Building Brains E-Book

David J. Price

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: New York Academy of Sciences
• Sprache: Englisch

Seitenzahl: 878

Veröffentlichungsjahr: 2017

Beschreibung

Provides a highly visual, readily accessible introduction to the main events that occur during neural development and their mechanisms

Building Brains: An Introduction to Neural Development, 2^nd Edition describes how brains construct themselves, from simple beginnings in the early embryo to become the most complex living structures on the planet. It explains how cells first become neural, how their proliferation is controlled, what regulates the types of neural cells they become, how neurons connect to each other, how these connections are later refined under the influence of neural activity, and why some neurons normally die. This student-friendly guide stresses and justifies the generally-held belief that a greater knowledge of how nervous systems construct themselves will help us find new ways of treating diseases of the nervous system that are thought to originate from faulty development, such as autism spectrum disorders, epilepsy, and schizophrenia.

• A concise, illustrated guide focusing on core elements and emphasizing common principles of developmental mechanisms, supplemented by suggestions for further reading
• Text boxes provide detail on major advances, issues of particular uncertainty or controversy, and examples of human diseases that result from abnormal development
• Introduces the methods for studying neural development, allowing the reader to understand the main evidence underlying research advances
• Offers a balanced mammalian/non-mammalian perspective (and emphasizes mechanisms that are conserved across species), drawing on examples from model organisms like the fruit fly, nematode worm, frog, zebrafish, chick, mouse and human
• Associated Website includes all the figures from the textbook and explanatory movies

Filled with full-colorartwork that reinforces important concepts; an extensive glossary and definitions that help readers from different backgrounds; and chapter summaries that stress important points and aid revision, Building Brains: An Introduction to Neural Development, 2^nd Edition is perfect for undergraduate students and postgraduates who may not have a background in neuroscience and/or molecular genetics.

“This elegant book ranges with ease and authority over the vast field of developmental neuroscience. This excellent textbook should be on the shelf of every neuroscientist, as well as on the reading list of every neuroscience student.” 
—Sir Colin Blakemore, Oxford University

“With an extensive use of clear and colorful illustrations, this book makes accessible to undergraduates the beauty and complexity of neural development. The book fills a void in undergraduate neuroscience curricula.”
—Professor Mark Bear, Picower Institute, MIT.

Highly Commended, British Medical Association Medical Book Awards 2012

Published with the New York Academy of Sciences

Leseprobe

Table of Contents

Cover

Title Page

Preface to Second Edition

Preface to First Edition

Conventions and Commonly used Abbreviations

Naming conventions for genes and proteins

Commonly used abbreviations

Significance of bold and blue bold terms

Introduction

Major Components of the Adult Human Brain

About the Companion Website

1 Models and Methods for Studying Neural Development

1.1 What is neural development?

1.2 Why research neural development?

1.3 Major breakthroughs that have contributed to understanding developmental mechanisms

1.4 Invertebrate model organisms

1.5 Vertebrate model organisms

1.6 Observation and experiment: methods for studying neural development

1.7 Summary

2 The Anatomy of Developing Nervous Systems

2.1 The nervous system develops from the embryonic neuroectoderm

2.2 Anatomical terms used to describe locations in embryos

2.3 Development of the neuroectoderm of invertebrates

2.4 Development of the neuroectoderm of vertebrates and the process of neurulation

2.5 Secondary neurulation in vertebrates

2.6 Formation of invertebrate and vertebrate peripheral nervous systems

2.7 Summary

3 Neural Induction: An Example of How Intercellular Signalling Determines Cell Fates

3.1 What is neural induction?

3.2 Specification and commitment

3.3 The discovery of neural induction

3.4 A more recent breakthrough: identifying molecules that mediate neural induction

3.5 Conservation of neural induction mechanisms in Drosophila

3.6 Beyond the default model – other signalling pathways involved in neural induction

3.7 Signal transduction: how cells respond to intercellular signals

3.8 Intercellular signalling regulates gene expression

3.9 The essence of development: a complex interplay of intercellular and intracellular signalling

3.10 Summary

4 Patterning the Neuroectoderm

4.1 Regional patterning of the nervous system

4.2 Patterning the anteroposterior (AP) axis of the Drosophila CNS

4.3 Patterning the AP axis of the vertebrate CNS

4.4 Local patterning in Drosophila: refining neural patterning within segments

4.5 Local patterning in the vertebrate nervous system

4.6 Summary

5 Neurogenesis: Generating Neural Cells

5.1 Generating neural cells

5.2 Neurogenesis in Drosophila

5.3 Neurogenesis in vertebrates

5.4 The regulation of neuronal subtype identity

5.5 The regulation of cell proliferation during neurogenesis

5.6 Temporal regulation of neural identity

5.7 Why do we need to know about neurogenesis?

5.8 Summary

6 How Neurons Develop Their Shapes

6.1 Neurons form two specialized types of outgrowth

6.2 The growing neurite

6.3 Stages of neurite outgrowth

6.4 Neurite outgrowth is influenced by a neuron’s surroundings

6.5 Molecular responses in the growth cone

6.6 Active transport along the axon is important for outgrowth

6.7 The developmental regulation of neuronal polarity

6.8 Dendrites

6.9 Summary

7 Neuronal Migration

7.1 Many neurons migrate long distances during formation of the nervous system

7.2 How can neuronal migration be observed?

7.3 Major modes of migration

7.4 Initiation of migration

7.5 How are migrating cells guided to their destinations?

7.6 Locomotion

7.7 Journey’s end – termination of migration

7.8 Embryonic cerebral cortex contains both radially and tangentially migrating cells

7.9 Summary

8 Axon Guidance

8.1 Many axons navigate long and complex routes

8.2 Contact guidance

8.3 Guidance of axons by diffusible cues – chemotropism

8.4 How do axons change their behaviour at choice points?

8.5 How can such a small number of cues guide such a large number of axons?

8.6 Some axons form specific connections over very short distances, probably using different mechanisms

8.7 The growth cone has autonomy in its ability to respond to guidance cues

8.8 Transcription factors regulate axon guidance decisions

8.9 Summary

9 Life and Death in the Developing Nervous System

9.1 The frequency and function of cell death during normal development

9.2 Cells die in one of two main ways: apoptosis or necrosis

9.3 Studies in invertebrates have taught us much about how cells kill themselves

9.4 Most of the genes that regulate programmed cell death in C. elegans are conserved in vertebrates

9.5 Examples of neurodevelopmental processes in which programmed cell death plays a prominent role

9.6 Neurotrophic factors are important regulators of cell survival and death

9.7 A role for electrical activity in regulating programmed cell death

9.8 Summary

10 Map Formation

10.1 What are maps?

10.2 Types of maps

10.3 Principles of map formation

10.4 Development of coarse maps: cortical areas

10.5 Development of fine maps: topographic

10.6 Inputs from multiple structures: when maps collide

10.7 Development of feature maps

10.8 Summary

11 Maturation of Functional Properties

11.1 Neurons are excitable cells

11.2 Neuronal excitability during development

11.3 Developmental processes regulated by neuronal excitability

11.4 Synaptogenesis

11.5 Spinogenesis

11.6 Summary

12 Experience‐Dependent Development

12.1 Effects of experience on visual system development

12.2 How does experience change functional connectivity?

12.3 Cellular basis of plasticity: development of inhibitory networks

12.4 Homeostatic plasticity

12.5 Structural plasticity and the role of the extracellular matrix

12.6 Summary

Glossary

Index

End User License Agreement

List of Tables

Chapter 02

Table 2.1 The approximate timing of human prenatal development after fertilization.

List of Illustrations

Chapter 01

Figure 1.1 A node of Ranvier: these highly organized structures, formed as a result of interactions between axons and glia, are essential for speeding up the transmission of electrical signals along axons. In this single fibre from the mouse spinal cord, sodium channels (blue) are sandwiched between the regions where axons and glia form junctions (called axoglial junctions) (green), which are, in turn, flanked by potassium channels (red). This picture is courtesy of Peter Brophy and Anne Desmazieres, University of Edinburgh, UK.

Figure 1.2 Schizophrenia, intellectual disabilities, autisms and epilepsies are neurological disorders affecting about 3–7% of people. Based on epidemiological and neurobiological evidence, schizophrenia is now believed to be a neurodevelopmental disorder with a large heritable component. Many possible susceptibility genes have been identified, but how abnormalities of these genes cause the symptoms of the disease is unknown. Similarly, autism spectrum disorders and intellectual disabilities are highly heritable and many of the known genetic causes seem to regulate the formation of synapses. Malformations of cerebral cortical development are among the commonest causes of epilepsy. Some are large defects that would be obvious to the naked eye whereas others would only be seen at a microscopic or molecular level. They are a consequence of a disruption of the normal steps of cortical formation, for example defective migration of neurons, and can be environmental or genetic in origin. A large number of malformations of cortical development have been described, each with characteristic pathological and clinical features. An example of a large congenital defect causing epilepsy is shown in the scan of a patient’s brain on the right (between the arrows): for comparison, a scan of the brain of a normal person is shown on the left. This picture is courtesy of Professor John S. Duncan and the National Society for Epilepsy MRI Unit, UK.

Figure 1.3 Two fruit flies face each other: the fly on the right is a normal (wild‐type) fly, the one on the left is a mutant. In the mutant, a gene that is essential for the formation of eyes is defective. Flies lacking this gene do not develop eyes. The gene in question,

Pax6

, can be found in virtually all animals: in humans, flies, molluscs and even very simple worms. The

Pax6

gene is also called

eyeless

in

Drosophila

, since

Drosophila

genes are often named after their mutant phenotype: thus, somewhat confusingly, the

eyeless

gene is

required

to make the eye. This striking image is reproduced here with permission and is the copyright of Jürgen Berger and Ralf Dahm, Max Planck Institute for Developmental Biology, Tübingen, Germany (www.ralf‐dahm.com).

Figure 1.4

Reverse genetics

RNA interference can be used to block gene function experimentally. (a) Inside normal cells, genes are transcribed to make single‐stranded messenger RNA (mRNA) that is translated by ribosomes to generate specific proteins. (b) To block gene function,

antisense RNA

molecules with sequences complementary to the sense sequences of specific mRNAs are introduced into cells where they interact with their target mRNAs and block their translation. Many types of antisense molecule have been developed. They fall into two broad groups: after binding to target mRNA, some cause its enzymatic degradation whereas others can block its translation. For example, antisense molecules called morpholinos, which have been exploited very successfully in studies of

Xenopus

and zebrafish development, are examples of the latter. As well as being experimental tools, antisense molecules have therapeutic potential for treatment of human diseases. The development of antisense methods to regulate gene function experimentally or therapeutically was followed by the discovery of a wide range of small RNA molecules called microRNAs that are generated naturally by cells and act as physiological antisense molecules (see Section 3.8.4 in Chapter 3). In this diagram, the antisense RNA is introduced by microinjection, but there are many other ways such as

electroporation

(see Figure 1.7 later) or the use of viruses. Each method has pros and cons and which one is best depends on factors such as the numbers and types of cells to be targeted, the species and the age of the organism. In the future, the use of inhibitory RNA molecules described here is likely to be replaced increasingly by the use of RNA‐based approaches that disrupt gene function by mutating the gene in the DNA (e.g. using the

CRISPR/Cas9

system; see Section 1.5.4).

Figure 1.5 A time‐lapse series of images of labelled neural cells in a live zebrafish embryo showing one dividing while the other remains quiescent, made by Paula Alexandre in Jon Clarke’s laboratory, King’s College London, UK. Green is green fluorescent protein (GFP) that is being used to label cell membranes. Red is red fluorescent protein (RFP), a different fluorescent protein that is being used here to show nuclei (see Box 1.4 for details of these fluorescent molecules).

Figure 1.6 Mouse embryonic stem (ES) cells are derived from the inner cell mass of a mouse blastocyst (here from a line of mice with a white coat). The inner cell mass is transferred into a culture medium in a plastic laboratory culture dish. The cells from the inner cell mass divide and spread over the surface of the dish to become ES cells. ES cells can be differentiated into many cell types in culture, including neurons, heart muscle cells and blood cells. They can be injected into new blastocysts, here from a line of mice with brown fur (although the blastocyst cells are shaded brown to help make the diagram clear, in reality they would be no different in colour to those from the blastocyst of mice with white fur). The injected blastocysts are implanted into the uterus to generate chimeric (Box 1.5) offspring comprising a mixture of cells derived from the ES cells and the host blastocyst (the chimeric mice would have a mixture of brown and white fur).

Figure 1.7

Reverse genetics

: generation of transgenic mice. This method allows the experimenter to manipulate specific genes so as to learn about their functions. For example, a normal gene might be replaced with a modified version to generate a knock‐out, or we might want to insert sequences to make conditional mutants (Figure 1.8). To do this, the genome of embryonic stem cells (see Figure 1.6) is manipulated in culture. (a) The experimenter constructs DNA molecules that have (i) stretches at each end identical to sequences in and/or around the gene that is to be mutated and (ii) a central portion whose incorporation into the target gene will prevent its function (red). (b) To enable the embryonic stem cells to take up these DNA molecules, they are put into the solution around the embryonic stem cells and a current is passed through the cells (this is called electroporation). In some cells the flanking sequences swap places with the identical sequences in the genome (a chance event called homologous recombination, indicated in (b) by the two crossed broken lines), carrying the central portion into the genome to prevent the function of the target gene. Nowadays, experimenters would greatly increase the efficiency of this process by using the CRISPR/Cas9 system to induce a break at a specific place in the targeted region of the DNA (yellow star). This is done by giving the cells RNA molecules linked to an enzyme called Cas9 nuclease that cuts DNA (not shown in the diagram). The sequence of the RNA is chosen by the investigator to guide it to the correct place in the DNA and so these RNAs are known as guide RNAs. If all the experimenters want is to disrupt the gene and not replace anything, they might not bother electroporating the new DNA molecules and just induce a break in their target, because when it heals it will usually do so imperfectly, leaving behind a disruptive mutation. (c) The mutated embryonic stem cells are then injected into blastocysts to generate chimeras (d) in which some of the animal’s cells are mutant, including some germ cells. (e) Since some germ cells in these chimeras should be mutant, subsequent breeding with normal mice will generate offspring in which all cells are heterozygous for the mutation as well as other mice that are normal. (f) A second round of breeding between heterozygotes will generate some mice that are homozygous for the mutation (double red dot), some that are heterozygous for it (single red dot) and some that are normal. Many variations of this method are possible: for example, DNA containing an entire gene controlled by appropriate regulatory elements might be added to the genome so as to overproduce a specific protein in a specific part of the animal.

Figure 1.8

Reverse genetics

: generation of conditional transgenic mice with the cre‐

loxP

system. (a) An example of a transgenic mouse that has been made to produce the bacteriophage enzyme cre recombinase specifically in its cerebral cortex when it was an embryo (purple). To achieve this, the cre recombinase gene is controlled by regulatory sequences in the DNA that are known, from previous work, to operate only in the cerebral cortex (regulatory sequences have right‐angled arrows above them, which is a common convention; they are discussed further in Chapter 3). (b) An example of a transgenic mouse that has

loxP

sites (triangles, again a common convention) flanking both copies of a gene that is present throughout the nervous system of the embryo; in this case the gene is called

Pax6

(orange: it codes for a protein that is a

transcription factor

; you can find more about these molecules in Chapter 3). (c) In the offspring produced by breeding these two lines of mice together,

only

those cells that make cre recombinase (i.e. cerebral cortical cells) will delete the target gene. In the photographs of thin slices of the brain,

Pax6

(which is stained brown using a method called immunocytochemistry, described in Chapter 3, Box 3.4) is deleted from the cerebral cortex of the conditional mutant. The cortex is the tissue underneath the double‐headed arrows. The mutant is on the right and a normal embryo is shown on the left for comparison. Pax6 is not affected in the regions beneath the cortex (asterisk). (d) Sometimes one finds that conditional mutants survive longer than mice in which all cells are affected by a mutation, extending the period over which the effects of mutation can be studied. This depends on what the mutated gene does. (You can find an example of how the cre‐

loxP

can be put to other uses in Chapter 7, see Box 7.2.)

Chapter 02

Figure 2.1 Key stages in the development of

C. elegans

: development is highly stereotypical and the lineages of all the cells of

C. elegans

are known. (a) This diagram charts the first divisions of the embryo. The zygote divides to generate an anterior blastomere (AB), which divides to generate more AB cells and will eventually form the nervous system, and a posterior cell (P

1

), which divides to generate other cell types (note that the cells derived from AB and P

1

are known simply by their letters and numbers, as here). Gastrulation is complete once derivatives of the AB blastomere have spread out to cover the derivatives of P

1

, which become internalized (making mesoderm and endoderm). (b) Arrows show the inward migration of AB‐derived ectodermal cells on the ventral side of the embryo. (c) These AB‐derived ectodermal cells form sensory and motor neurons along the body: only a single example of each is shown here.

Figure 2.2 Major stages of

Drosophila

development. (a) Blastoderm stage: arrows show the main directions of cell movements during gastrulation. The neuroectoderm (orange) is initially split into two domains along either side of the ventral part of the embryo and as the mesoderm involutes these domains coalesce ventrally (curved arrows). Endoderm is shown at the top right but, for clarity, is omitted in later pictures. (b) Subsequently, some of the cells in this ventral neurogenic region move inside the embryo to become neuroblasts; this process is called delamination. Other cells in the lateral ectoderm delaminate to become sense organ precursors: the development of these cells is shown down the left between (b) and (d) (see also Section 2.6.1). (c) The neuroblasts divide to generate the neurons and glia of the

Drosophila

’s CNS, many of which reside in the ventral nerve cord. This is achieved through the production of intermediate cells, the ganglion mother cells (GMCs), which divide to generate pairs of neurons or glia. (d) Sensory nerves from developing sense organs converge on the ventral nerve cord. Motor nerves exit the ventral nerve cord to innervate the body. Axons within the ventral nerve cord are organized into bundles running along the length of the embryo, called longitudinal fascicles, and bundles linking across the midline, called commissures.

Figure 2.3 Schematic of primary neurulation in a vertebrate: note how the lateral edges of the neural plate roll up and join dorsally, while cells that are medial in the neural plate end up ventrally. This establishes the dorsoventral axis of the vertebrate neural tube.

Figure 2.4 The main stages of development of

Xenopus

: (a) formation of the blastula, (b) gastrulation, which occurs about a day after fertilization, and (c) development of the neural plate and neural tube (orange) through primary neurulation. In (c), the neural plate is first shown in a transected embryo tilted towards the viewer along its anteroposterior axis. Cells along the lateral edges of the neural plate (brown) form the neural crest: its cells migrate laterally throughout the body (see later in this chapter, Section 2.6). The final drawings show the tadpole stage: brain structures such as the retinae and optic tectum, which are major components of the visual system, develop from the neural tube anteriorly and become connected by axons (an example is shown in red), forming a favourite model for the study of axonal guidance (discussed further in Chapter 8).

Figure 2.5 Main stages of chick development. (a) A small portion of the shell of the hen’s egg is removed to expose the early embryo (yellow), which is enlarged and viewed from above in the central drawing. A perpendicular section through the embryo’s posterior part reveals the cells beneath the epiblast around Koller’s sickle. (b) Arrows show the movement of cells in the epiblast posteriorly and then inside the embryo to form the primitive streak and, at its anterior end, Hensen’s node. (c) Gastrulation (which starts about 30 hours after fertilization) involves the movement of epiblast cells inside the embryo and the formation of the three germ layers, shown in a slice through the embryo. This is similar in principle to gastrulation in

Xenopus

but takes place in a flatter embryo. Hensen’s node retreats from the anterior end where the head process forms, elongating the neural plate as it moves. (d) Later stages show elongation of the nervous system, the formation of somites, the development of brain structures and the growth of nerves from the CNS (numerous small arrows emerging from the neural tube). The branchial arches are a set of mesodermal structures on either side of the developing

pharynx

.

Figure 2.6 Early zebrafish development. (a) Gastrulation gets underway with the blastoderm covering about half of the blastula, about 5 hours after fertilization. The top diagram shows a slice through the embryo. The bottom diagram is the embryo turned 90

o

so that the part that will become dorsal is facing you. Cells move under the advancing lip of the epiblast (arrows) to form the hypoblast. Cells stream toward the part of the germ ring that will become dorsal, producing a structure called the

embryonic shield

. This is the equivalent of the dorsal lip of the blastopore in amphibians (Section 2.4.1) and Hensen’s node in the chick (Section 2.4.2). All of these structures have important organizing functions (discussed in Chapter 3). (b) A slice through the embryo, at a later stage than the upper one in (a). The dorsal epiblast thickens along the midline near the end of gastrulation, producing the neural plate (orange). Below this the hypoblast generates mesoderm (purple) and endoderm (grey). (c) The notochord and somites form from the mesoderm as the embryo, which now starts to look like a little fish, grows bigger and the yolk starts to disappear.

Figure 2.7 Development of the mouse nervous system from (a) the blastocyst stage to (d) neural tube closure. (b) Cell movements at gastrulation (about 6 days after fertilization), indicated by arrows, result in the formation of the primitive streak. To help understand this stage, imagine the flat chick embryo on the left in Figure 2.5(c) rolled up with its right‐ and left‐hand edges joined and the primitive streak on the inside: essentially, this would give the diagram on the right in (b) here. (c) The notochord and the neuroectoderm form anterior to the node. Look at the diagram on the right in Figure 2.5(c): imagine looking at it along the surface of the page from its left‐hand side, in which case the neuroectoderm (orange) would be to the left, the primitive streak to the right and the node in the middle, and if it is bent upwards at its ends it will have the equivalent layout of structures to those on the left of (c) and (d) here. The neural plate is divided into domains that will form the forebrain, midbrain and hindbrain before it folds from its lateral edges, as indicated by arrows. (d) Further development of the embryo: the neuroectoderm is folding to form the neural tube; somites are forming from mesoderm; non‐neural ectoderm is shaded yellow. The origin of the neural crest is shown in a section off to the right. Growth of the neural tube is accomplished by proliferation on the side nearest the lumen, followed by migration to and differentiation on the other side. Further development continues in Figure 2.9.

Figure 2.8 Scanning electron micrographs of neural tube closure in mouse embryos.

Figure 2.9 Development of the mouse brain (continued from Figure 2.7). (a) A reversal in the embryo’s shape is caused by movements indicated by thick arrows. The drawing at the top is a simplified version of the drawing in Figure 2.7(d). The drawing on the right shows more detail of formation of the forebrain, midbrain, hindbrain and optic vesicles. (b) Subsequent growth and specialization increase brain complexity. Growth is achieved through continuing cell divisions in the brain’s ventricular and subventricular zones followed by migration of these cells towards the outer surface of the brain where they differentiate (see also Figure 2.11 later and Figure 5.16 in Chapter 5 for more details). (c) The mature brain, containing maps of sensory surfaces such as the skin (more on this in Chapter 10, see Sections 10.1 and 10.2).

Figure 2.10 Comparison of mouse and human brain development. (a) The diagrams down the left‐hand side are all taken from Figures 2.7 and 2.9, where they are explained. They are lined up opposite diagrams of human embryos undergoing equivalent events. From the top to the bottom takes a little over a week in mice and a little under 4 weeks in humans. (b) We start with the blastocyst, which is more rounded in humans. The developing epiblast (the primitive ectoderm) remains above the hypoblast (the primitive endoderm) in humans. The human embryo develops from a bilaminar disc‐shaped structure at the interface of the epiblast and the hypoblast, which is a different arrangement to that in the mouse embryo. From above, the human embryonic disc looks a bit like the early chick embryo (compare part (c) here to part (c) of Figure 2.5). (c) The onset of gastrulation is marked by the formation of the primitive streak, which has the node in front of it (we have encountered equivalent structures before, in Figures 2.5 and 2.7). Epiblast cells move into the primitive streak to form the mesoderm (purple). (d) The neural plate (orange) develops in front of the node and folds to form the neural tube (curved arrows). (e) The neural tube expands at its anterior end to form the brain. In this diagram, TS stands for Theiler Stage and CS for Carnegie Stage (see text for explanation).

Figure 2.11 Comparison of mouse and human cerebral neocortical development. (a) Developing mouse neocortex, reproduced from Figure 2.9. The arrows follow a radial glial cell as it goes through division to produce a cell that migrates up its radial process. It might get off in the subventricular zone, where it would become an intermediate progenitor that would divide again to make two neurons, or it might go straight through to the cortical plate to become a neuron. (b) Developing primate cortex has an inner subventricular zone (ISVZ) similar to the subventricular zone of the mouse, but has a large outer subventricular zone (OSVZ) populated by outer radial glia that have long processes connecting only to the outer edge of the cortex. The OSVZ is thought to be a major source of the huge numbers of neurons generated in the neocortex of primates. The marginal zone is described in Section 2.4.4.

Figure 2.12 Variations in neurulation at different positions along the neural tube and in different species. (a) Primary neurulation, involving the rolling or folding of the neuroectoderm around a central lumen, occurs along much of the length of the neural plate in most vertebrate species. (b) In posterior regions, secondary neurulation involves the initial formation of a rod of cells, which then cavitates. (c) In some species of fish such as zebrafish, the neural tube forms by a thickening of the neural plate into a so‐called neural keel due to movement of neuroectodermal cells towards the midline (arrows); the keel becomes a rod of cells that then cavitates.

7

Figure 2.13 Diagram illustrating the paths taken by neural crest cells in the trunk of a chick embryo. (a) The embryo seen from the side is cut as shown by the broken line to give (b) a cross‐section through the trunk. Some neural crest cells migrate under the skin and form melanocytes (pigment cells); others migrate in more ventral directions to regions alongside the neural tube (where they form dorsal root ganglia), near to the aorta (where they form autonomic neurons, aortic plexuses and adrenal medulla) and around the gut (where they form the enteric nervous system).

Figure 2.14 The cranial placodes, shown on a diagram of a chick embryo. Similar to SOPs (sense organ precursors) in

Drosophila

(Figure 2.2 and Section 2.6.1), the cranial placodes arise in lateral ectoderm. The olfactory placodes, which generate the olfactory epithelium and its sensory neurons, and the lens placodes are the furthest anteriorly. Further posteriorly lie the trigeminal, geniculate, petrosal and nodose placodes, all of which contribute neurons to cranial sensory ganglia, and the otic placode, which generates the inner ear including its sensory epithelium and the auditory and vestibular ganglia.

Figure 2.15 (a) The ear in humans showing its major components, including the vestibulocochlear nerve (also known as the auditory or eighth

cranial nerve

) responsible for transmitting information on sound and balance to the brain. (b to f) Steps in the early formation of the inner ear studied in the chick, from the otic placode through the formation of a cup and eventually a fluid‐filled chamber with the vestibulocochlear ganglion of the auditory nerve developing on its medial side.

Chapter 03

Figure 3.1 Transplantation experiments show that the dorsal lip of the amphibian blastopore can induce the formation of neural tissue (parts (a) to (c) are adapted from Figure 2.4 in Chapter 2). (a) The dorsal lip of the blastopore (shown in red) is taken from embryo 1 (the donor) and placed at a different location in embryo 2 (the host), which now has a new donor lip and its own lip. (b) This results in the generation of an additional neural plate in embryo 2; while some cells in this structure come from the donor (marked in red), most come from the tissues of the host embryo, showing that the donor tissue has not itself generated the additional neural plate but has induced the host cells to do so. (c) The additional neural plate neurulates (Section 2.4 in Chapter2) to generate a second nervous system. (d) The second nervous system is, in fact, a component of an entire second body axis. (e) Photograph of transplantation in progress: the arrow shows where the graft (held on a fine needle) that has been cut from the donor embryo (held in fine forceps) will be placed in the host embryo. (f) Photograph of the resulting embryo. Photographs are reproduced by permission from Macmillan Publishers Ltd from De Robertis, E.M. (2006) Spemann’s organizer and self‐regulation in amphibian embryos.

Nat. Rev. Mol. Cell Biol

.,

7

, 296–302, copyright 2006.

Figure 3.2 Evidence that inhibition of bone morphogenetic proteins (BMPs) is important in neural induction. (a) Normal eggs generate embryos with BMP receptors that respond to BMPs, causing them to signal to the cell to become epidermal (see Figure 3.5 for more on the BMP receptor’s signalling mechanism). Cultured animal cap cells (dissected from the region above the broken red line) form epidermis if they are kept together, probably because they trap BMP between them and they are removed from the influence of BMP antagonists present in the embryo. They form neurons if they are

dissociated

, probably because the BMP molecules around them are washed away. However, if BMPs are added to these dissociated cells, they form epidermis. (b) If embryos are manipulated experimentally so that they produce truncated non‐functional receptors for BMPs, then the truncated receptors block the response to BMPs. If animal cap cells from these manipulated embryos are cultured intact then they form neurons and not epidermis. (Note: see the Conventions and Commonly used Abbreviations section at the start of the book for information on how the names of proteins and genes are written.)

Figure 3.3 The default model of neural induction based largely on work in amphibians: mesoderm dorsal to the blastopore secretes molecules that antagonize BMPs present in the part of the overlying ectoderm dorsal to the blastopore, converting it to neuroectoderm.

Figure 3.4 Neural induction in

Drosophila

: (a) the oocyte (egg) and early embryo have a concentration gradient of Dorsal protein along the dorsoventral axis (grey shading). High, intermediate and low levels of Dorsal relative to two thresholds activate different sets of genes in ventral, intermediate and dorsal regions, respectively. High levels of Dorsal result in the specification of mesoderm ventrally. Intermediate levels of Dorsal result in the specification of neural tissue (orange) ventrolaterally on each side of the embryo. This lateral neural tissue contains SOG protein, a homologue of vertebrate chordin, which antagonizes DPP protein, a homologue of vertebrate BMPs. (b) and (c) As the embryo develops, mesoderm moves into the embryo’s interior, lateral neural tissues move ventrally and fuse, and dorsal tissue forms epidermis (see also Figure 2.2 in Chapter 2).

Figure 3.5 (a)

BMP signalling

: there are two types of BMP receptor (BMPR), Types I and II. BMP molecules bind together forming dimers and then bind to Type II receptors, causing the formation of complexes comprising four receptor proteins, two Type I and two Type II (complexes of this type are called heterotetramers). This causes phosphorylation of the Type 1 receptors and the activation of two pathways: (i) the SMAD pathway and (ii) the mitogen‐activated protein

kinase

(MAPK) pathway (phosphorylation causing activation is indicated by small green circles). BMPR Type I phosphorylation leads to the phosphorylation of SMADs 1, 5 or 8, which then combine to form complexes with SMAD4 and enter the nucleus to regulate the transcription of specific target genes. It also causes the sequential phosphorylation of three molecules, each of which is a type of kinase: first, a MAPKK kinase (MAPKKK), second a MAPK kinase (MAPKK) and finally a MAP kinase (MAPK) itself. Phosphorylated MAPK proteins can enter the nucleus to phosphorylate and hence activate transcription factors controlling the transcription of specific target genes. There are many different MAPKKKs, MAPKKs and MAPKs. (b)

FGF signalling

: there are several FGF receptors (FGFRs), which are receptor tyrosine kinases. Binding of FGFs causes FGFRs to dimerize,

autophosphorylate

on their tyrosines and activate the MAPK pathway. As well as regulating gene transcription, MAPKs can also have cytoplasmic functions, one of which is to phosphorylate SMAD1 in a central rather than a terminal position (small red circle), thereby inhibiting this pathway (this inhibitory pathway is in the figure’s centre and is in red). During neural induction, the net effect of BMP antagonism and FGF signalling is inhibition of SMAD1 activity.

Figure 3.6

Wnt signalling

: Wnts can signal through several pathways and only the best‐known is illustrated here. This is the canonical Wnt pathway (now often referred to as the Wnt/β‐catenin pathway) and its major components are conserved from

C. elegans

to mammals. Wnt proteins bind to cell‐surface receptors of the Frizzled family causing them to activate intracellular Dishevelled family proteins. When Dishevelled is activated it inhibits a complex of proteins including axin, glycogen synthase kinase 3 (GSK‐3) and adenomatous polyposis coli (APC). In the absence of Wnt signalling, this complex causes phosphorylation (small green circle) of β‐catenin, which is then

degraded proteolytically

; inhibition of the complex prevents phosphorylation and degradation of β‐catenin, which enters the nucleus where it displaces a repressor protein and, in combination with proteins of the TCF/LEF (T‐cell factor/lymphoid enhancer‐binding factor) family, affects the transcription of specific target genes.

Figure 3.7 (a) In this experiment, mRNA from a

Zic

gene and a

Sox

gene have been detected using labelled RNA probes, a process called

in situ

hybridization (see Box 3.4 for a description of the method), revealing transcription of each gene in the neuroectoderm (purple stain indicated by small arrows; large arrows mark the blastopore). (b) In this experiment, mRNA from a

Sox

gene has been detected using the same method, revealing transcription of this gene in the intact epiblast of the chick embryo (purple stain), which is competent to develop as neural tissue (for a description of the anatomy of epiblast development, see Figure 2.5 in Chapter 2). (c) In this experiment, the same method has been used on sections cut through the neural plate to show transcription of a

Sox

gene restricted to the developing neural plate. (Photographs are from Mizuseki, K.

et al.

(1998) Xenopus

Zic

‐related‐1 and

Sox

‐2, two factors induced by chordin, have distinct activities in the initiation of neural induction.

Development

,

125

, 579–587; Rex, M.

et al

. (1997) Dynamic expression of chicken

Sox2

and

Sox3

genes in ectoderm induced to form neural tissue.

Dev. Dyn

.,

209

, 323–332; Wakamatsu, Y.

et al

. (2004) Multiple roles of

Sox2

, an HMG‐box transcription factor in avian neural crest development.

Dev. Dyn

.,

229

, 74–86.) (Note: see the Conventions and Commonly used Abbreviations section at the start of the book for information on how the names of proteins and genes are written.)

Figure 3.8 An example of a gene regulatory network in mouse stem cells: this network appears complicated but is a relatively simple example. It includes a

Sox

gene that regulates itself as well as other genes, some directly and others indirectly.

Sox

genes are discussed in this chapter; the other genes are not discussed and are only named here to illustrate the nature of the network. Those genes in yellow encode proteins that interact with proteins encoded by genes in pink; orange and green lines represent protein interactions. Blue and pink arrows indicate regulatory interactions of a protein with genes that are its downstream targets; arrows from a dashed ellipse indicate that the targets are regulated by all of the regulators inside the ellipse. Some regulators appear multiple times in the network to reduce the number of intersecting arrows. Reproduced from Zhou, Q, Chipperfield, H, Melton, D.A. and Wong, W.H. (2007) A gene regulatory network in mouse embryonic stem cells.

Proc. Natl Acad. Sci. USA

,

104

, 16438–16443. Copyright 2007 National Academy of Sciences, USA.

Figure 3.9 Control of gene function by microRNAs: some sequences in the genome are transcribed to give mRNAs, which are exported to the cytoplasm for translation. Other sequences can be transcribed to give pre‐microRNAs, which form hairpin structures from which pre‐microRNAs and eventually microRNAs are cut following export from the nucleus. Processing involves the RNAase enzymes Drosha in the nucleus and Dicer in the cytoplasm. Mature microRNAs combine with the RNA‐induced silencing complex (RISC) to bind to and inhibit or block translation from specific mRNAs. Typically, microRNAs bind in the 3′ untranslated regions of target mRNAs (i.e. the regions of mature mRNAs that do not code for proteins).

Chapter 04

Figure 4.1 In both

Drosophila

and vertebrates, patterning the nervous system involves subdivisions in the two axes of the neuroectoderm (or neural plate). Its long axis is the anteroposterior axis. The major partitions of the nervous system originate as regions along this axis, with the brain at the anterior end. At right angles to this is the bilaterally symmetrical dorsoventral axis. This axis comprises mirror‐image left and right halves separated by a midline (the dashed line). A = anterior; P = posterior; D = dorsal; V = ventral. In vertebrates, upon neurulation, the neural plate midline becomes the ventral apex of the neural tube while the lateral edges of the neural plate join to form the dorsal apex, defining its DV axis. In

Drosophila

, the neural midline defines the ventral pole of the whole embryo (see Figure 3.4 to visualize the neuroectoderm in the context of the DV axis of the embryo). The

Drosophila

neuroectoderm does not undergo neurulation (see Figure 2.2 in Chapter 2 for a reminder), but the AP and DV axes define the positional identities of the neuroblasts that delaminate from the neuroectoderm.

Figure 4.2 Regional specializations emerge sequentially in the anteroposterior (AP) axis of the chick neural tube (often known as the rostrocaudal axis; see Chapter 2, Section 2.2). After neurulation, the anterior or rostral end of the neural tube becomes progressively subdivided early on into the regions that will form the brain. More posteriorly, the neural tube is rather uniform and gives rise to the spinal cord. (See the Introduction for a description of these brain regions in humans; note that the tectum is the chick equivalent of the human superior colliculus.)

Figure 4.3 Establishing the regional expression of gap genes. (a) Expression gradients of the morphogens, Bicoid and Hunchback (HB), are formed in the syncytial blastoderm embryo. These switch on the gap genes in different regions along the AP axis, including

knirps

(

kni

),

tailless

(

tll

),

giant

(

gt

) and

Krüppel

(

Kr

). (b) Regulation of the

Krüppel

(

Kr

) gene by HB protein is shown as a graph representing protein levels along the AP axis of the embryo. HB inhibits

Kr

expression at high concentrations (near the anterior) but activates at moderate concentrations (in the middle). (c) Initially, the gap genes (e.g.

gt

and

Kr

) are expressed in broad overlapping domains. Cross‐repressive interactions sharpen the boundaries between these domains. (d) Image of an embryo in which

Kr

mRNA has been detected by

in situ

hybridization (see Box 3.4 in Chapter 3 for an explanation of this method). Image from Haecker

et al.

(2007)

Drosophila

brakeless interacts with atrophin and is required for tailless‐mediated transcriptional repression in early embryos.

PLoS Biology

,

5

, e145 (Creative Commons Attribution Licence).

Figure 4.4 (a) Pair‐rule and Hox genes are activated by Hunchback and the gap genes. Pair‐rule genes such as

eve

and

ftz

are activated in seven stripes, marking the segments. The gap genes and pair‐rule genes combine to activate Hox genes in different groups of segments. Note that by the stage of Hox gene activation, the embryo has reached the ‘germ‐band extended’ stage where the posterior end of the embryo loops over the rest of the body. (b) Image of

ftz

mRNA expression as detected by

in situ

hybridization in a fixed embryo. The nuclei of the embryo glow brightly due to a fluorescent stain while

ftz

mRNA is seen as seven dark stripes corresponding to odd‐numbered segments (image courtesy of Ilan Davis, University of Oxford). (c) In this embryo, seven different Hox gene mRNAs have been detected using multiple fluorescent probes (image from Lemons, D. and McGinnis, W. (2006) Genomic evolution of Hox gene clusters.

Science

,

313

, 1918–1922.)

Figure 4.5 Hox genes underlie AP patterning. (a) and (c) show Hox gene activity domains in

Drosophila

and mouse embryos, respectively, while (b) shows the chromosomal arrangements of Hox genes (in ‘gene clusters’) in

Drosophila

, mouse and

C. elegans

. The genes are colour‐coded to show their sequence homology across species. In

Drosophila

, the cluster has been split into two. In the mouse, the cluster has been duplicated twice, and there have been subsequent losses or duplications of certain genes in each cluster. The cluster is also conserved in a simplified form in

C. elegans

even though this organism is not segmented. The embryos are coloured to represent the regions that are most affected by each gene – note that the order is the same as the order of genes on the chromosome. Next to each embryo is depicted the approximate expression domains of the genes. Note how in the mouse the area affected by each gene is coincident with the anterior limit of its expression domain. Also note how the Hox genes do not participate in patterning of the extreme anterior head/brain in either

Drosophila

or the mouse. Redrawn with permission from Macmillan Publishers Ltd: Pearson, J.C., Lemons, D. and McGinnis, W. (2005) Modulating Hox gene functions during animal body patterning.

Nature Reviews Genetics

,

6

, 893–904, copyright 2005.

Figure 4.6 As the neural tube develops, subdivisions appear. (a) As shown in the chick embryo, the hindbrain is characterized by swellings called rhombomeres, while the spinal cord is divided into cervical, thoracic and lumbar regions (the arrows represent nerve exits). (b) As shown for the mouse, rhombomeres are segmental units, each of which exhibits a similar pattern of neurons later in development. In addition, there are differences between odd‐ and even‐numbered rhombomeres, as illustrated by the arrangement of cranial nerve exits (centre) and also the formation of neural crest cells (not illustrated here). Note that the chick has eight rhombomeres, but only seven are visible in the mouse. The spinal cord regions are strongly apparent in the human (right) in terms of nerve exits. (c) Different nerves emanate from the different regions of the spinal cord. For instance, the cervical region has nerves that innervate the arms (it is also called the brachial region) while the thoracic region has nerves that innervate the intercostal muscles of the ribs.

Figure 4.7 Hox gene expression domains in the mouse neural tube. (a) Expression patterns of Hox genes coincide with rhombomeres of the mouse hindbrain. (b) Expression of

Hoxb4

mRNA in the mouse neural tube posterior to r6, as detected by

in situ

hybridization. (c) Hox gene misexpression alters identity along the AP axis. Image from Brend

et al.

(2003) Multiple levels of transcriptional and post‐transcriptional regulation are required to define the domain of

Hoxb4

expression.

Development

,

130

, 2717–2728.

Figure 4.8 In

Xenopus,

AP patterning of the neural plate is induced by the mesoderm. The black bracket denotes the AP axis of the neuroectoderm overlying the mesoderm. Induction signals (red) promote neural induction, while graded posteriorizing signals that arise from the later‐formed posterior mesoderm (blue) are required to confer posterior neural identity. In the absence of posteriorizing signals (in the anterior), the neural tissue adopts an anterior character. This posteriorizing signal may be produced either by the posterior mesoderm (short blue arrows) or by the late‐stage dorsal lip of the blastopore itself (curved blue arrow), or both. More recently, it has been found that a separate anterior signal is required for forebrain induction (green).

Figure 4.9 Regions in the anterior brain are defined by early domains of transcription factor expression in the neural plate. The bars represent the AP extent of expression for pairs of genes, such as

Six3

and

Irx3

. Each pair of genes defines a boundary between two future brain regions, as becomes visible later in development (on the right). Thus, the interface of

Otx2

and

Gbx2

expression defines the midbrain–hindbrain boundary. In a manner reminiscent of

Drosophila

gap genes, the genes of each pair are initially activated in broad, somewhat overlapping domains, and then their borders are sharpened by mutual repression.

Figure 4.10 Neuroblast identities in

Drosophila.

In the image is shown a ventral view of a

Drosophila

embryo in which antibody has been used to stain the neuroblasts. The dotted lines represent the ventral midline and the segmental boundaries. In each segment, neuroblasts arise from the neuroectodermal cells in stereotyped locations and have unique identities. They appear in several waves, and only the neuroblasts of the first waves are shown.

Figure 4.11 Local AP patterning in

Drosophila

segments. The embryo image shows that cells in the posterior halves of each segment express one of the segmentation genes,

engrailed

(EN), which encodes a transcription factor.

Engrailed

triggers local interactions across the boundary between anterior and posterior cells, which leads to the formation of stripes of cells expressing genes encoding two signalling molecules,

wg

and

hh

. Whilst their mRNAs are expressed in narrow abutting stripes, the secreted protein products diffuse through the rest of the segment, leading ultimately to activation of other segmentation genes in 1–2 cell‐wide stripes along the AP axis of the segment.

Figure 4.12 Activation of the

Drosophila

columnar genes. Morphogen gradients of DL and DPP expression combine to activate the expression of

vnd

,

ind

and

msh

in different domains (‘columns’) along the DV axis in the neuroectoderm. Lower left shows a schematic cross‐section through the columns, whereas on the right is a lateral view of a fixed embryo showing expression columns of mRNAs of

dpp, msh, ind

and

vnd

running from top to bottom. Below right shows some of the interactions that regulate the columnar genes;

vnd

is activated by high levels of DL but is very sensitive to inhibition by DPP. This confines its expression close to the ventral midline, while

ind

is activated by DL and is less sensitive to DPP inhibition, but is inhibited near the midline by

vnd

, ensuring that its expression is restricted to the middle column abutting

vnd

. In turn

ind

inhibits

msh

, but is itself more sensitive to DPP inhibition than is

msh

. As a result

msh

is expressed in the most dorsal (lateral) part of the neuroectoderm. Image from Kosman, D.

et al.

(2004) Multiplex detection of RNA expression in

Drosophila

embryos.

Science

,

305

, 846.

Figure 4.13 Patterning of neuroblast identities in

Drosophila

. This diagram brings together the functions of the AP and DV patterning genes shown in Figures 4.10 and 4.12. The cartoons show a single segment, with the neuroblasts as larger pale cells on the neuroectoderm (therefore this is a view

from inside

the embryo). Each neuroblast has a unique identity. These identities are assigned according to location by combined information from the AP and DV systems of patterning in the neuroectoderm. For example, the 5–2 neuroblasts express the unique combination of

vnd

and

gsb

. Similarly, the intersection of

wg

and

ind

defines neuroblast 4–2.

Figure 4.14 The boundaries between brain regions become new signalling centres for local patterning. (a) Two signalling centres are illustrated here: the zona limitans intrathalamica (ZLI) forms in the diencephalon and the isthmic organizer (ISO) forms at the midbrain–hindbrain boundary. (b) The ZLI forms at the boundary between

Six3

and

Irx3

expression domains. SHH signalling from the ZLI locally induces the expression of the

Dlx2

anteriorly and

Gbx2

posteriorly, leading to the specification of the prethalamus and thalamus, respectively. This difference in cellular response to SHH arises because of the expression of different transcription factor genes on each side of the ZLI (

Six3

and

Irx3

), which cause the SHH signal to be interpreted differently. (c) At the ISO, FGF8 diffuses into the midbrain and first rhombomere, where it activates the

En1

and

En2

genes. This results in the formation of the tectum and cerebellum. The image shows

Fgf8

mRNA expression in a

Xenopus tropicalis

embryo. The stripe of expression at the ISO is indicated by the arrow. Image from Lea, R.,

et al.

(2009) Temporal and spatial expression of FGF ligands and receptors during

Xenopus

development.

Developmental Dynamics

,

238

, 1467–1479.

Figure 4.15 Neural organization within the DV axis of the spinal cord. The photograph (lower left) shows a cross‐section through the mature cat spinal cord revealing its bilateral symmetry and dorsoventral (DV) structure. Motor axons exit ventrally while sensory axons from the dorsal root ganglia enter dorsally. Patterning begins in the mediolateral plane of the neural plate (top left), which becomes the DV axis of the neural tube (right). Different classes of neurons are produced at different levels along the DV axis.

Figure 4.16 Along the DV axis for the spinal cord, at least 11 domains of neural progenitor cells have been distinguished (from p3 ventrally to pd1 dorsally). Upon neurogenesis, these progenitor pools generate distinct classes of neurons, distinguishable by their expression of marker genes, morphology and contacts. For example, pMN contains progenitors of the motor neurons whereas p2 gives rise to the V2 interneurons. In addition, specialized glia form the floor plate (FP) and roof plate (RP) at the ventral and dorsal points of the neural tube, respectively. These will later be seen to be crucial for DV patterning.

Figure 4.17 Patterning of the DV axis of the vertebrate neuroectoderm begins in the neural plate. (a) Members of the BMP family (mainly BMP4 and 7) diffuse in from the lateral ectoderm to induce the formation of neural crest precursors and roof plate cells (RP). (b) SHH signalling from the notochord induces the formation of floor plate cells (FP) at the midline. (c) After neural tube closure, the floor plate and roof plate become new signalling centres for SHH and BMP respectively, as shown in (d).

Figure 4.18 The notochord induces ventral neural tube patterning. In the chick, transplantation of an ectopic notochord next to the lateral neural plate leads to the appearance in the neural tube of a second floor plate (FP) flanked by motor neurons (MN). Conversely, when the notochord is surgically removed, these ventral cell types fail to appear in the neural tube. The conclusion is that the notochord is both

necessary

and

sufficient

for the induction of ventral types in the chick neural tube. The direct effect of notochord signalling is to induce the floor plate. The induction of other ventral fates (e.g. motor neurons) occurs secondarily via signalling from the floor plate.

Figure 4.19 SHH signalling from the floorplate patterns the neural tube as a morphogen. (a) The concentration‐dependent effect of SHH on neural tube differentiation was demonstrated in chick explant experiments. Pieces of naïve neural plate were cultured

in vitro

in media containing different concentrations of SHH. Some 24 hours later, the types of neurons produced in each explant varied in a manner that correlated with the

in vivo

proximity of the neurons to the floor plate. In other words, explants treated with high concentrations of SHH produced neuronal subtypes that are normally found close to the floor plate, while explants exposed to lower SHH concentrations made neurons of a type normally found in more dorsal domains, further from the floor plate. (b) SHH defines neuronal progenitor domains via the activation and inhibition of homeodomain transcription factors. One group of factors (Class II) is activated at different threshold concentrations of SHH, leading to their expression in ventral regions of differing dorsal extent. Conversely, Class I genes are repressed at different SHH concentrations, leading to their expression in dorsal regions of differing ventral extent. Unique combinations of these factors thereby determine the identity of each progenitor domain. The expression of several of these genes is shown in the image on the bottom left. BMP has a similar role to SHH in the dorsal neural tube. Neural tube image courtesy of Vanessa Ribes, Institut Jacques Monod, France.

Chapter 05

Figure 5.1 In

Drosophila

, neuroblasts and sense organ precursors (SOPs) generate the CNS and PNS, respectively. (a) Neuroblasts and SOPs are ectodermal cells that leave the epithelial layer. (b) These cells can be detected in the developing embryo (seen here in side view) using antibodies against cell‐specific proteins that are often described as being ‘markers’. Note the segmentally repeated pattern of SOPs. (c) In an older embryo, these cells have divided and some of their progeny have differentiated into the neurons of the CNS and PNS. The CNS forms a ventral nerve cord (see Figure 2.2 in Chapter 2).

Figure 5.2 Proneural genes are expressed in clusters of cells in the ectoderm. (a) Shown here is a cartoon of one such cluster (light green) in the lateral ectoderm. The cells of each cluster have the potential (competence) to become SOPs. A single cell attains this fate (dark green) and it then inhibits the remainder of the cells (lateral inhibition). The same process produces neuroblasts in the neuroectoderm. (b) Expression pattern of a proneural gene (

scute

mRNA) in the developing wing of the fly. A number of proneural clusters are visible (e.g. black arrow). Expression is also observed in individual cells (white arrow), which represent already committed SOPs (Stage 3 of panel (a)) and

scute

mRNA was detected by

in situ

hybridization (Chapter 3, Box 3.4). (c) Each SOP will form a sensory bristle. These bristles can be seen projecting from the back of the fly. (d) A

scute

mutant fly, showing lack of bristles due to failure in SOP commitment. This mutation was first described in the 1930s. (e) When the

scute

gene is ectopically expressed throughout the ectoderm (using the GAL4/UAS system, Chapter 1, Box 1.2), many more ectodermal cells become SOPs and the fly becomes very bristly.

Figure 5.3 Initially, Notch signalling between all cells in the proneural cluster results in mutual inhibition (1). Proneural factors and their HES antagonists are co‐expressed at this stage, and all cells maintain their neural competence. Subsequently this shifts to lateral inhibition in which just one cell signals to the rest of the cluster (2). This cell (the SOP) expresses only proneural factor and the surrounding cells express only HES. The graph represents the transition in gene expression that occurs in the SOP.

Figure 5.4 Sensory cell formation in the mouse inner ear. Cells in the otic epithelium differentiate as sensory ‘hair cells’, which mediate auditory sensory transduction and non‐sensory support cells. A proneural gene,

Atoh1

, is expressed in the otic epithelium and promotes commitment to hair cell fate. In mouse knock‐out mutants of

Atoh1

, no hair cells are formed and all cells become support cells by default.

Jagged1

encodes a

Notch

ligand. In

Jagged1

mutants, some support cells differentiate as extra hair cells. This is due to a reduction of lateral inhibition. Loss of inner ear hair cells is a major cause of hearing loss – through trauma or age. There is much interest in attempting to reactivate

Atoh1

expression in the inner ear to provoke the formation of new hair cells in order to restore hearing. Image provided by Elizabeth Orton and Karen Steel, Sanger Institute, Cambridge, UK.

Figure 5.5 A generalized view of neurogenesis in the vertebrate neural tube. The neuroepithelium consists of neural stem cells called radial glia (brown), which stretch across the epithelium (1 in the lower cartoon). These undergo repeated divisions in the ventricular zone (2). Some of the daughter cells leave the cell cycle to become post‐mitotic neural precursors (3). These migrate along the remaining radial glia (4) into the mantle zone, where they differentiate (5). The image (upper right) is from a slice through a mouse neural tube. The nuclei of radial glia (magenta) are in the ventricular zone on each side, while maturing neurons (green) have moved out into the mantle zone. Image adapted by permission from Macmillan Publishers Ltd: Petersen, P.H.

et al

. (2002) Progenitor cell maintenance requires

numb

and

numblike

during mouse neurogenesis.

Nature

,

419

, 929–934, copyright 2002.

Figure 5.6 The progenitor cells (radial glial cells) in the ventricular zone of the neural tube express proneural factors (including Ascl1, Notch ligands (including Delta1) and HES antagonists (including Hes5)). This is apparent in the

in situ

hybridizations of sections of the developing mouse cerebellum on the left. In each case, mRNA (purple stain) is confined to the lower ventricular zone (VZ) rather than the upper zone of migrating neural precursor cells and differentiating neurons. There is strong evidence that expression of these genes oscillates within radial glial cells. This is represented in the right‐hand schematic, in which the temporal oscillations within a radial glial cell are represented before it commits to become a neural precursor cell. This is equivalent to the mutual inhibition phase in

Drosophila

neurogenesis. Over time, some of the cells cease oscillations and express a sustained high level of proneural genes, thereby becoming neural precursor cells. The proneural factor function is then able to commit these cells to exit the cell cycle, migrate and differentiate. Compare and contrast this to

Drosophila

SOP commitment in Figure 5.3. The

in situ

hybridization image is adapted from Machold, R.P., Kitell, D.J. and Fishell, G.J. (2007)

Neural Development

,

2

, 5.

Figure 5.7 Neurogenesis is triggered by different proneural factors in different regions of the nervous system. This contributes to regional differences in the production of neuronal subtypes. This example shows the expression of mRNAs for different proneural genes in distinct progenitor domains along the DV axis of the mouse neural tube. Refer back to Figure 4.19 in Chapter 4 for an explanation of the neural tube schematic. Images reproduced from Muroyama, Y.

et al.

(2002) Wnt signalling plays an essential role in neuronal specification of the dorsal spinal cord.

Genes Dev

.,

16

, 548–553.

Figure 5.8 Different combinations of bHLH (blue) and homeodomain (red) transcription factors define different classes of neuron and glial cell in the vertebrate retina. Evidence that these factors are important for subtype identity comes from analysis of mouse mutants and experimentally induced co‐expression of different bHLH and homeodomain protein combinations in retinal

explant

tissues grown in culture.

Figure 5.9 Examples of asymmetric cell division in the nervous system. (a) Such divisions can generate cellular diversity, as seen in the

Drosophila

PNS. Each SOP divides asymmetrically to give two intermediate cells (pIIa and pIIb), which divide asymmetrically again to give the sensory neuron and three support cells, which form a sensory bristle on the surface of the fly. The three support cells comprise: a glial cell that ensheathes the neuron, a cell that secretes the bristle shaft and a cell that makes the socket in which the shaft sits. Asymmetric cell division can also be a self‐renewal or stem cell type, and this is characteristic of progenitor cells in the CNS of both

Drosophila

and vertebrates. In the

Drosophila

CNS (b), neuroblasts undergo repeated asymmetric divisions. At each division one daughter retains the neuroblast’s characteristics while the second becomes a ganglion mother cell, which has very limited potential to divide further. It divides only once more to yield two neurons. (c) In the vertebrate neural tube, radial glial cells undergo repeated asymmetric cell divisions to generate neural precursors. Progenitors are theoretically able to carry on dividing indefinitely – they have high proliferative potential. If different types of neural cell are produced at different divisions, then the progenitor is also multipotent (e.g. the production of neurons and then astrocytes by the pMN radial glia). These are both key characteristics of stem cells.

Figure 5.10 The gene

numb

is required for asymmetric cell division in

Drosophila

sense organs. The four cells of each sensory bristle arise from asymmetric division. Outwardly, only the shaft and socket are visible (unfilled and filled arrowheads, respectively, indicate one example on this close‐up of a fly’s head). Mutations that affect the cell divisions are readily visible in genetic screens by virtue of changes in external appearance of the bristles. In the

numb

mutant fly (

numb

15

), groups of four socket cells are observed because the divisions have become symmetric. Several examples of socket clusters are indicated on the image but note that not all bristles are affected in this particular fly. Images adapted from Berdnik, D., Török, T., González‐Gaitán, M. and Knoblich, J.A. (2002) The endocytic protein α‐adaptin is required for numb‐mediated asymmetric cell division in

Drosophila

.

Dev. Cell

,

3

, 221–231, with permission from Elsevier.