Meiosis Versus Mitosis
Why do siblings look different from each other, even though they share the same two parents? The answer lies in a single specialized cell division program. This program runs only in the germline. Every somatic cell in your body divides by mitosis. Mitosis faithfully copies and distributes an identical genome to two daughters. But the cells that make sperm and eggs use a different program. This program is meiosis. Meiosis halves the chromosome number and reshuffles the genetic material before doing so. Meiosis solves two problems simultaneously. It must reduce a diploid genome to a haploid one. This ensures fertilization restores the correct chromosome number generation after generation. It must also generate genetic variation so that offspring are not clones of their parents. It does this with an elegant trick. There are two rounds of chromosome segregation following only one round of DNA replication. This is combined with a deliberate, enzymatically orchestrated exchange of DNA between homologous chromosomes. This exchange is called crossing over. The result is gametes that are genetically unique. This is a requirement for sexual reproduction. It is also a major engine of evolutionary variation. This article compares meiosis and mitosis side by side. It traces exactly where their molecular choreography diverges. Both processes rely on the same core segregation machinery. This includes spindle microtubules, kinetochores, cohesin, separase, and the anaphase-promoting complex. These were introduced in the context of mitosis. Meiosis repurposes and modifies every one of these components. Cohesin is protected in a stage-specific manner. Kinetochores are reoriented to attach sister chromatids to the same pole. An entirely new structure, the synaptonemal complex, physically zips homologous chromosomes together. By the end you should be able to explain this at the molecular level. You should understand why meiotic nondisjunction is common enough to cause a large fraction of human miscarriages. You should also understand its role in conditions like Down syndrome. You should be able to explain why every gamete you make is genetically distinct from every other. You should also understand how classic experiments proved this. These range from Boveri and Sutton's chromosome theory to modern spore-tetrad genetics. They proved that chromosomes, not some vaguer hereditary fluid, carry the units of Mendelian inheritance.
- Mitosis: one S phase, one division, two genetically identical diploid daughters — preserves the genome.
- Meiosis: one S phase, two divisions (meiosis I and II), four genetically distinct haploid daughters — reduces chromosome number and reshuffles alleles.
- Meiosis I is reductional (homologs separate); meiosis II is equational (sister chromatids separate), like a modified mitosis.
- Homologous chromosomes pair, synapse via the synaptonemal complex, and exchange DNA by crossing over, generating chiasmata that hold homologs together until anaphase I.
- Cohesin is removed in two stages: arm cohesin at anaphase I, centromeric cohesin at anaphase II — protected in between by shugoshin-PP2A.
- Independent assortment of homologs plus crossing over produces vastly more allele combinations than mitosis ever could; errors in this process cause aneuploidy.
Meiosis achieves genome halving and genetic shuffling by running two chromosome segregations — first homologs, then sister chromatids — after a single round of DNA replication, using recombination-generated chiasmata and stage-specific cohesin protection to control the order.
The core idea
Mitosis and meiosis both begin from the same starting material — a diploid cell with replicated chromosomes, each consisting of two sister chromatids joined by cohesin — but they end in profoundly different places. Mitosis produces two daughter cells, each diploid (2n) and genetically identical to the parent cell and to each other, barring rare replication errors or new mutations. Meiosis produces four daughter cells, each haploid (n), and each one is genetically unique both relative to the parent cell and to its three sibling products. The key conceptual distinction is reductional versus equational division. A division is equational when it separates sister chromatids — genetically identical copies — as in ordinary mitosis and in meiosis II. A division is reductional when it separates homologous chromosomes — the maternally and paternally inherited copies of each chromosome, which carry different alleles — as in meiosis I. Because meiosis I is reductional, chromosome number is halved after just one division; meiosis II then behaves mechanically like a mitosis, splitting sister chromatids in each of the two haploid cells produced by meiosis I. One S phase followed by two M phases (meiosis I and meiosis II) is the defining arithmetic of meiosis: 4n of DNA content is halved to 2n after meiosis I and to n after meiosis II, while chromosome number goes from 2n (46 in humans) to n (23). The second defining feature of meiosis is recombination. During prophase of meiosis I, homologous chromosomes find each other, align side by side (synapsis), and exchange DNA segments through programmed double-strand breaks repaired by homologous recombination. This crossing over physically links homologs at sites called chiasmata, which are essential not just for shuffling alleles but for holding homologs together so that the spindle can exert tension across them, ensuring accurate segregation. Combined with the independent, random orientation of each homolog pair on the meiosis I spindle — independent assortment — recombination guarantees that essentially every gamete produced by an organism carries a unique combination of maternal and paternal alleles.
Mechanism
Setting the stage — premeiotic S phase. Like mitosis, meiosis is preceded by a single round of DNA replication, converting each chromosome from one chromatid to two sister chromatids joined along their length by cohesin complexes (Scc1/Rec8-containing rings in meiosis). Critically, in meiotic S phase the kleisin subunit of cohesin is often the meiosis-specific paralog Rec8 rather than mitotic Scc1. This substitution matters enormously downstream, because Rec8 cohesin — unlike Scc1 — can be selectively protected from cleavage at the centromere during anaphase I, a molecular switch that has no counterpart in mitosis.
Prophase I — the longest, most elaborate stage. Mitotic prophase is brief; meiotic prophase I can last hours to days (in human oocytes, it is arrested for years to decades). Prophase I is conventionally divided into five substages:
- Leptotene: chromosomes begin condensing; programmed double-strand breaks are introduced by the meiosis-specific topoisomerase-like enzyme Spo11.
- Zygotene: homologous chromosomes begin to pair and synapse — the synaptonemal complex (SC), a proteinaceous zipper-like structure, assembles between them.
- Pachytene: synapsis is complete along the full chromosome length; recombination proceeds, converting a subset of Spo11-induced breaks into crossovers, resolved as reciprocal exchanges between non-sister chromatids.
- Diplotene: the synaptonemal complex disassembles, but homologs remain physically connected at chiasmata, the cytological manifestation of crossovers. In human oocytes, diplotene arrest (dictyate stage) can last from before birth until ovulation, decades later.
- Diakinesis: chromosomes condense further, chiasmata terminalize toward chromosome ends, and the nuclear envelope breaks down, transitioning into metaphase I.
Metaphase I — homolog pairs, not chromosomes, line up. Unlike mitotic metaphase, where individual chromosomes align with sister kinetochores facing opposite poles, in meiosis I each bivalent (a pair of homologs held by chiasmata) aligns at the metaphase plate. Crucially, the two sister kinetochores of each homolog are mono-oriented — captured by microtubules from the same pole — a phenomenon requiring the meiosis-specific protein complex monopolin (well characterized in budding yeast) and modified kinetochore geometry. This ensures that at anaphase I, whole homologs (each still consisting of two sister chromatids) move to opposite poles, not individual chromatids.
Anaphase I — reductional segregation. Separase cleaves cohesin, but only along the chromosome arms, not at the centromere. Centromeric Rec8 cohesin is protected by a shugoshin protein (SGO1/SGO2) that recruits the phosphatase PP2A, which keeps the cohesin ring dephosphorylated and resistant to separase cleavage. Arm cohesin loss releases the chiasmata, allowing homologs to separate, while centromeric cohesin holds sister chromatids together. This is the single most important molecular distinction between meiosis I and mitotic anaphase: mitosis cleaves cohesin everywhere at once, meiosis I cleaves it only on the arms.
Meiosis II — an equational division without an intervening S phase. After a brief interkinesis (no DNA replication), the two haploid cells each undergo a division mechanically resembling mitosis: sister kinetochores biorient toward opposite poles, and at anaphase II, shugoshin/PP2A protection is removed, separase cleaves the remaining centromeric cohesin, and sister chromatids finally separate. Four haploid cells result, each with unreplicated, single-chromatid chromosomes.
Cohesin protection — the switch that makes reductional division possible. This deserves emphasis because it is the crux of how meiosis diverges from mitosis. In mitosis, separase (activated once the spindle checkpoint is satisfied and the anaphase-promoting complex, APC/C, degrades securin) cleaves cohesin uniformly along the whole chromosome at once, so sisters separate together everywhere. In meiosis I, shugoshin-PP2A creates a centromere-specific sanctuary that keeps Rec8 intact there while arm cohesin is destroyed — so only homologs separate. Shugoshin itself is removed from centromeres between meiosis I and II (a process involving the kinase Aurora B/C and dephosphorylation events), unmasking centromeric cohesin so that meiosis II can finally cleave it and separate sisters. If shugoshin fails to be removed, sister chromatids never separate; if it is lost too early (as it is in aging human oocytes, correlating with increasing maternal age and rising trisomy rates), sisters can separate prematurely, causing errors.
The evidence
Boveri and Sutton's chromosome theory of heredity (1902–1903). Theodor Boveri, working with sea urchin embryos, and Walter Sutton, working with grasshopper spermatocytes, independently connected the cytologically observed behavior of chromosomes during meiosis to the abstract "factors" Mendel had proposed decades earlier. Sutton noted that homologous chromosome pairs separate from each other during meiosis I in a manner that mirrors the segregation of allele pairs in Mendelian crosses, and that different pairs assort independently — exactly matching Mendel's second law. Boveri showed experimentally, by manipulating sea urchin eggs to produce abnormal chromosome complements (using multiple sperm to fertilize a single egg, producing embryos with random, unbalanced chromosome sets), that normal development required a full, balanced set of chromosomes — different chromosomes were not interchangeable but carried distinct, necessary hereditary information. Together their arguments constitute the chromosome theory of heredity: chromosomes are the physical carriers of Mendelian genes, and their meiotic behavior explains segregation and independent assortment. This was confirmed genetically soon after by Thomas Hunt Morgan's Drosophila work and, most directly, by Calvin Bridges (1916), who showed that rare nondisjunction of X chromosomes in Drosophila produced exceptional offspring whose phenotypes and, crucially, whose chromosome complements (visible cytologically) matched precisely what nondisjunction would predict — an unambiguous physical link between a genetic phenomenon (exceptional segregation ratios) and a cytological one (an extra or missing chromosome).
Tetrad analysis in fungi — direct proof of recombination and segregation mechanics. Organisms like Neurospora crassa and Saccharomyces cerevisiae package all four products of a single meiosis into an ordered ascus (in Neurospora) or unordered tetrad (in yeast). This allows a researcher to examine the actual products of one meiotic event rather than inferring ratios statistically across a population. Ordered tetrad analysis in Neurospora (pioneered by Carl Lindegren and others in the 1930s–40s) directly demonstrated first-division segregation (an allele pair separates at meiosis I, no crossover between the gene and its centromere) versus second-division segregation (a crossover between the gene and centromere causes the allele pair to separate only at meiosis II) — visible as different spore-order patterns in the ascus. This was direct cytological/genetic proof that crossing over physically occurs between homologous chromatids during prophase I and that its frequency can be used to map gene-to-centromere distances, converting recombination frequency into a real physical and genetic map.
Creighton and McClintock (1931) — cytological proof that crossing over involves physical chromosome exchange. Harriet Creighton and Barbara McClintock studied maize plants heterozygous for both a genetic marker (colored versus colorless kernels, a pigmentation gene) and a visibly distinguishable cytological chromosome marker (an unusual knob and a translocated segment on one homolog of chromosome 9, making the two homologs distinguishable under the microscope). They showed that whenever the genetic markers recombined (new allele combinations appeared in offspring), the cytological markers on the chromosome also recombined in exact correspondence. This was the first direct proof — combining cytology and genetics in the same experiment — that genetic recombination corresponds to physical exchange of chromosome segments during meiosis, not some more abstract informational process. A similar demonstration was published independently the same year by Curt Stern working in Drosophila.
Modern cytology confirms the cohesin/shugoshin model. Live-cell imaging of cohesin subunits (fluorescently tagged Rec8 and Scc1) in yeast and mouse oocytes, alongside experiments depleting shugoshin (SGO1/SGO2) by RNAi or knockout, directly showed that loss of shugoshin causes premature separation of sister chromatids at anaphase I instead of anaphase II — confirming that centromeric cohesin protection, not some difference in the segregation machinery itself, is what converts an equational division into a reductional one.
Worked examples
Example 1 — a dihybrid cross reasoned through chromosome mechanics. Consider a pea plant heterozygous for seed shape (Rr, round dominant over wrinkled) and seed color (Yy, yellow dominant over green), where the two genes lie on different chromosome pairs (unlinked). During meiosis I, the R/r homolog pair assorts independently of the Y/y homolog pair, because each bivalent orients randomly on the metaphase I spindle with respect to the others. This produces four gamete classes in equal proportion: RY, Ry, rY, ry, each at frequency 1/4. Selfing this plant (RrYy × RrYy) gives the classic 9:3:3:1 phenotypic ratio (9 round-yellow : 3 round-green : 3 wrinkled-yellow : 1 wrinkled-green) among 16 offspring classes — a ratio that only makes mechanistic sense once you know that independent assortment of homolog pairs at metaphase I, not some blending process, generates the four equally frequent gamete classes. Now suppose the two genes are actually on the same chromosome, 20 map units (cM) apart. Instead of 1/4 each, you would predict mostly parental-type gametes (RY and ry) and a minority of recombinant gametes (Ry and rY) totaling 20%, i.e., 10% each, with 40% each of the parental classes — a direct read-out of the frequency of crossing over between the two loci during pachytene of meiosis I.
Example 2 — calculating nondisjunction risk and its clinical consequence. In human oogenesis, homologous chromosome pairs are held together by chiasmata formed during fetal life, but the cell arrests in prophase I (dictyate stage) for years before completing meiosis I at ovulation, sometimes decades later. Because centromeric cohesin (Rec8) is not renewed during this arrest and gradually degrades over the years, chiasmata can become too weak to resist the pulling force of the meiosis I spindle as maternal age increases — a leading explanation for the well-documented rise in trisomy risk with maternal age. Epidemiological data show maternal-age-related trisomy 21 (Down syndrome) risk rising from roughly 1 in 1,250 live births at maternal age 25 to roughly 1 in 100 at age 40, and further to about 1 in 30 by age 45. Mechanistically: if a chiasma near the telomere is lost or if cohesin has degraded too much to resist spindle tension, the two homologs may fail to segregate at anaphase I (nondisjunction), producing one gamete with two copies of chromosome 21 and one with none. Fertilization of the disomic gamete by a normal haploid sperm yields a trisomy-21 zygote. This example shows how a purely mechanical/biochemical feature of meiosis I — the duration of cohesin-dependent chiasma maintenance — has a direct, quantifiable clinical consequence.
Common misconceptions & traps
| Misconception | Correction |
|---|---|
| "Meiosis is just mitosis done twice." | Meiosis I is reductional (homologs separate); only meiosis II resembles mitosis (sisters separate). The mechanisms of chromosome pairing, recombination, and cohesin protection in meiosis I have no mitotic counterpart. |
| "Crossing over happens in mitosis too, just less often." | Programmed meiotic recombination initiated by Spo11-generated double-strand breaks is a meiosis-specific pathway; mitotic cells actively suppress homologous recombination between homologs and rely on sister chromatids as repair templates instead. |
| "Independent assortment and crossing over are the same source of variation." | They are two distinct mechanisms: independent assortment shuffles whole intact homologs between gametes; crossing over reshuffles alleles within a single chromosome by physical exchange between non-sister chromatids. |
| "All four products of meiosis are identical because they came from the same cell." | The whole point of meiosis is that recombination and independent assortment make all four products genetically distinct from one another (except for rare cases of no crossover and identical parental alleles). |
| "Down syndrome is caused by a mutation." | It is caused by a chromosome segregation error (nondisjunction) during meiosis, producing an extra whole chromosome — a numerical/karyotypic abnormality, not a sequence-level mutation. |
The scientist behind it
Origins · 1 minBarbara McClintock and how this was discovered
McClintock was a cytogeneticist whose meticulous chromosome-squash microscopy of maize let her see recombination happening, not just infer it statistically. Working with Harriet Creighton in 1931, she engineered maize strains in which one homolog of chromosome 9 carried both an unusual visible knob and a translocated chromosome arm — physical landmarks she could track under a light microscope — linked to Mendelian markers for kernel color and texture. By showing that genetic recombination of the visible traits always coincided with physical recombination of the chromosomal landmarks, she supplied the direct proof that crossing over is a real, physical exchange of chromosome material. She later discovered transposable elements in maize, work for which she was awarded the Nobel Prize in Physiology or Medicine in 1983, decades after she first proposed it, reflecting how far ahead of contemporary genetic thinking her cytogenetic evidence-based reasoning was.
Where this matters today
Understanding meiotic mechanics underlies prenatal genetic counseling and screening: noninvasive prenatal testing (NIPT) for trisomies 21, 18, and 13 relies on detecting the chromosomal signatures of meiotic nondisjunction in cell-free fetal DNA, and risk estimates given to patients are built directly from the age-dependent cohesin-degradation model described above. In assisted reproduction, understanding why oocyte aneuploidy rises with age informs decisions about egg freezing, preimplantation genetic testing for aneuploidy (PGT-A) in IVF embryos, and donor egg counseling. In agriculture and plant breeding, controlled recombination frequency (manipulated via mutants in genes controlling crossover number and placement) is used to break undesirable linkage between traits or to fix desirable combinations more quickly, and haploid-inducer lines that skip normal meiotic fertility are used to accelerate doubled-haploid breeding programs in maize and wheat. In cancer biology, meiosis-specific genes (the "cancer-testis antigen" family, including several genes normally restricted to germline meiotic cells) become aberrantly reactivated in some tumors, making them attractive, tumor-specific immunotherapy targets because they are essentially absent from normal somatic tissue. In conservation genetics, understanding recombination and independent assortment underlies estimates of effective population size and inbreeding risk in small, endangered populations, since reduced genetic shuffling correlates with reduced adaptive potential. Finally, meiosis is the conceptual bridge to Stem Cells and Cellular Potency: primordial germ cells must be specified and maintained in a state competent to later enter meiosis, distinguishing the germline's unique developmental trajectory from all somatic lineages.
Ask the tutor: Why does mono-orientation of sister kinetochores in meiosis I require a special protein complex, and what would happen to chromosome segregation if that complex were missing?
Why does meiosis have two divisions but only one round of DNA replication?
Because the biological goal is to halve chromosome number exactly once per generation while still separating both homologs and sister chromatids cleanly. A single round of replication produces chromosomes with two sister chromatids each; meiosis I then separates homologs (reductional), so each resulting cell is already haploid but each chromosome still has two chromatids. Meiosis II separates those chromatids (equational) without further replication, since replicating again would restore diploidy. This two-step process, built on one replication event, is precisely what yields four haploid cells instead of two diploid ones, and it depends entirely on the stepwise, two-stage removal of cohesin (arms first, centromeres second) described in the mechanism section.
Is crossing over required for meiosis to work, or is it just a bonus source of variation?
It is largely required, not optional. Chiasmata — the physical remnants of crossovers — are what hold homologous chromosomes together from diplotene through metaphase I, allowing the spindle to generate the tension needed to satisfy the spindle assembly checkpoint and orient homologs correctly. Chromosomes that fail to recombine (achiasmate) are at much higher risk of nondisjunction because nothing resists the pulling force of the spindle before anaphase I. Some organisms have evolved backup mechanisms (like achiasmate segregation systems in Drosophila males), but in humans and most sexually reproducing species, at least one crossover per chromosome pair is essentially obligatory for accurate segregation.
Why are gametes genetically different from each other if they came from the same original diploid cell?
Two independent shuffling mechanisms act during meiosis. First, independent assortment randomly distributes each pair of homologs to the two poles at meiosis I, independently of every other pair — with 23 human chromosome pairs, this alone yields 2²³, about 8.4 million, possible combinations. Second, crossing over during prophase I exchanges segments between non-sister chromatids of homologous chromosomes, so even chromosomes that assort together are not identical to the original parental chromosomes. Together these mechanisms make the probability that two gametes (barring identical twins from mitotic division of one embryo) carry the exact same genome vanishingly small.
Does meiosis only happen in animals?
No. Meiosis is a defining feature of essentially all sexually reproducing eukaryotes, including plants, fungi, protists, and animals, though the timing relative to the life cycle (its position relative to fertilization and mitotic growth phases) varies enormously. In animals, meiosis directly produces gametes (a diplontic life cycle). In many plants and fungi, meiosis instead produces haploid spores that undergo mitotic divisions to build a multicellular haploid generation before eventually producing gametes (an alternation of generations). The core molecular mechanism — homolog pairing, recombination, reductional then equational division — is deeply conserved across all these lineages, evidence of its ancient evolutionary origin alongside the origin of sexual reproduction itself.
How does the spindle checkpoint operate differently in meiosis I compared to mitosis?
The core surveillance machinery — the spindle assembly checkpoint (SAC), involving Mad and Bub proteins monitoring kinetochore-microtubule attachment — is conserved from mitosis into meiosis. But what counts as a "correct" attachment differs: in mitosis, correct means each sister kinetochore attached to opposite poles (biorientation); in meiosis I, correct means the two sister kinetochores of a homolog are co-oriented to the same pole (mono-orientation, enforced by monopolin-like complexes), while the two homologs' kinetochore pairs attach to opposite poles. The SAC in meiosis I is also generally considered less stringent than in mitosis, part of why meiotic nondisjunction, especially in human oocytes, is comparatively common relative to mitotic segregation errors.
What happens if DNA replication occurred twice before a single meiotic division, or if meiosis II were skipped?
Neither occurs naturally in standard sexual meiosis, but both illuminate why the system is built the way it is. Skipping meiosis II (which does happen in some parthenogenetic organisms via "endomitosis" variants) would leave gametes with unseparated sister chromatids, effectively keeping them diploid despite meiosis I having occurred — this is actually one mechanism some organisms use for automictic parthenogenesis, generating diploid offspring from unfertilized eggs. A double replication before a single reductional division, by contrast, is not known to occur naturally in standard meiosis; it would produce grossly aneuploid, non-viable gametes, underscoring why the strict coupling of exactly one replication event to exactly two carefully ordered divisions is so tightly enforced by checkpoint and cohesin-based controls.
Why is it clinically important to distinguish meiotic errors from mitotic errors in a developing embryo?
Because they produce different patterns of mosaicism and carry different risks. A nondisjunction event during meiosis I or II in the parent's germline produces a constitutional aneuploidy present in every cell of the resulting embryo (e.g., full trisomy 21). A mitotic error occurring after fertilization, in contrast, produces mosaicism — only some cells of the embryo carry the abnormal chromosome number, and outcomes can range from clinically silent to severe depending on the proportion and location of affected cells. Preimplantation genetic testing (PGT-A) in IVF must account for this distinction, since a biopsy showing aneuploid cells from mosaic mitotic errors does not necessarily mean the whole embryo, or even most of it, is abnormal — a very different clinical implication from a uniform meiotic-origin aneuploidy.