8.1 The same genome, different transcription #
8.1.1 Genomic equivalence #
If differentiation threw genes away, a nucleus from a specialised cell could not build any other cell. That was a serious possibility in the era of Weismann’s unequal nuclear determinants. It is not the general rule. Briggs and King transplanted nuclei from Rana blastulae into enucleated eggs and got development; nuclei from later stages did worse, which left open the idea that potency really does shrink. Gurdon’s transplants in Xenopus closed it for practical purposes. Nuclei from differentiated tadpole intestinal epithelium, placed in enucleated activated eggs, supported swimming tadpoles, and with serial transplantation fertile adults were recovered. An adult mammalian nucleus can do the same: the sheep Dolly was cloned from a cultured mammary epithelial cell. Induced pluripotent stem cells, returned toward an embryonic state by transcription factors rather than by an egg, are the modern form of the same evidence. The factors and the hazards of incomplete reprogramming are in the Cell Biology notes. The developmental conclusion is the one needed here. Differentiation is a regulatory state, not a deletion of the genome.
The rule has real exceptions, and they should be named so that “equivalence” stays a rule rather than a superstition. Lymphocytes rearrange and delete DNA at the immunoglobulin and T-cell receptor loci; Immunology owns that system. Mammalian erythrocytes enucleate. Some nematodes, including Ascaris, eliminate parts of the chromosomes in somatic cells, chromatin diminution, while the germline keeps the full set. None of these is how a muscle or a neuron is made. Organ identity is differential transcription of a retained genome, stabilised by chromatin states that Molecular Biology describes and by autoregulation: a transcription factor that turns itself on will persist after the signal that started it has gone. Determination, in the sense of Chapter 1, is often that kind of self-maintaining circuit.
8.1.2 MyoD, and the limit of the word master #
MyoD is the classical master regulator of skeletal muscle. It is a basic helix-loop-helix transcription factor, it dimerises with E proteins, and the dimer binds E-boxes in muscle genes. Expressing MyoD in many fibroblast and other non-muscle cultured cells is enough to turn on the muscle programme and, in favourable cases, to fuse the cells into myotubes that express contractile proteins. The experiment, from Davis, Weintraub, and Lassar, is the gain-of-function definition of a master regulator: a single introduced gene starts a recognisable differentiation programme in a cell that was not using it.
In the embryo one gene rarely does everything, and MyoD does not. The family has four members: MyoD, Myf5, Mrf4 (Myf6), and myogenin. MyoD and Myf5 are largely redundant for specifying the myoblast. Single mutants still make skeletal muscle; the double mutant does not make skeletal myoblasts. Myogenin acts downstream, in the differentiation step itself, and without it myoblasts are specified but fail to execute the contractile programme properly. Pax3 and Pax7 sit upstream in somitic and migratory progenitors, which is why limb muscle, born in the somite, still depends on the myogenic family after the cells arrive. Cardiac muscle and smooth muscle do not use MyoD. They use other networks, including Nkx2-5, GATA factors, and Tbx5 in the heart, and myocardin with serum-response factor in smooth muscle. A master regulator of skeletal muscle is not a master regulator of “muscle”.
The cultured fibroblast and the embryo also differ in competence. MyoD works well in cells whose chromatin can still be opened by that factor and its recruited enzymes. It works poorly or not at all in cells that cannot. The ectopic eye of Chapter 7 fails for the same reason outside a competent disc. Autoregulation then makes the successful switch stable: MyoD activates MyoD. The word “master” is a summary of leverage plus feedback. It is not a claim that the rest of the genome is silent scenery.
8.2 Stem cells in the embryo #
8.2.1 Potency, progenitors, and a pointer #
Potency is the range of cell types a cell can still produce. The zygote is totipotent: it can make the embryo and the extraembryonic membranes. In the mouse, blastomeres very early can also give rise to both classes of tissue; by the blastocyst that equivalence is gone. The inner cell mass is pluripotent, able to make all three germ layers of the body, and trophoblast is restricted to extraembryonic fates. Embryonic stem cells are cultured derivatives of that pluripotent population. Making and maintaining the cultures is a method these notes do not teach. The developmental fact is their origin.
Multipotent stem cells produce a limited set of related types: a haematopoietic stem cell, or the neural stem cells of the embryonic ventricular zone. Unipotent stem cells produce one differentiated type, as spermatogonial stem cells produce sperm. A progenitor is not a stem cell merely because it divides. Many embryonic progenitors are transient. They are specified, they divide a limited number of times, and they are used up in building the organ. The progress-zone mesenchyme was never a stem-cell niche, even in the old model. Calling every dividing embryonic cell a stem cell empties the term.
The niche is the local environment β neighbours, matrix, and signals β that keeps a stem cell from differentiating. Those signals are familiar: Wnt, Notch, BMP, and others, depending on the tissue. Cell Biology treats the niche as cell biology, including the intestinal crypt. Developmental biology’s addition is that the embryo built those niches with the same pathway grammar as Chapter 6, and that a daughter which leaves the niche loses the signals and differentiates. Lineage tracing is how a stem cell is recognised in the animal rather than in culture. One cell is marked permanently, its mark is inherited, and years later the marked clone still contains the differentiated cell types and at least one cell that is still the stem cell. A population average cannot say that. A clone can.
8.3 Regeneration #
8.3.1 Hydra: interstitial cells and a Wnt head #
Hydra is a freshwater cnidarian with an oralβaboral axis and no fixed adult endpoint. Tissue is continuously replaced. Two stem-cell systems run in parallel. Epithelial stem cells of the ectoderm and of the endoderm renew those layers. Interstitial cells, or I-cells, are the stem cells that produce neurons, nematocytes, secretory cells, and gametes. The head organiser is the hypostome, and it is a Wnt/Ξ²-catenin high point, recognisably the same pathway as a dorsal frog axis although the animal is not a bilaterian. Grafting a hypostome to the body column induces a second axis.
Cut a hydra and the piece regenerates according to the polarity it already had. The apical cut surface makes a head, the basal cut surface makes a foot, and a slice from the middle does both. Polarity is not re-rolled at random. Much of this regeneration is morphallaxis: existing tissue is re-patterned, with less dependence on a local mound of proliferation than a salamander limb shows. The old literature’s low-molecular-weight “head activator” is not a result these notes rely on. The Wnt centre at the hypostome is.
8.3.2 Planarian neoblasts #
Planarians regenerate from small pieces and can do it repeatedly. The only proliferating somatic cells are neoblasts, small cells scattered through the parenchyma, marked by piwi-family genes. Some of them, the clonogenic neoblasts, are pluripotent in a strong sense: a single transplanted neoblast can reconstitute an animal whose own neoblasts have been killed by irradiation. Not every neoblast has been shown to be that cell. Many are already lineage-biased progenitors. The word “neoblast” is a class, not a guarantee of totipotency.
Positional information is not stored inside the neoblast. Body-wall muscle expresses position-control genes: Wnt ligands posteriorly, the Wnt inhibitor notum at an anterior regeneration pole, and other signals for the other axes. Neoblasts supply cells; muscle supplies the address. Knock down Ξ²-catenin and a regenerating fragment makes a head at both ends. Force Wnt signalling up and it makes two tails. The French flag has been rewritten in an adult, with the same canonical Wnt pathway the embryo used, which is the deepest reason regeneration belongs in a developmental course rather than in a footnote about healing.
8.3.3 The salamander blastema #
Urodele amphibians β newts and the axolotl β regenerate a limb after amputation. Epidermis closes the wound and thickens into an apical epithelial cap, a functional analogue of the AER, dependent on FGFs. Beneath it, cells near the cut re-enter the cell cycle and form a blastema, a mound of proliferating progenitors. This is epimorphosis: regeneration through a proliferative bud, as opposed to hydra’s morphallaxis. Pattern in the blastema obeys the rule of distal transformation. A blastema at the wrist regenerates a hand. A blastema at the upper arm regenerates an upper arm, forearm, and hand. Structures proximal to the cut are not remade. Graft a distal blastema onto a proximal stump and the positional discrepancy is filled in by intercalation. Positional values, the same idea as Wolpert’s flag, survive in the adult tissues and are re-read after injury.
The older textbook sentence, that blastema cells dedifferentiate all the way back to pluripotency, is not what lineage tracing shows. In the axolotl limb, muscle makes muscle, epidermis stays epidermis, and dermal connective tissue can make connective tissue and cartilage but does not casually become muscle. The blastema is a collection of lineage-restricted progenitors, not an inner cell mass. Dedifferentiation is real and local. It is not a return to the zygote.
Nerves are required for the early blastema to proliferate. A denervated stump heals and fails to grow a limb. In newts, anterior gradient protein from the nerve can substitute for innervation in the experiments that defined it, acting through a receptor, Prod1, that is not a universal vertebrate gene. The result to remember is nerve dependence of blastema growth, not a single factor sold as the reason mammals cannot regenerate. Exogenous retinoic acid proximalises a blastema: a wrist-level blastema treated with it can regenerate structures more proximal than the cut, up to a whole limb pattern. That is a gain-of-function on positional value. It is not a measured map of endogenous concentrations, and Chapter 7’s caution about retinoic acid in the embryonic limb applies here as well.
8.3.4 Why mammals are limited #
Mammals have the limb toolkit. Embryos use AER-FGFs, Shh, and Wnt7a. Adults do not, after an amputation, assemble a blastema that reads those signals the way a salamander does. The wound response favours fibrosis: fibroblasts deposit a collagen scar, and the scar is a barrier as much as a failure of imagination. More proximal than the nail bed, a mouse or a human digit does not replace the lost phalanx. Distal to the nail bed, in mice and in young children, the fingertip can regenerate, and Msx1 marks the distal mesenchyme that is capable of it. The nail organ provides a local signalling environment. It is not a small version of an axolotl blastema at the shoulder.
The genes were not deleted from the mammalian genome. The African spiny mouse, Acomys, regenerates an ear punch, including cartilage, instead of sealing it with the scar a laboratory mouse makes. A mammal can therefore run a regenerative skin programme. What mammals do not generally do is re-pattern a whole limb. Honest limits on the explanation include the speed and fibrotic character of the inflammatory response, the lineage restrictions that even salamanders obey, and body size. A single “regeneration gene” that mammals lost is not a serious model.
8.4 Ageing #
8.4.1 Not a developmental stage #
Ageing is the decline of maintenance after an organism is already built. It is not a late Hox code, and it is not stage 47 of the programme that laid down the somites. Development is under strong selection because animals that do not build a working body do not reproduce. Selection on the years after reproduction is weak. Mutation accumulation, antagonistic pleiotropy β alleles that help early and harm late β and the disposable-soma idea that maintenance is traded against reproduction are the evolutionary accounts. They live, as theories of selection, in Evolutionary Biology. They are mentioned here only to stop a confusion. Shared molecular parts do not make ageing into embryology. Insulin/IGF signalling and mTOR alter lifespan in worms, flies, and mice when they are attenuated. Those pathways also control growth. Changing a growth pathway changes how long maintenance lasts. It does not reveal a genetic cascade whose function is to time death the way the segmentation clock times a somite. Cellular senescence and telomere erosion are Cell Biology. These notes stop at the boundary.
Eb Development And Tj Formation
8.5 Key terms, summary, and self-check #
Key terms. Genomic equivalence; MyoD; myogenin; potency; pluripotent; progenitor; niche; interstitial cell; neoblast; morphallaxis; epimorphosis; blastema; distal transformation; fibrosis.
Summary. Differentiated nuclei retain the genome, as nuclear transplantation and, later, induced pluripotency showed, with listed exceptions that really do lose DNA. MyoD can start skeletal myogenesis in competent cells, but in the embryo it shares the job with Myf5, and myogenin executes differentiation. Stem cells are defined by potency and by long-term self-renewal in a niche; many embryonic dividing cells are only progenitors. Hydra regenerates by re-patterning, using epithelial stem cells and interstitial cells under a Wnt head. Planarians rebuild from neoblasts whose positions are dictated by muscle. Salamanders build a lineage-restricted blastema that replaces only what was distal to the cut. Mammals scar, except at the digit tip and in a few special cases, and ageing is a failure of maintenance rather than a last developmental stage.
Self-check.
1. What did Gurdon’s intestinal-nucleus transplants show, and name one genuine exception to genomic equivalence?
2. Why do single mutants of MyoD or Myf5 still make skeletal muscle, while the double mutant does not? Which family member is required for differentiation rather than specification?
3. Why is a transient embryonic progenitor not a stem cell? What does a lineage-tracing clone have to contain?
4. A planarian fragment regenerates a head at one cut and a tail at the other. Which tissue stores the positional information, and which pathway is high at the posterior?
5. Why is a mutation that extends lifespan by weakening insulin/IGF signalling not evidence for a developmental programme of ageing?
—
—
—