Kaufman’s Atlas of Mouse Development Supplement: With Coronal Sections

Kaufman's Atlas of Mouse Development: With Coronal Sections continues the stellar reputation of the original Atlas by providing updated, in-depth anatomical content and morphological views of organ systems.The publication offers written descriptions of the developmental origins of the organ systems alongside high-resolution images for needed visualization of developmental processes. Matt Kaufman himself has annotated the coronal images in the same clear, meticulous style of the original Atlas. Kaufman's Atlas of Mouse Development: With Coronal Sections follows the original Atlas as a continuation of the standard in the field for developmental biologists and researchers across biological and biomedical sciences studying mouse development.

• Provides high-resolution images for best visualization of key developmental processes and structures
• Offers in-depth anatomy and morphological views of organ systems
• Written descriptions convey developmental origins of the organ systems

• Language: English
• Publisher: Academic Press
• Release date: Sep 23, 2015
• ISBN: 9780128009130

Book preview

USA

1

Introduction

Richard Baldock¹, Jonathan Bard², Duncan R. Davidson¹ and Gillian Morriss-Kay²,    ¹MRC Human Genetics Unit, IGMM, The University of Edinburgh, Edinburgh, UK,    ²Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK

The original Atlas of Mouse Development by Mathew Kaufman was published in 1989. In this chapter, we introduce the supplement to this definitive atlas, bringing the description of mouse embryo development up to date in light of the new knowledge and understanding of the molecular underpinnings.

Keywords

Atlas; mouse embryo development; coronal images

It is a now a little over 25 years since Matthew Kaufman published his comprehensive Atlas of Mouse Development. This was around the time that molecular biologists realized that if they were to understand the developmental roles of newly cloned genes, they would need to know mouse developmental anatomy in far greater detail than they had expected. As a result, many laboratories began working on mouse development, and this Atlas became a necessary presence in all of them. Since its publication our knowledge of mouse developmental anatomy has increased almost beyond measure, and it is now time to pull some of these advances together in a volume that is a supplement to the original Atlas.

The driving force behind this new endeavor has been the need to understand the functional roles of the genomic sequences and proteins whose cooperative activity drives embryogenesis; a major outcome of all this work is that we now understand mouse developmental anatomy and its genetic underpinnings much better than we did. These advances have come from gene function and transgenic mouse studies, combined with the use of novel microscopy techniques, which are often computer assisted. From the point of view of the developmental anatomist, however, the major advance has been in the understanding of cell lineage: the use of transgenic lineage markers has revolutionized the subject. The original Atlas no longer summarizes the state of the art.

When MK and the editors started to plan this book, the intention was clear: We wanted to include the advances made in every anatomical system in the mouse embryo over the past quarter century. The main emphasis would be on how microscopy and lineage studies had improved our understanding of the anatomy, but it would not have been possible (nor would it have been sensible) to do this without at least summarizing some of the key molecular drivers of tissue change. This was the anatomical brief that we provided to our authors, and their chapters comprise the bulk of this book. We are grateful to all of them for the work that they have put into their contributions.

There is, however, a further aspect of mouse developmental anatomy that was barely considered when the original Atlas was published. This was the production of computational resources to handle mouse anatomy and associated data. The first of these resources, made available in 1996, connected developmental anatomy and gene expression in a standard textual database (GXD supported by the Jackson laboratory). Because gene expression in embryos frequently fails to respect tissue boundaries, a graphical database that represented gene-expression data and anatomy pictorially was implemented a few years later (EMAP, supported by the MRC Human Genetics Unit). More recently, progress has been made in a much harder task-linking mutant gene expression to abnormal anatomy (the phenotyping problem). The editors felt that the supplement would not do justice to the current state of the art if this area was not included, therefore the last section of this supplement covers these aspects of computational anatomy.

Producing this supplement has given us the opportunity to stand back from the original Atlas to see if any other changes need to be included in this book, and we felt that two were justified. The first, suggested by an Atlas-user survey, was that the original Atlas lacked sufficient sets of coronal sections, the ones particularly useful for developmental neuroanatomy; this supplement now includes such sections. Second, we have altered the terminology for age from dpc (days post coitum) to E because the latter is now more widely used. In current usage, E means the same as dpc; that is, E is the number of days or fractions thereof since mating. Thus, the text of the supplement is directly comparable with the original Atlas, which used dpc throughout. Generally, the time of mating is assumed to be at midnight but is rarely known precisely, and, more importantly, neither is the time of implantation of individual embryos. Thus, unless matings are carefully timed, E or dpc simply represent the approximate embryonic age of individual embryos usually to within about 8 hours. It is also well known to embryologists that embryos develop at variable rates, as, indeed, do different organs within the same embryo. Thus, embryonic age, however precisely determined, is not a consistent indicator of the actual stage of development of individual embryos, even in the same litter. This must be borne in mind in any practical application of age-related information given in this book. It is becoming clear that the anatomy of the early implantation stages is strain dependent, and we are grateful to Kirstie Lawson for her detailed comparisons and revisions of Theiler stages 6–11, which we are sure will be useful to everyone working on early embryos.

This book does not end with the printed version: Readers should be aware that a considerable amount of the illustrations and associated data are available at the book’s online site (http://booksite.elsevier.com/9780128000434). The resolution of the printed images of coronal sections, for example, is limited by print technology. A low enough magnification to include images of whole sections of late embryos is effectively ~10 μm. Because web-based images can use digital tiling technology, users can seamlessly enlarge pictures so that the resolution is only limited by the resolving power of the lens used to digitize that image (<1 μm). The contributing authors were also given the opportunity to post any supplementary data on the book’s website, and readers may well enjoy, for example, videos of exploration of the embryonic heart.

In collaboration with Academic Press/Elsevier, new images of the histological sections of the original Atlas have been captured at high resolution (×20 objective) and in color to be made freely available to the community from the eMouseAtlas website (www.emouseatlas.org). The images include the original annotations and are indexed with the original plate numbering. In addition, the agreement has been extended to allow the new coronal images to be made available at the same site. We are grateful to the Editors at Elsevier, Janice Audet and Mary Preap, who have helped guide this agreement through from the original concept developed when MK was alive and for sharing the value of making the image data freely available to the community.

Finally, we are conscious of the debt that the field owes to Matt Kaufman, a key figure in the study of mouse developmental anatomy and whose work underpinned a great deal of experimental and computational research. He was a good friend to us all.

2

Coronal Sections

David J. Price¹, Elizabeth Graham², Julie Moss², Chris Armit² and Richard Baldock²,    ¹Centre for Integrative Physiology, The University of Edinburgh, Edinburgh, UK,    ²MRC Human Genetics Unit, IGMM, The University of Edinburgh, Edinburgh, UK

The original The Atlas of Mouse Development by Matt Kaufman (MK) included transverse and sagittal sections at multiple stages of development, but only two plates of coronal sections (stages E14.5 (plate 34) and E16.5 (plate 39)). When a revised version of the Atlas was suggested, a survey of users recommended that the series of coronal sections should be extended to include additional stages of development, particularly the brain at E11, E11.5, E12.5, E13.5, and E15.5. This chapter presents these new coronal sections, together with some new sections to extend the E14.5 images provided in the original Atlas. These sections include some original annotations from MK that are supplemented by more detailed brain annotations.

Keywords

Atlas; ontology; Matt Kaufman; mouse development; histology plates; coronal sections

Introduction

The Atlas of Mouse Development by Matt Kaufman (MK; Kaufman, 1994) is the defining standard for mouse development and remains an essential text for developmental biologists and biomedical scientists. The Atlas has a series of plates of histological sections through embryos that have been selected to represent each Theiler stage of development. Each image within a plate is carefully annotated with the tissues visible in each section clearly labeled. The original Atlas includes two embryos sectioned coronally at E14.5 and E16.5. When users were asked what should be added to a revised edition of the Atlas, there was a clear demand for additional coronal sections to complete the developmental series in order to make the development of the embryonic brain easier for the research community to understand.

This chapter provides these new coronal sections as a series of plates that can be used in conjunction with the original Atlas and the chapters of this volume. Before ceasing work on the revised Atlas (now presented as this Atlas Supplement), MK selected the embryos, specified the sections that were most illustrative and should be used for these new plates, and provided some preliminary annotation. Here, we have taken that selection and extended the annotation.

The printed version of these sections follows the style of the Atlas with each image labeled with lines and numbers, with each number linked to an annotation on the same two-page view. These coronal sections only show the head-region of the section so that the brain is printed at the highest possible resolution. We also provide the fully annotated, high-resolution images online within a zoom style viewer (see www.emouseatlas.org).

While MK’s annotations in the original Atlas were given as short descriptions, the plates here use the anatomical terms defined by the Edinburgh Mouse Atlas Project (EMAP) mouse developmental anatomy ontology (see Chapter 23 and Hayamizu et al., 2013). This ontology uses a controlled vocabulary, and this has been extended to include some new annotations shown here. An advantage of using the EMAP ontology descriptions for labeling tissues is that the online visualization of each section is directly linked (through the hidden ontology term IDs) to other database resources, and this allows a user to, for example, directly query gene-expression in any anatomical component. The EMAP ontology also describes the stage range for specific anatomical components, and this allows researchers to follow, in order of appearance, the various anatomical components of the developing mouse embryo. We hope that the use of EMAP ontology terms will allow researchers to explore embryonic development in new ways.

Methods

To generate the new coronal sections, we used C57Bl6J×CBA F1 embryos that were collected, fixed, and washed in the same way as those for The Atlas of Mouse Development (Kaufman, 1994). Embryos were dehydrated through an ethanol gradient, cleared in xylene, then infiltrated and embedded in paraffin wax (56°C melting point). They were then embedded on their side on a layer of setting wax in molds of suitable size, and this allowed the embryo to be oriented for coronal sectioning. The trimmed wax block was mounted on a wooden chuck and attached to the microtome (Leica RM2264 motorized with retraction and microscope attachment). Starting at the front of the embryo, serial sections were cut at 7 μm, using the motorized setting, directly onto distilled water (room temperature) in a trough specifically designed to keep section loss to a minimum (this was manufactured in house to fit the disposable knife holder).

Strings of sections were removed from the trough on a glass slide and carefully introduced into a floating-out bath at 42–45°C where they were allowed to stretch out for a few seconds. Time and temperature were carefully monitored to make the stretching consistent. The sections were then aligned on a cleaned glass slide, withdrawn from the bath, and dried vertically at room temperature overnight. Before staining, the slides were placed in an oven at 60°C for 2 h, dewaxed in xylene, and rehydrated through an ethanol gradient to water. The sections were stained in Mayer’s Haemalum for 5 min, and aqueous Eosin Y for 3 min, then dehydrated, cleared in xylene, and mounted in DPX (coverslip thickness No. 1.5: 0.13–0.17 mm).

Once the mountant had completely set, the slides were digitized using the Olympus DotSlide, which capture a complete image of each section by automated scanning over the area of tissue. To optimize the digital image focus, we manually set focus points for each section then the scanning was performed with either a ×10 or a ×20 objective resulting in image pixel resolutions of 0.32 and 0.16 μm, respectively.

The DotSlide images were output to tiff format and converted to multiresolution tiled pyramidal format using the VIPS image processing software (Martinez and Cupitt, 2005). This provides the required tiled format for a zoom-viewer interface using the IIP3D server (Husz et al., 2012). In addition, we developed a bespoke MySQL database to hold the plate information and annotations. The interface used to deliver the coronal image can be configured to allow editorial placement of points associated with any ontology term.

Results

The following table lists the stages covered by these additional coronal sections and some additional data for each embryo.

Discussion

These coronal sections extend the series of coronal sections provided in the original The Atlas of Mouse Development. Low-resolution images are available in the Supplement, while high-resolution versions are available from the eMouseAtlas Web site (www.emouseatlas.org) and these show the annotations identified by the matching number in the plates shown here. The same images available at much higher resolution on the Internet are an open-access community resource available freely for scientific and educational purposes. In this context, we hope to invite expert developmental anatomists to contribute to the resource to enhance the annotations available and so increase our knowledge of developmental anatomy.

Acknowledgments

We thank Allyson Ross for preparing and sectioning the embryos and Nick Burton for developing the editing viewer that enabled anatomical terms to be selected and annotations to be placed on images.

References

1. Hayamizu TF, Wicks MN, Davidson DR, Burger A, Ringwald M, Baldock RA. EMAP/EMAPA ontology of mouse developmental anatomy: 2013 update. J Biomed Semantics. 2013;4:15 In: http://dx.doi.org/10.1186/2041-1480-4-15.

2. Husz ZL, Burton N, Hill W, Milaev N, Baldock RA. Web tools for large-scale 3D biological images and atlases. BMC Bioinformatics. 2012;13:122 In: http://dx.doi.org/10.1186/1471-2105-13-122.

3. Kaufman MH. The Atlas of Mouse Development revised ed. London: Academic Press; 1994.

4. Martinez, K., Cupitt, J., 2005. VIPS—a highly tuned image processing software architecture. In: Conference Proceedings, IEEE International Conference on Image Processing 2, Genova, pp. 574–577.

3

A Revised Staging of Mouse Development Before Organogenesis

Kirstie A. Lawson¹ and Valerie Wilson²,    ¹MRC Human Genetics Unit, IGMM, University of Edinburgh, Edinburgh, UK,    ²School of Biological Sciences, MRC Centre for Regenerative Medicine, University of Edinburgh, Edinburgh, UK

Theiler’s system of staging mouse embryos, published in 1972, is based on 12-hour and 24-hour periods, but it has two limitations. First, the asynchrony between littermates and between litters of the same nominal gestational age makes it of limited use for early postimplantation development, when morphogenesis is rapid. Second, it does not capture the details of early mouse development that have been discovered over the past 40 or so years. This chapter amplifies the original Theiler stages 1–12 (E0–E8) by subdividing them into 21 stages on the basis of new morphological criteria. Each stage has a descriptive name as close as possible to those in existing systems except that the term Pre-Headfold replaces Neural Plate. The individual stage descriptions are supported by scaled, labeled schematic figures and accompanied by examples of gene expression that mark and define structures and cell populations at specific stages.

Keywords

Allantoic bud; early postimplantation stages; litter variation; molecular markers for early development; mouse embryo staging; preimplantation stages; Theiler stages

Introduction

Theiler’s staging system (Theiler, 1972), based on daily or half-day intervals, was used throughout the Atlas of Mouse Development (Kaufman, 1992). While giving uniformity to the whole of prenatal development, it was, even then, unsatisfactory with regard to early postimplantation development up to the onset of somitogenesis (TS12, E8). The main reason is litter variation, both between litters (the moment of conception, E0 or 0 dpc, is set by convention in the middle of the dark period before vaginal plug detection) and within litters (principally, variation in the time of fertilization and of implantation). The effect of such differences in timing on assigning a meaningful developmental stage to individual embryos was exacerbated by the rapid growth and morphogenesis during the approximately 24 hours of gastrulation (~E6.5 to E7.5). In addition, work on peri-implantation and pregastrulation events, such as work on the role of the anterior visceral endoderm (AVE) in defining the anteroposterior embryonic axis and the subsequent change in shape with respect to the same axis (reviewed in Rivera-Pérez and Hadjantonakis, 2014), has revealed morphological detail that justifies an upgrading of the staging system.

The purpose of this chapter is to provide a more detailed staging system up to TS12, by creating TS substages, consistent with the rest of the Atlas, and with descriptive labels as far as possible in accordance with existing systems (e.g., Fujinaga et al., 1992; Downs and Davies, 1993; Rivera-Pérez et al., 2010; e-MouseAtlas [www.emouseatlas.org]). The revised system from zygote to early somitogenesis contains 21 stages, with descriptive names. The stages are illustrated with scaled, labeled drawings of sagittal sections accompanied by images of whole embryos stained for molecular markers representative of the stage. A tabulated summary of the revised stages with emphasis on distinguishing features is provided in Table 3.1.

Table 3.1

The Revised Staging System

Methods

The embryos used for the schematic figures of the stages were [C57BL6 × CBA] F1 or F2 (hereafter referred to as BL6CBA) for postimplantation stages from TS7d, except for TS9a, TS10a, and TS10c/11a, which were CD1 (Rivera-Pérez and Magnuson, 2003; Pereira et al., 2011). Midsagittal histological sections or midsagittal slices of three-dimensional reconstructions from 2 μm sections (see the e-MouseAtlas website or material prepared for emap) were traced to show cell layers. A paraxial section was underlaid at the headfold and early somite stages. In order to check that the sizes of the embryos used were representative for the stage, the total length (from the junction with Reichert’s membrane to the distal tip), and the embryonic height and width were compared with these dimensions in freshly dissected, staged E5.5–E8 BL6CBA embryos (N = 387). The scale of the drawing was adjusted if all three dimensions differed by more than one standard deviation (SD) from the stage mean ×0.8 in the freshly dissected series. Adjustment was only necessary for TS7c (nondehydrated embryo; Rivera-Pérez and Magnuson, 2003), TS8 (emap; large embryo for stage), TS10a and TS10c/11a (glycol methacrylate sections causing 11% expansion; Pereira et al., 2011) and TS11c (nondehydrated embryo). No adjustments were made to TS1-7b (nondehydrated embryos).

The images of marker gene expression were taken from the following (with permission, and with annotations digitally removed): Bmp4 (Lawson et al., 1999), Hesx1 (Thomas and Beddington, 1996; Cajal et al., 2012), Lefty1 (Takaoka et al., 2007), Nanog (Chambers et al. (2003), Nkx2.5 (Liberatore et al., 2000; Briegel, 2001; Y. Huang, unpublished data), Noto (Plouhinec et al., 2004), Runx1 (Samokhvalov, 2007), Sox2 (Wood & Episkopou, 1999; Avilion et al., 2003; Cajal et al., 2012), T(bra), (Herrmann, 1991; Rivera-Pérez & Magnuson, 2005; F. Wong and V. Wilson unpublished data), Tcf15 (A. Pegg and V. Wilson, unpublished data), Tcfap2c (Cajal et al., 2012).

Overview of the Revised Staging System, TS1–12

A summary of the revised staging with emphasis on distinguishing features is shown in Table 3.1. Timing, ranges and stages are modified and expanded from the Theiler stages available on the e-MouseAtlas website.

TS7 has been divided into substages to distinguish between the changes in the visceral endoderm. There is also a change in terminology in TS11. The stage after the closure of the amnion has long been called the neural plate stage (Fujinaga et al., 1992; Downs and Davies, 1993). There is, however, no morphological neural plate in the mouse embryo at this time, in the sense of a thickened region of columnar epithelium, distinguishable from the surrounding ectoderm and destined to form neural tissue. Rather than thickening after amnion closure, the epiblast is a columnar epithelium from the time of its formation at TS7b and maintains a fairly uniform thickness of ~40 μm from TS8 (advanced egg cylinder stage) through to TS12 (at least two somites). The most proximal epiblast does become thinner after amnion closure, but this change in morphology does not define the rostral boundary between the neural ectoderm and the non-neural ectoderm; that boundary lies more distally in the thicker ectoderm (Cajal et al., 2012; K.A. Lawson, unpublished data). We therefore consider the neutral term pre-headfold a more suitable description for the period after amnion closure and before the headfold stages and have renamed the stages Neural Plate (NP) and Late Neural Plate (LNP) as Early Pre-Headfold (EPHF) and Late Pre-Headfold (LPHF), respectively.

Anatomy of Stages with Representative Molecular Markers

The revised staging is illustrated in Figures 3.1–3.4 in a series of drawings to scale of midline views of embryos, showing the positions of essential anatomical features. The embryos used for the postimplantation stages were mainly BL6CBA (see Section on Methods). A key to the figure labels is in Table 3.2. The accompanying examples of gene expression were selected to demonstrate and emphasize the anatomy. Whereas most can be considered cell lineage markers at the stages shown, early regional expression does not necessarily identify the cell progenitors of the later structure, for example, Tcfap2c (Cajal et al., 2012; K.A. Lawson, unpublished data).

Fig. 3.1 Staging diagrams and gene expression markers for TS1-5, preimplantation stages.

Fig. 3.2 Staging diagrams and gene expression markers for TS6-9a, implanted blastocyst stage to prestreak stage.

Fig. 3.3 Staging diagrams and gene expression markers for TS9b-11b, early primitive streak stage to late pre-headfold stage.

Fig. 3.4 Staging diagrams and gene expression markers for TS11c-12, early headfold stage to early somitogenesis stage.

Table 3.2

Labels for Figures

The drawings in Figures 3.1–3.4 are oriented with the animal pole (zygote to blastocyst) and mesometrial side (implanted embryos) to the top and the anterior facing left (TS7c onward). The bar at the foot of each figure is 100 μm. Images of embryos showing molecular markers are similarly oriented, except for TS11c, Runx1, and TS11d, Nkx2.5, which are laterofrontal and frontal views, respectively.

Preimplantation Stages: TS1–5, E0 to E3.5 (Figure 3.1)

One cell zygote: TS1, E0.5. The second polar body is visible. Male and female pronuclei are transiently visible.

2–4 cell embryo: TS2, E1.5. The first polar body is now lost.

Morula: TS3, E2.5. The embryo has 5–32 cells but is not yet cavitated. Compaction normally occurs at the eight-cell stage.

Marker gene expression: Nanog, a marker for the inner cell mass (ICM) and a pluripotency indicator, first appears in inner cells of the compacted morula.

Early blastocyst: TS4, E3. The embryo has 32–64 cells. The outer trophectoderm encloses the inner cell mass (ICM) and a nascent blastocyst cavity.

Marker gene expression: Nanog is expressed in the ICM.

Expanded blastocyst: TS5, E3.5. The embryo has 64–130 cells. The blastocyst cavity fills most of the volume of the embryo. Toward the end of this stage, the embryo hatches from the zona pellucida. The ICM segregates into epiblast and primitive endoderm.

Marker gene expression: Nanog and Sox2 are expressed in the ICM. Lefty 1, an inhibitor of Nodal and a marker for the DVE/AVE, is expressed sparsely in the ICM.

From Implantation to the Prestreak Stage: TS6–9a, E4.5–6.5 (Figure 3.2)

Implanted blastocyst: TS6, E4.5. The embryo has more than 150 cells. The primitive endoderm is segregated into visceral endoderm (VE) over the epiblast and parietal endoderm spreading over the mural trophectoderm (trophoblast). The epiblast is tilted with respect to the long (embryonic-abembryonic) axis of the embryo.

Marker gene expression: Nanog is downregulated in the epiblast. Lefty 1 is expressed in the primitive endoderm on the side nearest the embryonic pole, possibly as the earliest indicator of the future anteroposterior axis.

Early egg cylinder: TS7a, E5. The epiblast (embryonic ectoderm) and the extraembryonic ectoderm are distinguishable. Growth of the polar trophectoderm, now called extraembryonic ectoderm, and the epiblast produces the early egg cylinder, which projects into the blastocyst cavity.

Marker gene expression: Lefty 1 is expressed in the visceral endoderm towards the distal pole of the embryo.

Cavitating egg cylinder: TS7b, E5.25. The ectoplacental cone forms mesometrial to the junction between parietal and visceral endoderm. Reichert’s membrane is produced by parietal endoderm in contact with mural trophoblast. The proamniotic cavity is initiated after the epiblast is epithelialized by rosette formation (Bedzhov and Zernicka-Goetz, 2014).

DVE egg cylinder: TS7c, E5.5. The embryonic visceral endoderm at the distal pole of the egg cylinder, the distal visceral endoderm (DVE), is thickened as columnar epithelium.

Marker gene expression: Lefty1 is expressed in the DVE.

AVE egg cylinder: TS7d, E5.75. DVE cells move proximally to form the early anterior visceral endoderm (AVE) covering the anterior third of the embryonic part and establishing the anteroposterior polarity of the embryo (for a review, see Rivera-Pérez and Hadjantonakis, 2014). The proamniotic cavity begins to extend into the extraembryonic ectoderm after earlier loss of the basal membrane between the epiblast and the extraembryonic ectoderm following rosette formation (Bedzhov and Zernicka-Goetz, 2014).

Marker gene expression: Lefty 1 is expressed in the AVE. Bmp4, a marker of the chorionic ectoderm and the extraembryonic mesoderm, is expressed in the extraembryonic ectoderm, as is T(bra), a marker for extraembryonic ectoderm, the primitive streak, and the notochord.

Advanced egg cylinder: TS8, E6. The proamniotic cavity is continuous between the epiblast and the extraembryonic ectoderm. The thickened AVE defines the anterior aspect. The epiblast is oval in cross-section; the short diameter is initially aligned with the anteroposterior axis of the embryo but changes throughout the stage towards an orthogonal position, producing an increasingly asymmetrical proamniotic cavity (see Chapter 4 and references therein).

Marker gene expression: The expression of T(bra) in the distal extraembryonic ectoderm is intensified.

Prestreak: TS9a, E6.5. The embryo has increased in size from the previous stage, but mesoderm is not yet present. In cross-section, the long diameter of the epiblast is oblique to the anteroposterior axis.

Marker gene expression: Lefty 1 expression defines the AVE. Sox2 is expressed in the epiblast and throughout the extraembryonic ectoderm. T(bra) is expressed not only in the distal extraembryonic ectoderm but also in a wedge of posterior, proximal epiblast heralding the early primitive streak of the next stage. Hesx1, a positional marker for the future anterior neural ectoderm, the adjacent non-neural ectoderm, and the underlying anterior endoderm, is expressed in the AVE.

Early Primitive Streak to Late Pre-Headfold. TS9b–11b, E 6.5–E7.5 (Figure 3.3)

Early streak: TS9b, E6.5. Gastrulation begins with the initiation of the primitive streak in the most proximal posterior midline epiblast. The streak is present continuously throughout gastrulation and early organogenesis and consists of cells undergoing epithelial-to-mesenchymal transition as they pass from the epiblast to the mesoderm and the endoderm. Although the structure is stable, its cellular composition is dynamic. The streak grows in length by progressive incorporation of descendants of cells from the lateral and more anterior regions of the expanding epiblast, as well as by proliferation within itself. It extends up to half the distance to the distal tip of the egg cylinder during this stage. The first mesoderm to be formed initiates the amniochorionic fold by inserting between the posterior extraembryonic ectoderm and the extraembryonic visceral endoderm. The embryonic mesoderm emanating from the streak appears as a pair of expanding wings between the epiblast and the endoderm, with the base of the wings in the extending primitive streak. Up to this stage, the length of the embryonic part of the egg cylinder (H), that is, from the junction of the epiblast with the extraembryonic ectoderm to the distal tip of the cylinder, equals its width (W), measured at the same junction over the longer