Biology › Inheritance and population genetics › Meiosis: two divisions, and why no two gametes come out the same

Meiosis: two divisions, and why no two gametes come out the same

Mitosis makes copies. Meiosis does the opposite job on purpose: it halves the chromosome number and shuffles the deck twice on the way, so that four cells leave a division that started with one, and no two of them carry the same set of alleles.

Before this The cell cycle and mitosis · DNA structure and replication · Eukaryotic cell ultrastructure

What you should be able to do

• Say which chromosomes separate in the first division and which in the second, and why that order matters.
• Describe how a bivalent forms and what crossing over does to the chromatids in it.
• Work out how many chromosome combinations a gamete could carry, given the haploid number.
• Track chromosome number and chromatid number through both divisions without losing count.
• Set out the differences between mitosis and meiosis in terms an examiner can mark.

The division that has no equivalent in mitosis

A body cell of yours carries 46 chromosomes, and they are not 46 unrelated objects. They are 23 homologous pairs: two copies of chromosome 1, two of chromosome 2, and so on, one member of each pair inherited from each of your parents. The two members of a pair are the same length, carry the same genes in the same order, and differ only in which alleles sit at those loci. Mitosis takes no notice of the pairing at all. Meiosis is built around it.

The problem meiosis solves is arithmetical. If a sperm carrying 46 chromosomes fertilised an egg carrying 46, the zygote would have 92, and its children 184. Gametes therefore have to be haploid: one chromosome from each pair, 23 in total. Halving the number is not a matter of throwing chromosomes away at random — a gamete missing chromosome 7 and carrying two copies of chromosome 12 is no use to anyone. Each gamete needs exactly one representative of each pair, and the only reliable way to arrange that is to line the pairs up and pull them apart.

DNA replication happens once, in the interphase before meiosis begins, and not again between the two divisions. So a cell entering meiosis has 46 chromosomes and 92 chromatids; the cells leaving meiosis I have 23 chromosomes and 46 chromatids each; the four cells leaving meiosis II have 23 chromosomes and 23 chromatids each. The chromosome number halves at the first division and the DNA content halves at both.

Homologous pair

Two chromosomes of the same length carrying the same genes at the same loci, one from each parent, though not necessarily the same alleles.

Diploid

Having chromosomes in homologous pairs, written 2n.

Haploid

Having one chromosome from each homologous pair, written n.

Bivalent

The structure formed when a homologous pair comes together in prophase I: two chromosomes, four chromatids.

Reduction division

A division that halves the chromosome number. Meiosis I is one; meiosis II is not.

Prophase I: pairing, and the exchange that follows

Prophase I is the longest and the most eventful stage in the whole process. The chromosomes condense as they do in mitosis, but then something else happens: each chromosome finds its homologue and the two lie down alongside each other along their whole length. That pairing is called synapsis, and the paired structure is a bivalent. Because both chromosomes were replicated in interphase, a bivalent contains four chromatids, which is why some books call it a tetrad.

Lying that close, chromatids break and rejoin. A chromatid of the maternal chromosome and a chromatid of the paternal chromosome — non-sister chromatids, and the word non-sister is worth a mark on its own — break at the same point and each joins to the other's cut end. The point where they cross is a chiasma, plural chiasmata, and it is visible down a good microscope. When the bivalent later comes apart, the two chromatids that took part leave carrying a stretch of DNA from the other parent's chromosome.

The consequence is easy to state and easy to underrate: after crossing over, sister chromatids are no longer genetically identical. Everything you learned about mitosis assumed they were. Here, a chromatid can carry the maternal allele of one gene and the paternal allele of a gene further along the same chromosome — a combination that existed on neither chromosome the cell started with. Crossing over is the only mechanism in meiosis that makes new combinations of alleles within a chromosome, and it is what stops genes on the same chromosome being locked together for ever.

Crossing over

The exchange of equivalent sections of DNA between non-sister chromatids of a homologous pair during prophase I.

Chiasma

The visible point at which two non-sister chromatids cross during a bivalent's crossover.

Recombinant

A chromatid, or an offspring, carrying a combination of alleles that neither parental chromosome carried.

Metaphase I, and an arithmetic worth doing

At metaphase I the bivalents move to the equator, and this is where the second source of variation appears. Each bivalent settles with one member facing one pole and the other facing the opposite pole, and which member faces which pole is decided independently for every bivalent. That is independent assortment. Nothing keeps the chromosomes you inherited from your mother together; they are dealt out pair by pair, and each deal is a fresh coin toss.

number of chromosome combinations in a gamete = 2ⁿn is the haploid number: the number of homologous pairs. Crossing over is not included in this count, and makes the true figure much larger.

With n = 23, that is 2²³ = 8 388 608 different chromosome combinations a single human gamete could carry. Fertilisation is random, so any one of those sperm could meet any one of those eggs: 2²³ × 2²³ = 2⁴⁶, or about 7 × 10¹³ different zygotes from one couple, before crossing over is allowed to complicate anything. The number is not the point; the reason it gets that large from such a simple mechanism is.

One subtlety, in case a question asks for the number of arrangements rather than the number of gametes. A bivalent that sits with the maternal chromosome facing left produces the same pair of daughter cells as one facing right with the whole plate mirrored, so the number of genuinely distinct metaphase I arrangements is 2ⁿ⁻¹. Each arrangement sends a different combination to each of the two poles, so 2ⁿ⁻¹ arrangements still give 2ⁿ kinds of gamete. Answer whichever the question asked for, and say which you have counted.

The second division, and keeping count

Meiosis II starts with two haploid cells whose chromosomes are still two chromatids each, and it does to them what mitosis would: chromosomes line up singly on a new equator, centromeres divide, chromatids are pulled apart. The outcome is four haploid cells. Because crossing over happened back in prophase I, the two cells produced from each meiosis I product need not be identical to each other either.

Counting is where marks are lost, so it is worth a table. Take a cell with a diploid number of 8.

The chromosome number halves once, at the first division. The chromatid number halves twice — once because the cell divides in two and once because the centromeres finally split. A cell at the end of meiosis I is already haploid, even though it contains as much DNA as an unreplicated diploid cell, and that pair of facts is what makes questions using DNA mass against time so effective at separating candidates.

Set against mitosis, which you have already met, the differences line up cleanly. Learn them as a table rather than as prose, because that is the form the question takes.

Three sources of variation, and one way it goes wrong

Sexual reproduction generates variation at three points, and a question asking you to explain how meiosis produces genetic variation is asking for the first two by name, with the third as a bonus if the wording allows it.

Crossing over in prophase I makes new combinations of alleles along a single chromosome. Independent assortment at metaphase I makes new combinations of whole chromosomes. Random fertilisation then combines any one gamete with any other, which is not part of meiosis but is part of the answer whenever the question says sexual reproduction rather than meiosis. Mutation supplies the alleles that all three of these shuffle, but it is a separate process and claiming it as a product of meiosis will not score.

Occasionally the machinery fails. If a homologous pair does not separate at anaphase I, or a pair of sister chromatids does not separate at anaphase II, the result is non-disjunction: gametes with one chromosome too many and gametes with one too few. A zygote formed from a gamete carrying an extra chromosome 21 has three copies of it, which is the origin of Down syndrome. The same failure at anaphase II affects only two of the four products rather than all four, which is a distinction some specifications ask you to draw.

In the exam

• Homologous chromosomes separate in meiosis I; sister chromatids separate in meiosis II. Almost every mark on this topic hangs off getting those two the right way round.
• Crossing over is between non-sister chromatids of a homologous pair. Writing 'between chromatids' loses the mark because sister chromatids are identical and swapping between them changes nothing.
• For the number of gamete combinations, the exponent is the number of pairs. Halve the diploid number first, every time.
• A cell at the end of meiosis I is haploid but its chromosomes are still two chromatids each. Questions plotting DNA content against time are testing exactly that.
• If a question says 'explain how meiosis produces genetic variation', give crossing over and independent assortment and name the stage each happens in. Random fertilisation belongs in the answer only when the question says sexual reproduction.

Questions

Written to the command words the boards use. Try them on paper before opening a scheme: the marks go to points made, not to length.

← Plant responses: what the classic experiments actually show · One gene at a time: genetic diagrams that actually score →