1.1 Four problems #
1.1.1 Differentiation #
Differentiation is a cell becoming structurally and functionally specialised: a myotube full of contractile proteins, an epidermal cell full of keratin, a neuron with an axon. The central fact is genomic equivalence. With defined exceptions, differentiated cells of one individual contain the same genes. They differ in which of those genes they transcribe and in the chromatin state that keeps the choice stable. Chapter 8 treats the evidence and the limits of “master” transcription factors. Molecular Biology owns the histone marks and DNA methylation that record the choice.
Specialisation is not the same thing as commitment. A cell can already be restricted to a muscle fate while it still looks like a plain embryonic cell. Differentiation is the overt execution of a decision that specification and determination, defined below, have already begun. Equating “looks different” with “fate is fixed” is the usual first mistake.
1.1.2 Morphogenesis #
Morphogenesis is the shaping of tissues and organs: sheets fold, cells migrate, tissues elongate, cavities open, and some cells die so that others can separate. It is not a second kind of genetics. It is cell behaviour organised in space. The behaviours themselves β apical constriction, convergent extension, epithelial-to-mesenchymal transition, crawling on extracellular matrix β use the cytoskeleton, adhesions, and motors described in Cell Biology. These notes ask which cells do them, in which direction, and under which signal.
A fate map says where cells will end up. It does not say whether they got there by moving, by being pushed, or by staying still while their neighbours moved. Gastrulation, neurulation, neural-crest migration, and the folding of the optic cup are morphogenesis. Patterning, the assignment of different identities to different places, is what tells morphogenesis which cells should move and which should stay.
1.1.3 Growth #
Growth is an increase in size, by cell proliferation, by cell enlargement, or by deposition of extracellular matrix. Cleavage, in Chapter 3, is not growth. The early embryo divides the egg’s volume into more cells; the embryo does not get larger until it can feed, or, in mammals, until the mother supplies it through extraembryonic tissues.
Growth is patterned. A limb bud does not proliferate equally everywhere, and a tube bends if one side grows faster than the other. Final organ size is also restrained. The Hippo pathway, which phosphorylates and holds the transcriptional regulators YAP and TAZ out of the nucleus when cells are crowded, is one widely used brake on proliferation. Contact inhibition in Cell Biology is the same class of problem. These notes will not re-derive that pathway. They will treat growth as the reason a pattern has to be scaled, maintained while cells divide, and eventually stopped.
1.1.4 Reproduction and the germline #
A body that cannot make gametes is a dead end. Reproduction, as a developmental problem, is the setting aside of a germline, the construction of eggs and sperm, and the return at fertilisation to a single diploid cell. The soma can differentiate aggressively because the germline keeps a genome that has not been spent on one specialised job.
A Putative Model Of How The Organizer Induces Germ Layer Specification And Patterning
Animals do this in two ways, both real. In Drosophila, Xenopus, and many other species, a localised germ plasm in the egg makes whatever nuclei enter it into primordial germ cells. This is inheritance of a ready-made cytoplasmic determinant. In the mouse, primordial germ cells are induced in the epiblast by BMP signals from extraembryonic tissue, and only afterwards migrate into the gonad. The germline is set aside early in both cases, but the mechanism is autonomous in the first and conditional in the second. Those two words are defined properly in the next topic. Gametogenesis itself is Chapter 2.
1.2 Maps, commitment, and the two tests #
1.2.1 Fate maps #
A fate map records what each region of the embryo becomes in normal development. It is a prediction under normal circumstances, not a test of whether that future is fixed. The classical maps were made by marking the surface of an amphibian blastula with vital dyes on agar chips and watching where the colour went, as Vogt did. Later maps used fluorescent lipophilic dyes, or a graft of quail tissue into a chick host. Quail cells carry a dense nucleolar heterochromatin clump, so Le Douarin and colleagues could score, cell by cell, what a grafted neural crest or neural tube had become. A genetic mark does the same job inside one species: a recombinase switched on in a chosen cell changes a reporter permanently, and every daughter inherits the change.
Marks dilute, die, or spread if the method is poor, so a fate map is only as good as the mark. Even a perfect map does not say whether the cell would have done the same thing in isolation. That requires the two tests in the next section.
| Term | Claim | Test |
|---|---|---|
| Fate | What the cell becomes if left in place | Label it and do not move it |
| Specification | It follows that fate in a neutral environment | Isolate it in simple culture |
| Determination | It follows that fate even in a new place in an embryo | Transplant it to a different site |
| Commitment | The path is restricted; specification is labile commitment, determination is stable commitment | Whichever of the two tests is meant |
1.2.2 Specification and determination #
A cell or a tissue is specified if, isolated from the embryo and placed in a neutral medium, it still develops according to its fate-map fate. A cell or a tissue is determined if it still does so after it is grafted to a different place in a host embryo, where new neighbours offer new signals. Determination is the stronger statement. Specified tissue can be respecified by a determined neighbour; determined tissue resists.
“Neutral” is an idealisation. An explant still contains its own internal neighbours, so local signals can survive the isolation. Dissociating the explant into single cells is the stricter test of whether a cell, rather than a small community, already carries its fate. The animal-cap experiments in Chapters 5 and 6 only make sense if this is kept straight: an intact cap, a dissociated cap, and a cap paired with another tissue are three different questions.
Commitment usually tightens with time. Early gastrula ectoderm of an amphibian can still be neuralised by an organiser graft. Later ectoderm has lost that competence and continues as epidermis even if the organiser is presented. Competence, the ability to respond, is itself developmental and is taken up in Chapter 6. A fate map of that later ectoderm and a fate map of the early ectoderm can look similar while the state of commitment has changed completely.
1.2.3 Autonomous and conditional specification #
Autonomous specification means the cell inherited a cytoplasmic determinant and differentiates according to it even when neighbours are absent. Ascidian embryos show it cleanly. Conklin watched a yellow cytoplasm, the myoplasm, segregate into particular blastomeres that become muscle. That cytoplasm contains macho-1 mRNA, encoding a transcription factor sufficient to push a blastomere toward muscle. Isolate that blastomere and muscle markers still appear.
Conditional specification means the fate depends on interactions. Separate the blastomeres of a four-cell sea-urchin embryo, as Driesch did, and each can make a whole, smaller larva. The fate those cells would have followed inside the intact embryo was not fixed inside each one. It was negotiated.
Most embryos use both, often in the same hour. The ascidian blastomere that makes muscle autonomously still takes part in later inductions, including neural induction. The sea-urchin embryo that regulates so impressively also localises Ξ²-catenin in the vegetal micromeres, and those micromeres signal to their neighbours to organise endomesoderm. Remove them and gastrulation suffers; transplant them to another position and they can induce a second gut. Regulation is not the absence of localised information. It is the ability of remaining cells to rebuild interactions when a part is removed.
1.2.4 Mosaic and regulative embryos #
Mosaic development is the historical name for embryos whose early blastomeres behave as if each already contains its piece of the body. Kill one cell and that structure is missing. Regulative development is the name for embryos that fill in. The words describe results of experiments. They are not natural kinds, and they are not synonyms for “has genes” and “does not”.
Early Gastrulation In Amphibian Embryos
Roux’s result is the cautionary tale. In 1888 he killed one blastomere of a two-cell frog embryo with a hot needle and saw a half-embryo. He concluded that development was mosaic. The dead cell was still sitting in the way. When the two blastomeres are fully separated, or when Spemann constricted a newt egg with a fine hair loop so that the two halves could develop without a corpse between them, the result depended on the plane of the cut. If both halves received part of the grey crescent, the mark of the future dorsal side, each could make an embryo. If the crescent went entirely to one half, that half made an embryo and the other made a disorganised ventral mass, the “belly piece”. Amphibians have localised dorsal information and still regulate. Roux had measured the obstruction as well as the fate.
| Emphasis | Isolation of an early blastomere | Classical case | What is still true of the other kind |
|---|---|---|---|
| Mosaic | Makes a partial embryo, the part it would have made | Ascidian muscle lineage; much of the early C. elegans lineage; snail blastomeres that inherit the polar lobe | Later conditional signals still occur |
| Regulative | Makes a whole, smaller embryo | Sea urchin at the four-cell stage; early mouse blastomeres; amphibian halves that both receive dorsal cytoplasm | Localised determinants are still present |
C. elegans is the cleanest place to see the mixture. The somatic lineage is nearly invariant, and cytoplasmic determinants (including PIE-1 for the germline and SKN-1 for pharyngeal and other fates) are segregated into particular blastomeres. That is mosaic logic. Yet the vulva is conditional. An anchor cell in the gonad secretes LIN-3, an EGF-like signal. The nearest vulval precursor, P6.p, becomes the central fate; its neighbours, receiving less LIN-3 and a lateral Notch signal, become the side fates. Kill the anchor cell and no vulva forms. A “mosaic animal” is one in which autonomous specification is easy to see, not one in which induction has been abolished.
1.3 The organisms these notes use #
1.3.1 What a model is for #
A model organism is a practical choice, not a claim that other species develop differently in every respect. The useful properties are some combination of accessible embryos, short generation time, genetics, a small cell number, or large cells that survive surgery. The cost is bias. A mechanism discovered in a syncytial fly embryo, or in a plant meristem, has to be re-tested before it is declared general. These notes name that limit wherever a conclusion depends on it. The worst limit in the set is the early Drosophila embryo, which patterns its axes before it has cell membranes. Chapter 4 is written so that this cannot be missed.
1.3.2 Seven organisms #
| Organism | Why it is used | Limit worth remembering |
|---|---|---|
| Caenorhabditis elegans | The whole somatic lineage was followed cell by cell in the hermaphrodite: 1090 somatic cells are born and 131 die, leaving 959. Transparent, fast, and suited to RNA interference | Few cells, and a body plan that is not a vertebrate’s. Invariance tempts people to forget inductions such as the vulva |
| Drosophila melanogaster | Saturated genetic screens for embryonic pattern. Maternal-effect and zygotic mutations can be told apart. Polytene chromosomes once made the genome physically visible | The first thirteen nuclear cycles are syncytial. There is no vertebrate organ to point at |
| Xenopus laevis, and X. tropicalis | Eggs about 1 mm across survive grafts and explants. The organiser was worked out by surgery on amphibians. X. tropicalis is diploid and more useful genetically | X. laevis is pseudotetraploid, with a long generation. A frog is not a mammal |
| Danio rerio | Vertebrate genetics in a transparent embryo that develops outside the mother. Large screens recovered mutations in somitogenesis, crest, and organs | Teleosts had an extra genome duplication, so many developmental genes exist as two paralogues |
| Chick, Gallus gallus | The embryo is a flat disc on a huge yolk and can be windowed. Limb buds, the node, and neural crest are grafted, including quailβchick chimeras | Classical genetics is weak compared with fly or mouse |
| Mouse, Mus musculus | Targeted mutations, conditional alleles, and chimeras. The only model here that implants and builds a placenta | Development is internal. A graft that is routine in a chick is microsurgery against a small, hidden embryo. How the alleles are engineered belongs to Biotechnology and Genetic Engineering |
| Arabidopsis thaliana | A small flowering plant with a short life cycle and a rich set of patterning mutants. Chapter 9 uses it for the embryo, the meristems, and floral organs | Walls and meristems, not gastrulation. It is not a model of a tree, a cereal grain, or an animal |
The list is short on purpose. Echinoderms appear when regulation or micromeres need them. Ascidians appear when autonomous muscle specification needs them. Salamanders, hydra, and planarians appear in Chapter 8 because regeneration, not convenience of genetics, is the question. Snails appear once, for spiral cleavage and a maternal effect on its direction. None of these is “the” organism. Each is an assay.
1.4 Key terms, summary, and self-check #
Key terms. Differentiation; morphogenesis; growth; germ plasm; fate map; specification; determination; commitment; competence; autonomous specification; conditional specification; mosaic; regulative.
Summary. Development has to differentiate cells, arrange them, control size, and keep a germline. A fate map is not a commitment. Specification is shown by isolation, determination by a graft that fails to respecify, and both autonomous determinants and conditional signals operate in embryos that textbooks call mosaic and in embryos they call regulative. The model organisms are chosen for lineage, genetics, grafts, or plant meristems, and the early fly is a special case because it is a syncytium.
Self-check.
1. A piece of tissue follows its fate-map fate in culture, but changes fate when grafted to a new position. Is it specified, determined, both, or neither?
2. Why did Roux’s killed frog blastomere overstate the case for mosaic development?
3. Give one autonomous and one conditional event in C. elegans or in the sea urchin, and say which experiment distinguishes them.
4. Why is the chick used for neural-crest grafts and the mouse for gene knockouts, rather than the other way round?
5. Which of the four problems is a fate map able to answer, and which of them requires knowing whether cells were talking to each other?
—
—
—