In nearly all diploid eukaryotes, meiosis generates haploid germ cells (eggs and sperm), which can then fuse with a germ cell from another individual to generate a diploid zygote that develops into a new individual. Meiosis is a fundamental aspect of the biology and evolution of all eukaryotes because it results in the reassortment of the chromosome sets received from an individual’s two parents. Chromosome reassortment and homologous recombination between parent DNA molecules during meiosis guarantee that each haploid germ cell generated will receive a unique combination of alleles that is distinct from each parent as well as from every other haploid germ cell formed.
The mechanisms of meiosis are analogous to those of mitosis. However, several key features of meiosis allow this process to generate genetically diverse haploid cells. In the mitotic cell cycle, each S phase is followed by chromosome segregation and cell division. In contrast, during meiotic cell division, one round of DNA replication is followed by two consecutive chromosome segregation phases. This process leads to the formation of haploid, rather than diploid, daughter cells. During the two divisions, maternal and paternal chromosomes are shuffled and divided so that the daughter cells are different in genetic makeup from the parent cell. In this section, we dis cuss the similarities between mitosis and meiosis as well as the meiosis-specific mechanisms that transform the canonical mitotic cell cycle machinery so that it brings about the unusual cell division that leads to the formation of haploid daughter cells.
Extracellular and Intracellular Cues Regulate Germ Cell Formation
The signals triggering entry into meiosis in metazoans are a very active area of research, and much is still unknown. However, the same basic principles govern the decision to enter the meiotic program in all organisms in which this transition has been studied. Extracellular signals induce a transcriptional program that produces meiosis-specific cell cycle factors that bring about the unusual meiotic cell divisions. This modification of the cell cycle goes hand in hand with a developmental program that induces the features characteristic of gametes, such as the development of a flagellum in sperm or the production of a stress-resistant cell wall during spore formation in fungi. At least one of the extracellular signals inducing entry into meiosis in mammals is retinoic acid, a signaling molecule that, by binding to the transcription factor retinoic acid receptor (RAR), functions in many different developmental processes. The cellular targets of this hormone, and how it functions to specify the meiotic fate, remain to be discovered.
The molecular mechanisms underlying the decision to enter meiosis are well understood in S. cerevisiae. The decision to enter the meiotic divisions is made in G1. Depletion of nitrogen and carbon sources induces diploid cells to undergo meiosis instead of mitosis, yielding haploid spores. During the meiotic divisions, budding is repressed. Pre-meiotic S phase and the two meiotic divisions thus occur within the confines of the parent cell. Spore walls are then produced around the four meiotic products. Recall that budding and the initiation of DNA replication are induced by G1/S phase CDKs. Their expression needs to be inhibited to prevent budding. Nutritional starvation represses expression of G1/S phase cyclins, thereby inhibiting budding. However, DNA replication also relies on G1/S phase CDKs. How can pre-meiotic DNA replication occur in the absence of G1/S phase CDKs? The sporulation-specific protein kinase Ime2 takes over the role of G1/S phase CDKs in promoting DNA replication. Ime2 promotes (1) phosphorylation of the APC/C specificity factor Cdh1, inactivating it so that S phase and M phase cyclins can accumulate; (2) phosphorylation of transcription factors to induce ex pression of genes whose products are required for S phase, including DNA polymerases and S phase cyclins; and (3) phosphorylation of the S phase CDK inhibitor Sic1, leading to release of active S phase CDKs and the onset of pre-meiotic DNA replication.
Several Key Features Distinguish Meiosis from Mitosis
The meiotic divisions differ in several fundamental aspects from the mitotic divisions. These differences are summarized in Figure 1. During the meiotic cell division, a single round of DNA replication is followed by two cycles of cell division, termed meiosis I and meiosis II (Figure 2). Meiosis II resembles mitosis in that sister chromatids are segregated. However, meiosis I is very different. During this division, homologous chromosomes—the chromosome inherited from your mother and the same chromosome inherited from your father—are segregated. This unusual chromosome segregation requires three meiosis-specific modifications to the chromosome segregation machinery. In what follows, we discuss these modifications and explain why they are needed.
Fig1. Comparison of the main features of mitosis and meiosis.
Fig2. Pre-meiotic cells have two copies of each chromosome (2n), one derived from the paternal parent and one from the maternal parent. For simplicity, the paternal and maternal homologs of only one chromo some are diagrammed. Step 1: All chromosomes are replicated during S phase before the first meiotic division, giving a 4n chromosomal complement. Cohesin complexes (not shown) link the sister chromatids composing each replicated chromosome along their full lengths. Step 2: As chromosomes condense during the first meiotic prophase, the replicated homologs pair and undergo homologous recombination, leading to at least one crossover event. At metaphase I, shown here, both chromatids of one chromosome associate with microtubules emanating from one spindle pole, but each member of a homologous chromosome pair associates with microtubules emanating from opposite poles. Step 3: During anaphase of meiosis I, the homologous chromosomes, each consisting of two chromatids, are pulled to opposite spindle poles. Step 4: Cytokinesis yields two daughter cells (now 2n), which enter meiosis II without undergoing DNA replication. At metaphase of meiosis II, shown here, the sister chromatids associate with spindle micro tubules from opposite spindle poles, as they do in mitosis. Steps 5 and 6: Segregation of sister chromatids to opposite spindle poles during anaphase of meiosis II, followed by cytokinesis, generates haploid gametes (1n) containing one copy of each chromosome. Micrographs on the left show meiotic metaphase I and metaphase II in developing gametes from Lilium (lily) ovules. Chromosomes are aligned at the metaphase plate. [Photos courtesy of Ed Reschke/Peter Arnold, Inc./Photolibrary/Getty Images.]
The tension-based sensing mechanism responsible for accurately attaching chromosomes to the spindle during mitosis is also responsible for segregating chromosomes during meiosis I. Thus homologous chromosomes must be linked so that this tension-based mechanism can function. Homologous recombination between homologous chromosomes creates these linkages (see Figure 1). Here we restrict our discussion to the importance of homologous recombination to successful meiotic divisions.
In G2 and prophase of meiosis I, the two replicated chromatids of each chromosome are linked together by cohesin complexes along the full length of the chromosome arms, just as they are following DNA replication in a mitotic cell cycle. In prophase of meiosis I, homologous chromosomes (i.e., the maternal and paternal chromosome 1, the maternal and paternal chromosome 2, etc.) pair with each other and undergo homologous recombination. Significantly, at least one recombination event occurs between a maternal and a paternal chromosome. The crossing over of chromatids produced by recombination can be observed microscopically in the first meiotic prophase and metaphase as structures called chiasmata (singular, chiasma). In contrast, no pairing between homologous chromosomes occurs during mitosis, and recombination between nonsister chromatids is rare. Concomitant with homologous recombination, homologous chromosomes associate with each other in a process known as synapsis. In most organisms, this synapsis is mediated by a proteinaceous complex known as the synaptonemal com plex (SC). Homologous chromosomes linked through chiasmata are called bivalents (see Figure 1). The chiasmata and the cohesin molecules distal to them now provide the resistance to the pulling force exerted by microtubules on the metaphase I spindle (see Figure 2).
The recombination between homologous chromosomes that occurs in prophase of meiosis I has at least two functional consequences: First, it connects homologous chromosomes during meiosis I metaphase. Second, it contributes to genetic diversity among individuals of a species by ensuring new combinations of alleles in different individuals. (Note, however, that genetic diversity primarily arises from the independent reassortment of maternal and paternal homologs during the meiotic divisions.) The homologs, now connected through at least one chiasma, must align on the spindle in metaphase of meiosis I so that maternal and paternal chromosomes are segregated away from each other during anaphase of meiosis I. This requires that the kinetochores of sister chromatids at tach to microtubules emanating from the same spindle pole rather than from opposite spindle poles, as in mitosis (Figure 3). Sister chromatids attached in this way are said to be co-oriented. However, the kinetochores of the maternal and paternal chromosomes of each bivalent attach to spindle microtubules from opposite spindle poles; they are bi-oriented.
Fig3. Chiasmata and cohesins distal to them link homologous chromosomes in meiosis I metaphase. Connections between chromosomes during meiosis I are most easily visualized in organisms with acrocentric centromeres, such as the grasshopper. The kinetochores at the centromeres of sister chromatids attach to spindle microtubules emanating from the same spindle pole, with the kinetochores of the maternal (red) and paternal (blue) chromosomes attaching to spindle microtubules from opposite spindle poles. The maternal and paternal chromosomes are attached to each other by chiasmata, which are formed by recombination between them, and by the cohesion between sister chromatid arms that persists until metaphase I. Note that elimination of cohesion between sister chromatid arms is all that is required for the homologous chromosomes to separate at anaphase. See L. V. Paliulis and R. B. Nicklas, 2000, J. Cell Biol. 150:1223.
Finally, to facilitate two consecutive chromosome segregation phases, cohesins must be lost from chromosomes in a stepwise manner. Recall that during mitosis, all cohesins are lost by the onset of anaphase (Figure 4a). In contrast, during meiosis, cohesins are lost from chromosome arms by the end of meiosis I, but a pool of cohesins around kinetochores is protected from removal (Figure 4b). This pool of cohesins persists throughout meiosis I, but is removed at the onset of anaphase II. As we will see next, loss of cohesins from chromosome arms is required for homologous chromosomes to segregate away from each other during meiosis I.
Fig4. Cohesin function during mitosis and meiosis. (a) During mitosis, sister chromatids generated by DNA replication in S phase are initially linked by cohesin complexes along the length of the chromatids. During chromosome condensation, cohesin complexes (yellow) become restricted to the region of the centromere at metaphase. Mei-S332/Shugoshin (purple) recruits PP2A to centro meres, where they antagonize Polo kinase and Aurora B, preventing the dissociation of cohesins from centromeric regions. Dissociation of Mei-S332/Shugoshin from centromeres and activation of separase leads to removal of cohesins at the centromere. Sister chromatids now separate, marking the onset of anaphase. (b) In prophase of meiosis I, maternal and paternal chromatids establish linkages between each other by homologous recombination. By metaphase I, the chromatids of each replicated chromosome are cross-linked by cohesin complexes along their full length. Rec8, a meiosis-specific homolog of Scc1, is cleaved along chromosome arms but not around the centromere, allowing homologous chromosome pairs to segregate to daughter cells. Centromeric Rec8 is protected from cleavage by PP2A, recruited to centromeric regions by the PP2A regulator Mei-S332/Shugoshin (shown in purple). By metaphase II, the Mei-S332/Shugoshin-PP2A complex dissociates from chromosomes. Cohesins can now be cleaved during meiosis II, allowing sister chromatids to segregate. See F. Uhlmann, 2001, Curr. Opin. Cell Biol. 13:754.
The mechanisms that remove cohesins during meiosis are the same as during mitosis. Securin degradation releases separase, which then cleaves the cohesins holding the chromosome arms together, but leaves the cohesins around centromeres in tact. This allows the recombined maternal and paternal chromosomes to separate, but each pair of chromatids remains associated at the centromere. During metaphase II, sister chromatids align on the metaphase II spindle, and separase is activated yet again, cleaving the residual cohesin around centromeres, facilitating anaphase II (see Figure 4b).
Recombination and a Meiosis-Specific Cohesin Subunit Are Necessary for the Specialized Chromosome Segregation in Meiosis I
As we have seen, in metaphase of meiosis I, both sister chromatids in one (replicated) chromosome associate with micro tubules emanating from the same spindle pole, rather than from opposite poles as they do in mitosis (see Figure 19-36). Two physical links between homologous chromosomes re sist the pulling force of the spindle until anaphase: (1) the chiasmata that result from crossing over between chromatids, and (b) cohesins distal to the crossover point (see Figure 4b, top). Evidence for the linking function of recombination during meiosis comes from the observation that when recombination is blocked by mutations in proteins essential for the process, chromosomes segregate randomly during meiosis I; that is, homologous chromosomes do not necessarily segregate to opposite spindle poles.
At the onset of meiotic anaphase I, cohesins between chromosome arms are cleaved by separase. This cleavage is required for homologous chromosomes to segregate. If co hesins were not lost from chromosome arms, the recombined sister chromatids would be torn apart during anaphase I. The maintenance of centromeric cohesion during meiosis I is necessary for the proper segregation of sister chromatids during meiosis II.
Studies in many organisms have shown that a specialized cohesin subunit, Rec8, is necessary for the stepwise loss of cohesins from chromosomes during meiosis. Expressed only during meiosis, Rec8 is homologous to Scc1, the cohesin subunit that closes the cohesin ring in the cohesin complex of mitotic cells. Immunolocalization experiments revealed that during early anaphase of meiosis I, Rec8 is lost from chromosome arms but is retained at centro meres. However, during early anaphase of meiosis II, centromeric Rec8 is cleaved by separase, so the sister chromatids can segregate, as they do in mitosis (see Figure 4b, bottom). Consequently, understanding the regulation of Rec8-cohesin complex cleavage is central to understanding chromosome segregation in meiosis I.
The mechanism that protects Rec8 from cleavage at cen tromeres during meiosis I is similar to the mechanism that protects Scc1 at centromeres during mitosis. Recall that during mitotic prophase, protein kinases, chief among them Polo kinase, phosphorylate cohesins in the chromatid arms, causing them to dissociate and eliminating cohesion on chromatid arms by metaphase. However, cohesion at the centromeres is maintained because a specific isoform of protein phosphatase 2A (PP2A) is localized to centromeric chroma tin by members of a family of proteins known as Mei-S332/ Shugoshin. PP2A keeps cohesin in a hypophosphorylated state that does not dissociate from chromatin until meiosis II (see Figure 19-37a). During metaphase II, Mei-S332/Shu goshin dissociates from chromosomes. In addition, when the last kinetochore is properly associated with spindle microtu bules, APC/CCdc20 is derepressed, causing ubiquitinylation of securin. The inactivation of securin releases separase activity, which cleaves Scc1 whether it is phosphorylated or not, eliminating cohesion at the centromere and allowing chromatid separation in anaphase (see Figure 4a).
Cohesin removal differs for meiosis I because when Rec8 replaces Scc1 in the cohesin complex, the complex does not dissociate in prophase when it is phosphorylated. The meiotic cohesin complex can be removed from chromatin only by the action of separase. Rec8 also differs from Scc1 in that it must be phosphorylated by several protein kinases to be cleaved by separase. During meiosis I, the centromere specific isoform of PP2A targeted to centromeric chromatin by Mei-S332/Shugoshin prevents this phosphorylation. The PP2A targeting factor and PP2A then dissociate from chromosomes by metaphase II, allowing separase to cleave Rec8.
Meiosis I is much more error-prone than mitosis. It is estimated that 10 percent of all conceptions in hu mans are aneuploid. These aneuploidies are largely caused by chromosome mis-segregation in meiosis I, also known as chromosome nondisjunction. When recombination fails to take place between homologous chromosomes, or when the chiasmata are too close to the end of the chromosomes, either no tension or too little tension is exerted on kinetochore–microtubule attachments, and they are destabilized, leading to nondisjunction. As a result, eggs or sperm receive too few or too many chromosomes. All monosomies (lack of a chromosome) and most trisomies (gain of a chromosome) result in embryonic lethality or death shortly after birth. Only trisomy 21, also known as Down syndrome, results in viable offspring, but it causes developmental abnormalities and intellectual disabilities.
Nondisjunction in meiosis I increases dramatically with maternal age. Whereas fewer than 0.1 percent of babies born to women under the age of 30 have Down syndrome, this rate increases to 3.5 percent at age 45. This increase in chromosome nondisjunction with maternal age is not specific to chromosome 21. The incidence of trisomy in clinically recognized pregnancies is less than 5 percent in women under 30, but as high as 35 percent in women over the age of 42 (Figure 5)! The reason for this dramatic increase in nondisjunction with maternal age lies in the biology of female meiosis in vertebrates. In all vertebrates, pre-meiotic DNA replication and recombination occur in the female embryo. The oocytes then become arrested in G2 of meiosis I until the female reaches sexual maturity, which in hu mans is between the ages of 12 and 16. It is at this time that the first oocyte enters meiosis and progresses to the second meiotic metaphase, where it arrests and awaits fertilization. The oocytes that enter the meiotic divisions in a 40-year-old woman have been arrested in G2 for 40 years, during which time chiasmata close to the ends of chromosomes can slip off, or cohesins that hold the homologous chromosomes together during this very long G2 phase can deteriorate, causing the homologous chromosomes to dissociate from each other. Both events can cause the homologous chromosomes to mis-segregate, leading to the formation of aneuploid eggs.
Fig5. Fetal aneuploidy increases with maternal age. The percentage of trisomic embryos identified among clinically recognized pregnancies is shown as a function of maternal age. [Data from T. Hassold and P. Hunt, 2001, Nature Rev. Genet. 2:280.]
Co-orienting Sister Kinetochores Is Critical for Meiosis I Chromosome Segregation
As discussed earlier, at metaphase in mitosis and meiosis II, sister kinetochores attach to spindle microtubules emanating from opposite spindle poles; the kinetochores are said to be bi-oriented. This is essential for segregation of sister chromatids to different daughter cells. In contrast, at meiosis I metaphase, sister kinetochores attach to spindle micro tubules emanating from the same spindle pole; these sister kinetochores are said to be co-oriented (see Figure 3). Obviously, attachment of sister kinetochores to the proper microtubules in meiosis I and II is critical for correct meiotic segregation of chromosomes.
Proteins required for meiosis I sister kinetochore co- orientation were first identified in S. cerevisiae. In this organ ism, a single microtubule attaches to each kinetochore. We now know that a protein complex known as the monopolin complex associates with sister kinetochores during meiosis I and fuses them into a single kinetochore unit, to which one microtubule attaches. In all other organisms, kinetochores attach to multiple microtubules. In these organisms, Rec8 containing cohesins are essential for sister kinetochore co orientation. These meiosis-specific cohesins impose a rigid kinetochore structure, restricting the movement of sister ki netochores and thereby favoring their attachment to micro tubules from the same spindle pole.
Correct attachment of meiosis I chromosomes is mediated by a tension-based mechanism, as it is during mitosis and meiosis II. During meiotic metaphase I, kinetochore associated microtubules are under tension (even though the co-oriented kinetochores of sister chromatids attach to microtubules coming from the same spindle pole) because chiasmata generated by recombination between homologous chromosomes and the cohesins distal to the chiasmata pre vent them from being pulled to the poles (see Figure 3). Since kinetochore-microtubule attachments are unstable in the absence of tension (due to Aurora B–mediated phosphorylation), kinetochores that attach to microtubules from the wrong spindle release those microtubules, which enables them to bind microtubules again until attachments are made that generate tension. As in mitosis, once tension is generated, microtubule attachment to the kinetochores is stabilized.
DNA Replication Is Inhibited Between the Two Meiotic Divisions
The mechanism by which DNA replication is suppressed be tween meiosis I and II is currently an active area of investigation, but it is thought that a change in the regulation of CDK activity is at least in part responsible for this suppression. The same S phase CDKs that promote DNA replication prior to mitosis are needed for pre-meiotic DNA replication. And the same mitotic CDKs that promote mitosis also pro mote the meiotic divisions, except that in this case we call mitotic CDKs meiotic CDKs since they promote the meiotic divisions and not mitosis.
So how is DNA replication prevented between the two meiotic divisions? Following anaphase of meiosis I, meiotic CDK activity does not fall as low as it does following mitotic anaphase. This partial drop in CDK activity is thought to be sufficient to promote the disassembly of the meiosis I spindle, but insufficient to promote MCM helicase loading (re call that a state of very low or no CDK activity is needed to load MCM helicases). During prophase of meiosis II, meiotic CDK activity rises again, and the meiosis II spindle forms. After all sister kinetochores have attached to microtubules from opposite spindle poles, separase is activated and cells proceed through meiosis II anaphase, telophase, and cytokinesis to generate haploid germ cells.
