Stage specificity for producing effects on anterior neural pattern

After performing significant numbers of lower layer removals, it became apparent that the stage of the embryo was critical for producing the effects outlined above (see Fig6.1). If the lower layer was removed at stage 4, no apparent effect was observed, and the majority of embryos (166/221; 75%) developed normally. However, endoderm removal between stage 4+ and 5, therefore just as the first axial mesoderm cells were being laid down and extending into a full head process, produced all aspects of the syndrome in the majority of cases (171/235; 73%). Interestingly, in the cases where endoderm was successfully removed cleanly between stage 5+ and 6, forebrain pattern appeared to be complete, even though the embryos still lacked a foregut tunnel and the heart remained laterally. The forebrain patterning defects seen after stage 4+ and 5 removals are therefore not just disrupted morphology due to the brain remaining flattened out. In essence, the period during which endoderm removal causes defects in subsequent forebrain régionalisation is restricted to stages 4+ and 5.

6.3.1 Immediate response to lower layer removals

In order to examine why most embryos developed normally following removal at stage 4, embryos were fixed either immediately or a very short time after operations (Fig6.2). As shown in Fig6.2A-D, within about 6 hours of further culture, the hole in the endoderm layer had been reduced to almost nothing. The exposed area appeared to be filled in from the posterior end (node region) first, moving anteriorly so that before any prechordal mesoderm had left the node, the neural ectoderm was once again covered by endoderm. This was not the case following a stage 5 removal, however (Fig6.2E-G). The hole in the endoderm layer did not close up to any great extent (although it did change shape due to convergent extension movements of the embryo), so that any axial mesoderm that had left the node prior to the operation remained permanently exposed.

%

+ 0 h r s + 2 h r s + 4 h r s + 6 h r s

Figure 6.2 Behaviour of lower laver after operations

Dark field views o f em bryos from the ventral aspect, as set up in ring culture, fixed

shortly after low e r later rem ovals. A - D E m bryos fixed at successive tim epoints

following removal o f lower layer at stage 4. W indow o f absent lower layer quickly closes up, from the posterior end (node region), so that by the time any prechordal

m esoderm em erges, the lower layer is again alm ost com plete (D). E - G Em bryos

fixed at successive tim epoints follow ing rem oval of low er layer at stage 5. The

window of absent low er layer remains, so that all axial m esoderm em erged prior to the operation rem ains abnormally exposed and lacks contact with foregut endoderm.

As the node continued to regress, it moved back under the remaining endoderm, so the more posterior mesoderm laid down later was closely apposed by endoderm as normal.

6.3.2 Cell movements confirmed by D il labelling

The origin of cells that move to fill in the gap of endoderm following a stage 4 removal, was examined further using Dil labelling. Lower layer was removed as usual and Dil was then injected onto the exposed ‘mesodermal’ cells of the node. The embryos were subsequently cultured for about 6 hours, until the lower layer was again complete. As shown in Fig6.3, labelled cells were now found in the endoderm anterior to the node,

i.e. part of the region that had filled in the area removed. This suggests that cells were continuing to move down through the node into the endoderm layer after the operation was performed, and so healing of the exposed area was observed. My labelling studies on normal embryos support the hypothesis that node cells are normally still entering the endoderm at this stage (see Fig5.1); however, response to the removal operation may also be a contributory factor.

6.3.3 Gene expressions in definitive endoderm

Endodermal gene expressions were examined a short time after lower layer removal, to assess for restoration of normal patterns. Hex is a homeobox gene that plays an important role in anterior brain patterning in mouse. It is first expressed in the mouse AVE, but chimaeiic analysis shows that the essential requirement for hex function in forebrain patterning resides in the definitive endoderm (Martinez-Barbera et a l, 2000a). In the chick, hex is also seen in the hypoblast at early stages, being swept to the periphery as the cells are moved by intercalation of definitive endoderm (Yatskievych et a l, 1999). From stage 5 onwards, expression is seen in a restricted patch of endoderm directly underlying the prechordal mesoderm, and an anterior crescent of pharyngeal endoderm that later becomes associated with the forming heart tube (Fig6.4A). Hex

Immediately after lower layer removal at stage 4, the deeper ‘mesodermal’ layer of the node was labelled with Dil. Embryos were cultured further for 6 hours, during which time healing of the window in absent lower layer was mostly completed. A, B Fluorescent and dark field images of the same embryo (now stage 4+). C - E Fluorescent, dark field and combined images of the same embryo (also now sage 4+). Labelled cells now populate the now intact lower layer around and ahead of the node. This area is much more extensive than that occupied by the newly emerging prechordal mesoderm (indicated by dotted white lines on B and D). Scale bar = 300p,m throughout.

has been difficult to record decisively on saggital sections (see also Yatskievych et at.,

1999). Following lower layer removal at stage 4, hex expression is fully restored by headfold stages (Fig6.4B). However, endoderm removal between stages 4+ and 5 leads to permanent loss of hex expression in 8/10 cases (Fig6.4C,H). In addition, the complete absence of prechordal expression in these embryos suggests that any normal mesodermal hex expression, whether imported by cell intercalation or induced by underlying endoderm, has also been prevented. Therefore, conditions of endoderm removal that lead to forebrain pattern loss clearly result in failure to recover hex

expression.

Crescent is a chick Frzb homologue strongly expressed in the hypoblast before

gastrulation, and later in anterior definitive endoderm and head process mesoderm (Fig6.4I,F; Pfeffer gr ah, 1997). Frzb family members are secreted proteins thought to play a role in head formation, since they can act as local antagonists of the Wnt

signalling pathway (Leyns et al., 1997; Wang et al., 1997; Niehrs, 1999). Crescent is expressed in a broad, intense domain in the definitive endoderm, centred on but more widespread than the site of hex expression. Seen in plan view, the area of crescent

expression at stage 4+ and 5 closely correlates with the region of endoderm that causes forebrain pattern deletions when removed. Indeed, I find that when removals are carried out at stage 4, normal crescent expression is restored in the endoderm by headfold stages (Fig6.4J). However, following a stage 4+ to 5 removal, the great majority of endodermal crescent expression is absent, and only axial mesodermal expression remains (8/8 embryos; Fig6.4G,K).