Evolution of the Ectodermal Expression Pattern of X-Delta-2
X-Delta-2 expression in the anterior part of the embryo at early neurula stages (stage 14) consists of two concentric broken rings surrounding the basal plate and the neural plate (arrow and arrowheads, respectively, in Fig. 1A). Of the six proneural domains (three on each side of the midline) where the primary neurons arise from, X-Delta-2 is only expressed in the two middle domains (arrows in Fig. 1B), whereas X-Delta-1 is expressed in all six proneural domains (Chitnis et al.,1995). Between neurula stages and tail bud stages, this simple early pattern evolves into a more complex pattern (Fig. 1A–G) and at stage 25–27 (Fig. 1H,I) X-Delta-2 is expressed in the entire CNS, in the optic vesicle and in several neurogenic placodes, namely the olfactory (OP), profundal (pPr), trigeminal (pV), lateral line (pAD, anterodorsal; pM, middle; pP, posterior), otic and epibranchial placodes (epVII, facial; epIX, glossopharyngeal; epX2/X3, fused second and third vagal). In this study, we adopted the abbreviations for the placodes used by Schlosser and Northcutt (2000) and the abbreviation (pDL) for the dorsolateral placodal area.
Figure 1. Evolution of X-Delta-2 anterior neural expression. A–G: Albino embryos from stage 14 (A) to 22 (G) shown from the anterior side with dorsal to the top. H–J: Lateral view of stage 25, 27, and 36 albino embryos; anterior to the left. The first appearance of X-Delta-2 expression in a certain region is marked in the figure, except in G and H, where all the regions are marked. Arrowheads in A indicate X-Delta-2 expression in the neural plate border, and the arrow indicates the basal plate border. Arrows in B indicate the expression in the neural tube. The arrow in H indicates the otic vesicle. Arrowheads in K indicate forebrain–midbrain and midbrain–hindbrain boundary. Arrows in L indicate the X-Delta-2 expression in the middle of the hindbrain. In M, white arrows indicate the expression in the forebrain, black arrows in the tectum of the midbrain, and the arrowheads the expression in the hindbrain. The dotted line in M represents the level of the cut shown in N. Arrowheads in N mark the rhombomere boundaries. Dorsal and ventral are shown by the arrow. E, eye; epVII, facial epibranchial placode; epIX, glossopharyngeal epibranchial placode; epX2/X3, fused second and third vagal epibranchial placodes; FB, forebrain; HB, hindbrain; MB, midbrain; OP, olfactory placode; pAD, anterior dorsal lateral line placode; pDL, dorsolateral placodal area; pM, middle lateral line placode; pP, posterior lateral line placode; pPr, profundal placode; pPrV, profundal–trigeminal placode area; pV, trigeminal placode.
Download figure to PowerPoint
To illustrate the evolution of the X-Delta-2 anterior pattern more clearly, a video was made, using pictures of embryos from stage 14 to 22 (see Supplementary Data, which can be viewed at http://www.interscience.wiley.com/jpages/1058-8388/suppmat). The ectodermal X-Delta-2 expression during embryogenesis will now be described in detail, dealing first with expression in the anterior CNS, and later with the expression in the placodes.
The first X-Delta-2 expression in the CNS, besides the neural tube, was detected in the region of the presumptive midbrain (MB in Fig. 1D), before neural fold closure. Soon afterward, expression is also visible in the hindbrain (HB in Fig. 1E). At early stages, it is hard to distinguish whether X-Delta-2 is expressed in the forebrain, or simply in the olfactory placodes. However, dorsal views of stage 27 and 36 embryos indicate that X-Delta-2 is indeed expressed in the forebrain at these later stages (Fig. 1K,L). It is interesting to note that the forebrain–midbrain and the midbrain–hindbrain boundaries do not express X-Delta-2 at stage 27 (arrowheads in Fig. 1K). The evolution of the X-Delta-2 expression pattern also reflects the morphological changes during neurulation, when lateral tissue becomes dorsal. Thus, X-Delta-2 changes from being expressed in two lateral stripes throughout the midbrain and hindbrain at stage 27 (Fig 1K), to being expressed in the middle of the midbrain and hindbrain at stage 36 (arrows in Fig 1L).
To get a more detailed picture of X-Delta-2 expression in the brain, we isolated brains from stage 48 embryos and looked at X-Delta-2 expression (Fig. 1M,N). At this stage, we were able to detect weak expression in the forebrain (white arrows in Fig. 1M), and in the midbrain, where the expression was restricted to the border of the tectum (black arrows in Fig. 1M). In the hindbrain, X-Delta-2 is expressed in several transversal stripes (arrowheads in Fig. 1M) and from a sagittal section of the isolated brain it is possible to see that X-Delta-2 is excluded from the rhombomere boundaries (arrowheads in Fig. 1N) and the expression is also restricted to the dorsal side of the hindbrain (Fig. 1N). This change of pattern could be related to the formation of rhombomere boundaries, which occurs at approximately stage 43, suggesting that X-Delta-2 could be involved in hindbrain segmentation (see next section), as was recently shown for deltaC and deltaD in zebrafish (Cheng et al.,2004).
X-Delta-2 is also expressed in the cranial placodes during embryogenesis. From neurula stages onward, it is possible to detect X-Delta-2 expression in the profundal–trigeminal placode area (pPrV in Fig. 1B). At around stage 21, this area is separated into two individual placodes: profundal and trigeminal (pPr and pV, respectively, in Fig. 1F). After stage 26, the X-Delta-2 expression in these two placodes is no longer detected (Fig. 1H–J). The expression of X-Delta-2 in the prospective olfactory placodes also starts at neurula stages as two lateral spots next to the border of the neural plate (OP in Fig. 1B). These two spots come together with the closure of the neural tube (Fig. 1B–E).
The lateral line placodes together with the otic placode initially form a common dorsolateral placode area (pDL). This area expresses X-Delta-2 from its formation at around stage 20 (pDL in Fig. 1E). At stage 25, it is already possible to distinguish the individual lateral line placodes: anterodorsal, middle, and posterior lateral line placode (pAD, pM, and pP, respectively, in Fig. 1H) all of which express X-Delta-2. X-Delta-2 is also expressed transiently in the epibranchial placodes; epVII and epIX are visible at around stage 27 (Fig. 1I) and epX2/X3 only at a later stage (stage 36; Fig. 1J). The expression in the otic vesicle is weaker and was only detected at stage 25 (arrow in Fig. 1H). X-Delta-2 is expressed in the eye vesicle from the beginning of its formation at around stage 21 (E in Fig. 1F), but its expression is excluded from the lens placode formed at around stage 28 (Fig. 1J).
X-Delta-2 expression shares similarities with the expression pattern of several other Notch ligands in different organisms (Bettenhausen et al.,1995; Myat et al.,1996; Dunwoodie et al.,1997; Haddon et al.,1998; Schlosser and Northcutt,2000; Kiyota et al.,2001). This finding raises interesting questions about the redundant or complementary function of the different genes (X-Delta-2, X-Delta-1, and X-Serrate-1 in Xenopus) in the many processes that Notch signalling is involved in. One example is the generation of neural diversity in the retina. It is known that X-Delta-1 is involved in this process (Dorsky et al.,1997); therefore, it would be interesting to investigate whether X-Delta-2 and X-Delta-1 have a complementary or redundant role in the neural development of the retina.
X-Delta-2 activates Notch signalling in adjacent cells and it is expressed in regions where boundaries are to be formed, such as in the somitomeres. The expression next to the basal plate and neural plate is in accordance with the role of X-Delta-2 in separating different tissues by conferring distinct identities. In these two regions, the downstream targets of the Notch receptor, Hairy-2a and Hairy-2b, are also expressed (Lopez et al.,2005; Tsuji et al.,2003), supporting the idea of a role for X-Delta-2 in these regions. Furthermore, that the forebrain–midbrain boundary, the midbrain–hindbrain boundary, and later on the rhombomere boundaries, are free of X-Delta-2 expression, raises the possibility that it activates the Notch pathway in the adjacent boundary cells and suppresses neurogenesis in these boundary cells, as was described recently in the zebrafish hindbrain (Cheng et al.,2004).
This expression analysis shows that, in the anterior ectoderm, X-Delta-2 is expressed from early development in several neurogenic regions (the CNS and cranial placodes) and suggests a possible role in establishing boundaries in several tissues.
X-Delta-2 Function in the Anterior CNS and Eye
To study the function of X-Delta-2, in the different neural tissues where it is expressed, we used the MO knockdown approach. We injected the X-Delta-2 MO on the left-hand side (LHS) to use the noninjected side as an internal control (Fig. 2). The analysis of the gross phenotype revealed malformations in the CNS, the most obvious being a reduction in the size of the eye and in the length of the hindbrain (Fig. 2A′). In addition to the effects on the CNS, the nasal pit and the otic vesicle were severely reduced (arrowhead and arrow, respectively, in Fig. 2A′). A second, nonoverlapping MO was also used and the same phenotype was obtained.
Figure 2. X-Delta-2 loss of function affects the central nervous system. A′: X-Delta-2 morpholino (MO) was injected in the two left blastomeres at the four-cell stage, and the phenotype was analyzed in embryos at stage 48. B′–D′: The expression of X-Delta-2 at stage 21 (B′) and Nrp-1 at stage 21 (C′) and stage 25 (D′) were also analyzed. A–D: Noninjected (Nic) embryos are also shown. Embryos at stage 48 are shown from the dorsal side with anterior to the top (A), and embryos at stage 21 and stage 25 are shown from the anterior side with dorsal to the top (B–D). Brackets and dotted circles in A′ indicate the olfactory pits and the otic vesicle, respectively. The asterisk indicates the injected side.
Download figure to PowerPoint
To analyze the X-Delta-2 phenotype in the CNS in more detail, we decided to look at the expression of X-Delta-2 and Nrp-1, a general neural marker (Fig. 2B′–D′). During somitogenesis, it has been reported that X-Delta-2 down-regulates its own expression in the posterior half of the somitomeres (Jen et al.,1997), explaining the up-regulation observed in the PSM when X-Delta-2 is knocked down (Jen et al.,1999, and data not shown). However, the opposite seems to be true in the CNS, where, at stage 21, the X-Delta-2 expression is severely down-regulated by injection of X-Delta-2 MO (Fig. 2B′; 100%, n = 7). This finding could be due to a general effect on neural tissue in the brain and neural tube, as suggested by the strong down-regulation of Nrp-1 in these regions (Fig. 2C′; 78%, n = 9). This down-regulation is maintained until at least stage 25 (Fig. 2D′; 100%, n = 16). Other genes expressed in the CNS however, such as Pax-6 (Fig. 3A′,B′), Engrailed-2 (Fig. 5A′), Krox-20 (Fig. 5A′), and Gbx-2 (Fig. 5B′), are not down-regulated, suggesting that neural tissue is still present but with an altered identity.
Figure 3. X-Delta-2 role in the establishment of the identity of the forebrain. A′–E′: Embryos injected with X-Delta-2 morpholino (MO) in the left-hand side were analyzed for the expression of the forebrain marker Pax-6 (A′,B′), the telencephalon marker Emx-2 (C′,D′), and the ventral forebrain marker Nkx2.1 (E′). Pax-6 and Emx-2 were analyzed at stage 21 (A,A′ and C,C′, respectively), and at stage 25 (B,B′ and D,D′, respectively), whereas Nkx2.1 was only analyzed at stage 25 (E,E′). Noninjected (Nic) controls are shown for each marker (A–E). All embryos are shown from the anterior side with dorsal to the top. The arrow in B′ indicates the loss of Pax-6 expression in the olfactory placode. Dotted lines in E and E′ indicate the midline. Asterisk indicates the injected side.
Download figure to PowerPoint
Figure 5. X-Delta-2 is involved in the segmentation and patterning of the hindbrain. A′–C′: X-Delta-2 morpholino (MO) was injected in the left-hand side, and the embryos were analyzed for the expression of different hindbrain markers at stage 21. A–C: Noninjected (Nic) embryos are also shown for all the markers. The expression of Engrailed-2 in the midbrain–hindbrain boundary (En in A,A′), Krox-20 in rhombomeres (r) 3 and 5 (A,A′), Gbx-2 in the r1 (bracket in B,B′), and Hoxb-3 (C,C′) in r5 were analyzed. The MyoD expression shows loss of segmentation in the injected side, confirming the loss of X-Delta-2 function. D: The morphology of the brain is shown using the neural antibody 2-G9 in stage 48 embryos. A–D: Embryos at stage 21 are shown from the anterior side with dorsal to the top (A–C), and the brain of a stage 48 embryo is shown from the dorsal side with anterior to the left (D). E: The distance between the posterior border of the Engrailed-2 stripe and the anterior border of the r3 Krox-20 stripe (r1–r2 region) and the distance between the anterior border of the r3 Krox-20 stripe and the posterior border of the r5 Krox-20 stripe (r3–r5 region) was measured in stage 21 (n = 6) and stage 25 (n = 5) embryos. E: The hindbrain regions measured are indicated by brackets in a stage 21 embryo stained for Engrailed-2 and Krox-20. E: Both sides were measured, and the ratio left-hand side/right-hand side (LHS/RHS; Nic/Nic or ΔMO/NIC) was plotted on a graph. Values above 1.0 indicate an increase in the length of the region on the left (injected) side, and values below 1.0 indicate a decrease. In both stages, the r1–r2 region is increased on the X-Delta-2 MO-injected side and the r3–r5 region is decreased. The arrow in D indicates the forebrain, and arrowheads indicate the rhombomere boundaries in the hindbrain. The asterisk indicates the injected side.
Download figure to PowerPoint
The forebrain comprises dorsally the telencephalon and the eyes, ventrally the hypothalamus, and more caudally the diencephalon (Wilson and Houart,2004). To further analyze the role of X-Delta-2 in the forebrain, we decided to look at the expression of Pax-6, Emx-2, and Nkx2.1, expressed in the entire forebrain, in the telencephalon and in the ventral forebrain (telencephalon and diencephalons), respectively (Pannese et al.,1998; Hollemann and Pieler,2000; Franco et al.,2001). At stage 21, the expression of Pax-6 in the forebrain is not affected (Fig. 3A′; 100%, n = 9), but at a later stage expression in the olfactory placode is reduced on the injected side (Fig. 3B′; 100%, n = 6). It is known that Pax-6 is involved in the development of the olfactory system in Xenopus (Reiss and Burd,1997; Franco et al.,2001); therefore, the reduction of the nasal pits observed at late stages (brackets in Fig. 2A′) could be by means of the down-regulation of Pax-6 expression in the olfactory placode. Evidence of a link between Notch signalling and Pax-6 has been demonstrated already in explant experiments, where Notch signalling induced ectopic Pax-6 (Onuma et al.,2002). Although the Pax-6 expression is not affected in the forebrain, the expression of Emx-2 (Fig. 3C′; 75%, n = 8; and 3D′; 78%, n = 9) and Nkx2.1 (Fig. 3E′; 71%, n = 7), is down-regulated. These results suggest that X-Delta-2 affects the patterning of the forebrain at the level of Emx-2 and Nkx2.1, but downstream of Pax-6. To confirm the specificity of the X-Delta-2 MO (Fig. 4C; 100%, n = 5), we tried to rescue the Emx-2 expression by co-injecting an X-Delta-2 mRNA construct that lacks the 5′ untranslated region (UTR), and, therefore, is not recognized by the MOs. We were able to partially rescue the expression of Emx-2 in the injected side (Fig. 4D; 67%, n = 6). When X-Delta-2 mRNA alone is injected, the Emx-2 expression seems to be expanded caudally and the domain of expression is wider (Fig. 4B; 86%, n = 7). This caudal expansion is also visible in the rescued embryos (Fig. 4D). These results confirm that X-Delta-2 MO specifically affects forebrain patterning.
Figure 4. Rescue of Emx-2 expression in the telencephalon. A–D: Noninjected controls (A) and embryos injected on the left-hand side with X-Delta-2 RNA (B), X-Delta-2 morpholino (MO; C), or a combination of X-Delta-2 RNA and X-Delta-2 MO (D) were analyzed for the expression of Emx-2 at stage 21. Embryos are shown from the anterior side, with the dorsal side to the top. D: Overexpression of X-Delta-2 RNA was able to rescue the telencephalon expression of Emx-2 in X-Delta-2 MO-injected embryos. B: Emx-2 expression is slightly expanded when X-Delta-2 RNA is overexpressed. Asterisk indicates the injected side of the embryo.
Download figure to PowerPoint
The effect on gene expression in the forebrain is reflected in the later morphology of the forebrain (Fig. 2A′), where the forebrain appears shorter and thinner on the X-Delta-2 MO-injected side. This forebrain phenotype is easier to observe in a confocal image where a neural specific antibody (Jones and Woodland,1989) was used to visualize the neural tissue (arrow in Fig. 5D).
Looking at the gross phenotype, it is also possible to observe a reduction in the size of the eye (Fig. 2A′). This reduction is already observed at stage 25, where the domains of expression of Nrp-1, Pax-6, and Emx-2 are smaller than in the control side, although the levels of expression are not affected (Fig. 2D′, 3B′, and 3D′, respectively). These results suggest that X-Delta-2 is not involved in the specification or patterning of the eye placode, but in the determination of the eye size. It has been reported that Notch signalling can activate Pax-6 expression and, thus, induce ectopic eye-like structures (Onuma et al.,2002), but in our study, despite down-regulation of Pax-6 in the hindbrain, we did not observe any down-regulation of Pax-6 expression in the eye with the X-Delta-2 loss of function (Fig. 3B′).
The hindbrain was also affected by the X-Delta-2 loss of function, as shown by the simple analysis of the gross phenotype (Fig. 2A′) and by the down-regulation of X-Delta-2 and Nrp-1 (Fig. 2B′ and 2D′, respectively). Pax-6 is down-regulated at stage 25, whereas its expression at stage 21 does not change (Fig. 3A′,B′). To further analyze the effect on hindbrain of losing X-Delta-2 function, we carried out a detailed marker analysis in embryos injected in the LHS with X-Delta-2 MO (Fig. 5). Certain aspects of hindbrain patterning are not affected, as shown by the normal Krox-20 stripes in rhombomeres 3 and 5 (r3 and r5; Fig. 5A′; 100%, n = 6). However, the Krox-20 stripes, together with that of Engrailed-2, seem closer together (Fig. 5A′; 83%, n = 6). To further investigate changes in the size of the hindbrain, we measured the length of the r1–r2 and r3–r5 regions in embryos injected on the left-hand side with X-Delta-2 MO. We performed this analysis at two different stages: stage 21 and stage 25. The comparison between the injected side and the control side shows that, at both stages, the r1–r2 region is enlarged and the r3–r5 region is reduced (Fig. 5E). The total length of the hindbrain is not severely affected, and the apparent shortened hindbrain observed in Figure 2A′ is due more to a twist in the neural tissue rather than a change in length. The analysis of the r1 marker Gbx-2 showed that its expression is expanded posteriorly (Fig. 5B′; 73%, n = 11), whereas Hoxb-3, expressed in r5, is down-regulated (Fig. 5C′; 100%, n = 9). The expansion of the Gbx-2 domain and the r1–r2 region on one hand, and the loss of Hoxb3 expression and reduction of the r3–r5 region on the other hand, suggest that the hindbrain is rostralized to a certain extent. However, the hindbrain still maintains an anteroposterior patterning as the Krox-20 stripes are still present. The involvement of X-Delta-2 in the patterning process seems to occur at a relatively late stage, as Krox-20 expression is unaffected.
The X-Delta-2 expression at stage 48 in the rhombomeres (Fig. 1M,N) suggests a possible role of X-Delta-2 in hindbrain segmentation. To address this question we analyzed the morphology of the brain using a neural specific antibody (Jones and Woodland,1989). Confocal imaging showed that, in the X-Delta-2 MO injected side (asterisk in Fig. 5D), no clear boundaries were formed (75%, n = 8), despite that the brain still retains its neural identity. This result together with the mis-patterning of the hindbrain shown by the expansion of Gbx-2 (Fig. 5B′) and the down-regulation of Hoxb-3 (Fig. 5C′) indicates that X-Delta-2 is necessary for the segmentation as well the patterning of the hindbrain. The exclusion of X-Delta-2 from the forebrain–midbrain and midbrain–hindbrain boundaries suggests that these boundaries could also be affected by X-Delta-2 loss of function. However, this does not seem to be the case, as the boundaries appear unaffected (Fig. 5D).
X-Delta-2 Role in Neurogenesis and Cell Migration
It is known that X-Delta-1 is involved in primary neurogenesis, regulating the number of cells that will become neurons by lateral inhibition (Chitnis et al.,1995). To analyze a possible role for X-Delta-2 in primary neurogenesis, we analyzed the expression of N-tubulin, an early marker for neural differentiation in the primary neurons (Oschwald et al.,1991). Instead of an increase in the number of cells expressing N-tubulin, as would be expected if lateral inhibition is blocked, N-tubulin was strongly down-regulated in the X-Delta-2 MO-injected side (Fig. 6A′), indicating that fewer primary neurons have been formed. However, when we look at the expression of X-Ngnr-1, a marker for neuralized but undifferentiated tissue (Ma et al.,1996), there is no effect on its expression in the midline (Fig. 6B′), showing that the tissue remains in an undifferentiated state. This finding suggests that X-Delta-2 is not involved in the process of lateral inhibition but is essential for the differentiation of the primary neurons. The X-Delta-2 role in neuronal differentiation does not extend to the pPrV placode, where the expression of N-tubulin is not affected (arrow in Fig. 6B′).
Figure 6. A–D′: X-Delta-2 role in the neurogenic placodes. Noninjected (Nic) controls (A–D) and embryos injected on the left-hand side with X-Delta-2 morpholino (MO; A′–D′) were analyzed for the expression of N-tubulin (A,A′) X-Ngnr-1 at stage 21 (B,B′) and stage 25 (C,C′), and for the expression of X-Delta-2 at stage 25 (D,D′). Embryos at stage 21 are shown from the anterior side with dorsal up, and the ones at stage 25 are shown from the lateral side with dorsal to the top, Nic with anterior to the right side, and injected embryos with anterior to the left side. Arrows in A′ and B′ indicate the N-tubulin and X-Ngnr-1 expression, respectively, in the profundal–trigeminal placodal area (pPrV). pAD, anterodorsal lateral line placode; pM, middle lateral line placode. Asterisk indicates the injected side. Note that the stage 25 Nic inC and D for X-Ngnr-1 and X-Delta-2 are the noninjected side of the same embryo shown in C′ and D′, respectively.
Download figure to PowerPoint
X-Delta-2 is expressed in all of the neurogenic placodes and the observation of the gross phenotype showed that at least the olfactory pits are affected by the X-Delta-2 loss of function (Fig. 2A′), possibly due to the loss of Pax-6 expression in the olfactory placode (Fig. 3B′). To analyze if other placodes were also affected we looked at the expression of X-Ngnr-1 and X-Delta-2 (Fig. 6). Around stage 21, X-Ngnr-1 is expressed in the profundal–trigeminal placode area (pPrV; Fig. 6B and Schlosser and Northcutt,2000). In the X-Delta-2 MO-injected side the expression of X-Ngnr-1 in the pPrV placode is enhanced and expanded, indicating that there is an increase in the number of cells expressing X-Ngnr-1 (arrow in Fig. 6B′; 100%, n = 13). This increase in the number of cells expressing X-Ngnr-1 is not a consequence of blocking lateral inhibition, as the expression of N-tubulin in the placode does not change (arrow in Fig. 6A′), but is likely to be due to the fact that the cells maintain their undifferentiated state, resulting in a failure of cell migration. It has been shown that cells stop expressing X-Ngnr-1 immediately after they leave the placode (Schlosser and Northcutt,2000), so a block in cell migration could explain the increase number of cells expressing X-Ngnr-1 in the placode. These results suggest that X-Delta-2 could be involved in regulating neuronal differentiation and cell migration in the placodes.
To study the possible role of X-Delta-2 in placodal cell migration in more detail, we analyzed the expression of X-Ngnr-1 and X-Delta-2 at around stage 25, when the placodes have extended ventrally. At this stage, X-Ngnr-1 is expressed in the anterodorsal lateral line placode (pAD) and in the middle lateral line placode (pM; Fig. 6C and Schlosser and Northcutt,2000). It is hard to say if there are more cells expressing X-Ngnr-1 when X-Delta-2 is knocked down, or if it is only the migration of the cells that is perturbed, as the domain of expression extends less ventrally in the injected side than in the control side (Fig. 6C′; 80%, n = 31). Similarly, X-Delta-2 expression in the pAD and pM does not extend as far ventrally as in the noninjected side (Fig 6D′; 92%, n = 26). These observations suggest that, in the X-Delta-2 knockdown, more cells stay in an undifferentiated state and this finding is followed by a failure of the cells to migrate ventrally. It is possible that the cells do not migrate because they have not differentiated into neurons.
To further analyze a possible role of X-Delta-2 in cell migration, we decided to look at the cranial neural crest cells (Fig. 7), known to migrate ventrally from the midline (reviewed in Kulesa et al.,2004). Xslug and Xsnail are expressed in the premigratory and migrating neural crest cells (Linker et al.,2000), and both of these markers are restricted to the region next to the hindbrain when X-Delta-2 is down-regulated (Fig. 7A′,B′). However, whereas the levels of Xsnail do not seem to be altered (Fig. 7B′; 80%, n = 10), Xslug expression is strongly up-regulated compared with the control side (Fig. 7A′; 100%, n = 6). This difference could be due to the fact that Xslug, but not Xsnail, is down-regulated when the neural crest cells reach their target (Linker et al.,2000). Therefore, the neural crest cells staying close to the midline do not have their Xslug expression down-regulated, so the apparent up-regulation of Xslug expression is the result of the increase of the number of neural crest cells close to the midline. The expression of X-dll4 in the branchial arches (Papalopulu and Kintner,1993) is either absent (Fig. 7C′; 38%, n = 21) or strongly down-regulated (Fig. 7D′; 52%, n = 21), with the expression domain not expanding as ventrally as in the noninjected side. This finding suggests that the branchial arches, derived from the cranial neural crest cells, fail to form. To confirm this hypothesis, we performed a cartilage staining, which showed that the gills, derived from the branchial arches, were missing on the X-Delta-2 MO-injected side (Fig. 7E′, 80%, n = 10). Thus, X-Delta-2 function is necessary for neural crest cell migration, despite that it does not seem to be expressed in the neural crest region. This effect, therefore, could be due to the function of X-Delta-2 in the adjacent hindbrain, It is interesting to note that, in the mouse delta-1 knockout, neural crest cell migration is also perturbed (De Bellard et al.,2002).
Figure 7. A–E′: X-Delta-2 loss of function affects cranial neural crest cell migration. Noninjected (Nic) controls (A–E) and embryos injected on the left-hand side with X-Delta-2 morpholino (MO; A′–E′) were analyzed for the expression of cranial neural crest markers or stained with Alcian blue to highlight the cartilage. A,A′,B,B′ show stage 21 embryos from the anterior side with dorsal to the top; C,C′,D,D′ show stage 25 embryos from the lateral side with dorsal to the top; and in E,E′ show stage 48 embryos from the ventral side with anterior to the top. Cranial neural crest cells fail to migrate as shown by the expression of X-Slug (A′), Xsnail (B′), and X-dll4 (C′,D′). The cartilage staining shows that the gills, derived from the branchial arches, do not form in the X-Delta-2 MO-injected side (arrow in E′). The asterisk indicates the injected side. Nic for X-dll4 in C and D is the noninjected side of the same embryos shown in C′,D′, respectively.
Download figure to PowerPoint
This effect on the neural crest cells is very similar to that seen when the paralogous group 1 (PG1) Hox genes are down-regulated (McNulty et al.,2005), and results from our group indicate that X-Delta-2 is also down-regulated in these embryos, suggesting that the effect of the PG1 knockdown on neural crest migration could be by means of a down-regulation of X-Delta-2 (Peres et al, unpublished observations).