nevertheless, they could be similar.
|
SIGNIFICANCE OF OOCYTE CENTROSOME REDUCTION |
|---|
Probably as a direct consequence of centriolar degeneration, oocytes lose the structural integrity of the centrosomes. Due to the absence of centrioles, the oocyte's centrosomal material does not aggregate into unified foci. The transitional stages of centrosomal protein dispersal due to centriole loss have been shown by rapidly fixing the mouse oocytes and pursuing correlative light and electron microscopic studies [123]. The pre-GVBD oocytes possess two large multivesicular aggregates (MVA) containing -tubulin. At the onset of maturation, the MVA are fragmented, transformed into smaller MTOCs, and translocated to the GV [123]. In dividing stage oocytes, the proximal ends of meiotic spindle microtubules coalesce into multiple bundles around the poles. Hence the spindle poles appear as rings [124, 125] or flat structures [126–128], and the spindles appear barrel-shaped [126, 129]. Mouse oocytes do not acquire centrioles from spermatozoa after fertilization, and hence the centrosomes remain dispersed as the cytoplasmic MTOCs, forming multiple asters during the pronuclear stage. The centrosomal proteins like 5051, pericentrin, and -tubulin are localized to the astral foci [124, 126, 128]. At mitotic entry, the asters disappear and the mitotic spindle forms from the perinuclear microtubules [126, 129]. In nonrodent mammalian zygotes and sea urchins, the sperm centrioles organize a dominant MTOC by recruiting dispersed oocyte centrosomal proteins [2, 63, 88, 130]. Oocytes that normally form a sperm aster after fertilization display cytoplasmic asters or microtubular network if activated parthenogenetically [130–132].
In Drosophila, the meiosis II spindles are formed in tandem [133]. The inner pole of the outer spindle and the outer pole of the inner spindle form a common central body analogous to a merged common pole [134]. The central body is a disk-shaped structure formed by accumulation of several centrosomal proteins like -tubulin [133, 135]; centrosomin [136]; and CP190 [137], but it lacks a centriole. None of these centrosomal proteins are observed in the proper spindle polar region [69].
Despite the lack of definite centrosomes, animal oocytes possess a variety of centrosomal proteins [reviewed in 2, 3]. Investigation of the oocyte's centrosomal proteins has been mostly done by immunocytochemistry using mono- and polyclonal antibodies. Positive detection of putative centrosomal proteins relies upon the presence of microscopically recognizable punctate structures in the oocytes. The soluble cytosolic proteins or proteins that are extracted during processing would not be revealed by immunocytochemistry. The oocyte's centrosomal proteins may exist in a dispersed form due to the lack of centriole and pericentriolar lattice. The majority of cellular -tubulin and centrin are present in soluble form [138, 139] that can be detected by immunoblotting or other molecular and biochemical techniques. Specificity of antibody reaction poses further problems in immunocytochemistry. An antibody produced against a particular isolated organelle or peptide from one organism may not cross-react with the homologous structures of other organisms. Most of the antibodies have been studied in one or few animal species, but have not covered phylogenetically diverse species [3], thus precluding a generalization.
Loss of Centrosomal Reproductive Capacity
Centriole degeneration in oocytes serves an important purpose by eliminating the reproductive capacity of the maternal centrosome that in turn helps to balance the centrosome number in zygotes and embryos. The reproductive capacity of centrosomes is correlated to the number of centrioles they contain. Two centrioles in the centrosome ensure full reproductive capacity (i.e., capability of producing a bipolar spindle in the subsequent division). The presence of one centriole results in half reproductive capacity that can produce only a monaster. These conclusions have been derived from a series of experiments in sea urchin [140] and starfish oocytes/zygotes [107–109]. By treating with mercaptoethanol, the dividing stage sea urchin zygotes can be blocked at a disorganized metaphase stage [140]. If the blockage is continued for the duration of one cell cycle, the mother and daughter centrioles of each spindle pole split apart and acquire independent centrosomal function. Upon reversal of the blockage, the zygotes re-enter mitosis and the four centrioles together organize a tetrapolar spindle with one centriole at each pole. The zygote divides into four blastomeres. Since each blastomere inherits a single centriole, their centrosomes cannot split into two, consequently forming monopolar spindles in the next mitosis. They enter into interphase without undergoing division; however, the centrioles subsequently replicate and form bipolar spindles in the following mitosis [107].
A simple correlation between the loss of centrioles and reproductive capacity of the centrosome holds true in sea urchin–type oocytes [118]. The meiosis I spindle poles contain two centrioles capable of producing two centrosomes by splitting, leading to the formation of a bipolar spindle in meiosis II division. Since the centrioles do not replicate before meiosis II division, each meiotic II spindle pole possesses only one centriole (Fig. 2G). The mature egg retains one centriole of the inner pole of the meiosis II spindle. One centriolar centrosome of the meiosis II spindle poles cannot form bipolar spindles. Moreover, unlike zygotic centrioles, the oocyte centrioles lose the ability to replicate. As a consequence, they are capable of forming only monasters in the subsequent division cycles [118]. The reproductive capacity of meiotic spindle poles was directly demonstrated by microinjecting them into pronuclear-stage zygotes [106]. The centrosomes of meiosis I spindle poles containing two centrioles double in the subsequent mitosis and produce bipolar spindles, whereas the meiotic II spindle pole centrosome containing one centriole fails to reproduce, resulting in monopolar spindles. These observations also indicate that the oocyte centrosomes lose reproductive capacity between meiosis I and II divisions [108].
A correlation between the centriole number and centrosome replication is not valid in all cases. Even in sea urchins, the microinjected metaphase I spindles sometimes failed to produce bipolar spindles [106], indicating degradation of the reproductive capacity of their centrosomes. When injected into zygotes, the first polar bodies containing two centrioles assemble monopolar spindles in a similar manner as the second polar bodies possessing one centriole [141]. Apparently, functional degeneration (ability to duplicate) of centrioles takes place before the structural degradation (reduction in number).
Loss of centriole and elimination of the reproductive potential of oocyte centrosome are crucially important for successful fertilization and cleavages. The zygotic centrosome organized around the sperm centrioles ensures pronuclear apposition, leading to close congregation of the male and female chromosomes into a single metaphase plate and bipolar spindle formation during embryonic mitoses. As evident, any extra centriole if retained in the oocyte would organize superfluous centrosomes interfering with pronuclear apposition and normal bipolar mitosis. On the other hand, the evolutionary strategy to retain centrioles in spermatozoa is justified by the fact that they need centrioles until the late spermiogenesis stage to generate axonemes of the tail. Centriole degeneration in oocytes confers an evolutionary advantage by accommodating exogenous centrioles introduced by the spermatozoa while maintaining the balance of centrosome number in the zygote. Mouse oocytes do not receive centrioles from spermatozoa after fertilization, yet their centrioles are completely eliminated as in other mammals. Independent evolutionary pathways of oocyte and sperm centriole degenerations as seen in mice suggest that the oocyte centriole degeneration might be involved in other vital functions (described below) in addition to accommodating sperm centrioles after fertilization.
Checkpoint Control of Parthenogenesis
Besides delivering the haploid male genome, spermatozoa perform additional vital functions by triggering the embryonic cell cycle and providing essential centrosomal components to the oocytes. Unfertilized oocytes can be activated by various physical or chemical stimuli [142, 143] and their female haploid genome can be diploidized, but the embryonic development does not proceed to term. One of the reasons could be the inability of the parthenogenetic oocytes to organize normal spindles due to the lack of a functional centrosome [144]. In Xenopus, microinjected exogenous centrosomes can function as zygotic centrosomes and induce successful parthenogenesis. It was shown that the centrosomes from various cell types are capable of inducing parthenogenesis [145–147] lest they contain intact and replication-competent centrioles [148, 149]. These observations imply that the oocyte centrosome reduction has evolved as a control mechanism to suppress parthenogenetic development. Due to the loss of functional autonomy, the oocyte centrosome cannot initiate or successfully complete normal embryonic cleavages without being supplemented by a fertilizing spermatozoon.
Yet many insect species are obligatory or facultative parthenotes [150]. Successful parthenogenesis depends upon the oocyte's ability to generate complete and functional centrosomes in the absence of one supplied by a spermatozoon. Parthenogenetically activated oocytes form multiple cytoplasmic asters [69, 150–152] containing centrosomal proteins and centrioles [100]. The asters behave like typical centrosomes by replicating and splitting [100]. Among the multiple astral centrosomes, two become associated with the female pronucleus and form the mitotic spindle, whereas the others degenerate. In some stick insect species, the spermatozoa do not contribute centrioles [153], so eggs replenish all of the components of the zygotic centrosome. For this reason, parthenogenetic and fertilized mitotic spindles are assembled in a similar manner and with equal efficiency [154]. This in turn has resulted in a widespread occurrence of parthenogenesis in stick insects.
Parthenogenetic development is not an evolutionarily preferred pathway of species propagation because it leads to genomic homogeneity that in turn results in the accumulation of genetic anomalies in the population. A typical example of anomalous features of parthenogenetic development is seen in Sciara embryos. Normally a spermatozoon introduces a giant basal body containing 60–90 singlet microtubules [155]. During fertilization they are broken down to give rise to MTOCs producing cytoplasmic asters. The cleavage spindles, however, originate from the perinuclear microtubules. The parthenogenetic oocytes lack asters; nevertheless, they assemble mitotic spindles from the perinuclear microtubules in a similar way as the fertilized oocytes. There is no visible anomaly during the first mitosis, but the parthenogenetic syncytial embryos formed after several division cycles are distinctly abnormal and nonviable. The abnormalities are attributed to the dysfunction of the centrosomal apparatus failing to separate and translocate the daughter nuclei to the cortex [156].
The oocyte centrosome becomes active only when sperm penetration fails. Although cytoplasmic asters form in the fertilized and unfertilized Nasonia oocytes, the sperm centrosomes always take over the role of the zygotic centrosome in the fertilized oocytes. In the absence of sperm, the oocyte asters participate in the cleavage spindle formation [152]. When parthenogenetically activated Drosophila mercatorum oocytes are fertilized, the oocyte centrosomes remain in the cortex forming small asters, whereas the sperm centrosomes recruit centrosomal proteins, nucleate astral microtubules, and establish efficient interaction with the female pronucleus [100], thus directing the embryonic development through the fertilized pathway.
Mouse oocytes do not receive centrosomes from spermatozoa, yet they do not develop through parthenogenesis. According to a viewpoint put forth by Gerald Schatten [2], eutherian mammals have evolved genomic imprinting as a strategy to ensure biparental fertilization, and mice represent the vanguard. Genomic imprinting is the differential methylation of maternal and paternal genomes [157–159] so that their expressions in the zygotes complement each other. In other words, without paternal genes the expression of the maternal genome alone cannot ensure normal embryonic development. According to G. Schatten [2], this strategy might have loosened a stringent control over the requirement of a paternal centrosome to initiate embryonic development, which is why parthenogenetic activation or early cleavages are not so uncommon in higher mammals [2]. The parthenogenetic embryogenesis does not proceed to term, possibly because of genetic imbalance [160], rather than as the result of centrosomal abnormality.
Acentrosomal Mode of Spindle Organization
The oocytes switch to acentrosomal mode of spindle organization due to centrosome reduction. Acentrosomal spindle assembly is accomplished in two steps: spontaneous nucleation of microtubule around the condensed chromatin followed by sorting of the randomly oriented microtubules into the bipolar spindle. Spontaneous microtubule nucleation is due to localized elevated activity of Ran GTPase protein [161]. The active form of Ran GTPase is GTP-bound and is associated with chromatin because it is generated by the GTP exchange factor, RCC1 [162], that colocalizes with chromosomes. In the downstream of Ran-GTPase activity, TPX2 (target protein for Xenopus kinesin-like protein 2) and NuMA are released from the inhibitory association with 
- and ß-importins [163, 164]. The TPX2 in turn stimulates Aurora A kinase, Eg2 [165], resulting in enhanced spindle assembly [165, 166]. Eg2 may result in phosphorylation of multiple substrates, including kinesin-like protein Eg5 [167] that in turn helps to establish the bipolar spindle. The inactive form of Ran GTPase is GDP-bound, and is distributed in the cytoplasm [168, 169]. The GTPase activating protein found in cytoplasm inactivates Ran by hydrolyzing GTP [170].
Various microtubular motors play a role in generating characteristic bipolar spindles by bundling, polar sorting, and polar focusing of the randomly nucleated microtubules. Cross-linking BimC motor proteins like Eg5 and Klp61F help in bundling and parallel-orienting the microtubules [171, 172]. Kinesin-like proteins (plus-end directed) including Xklp1 and Klp2 orient the microtubular plus-ends toward the chromosomes and minus-ends away [173, 174], thus contributing to bipolar spindle assembly. Centromere-associated kinesin-like protein (CENP-E) localized to the kinetochore corona fibers [175, 176] captures the microtubular plus-ends and stabilizes them as kinetochore microtubules [177]. Due to the pushing force of the kinetochore microtubules against the chromosomes and pulling force of the CENP-E on microtubules, the chromosomes congregate at the metaphase plate [178]. Dynein moves NuMA and various other molecular cargos along the microtubules toward the minus-ends and accumulates in the polar regions [179, 180]. Due to microtubule cross linking characteristics of NuMA [180, 181] and dynein, the minus-ends of the spindle microtubules are tethered together forming foci in the polar region [180, 182, 183]. Some other minus-end– directed motor proteins like Drosophila Ncd kinesins and Xctk2 are also required for maintaining integrity of the spindle poles [184, 185]. Once the polarized spindles are established, the centrosomal proteins are recruited to the spindle poles [100, 186] that may further promote microtubule nucleation.
The mechanism of acentrosomal spindle organization as described above is an attractive hypothesis; however, such evidence is mainly based on experiments done in cell-free Xenopus egg extracts. Spindle organization in animal oocytes may be similar to that of in vitro–assembled spindles in Xenopus egg extracts [187, 188], but a strict analogy is not justified considering the fact that they are very different systems [189; see also 190]. Systematic investigations are necessary to understand how acentrosomal spindles are organized in oocytes.
Centrosome reduction and switching to acentrosomal mode of spindle organization in oocytes have multiple evolutionary implications. Acentrosomal spindle organization serves as a default mode that ensures normal meiotic nuclear divisions and cytokineses in the absence of centrosomes. The acentrosomal spindles have lower fidelity; they can complete meiotic divisions but are unable to initiate or execute normal embryonic cleavages except in some rodent species. Hence the oocytes are unable to undergo embryogenesis until the centrosomes of fertilizing spermatozoa restore the centrosomal mode of spindle organization.
|
CONCLUSIONS |
|---|
Understanding of the gametic centrosome would bear immediate medical benefit in resolving male infertility related to putative centrosomal dysfunction. The missing centrosomal component of the infertile spermatozoa is not known yet. When properly characterized, it could be biochemically synthesized and injected into the oocytes in conjunction with in vitro fertilization or intracytoplasmic sperm injection therapy.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: G. Manandhar, University of Missouri, S-141 ASRC, 920 East Campus Drive, Columbia, MO 65211. FAX: 573 884 5540; manandharg{at}missouri.edu
Received: 23 April 2004.
First decision: 25 May 2004.
Accepted: 7 September 2004.
|
REFERENCES |
|---|
TOP
ABSTRACT
INTRODUCTION
