THREE TO ONE
Mendel’s Peas
5,474 round : 1,850 wrinkled
Mendel’s second generation of round × wrinkled peas gave 5,474 round seeds and 1,850 wrinkled, 2.96 : 1: one gene, two alleles, round dominant. On the hybrid plants both kinds lie side by side in the same pod. The wrinkled allele, cloned in 1990, is a 0.8 kb transposon-like insertion in the gene for starch-branching enzyme I.
THREE TO ONE
The Dihybrid Cross
315 : 108 : 101 : 32
Crossing round yellow peas with wrinkled green, Mendel’s second generation came out 315 round yellow, 108 round green, 101 wrinkled yellow and 32 wrinkled green: 9 : 3 : 3 : 1, two genes assorting independently. Punnett’s square shows why: four kinds of pollen meet four kinds of egg in sixteen equally likely ways.
THREE TO ONE
Hardy–Weinberg
p² + 2pq + q² = 1
With random mating and no selection, migration, mutation or drift, genotype frequencies settle in one generation at p², 2pq and q² and stay there. Hardy and Weinberg showed it independently in 1908, answering the idea that a dominant allele should spread until it is common: dominance alone changes nothing.
THREE TO ONE
The ABO Blood Groups
three alleles, one gene
Landsteiner found three of the groups in 1901 by mixing people’s serum with each other’s red cells, and AB followed in 1902. In 1924 Bernstein showed they are three alleles of one gene: A and B codominant, O recessive. Anti-A clumps A and AB cells, anti-B clumps B and AB, and O cells clump with neither.
THREE TO ONE
Johannsen’s Beans
the word gene, 1909
Johannsen weighed the seeds of self-fertilized Princess beans. Within one pure line, heavy and light beans gave offspring of the same average weight: the spread inside a line was not inherited, and selecting on it did nothing. In 1909 he named the gene, the genotype and the phenotype.
SUTTON–BOVERI
Sturtevant’s Map
the first genetic map, 1913
In 1913, working in Morgan’s fly room, Sturtevant turned the recombination frequencies between six sex-linked genes into distances along a line: yellow 0.0, white and eosin 1.0, vermilion 30.7, miniature 33.7, rudimentary 57.6. One percent recombination became one map unit, later named the centimorgan.
SUTTON–BOVERI
The Chiasma
crossing over, seen
In meiosis each pair of matching chromosomes, four chromatids in all, is held together by chiasmata, where chromatids of the two partners cross. Janssens proposed in 1909 that they mark exchanges; in 1931 Creighton and McClintock in maize, and Stern in Drosophila, proved that crossing over is a physical swap of chromosome pieces.
SUTTON–BOVERI
The Polytene Chromosome
five arms and a chromocenter
In the larval salivary glands of Drosophila the chromosomes copy themselves about ten times without dividing and keep the copies in register: each becomes a cable about a thousand strands thick, crossed by a fixed pattern of bands. Painter used the bands to place genes in 1933, and Bridges drew the standard maps.
SUTTON–BOVERI
McClintock’s Maize
Ac and Ds, 1950
McClintock found genes that move. In her maize the Dissociation element, sitting in a pigment gene, keeps a kernel colorless; when Activator lets it jump out in one cell, that cell’s descendants make purple. Early jumps give big purple sectors, late ones small specks. She described it in 1950 and was awarded the 1983 Nobel Prize.
SUTTON–BOVERI
The Calico Cat
one X switched off per cell
In a female mammal one X chromosome in every cell is silenced early in development, at random, and stays silent in that cell’s descendants. A cat’s orange gene is on the X, so a female carrying orange and non-orange wears both in patches. Lyon proposed it in 1961; Barr and Bertram had seen the silent X, as a dark body, in female cat neurons in 1949.
5′ TO 3′
The Base Pairs
C1′ to C1′: 10.5 Å in both
Adenine pairs with thymine by two hydrogen bonds and guanine with cytosine by three, and the two pairs have the same width: their sugar attachment points sit about 10.5 Å apart at the same angles, so any sequence fits one regular helix. Watson and Crick drew G·C with two bonds in 1953; Pauling and Corey added the third in 1956.
5′ TO 3′
The Replication Fork
Okazaki, 1968
DNA polymerase adds only to a 3′ end, and the two strands run in opposite directions. So at a fork one new strand is made continuously while the other is made backward in short pieces, each started from an RNA primer and later joined. Okazaki’s group found the pieces in 1968: 1,000 to 2,000 nucleotides long in bacteria, 100 to 200 in eukaryotes.
5′ TO 3′
The Holliday Junction
Holliday, 1964
Holliday proposed in 1964 that recombining DNA molecules swap strands through a four-way junction, which can slide along the DNA and be cut either way to finish the exchange. With magnesium it folds into a stacked X: two continuous double helices crossing at about 60°, joined by the two strands that cross between them.
5′ TO 3′
Telomeres
(TTAGGG)n at every end
Each chromosome end carries a long run of the repeat TTAGGG in vertebrates, capped by proteins so the cell does not mistake it for a break. Ordinary replication cannot copy the very end, so telomeres shorten unless telomerase rebuilds them. Blackburn and Gall found the repeats in 1978; Greider and Blackburn found telomerase in 1985.
5′ TO 3′
A, B and Z
11, 10.5 and 12 bp a turn
The same double helix has three classic forms. B-DNA, the form in cells, is right-handed with 10.5 base pairs a turn. A-DNA, the form of dehydrated DNA and of RNA duplexes, is right-handed, wider and squatter, with about 11. Z-DNA, found by Wang, Rich and colleagues in 1979 in alternating CG, is left-handed and slim, 12 a turn, its backbone zigzagging.
CIS AND TRANS
The Lac Operon
Jacob and Monod, 1961
Lac repressor, four identical chains, grips two operators at once and loops the DNA between them, so the lactose genes stay off; when allolactose binds, it lets go and they switch on. Jacob and Monod deduced the circuit from bacterial crosses in 1961, the first mechanism of gene regulation; the looped complex was solved in 1996.
CIS AND TRANS
The Christmas Tree
Miller spreads, 1969
Spread for the electron microscope, genes being transcribed look like feathers: a DNA axis with RNA chains growing from it, short at the start of the gene and long at the end, a polymerase at the base of each. Miller and Beatty showed it for the ribosomal RNA genes of newt oocytes in 1969: identical trees in a row, bare spacers between.
CIS AND TRANS
The Lariat
a 2′–5′ branch
Splicing removes introns from a gene’s RNA in two steps. First the 2′ hydroxyl of an adenosine inside the intron attacks the intron’s 5′ end, tying it into a loop with a tail; then the two exons are joined and the lariat is released. The branch adenosine carries three phosphodiester bonds. Split genes were found in 1977, the lariat in 1984.
CIS AND TRANS
The Nucleosome
147 bp in 1.65 turns
DNA in a nucleus wraps 1.65 left-handed turns, 147 base pairs, around a core of eight histones, two each of H2A, H2B, H3 and H4. Kornberg proposed the repeating unit in 1974. The core particle’s structure, solved by Luger and colleagues in 1997, shows the histone tails reaching out between the turns, where chemical marks are written.
CIS AND TRANS
Seven Stripes
eve: every other segment
About three hours after laying, a fruit-fly embryo is a single layer of cells, and the pair-rule gene even-skipped is already on in seven sharp stripes, one for every other future segment. Each stripe is set by its own enhancer reading gradients of other gene products. Nüsslein-Volhard and Wieschaus found the pair-rule genes in 1980.
MINIPREP
PCR
doubling every cycle
Heat separates the strands, cooling lets two primers bind, and a heat-stable polymerase copies from each: every cycle can double the stretch between the primers, so thirty cycles turn one molecule into about a billion. Mullis conceived it in 1983; Taq polymerase, from a hot-spring bacterium, made it routine in 1988.
MINIPREP
The Agarose Gel
λ/HindIII: 23,130 to 125 bp
DNA is negatively charged, so in an electric field it runs through agarose toward the anode, short pieces fastest, and ethidium makes it glow orange in ultraviolet light. Phage λ's 48,502 base pairs cut with HindIII give the classic marker: 23,130, 9,416, 6,557, 4,361, 2,322, 2,027, 564 and 125. Plasmid pBR322, also 4,361 base pairs, runs level with the 4,361 band once cut open.
MINIPREP
The Sanger Ladder
ATG GTG CAT CTG ACT CCT GAG
Sanger’s 1977 method copies a template with a polymerase and a little chain-ending dideoxynucleotide, one base per reaction, so four lanes of fragments each end at one base. Read from the bottom, the bands spell the sequence. This film reads the start of human β-globin; its seventh codon, GAG, becomes GTG in sickle-cell disease.
MINIPREP
CRISPR–Cas9
a 20-nt guide, then NGG
Cas9 from Streptococcus pyogenes cuts DNA wherever a 20-nucleotide guide RNA pairs with it, but only if the next three bases on the other strand read NGG, the PAM. It unwinds the DNA, the guide pairs with one strand, and two nuclease domains cut both strands three base pairs before the PAM, leaving blunt ends. Jinek and colleagues made it programmable in 2012.
MINIPREP
Blue–White Screening
X-gal: blue = empty vector
The plasmid carries a short piece of lacZ that completes the host’s broken β-galactosidase, which turns X-gal blue. Cloning a gene into that piece breaks it, so colonies with an insert stay white. It rests on α-complementation, found by Ullmann, Jacob and Monod in 1967, and on the pUC vectors of Vieira and Messing (1982).