The series of visible changes that occur in the nucleus and chromosomes of non-gamete-producing plant and animal cells as they divide. During mitosis, the replicated genes, packaged within the nucleus as chromosomes, are precisely distributed into two genetically identical daughter nuclei (Fig. 1). The series of events that prepares the cell for mitosis is known as the cell cycle. When viewed in the context of the cell cycle, the definition of mitosis is often expanded to include cytokinesis, the process by which the cell cytoplasm is partitioned during cell division. Although the continuity of heredity by cell division was first predicted by Rudolf Virchow in 1858, the central role of mitosis was not fully understood until the mid-1900s, when the deoxyribonucleic acid (DNA) within chromosomes was proven to contain the hereditary blueprints for life. See also: Cell (biology); Cell cycle; Cell division; Cell nucleus; Chromosome; Deoxyribonucleic acid (DNA); Gene
Chromosome segregation is mediated in all nonbacterial cells (that is, eukaryotes) by the transient formation of a complex structure known as the mitotic spindle (Fig. 2). During mitosis in most higher plants and animals, the nuclear membrane surrounding the replicated chromosomes breaks down, and the spindle is formed in the region previously occupied by the nucleus (open mitosis). In lower organisms, including some protozoa and fungi, the spindle is formed and functions entirely within the nucleus, which remains intact throughout the process (closed mitosis). See also: Mitosis and the spindle assembly checkpoint
All spindles are bipolar structures, having two ends or poles. In animal cells, each spindle pole contains an organelle, the centrosome (Fig. 3), onto which the spindle focuses. As a result, the spindle in animal cells looks like an American football. The polar regions of plant spindles lack centrosomes and, as a result, are much broader. In animals, the bipolar nature of the spindle is established by the separation of the centrosomes, which is critical for a successful mitosis; the presence of only one pole produces a monopolar spindle in which chromosome segregation is inhibited. The presence of more than two poles produces multipolar spindles that distribute the chromosomes unequally among three or more nuclei. Centrosomes are duplicated during interphase near the time that the DNA is replicated, but then act as a single functional unit until the onset of mitosis. In plants, and during meiosis in some animals, the two spindle poles are organized by the chromosomes and by molecular motors that order randomly nucleated microtubules into parallel bundles. See also: Centrosome; Meiosis; Plant cell
Microtubules are the primary structural components of the mitotic spindle and are required for chromosome motion (Fig. 3). These 25-nanometer-diameter, hollow, tubelike structures are formed from the polymerization of protein subunit dimers composed of alpha and beta tubulin. Microtubules are polarized structures that grow much faster at one end than at the other; the fast-growing end is referred to as the positive (+) end, whereas the slow-growing end is referred to as the negative (−) end. During interphase, microtubules are distributed throughout the cytoplasm, where they serve to maintain cell shape and also function as polarized roadways for transporting organelles and cell products. As the cell enters mitosis, the cytoplasmic microtubule network is disassembled and replaced by the mitotic spindle. The microtubules in animal cells originate from the centrosome that, like the chromosomes, was inherited during the previous mitosis, where it functioned as a spindle pole. The centrosome contains a unique type of microtubule protein (gamma tubulin) that is involved in seeding microtubule assembly. Once growth is initiated, the positive ends of centrosomal microtubules grow away from the centrosome, whereas the negative ends either remain associated with the centrosome or are shed from it. At any one time, many microtubules within the cell are growing or shrinking as subunits are added or removed from their positive ends. Changes in the parameters that modulate this dynamically unstable behavior of microtubule ends lead to changes in the numbers and average length of microtubules.
The motion associated with microtubules is mediated by several families of molecular motors that bind to and move along the wall of the microtubule. These motors include kinesins that move cargo along microtubules toward their positive ends away from the centrosome, and cytoplasmic dynein (and also some members of the kinesin family) that transports cargo along individual microtubules toward their centrosomal or negative ends. See also: Cytoskeleton
As mitosis begins, each replicated chromosome consists of two identical sister chromatids that are joined along their length. In most cells, chromosomes possess a unique region of highly condensed chromatin (DNA plus protein), known as the centromere, which forms an obvious constriction on the chromosome, referred to as the primary constriction. Spindle microtubules attach to a small specialized structure on the surface of the centromere known as the kinetochore (Fig. 3). Fragments of chromosomes lacking a kinetochore do not move poleward; it is always the kinetochore that leads in the poleward motion of the chromosome. Kinetochores in most plant cells are shaped like a ball, whereas those in animal cells resemble a plate. The centromere region of each replicated chromosome contains two sister kinetochores (one attached to each chromatid) that lie on opposite sides of the primary constriction. See also: Sister chromatid cohesion
Regulation of mitosis
Cells that are in the cell cycle contain an inactive kinase, known as p34cdc2. Once chromosome and centrosome replication is completed at the end of S phase, two cyclin proteins (A and B) are synthesized that rapidly bind to and activate p34cdc2. This activated enzyme is called cyclin-dependent kinase 1 (CDK1); it was formerly known as mitosis or maturation promoting factor (MPF). CDK1 then drives the cell into mitosis by phosphorylating specific proteins involved in the interphase/mitosis transition. In vertebrates, cyclin A is synthesized before cyclin B; then, once active CDK1/cyclin A accumulates in the nucleus, it initiates the early events of mitosis, including chromosome condensation. The subsequent accumulation of active CDK1/cyclin B then commits the cell to mitosis, and the cell will stay in mitosis as long as CDK1/cyclin B activity remains high.
Stages of mitosis
Once initiated, mitosis is a continuous process that, depending on the temperature and organism, requires several minutes (Drosophila and yeast) to many hours (newts) to complete. Traditionally, it has been subdivided into five consecutive stages that are distinguished primarily by chromosome structure, position, and behavior. These stages are prophase, prometaphase, metaphase, anaphase, and telophase (Fig. 4).
The first visible sign of an impending mitosis occurs within the nucleus as the DNA-containing chromatin begins to condense into chromosomes. This condensation is the result of the phosphorylation of chromatin-associated proteins, including histone H1 and H3, by CDK1/cyclin A and the aurora B kinase within the nucleus. Up to a point, this is a reversible process, and cells in early to mid-prophase can be induced to decondense their chromosomes and return to interphase by a variety of treatments that stress the cell and/or damage the DNA. However, once active CDK1/cyclin B accumulates in the nucleus at late prophase, the cell becomes irreversibly committed to the division process. See also: Histone
By late prophase, the individual chromosomes are well defined and the nucleoli begin to dissipate. Near this time, the cytoplasmic network of microtubules breaks down and is replaced in animal cells by two radial astral arrays of microtubules growing from the two spindle poles (centrosomes). When compared to interphase, the microtubules associated with mitotic centrosomes contain more, but significantly shorter, microtubules. The functional changes that occur in the centrosome as the cell enters mitosis are mediated by the activity of CDK1 and other kinases that modify both centrosome chemistry and the proteins that control microtubule dynamic instability.
As the asters form, they separate and move toward opposite sides of the nucleus. During this separation, the microtubule arrays from each aster exhibit considerable overlap; within this overlap, the microtubules from opposing asters are of opposite polarity. In some lower organisms, including yeast and diatoms, the spindle poles are pushed apart by the action of microtubule plus-end motors that bind microtubules of opposite polarity within the region of overlap, and then slide them away from one another. However, in higher animal cells, including vertebrates, the asters continue to move apart, even when their microtubule arrays no longer overlap. Thus, it is likely that the separation of spindle poles in higher animals occurs via multiple mechanisms, including one that involves pulling.
The sudden activation of CDK1/cyclin B during late prophase results in the rapid phosphorylation of various nuclear and cytoplasmic proteins. The most obvious consequence of this activity is that the nuclear envelope surrounding the chromosomes suddenly swells and dissolves, as many of its constituent proteins become phosphorylated and soluble. At the same time, CDK1 activity also induces other membrane systems within the cell, including the Golgi apparatus and the endoplasmic reticulum, to fragment into numerous small vesicles that then become randomly distributed. See also: Endoplasmic reticulum; Golgi apparatus
The breakdown of the nuclear envelope initiates the prometaphase stage of mitosis. During this stage, the chromosomes attach to the separating centrosomes as their kinetochores capture dynamically unstable astral microtubules that are randomly probing the cytoplasm. Once an astral microtubule is captured by a kinetochore, it becomes known as a kinetochore microtubule, and its positive end becomes firmly anchored within the kinetochore. A kinetochore begins to move as soon as it encounters its first microtubule. However, as a rule, kinetochores continue to capture as many astral microtubules as their surface area allows and, throughout the capture process, they induce these microtubules to bundle into a kinetochore fiber. In vertebrates, kinetochore fibers rarely form at the same time on sister kinetochores. Instead, the tendency is for a chromosome to attach first to the closest spindle pole as the kinetochore facing that pole acquires microtubules. Because only one of its kinetochores is attached to the spindle, such chromosomes are mono-oriented, and they rapidly move toward the pole to which they have become attached. Over time, the opposing sister kinetochore on a mono-oriented chromosome, which lacks an attachment to microtubules, must become associated with the positive ends of microtubules generated from the opposing half-spindle (that is, the far pole). This occurs by chance when it captures one or more microtubule ends growing from the far pole. To increase the chances of this happening in a reasonable time, the unattached kinetochore often becomes laterally associated with the surface of adjacent microtubules residing in the same half-spindle as the mono-oriented chromosome. Once this lateral association is established, microtubule positive-end centromere-associated protein E (CENP-E) motors on the surface of the unattached kinetochore transport it, and its associated chromosome, along the surface of the microtubules toward the spindle equator (that is, away from the closer pole). As the kinetochore approaches the equator, its chances of encountering a positive microtubule end growing from the opposing half-spindle dramatically increase. Once both sister kinetochores are attached to the spindle, so that each kinetochore is attached to different and opposing poles, the chromosome is considered to be bioriented. The failure of a chromosome to acquire this proper biorientation leads to the production of aneuploid cells with unequal numbers of chromosomes, with corresponding detrimental consequences to the organism.
Once a chromosome has achieved biorientation status, it undergoes a complex series of motions (collectively termed congression) that ultimately position its centromere on the spindle equator midway between the poles. In order for congression to occur, the kinetochore fiber microtubules that terminate in the sister kinetochores must elongate and shorten in a coordinated fashion. For example, as a bioriented chromosome moves toward the equator, those kinetochore microtubules that attach the chromosome to the proximal pole must elongate, whereas those that attach it to the distal pole must shorten. In vertebrates, the elongation and shortening of kinetochore fiber microtubules during chromosome motion occur primarily by the addition and deletion of tubulin subunits into microtubules at the kinetochore, but shortening also occurs to a lesser extent from subunit loss at the microtubule minus end associated with the pole.
Mono-oriented chromosomes exhibit constant-velocity, oscillatory motions toward and away from their associated spindle pole. The tendency of the only attached kinetochore on a mono-oriented chromosome to autonomously switch between poleward and away-from-the-pole motion means that the motility of kinetochores is directionally unstable. When the unattached kinetochore of a mono-oriented chromosome attaches to the distal pole, it begins to move toward that pole. In order for this congression motion to continue, its sister kinetochore, attached to the closer pole, must switch into a prolonged movement away from the pole state of motion. The factors that coordinate the motile behavior of sister kinetochores, that is, to allow for congression, remain unknown.
When the last chromosome becomes positioned on the spindle equator, the cell is considered to be in metaphase of mitosis (Fig. 5). By this time, many of the astral microtubules have been sequestered into the spindle, which is now a fully mature structure, and cyclin A has been degraded. The metaphase spindle consists of two half-spindles, tethered to one another by the sister kinetochores on the chromosomes and by interactions between their associated overlapping (antipolar) microtubule arrays. Each half-spindle contains the same number of kinetochore fiber microtubule bundles interspersed among more numerous microtubules growing from the pole or free within the spindle.
The sudden and largely synchronous separation (disjunction) of sister chromatids initiates the anaphase stage of mitosis. This disjunction is not dependent on forces generated by the spindle, as evidenced by the fact that it also occurs in some cells when spindle formation is inhibited by drugs. In all but embryonic cells, entry into anaphase is controlled by a cell-cycle checkpoint that delays chromatid disjunction until all of the kinetochores have achieved a stable attachment to the spindle. In these cells, unattached or weakly attached kinetochores trigger a signal transduction cascade that delays anaphase by inhibiting the activity of a macromolecular assembly known as the anaphase promoting complex (APC). These large, spindle-associated complexes selectively target several proteins that are involved in the metaphase/anaphase transition (including securin, which maintains the cohesion between the sister chromatids and cyclin B) for proteolytic destruction by ubiquitinating them. This “kinetochore attachment” or “spindle assembly” checkpoint ensures that anaphase does not normally start until all of the chromosomes have achieved a bipolar orientation, which is the minimum requirement for equal chromosome segregation. See also: Signal transduction; Ubiquitination
Once the chromatids disjoin in non-drug-treated cells, they slowly (1–2 μm/min) move toward their respective poles in a process referred to as anaphase A. During this time, the once highly organized spindle begins to disassemble. After anaphase A is under way, the poles themselves begin to move farther apart in a process known as anaphase B.
The activation of APCs at the metaphase/anaphase transition not only leads to chromatid disjunction, but it also inactivates the mitotic (CDK1) kinase by destroying its associated cyclin B. As CDK1 activity falls, the various proteins that were phosphorylated during the initial stages of mitosis become dephosphorylated, and pathways are initiated that ultimately return the dividing cell to interphase. During this telophase stage of mitosis, the groups of separated sister chromosomes, which are now positioned near their respective poles, begin to swell and stick to one another. As this occurs, a nuclear envelope is deposited on the surface of the decondensing chromatin, and a Golgi apparatus reforms in association with each of the two centrosomes.
During telophase, new microtubule-based structures, known in animals as stem bodies and in plants as phragmoplasts, also form between the now-separated daughter nuclei. The phragmoplasts are responsible for directing the construction of a new cell wall that ultimately partitions the dividing plant cell into two separate (but connected) entities. In animals, the stem bodies act as a template and catalyst for forming and stabilizing the cytokinetic furrow, which constricts the cell into two independent daughters. At the end of telophase, the centrosome inherited by each daughter cell nucleates another interphase complex of cytoplasmic microtubules. See also: Plant cell wall
In some cells, a kinetochore (and thus its associated chromosome) moves toward a spindle pole because the fiber to which it is attached is pulled toward the pole. This “traction fiber” mechanism accounts for 100% of chromosome poleward motion in some systems, including insect spermatocytes, various oocytes, and perhaps plant cells. It remains to be resolved whether the pulling force for this motion is produced along the entire length of the fiber by microtubule plus-end motors anchored in the spindle matrix, or only at its end in the spindle pole. For poleward motion to occur, however, the kinetochore microtubules must shorten as they slide poleward, and this shortening clearly occurs by subunit loss at the microtubule minus ends at the pole. In contrast to a traction fiber mechanism, only 25% of kinetochore fiber shortening during poleward chromosome motion in vertebrate somatic cells can be attributed to subunit loss at the spindle pole. The other 75% occurs by subunit loss at the microtubule plus ends within the kinetochore. Immunological studies show that kinetochores in vertebrates contain cytoplasmic dynein, CENP-E, and other microtubule-based motor molecules, and that they can move rapidly along the surface of a microtubule toward a centrosome (spindle pole) during prometaphase. It remains to be determined if the kinetochore-based force for poleward chromosome motion in these cells is produced by microtubule minus-end molecular motors associated with the kinetochore or simply from the disassembly of microtubule minus ends within the kinetochore. Regardless, in humans, the kinetochore can be viewed as a complex biological machine that uses its associated kinetochore fiber microtubules to generate the force for chromosome poleward motion, while allowing the plus ends of these microtubules to grow and shorten within its confines.
The mechanism that ultimately positions the sister kinetochores of a bioriented chromosome stably on the spindle equator, halfway between the two spindle poles, remains to be solved. For this to occur, the behavior of sister kinetochores must somehow be coordinated; thus, when one is moving toward its associated pole, the other is moving away from its pole. At the moment, the best guess for how this occurs is that kinetochores somehow "sense" their position on the spindle, and regulate their behavior accordingly.