Retinoic acid-binding protein: a plasma membrane component

Soluble protein extracts from lyophilized plasma membranes prepared from chick embryo skin and transplantable murine carcinomas were found to contain a specific retinoic acid-binding component. This binding component showed high affinity for biologically active analogs of retinoic acid. The plasma membrane component exhibited similar physicochemical properties to those of the retinoic acid-binding protein described earlier in the cytosol. The protein exhibited mercurial-sensitive thiol functions in ligand binding; the mercurial-inhibition was reversed on treatment with thiol compounds. The plasma membrane binding component may be involved in the cellular uptake of retinoic acid.     

A paraxial exclusion zone creates patterned cranial neural crest cell outgrowth adjacent to rhombomeres 3 and 5.

Cranial neural crest cell migration is patterned, with neural crest cell-free zones adjacent to rhombomere (R) 3 and R5. These zones have been suggested to result from death of premigratory neural crest cells via upregulation of BMP-4 and Msx-2 in R3 and R5, consequent to R2-, R4-, and R6-derived signals. We reinvestigated this model and found that cell death detected by acridine orange staining in avian embryos varied widely numerically and in pattern, but with a tendency for an elevated zone centered at the R2/3 boundary. In situ hybridization of BMP-4 mRNA resolved to centers at R3 and R5 but Msx-2 resolved to the R2/3 border with only a faint smear from R5 to R6. Outgrowth of neural crest cells was less in isolated R3 cultures than in R1+2, R2, and R4 cultures, but R3 showed neither a decrease in outgrowth of neural crest cells nor an increase in cell death when cocultured with R1+2, R2, or R4. In addition, in serum-free culture, exogenous BMP-4 strikingly reduced neural crest cell outgrowth from R1+2 and R4 as well as R3. Thus we cannot confirm the role of intraneural cell death in patterning rhombomeric neural crest outgrowth. However, grafting quail R2 or R4 adjacent to the chick hindbrain demonstrated a neural crest cell exclusion zone next to R3 and R5. We suggest that one important pattern determinant for rhombomeric neural crest cell migration involves the microenvironment next to the neural tube.     

Retinoic acid-dependent eye morphogenesis is orchestrated by neural crest cells

Using genetic approaches in the mouse, we show that the primary target tissue of retinoic acid (RA) action during eye morphogenesis is not the retina nor the corneal ectoderm, which both express RA-synthesizing retinaldehyde dehydrogenases (RALDH1 and RALDH3), but the neural crest cell-derived periocular mesenchyme (POM), which is devoid of RALDH. In POM, the effects of the paracrine RA signal are mediated by the nuclear RA receptors heterodimers RXRalpha/RARbeta and RXRalpha/RARgamma. These heterodimers appear to control: (1) the remodeling of the POM through activation of Eya2-related apoptosis; (2) the expression of Foxc1 and Pitx2, which play crucial roles in anterior eye segment development; and (3) the growth of the ventral retina. We additionally show that RALDH1 and RALDH3 are the only enzymes that are required for RA synthesis in the eye region from E10.5 to E13.5, and that patterning of the dorsoventral axis of the retina does not require RA.     

Retinoic acid-binding protein in rat tissue. Partial purification and comparison to rat tissue retinol-binding protein.

When the 100,000 X g supernatant fractions of several rat organs are incubated with all-trans-[3H]retinoic acid, a binding component for retinoic acid with a sedimentation coefficient of 2 S can be detected by sucrose gradient centrifugation. This tissue binding protein for retinoic acid is distinct from the tissue binding protein for retinol which has been previously described. The tissue retinoic acid-binding protein has been partially purified from rat testis and this partially purified protein would appear to have a molecular weight of 14,500 as determined by gel filtration and high binding specificity for all-trans-retinoic acid. Binding of [3H]retinoic acid is not diminished by a 200-fold molar excess of retinal, retinol, or oleic acid but is reduced by a 200-fold excess of unlabeled retinoic acid. Tissue retinoic acid-binding protein can be detected in extracts of brain, eye, ovary, testis, and uterus but is apparently absent in heart muscle, small intestine, kidney, liver, lung, gastrocnemious muscle, serum, and spleen. This distribution is different than that observed for the tissue retinol-binding protein. Tissue retinol-binding protein was also purified extensively from rat testis. The partially purified protein has an apparent molecular weight of 14,000 and high binding specificity for all-trans-[3H]retinol as only unlabeled all-trans-retinol but not retinal, retinoic acid, retinyl acetate, retinyl palmitate, or oleic acid could diminish binding of the 3H ligand under the conditions employed. The partially purified protein has a fluorescence excitation spectrum with lambda max at 350 nm. In contrast, the retinol-binding protein isolated from rat serum and described by others has a fluorescence excitation spectrum with lambda max at 334 nm and an apparent molecular weight of 19,000. When partially purified tissue retinol-binding protein is extracted with heptane, the heptane extract has a fluorescence excitation spectrum similar to that of all-trans-retinol.     

Domains of cellular retinoic acid-binding protein I (CRABP I) expression in the hindbrain and neural crest of the mouse embryo.

We describe here the distribution of cellular retinoic acid-binding protein I (CRABP I) in the head of the early mouse embryo from day 8 to day 13 of gestation, using both in situ hybridisation to localise mRNA and immunocytochemistry to localise protein. The distribution of mRNA and protein was found to be identical. CRABP I first appeared in part of the presumptive hindbrain of the presomite embryo and then became localised to rhombomeres 2, 4, 5 and 6. The only other area of expression in the cephalic neuroepithelium was in a part of the midbrain roof. The neural crest and its mesenchymal derivatives, the branchial arches, expressed CRABP I and crest could be seen streaming from the neuroepithelium of individual rhombomeres into particular branchial arches. This suggested a fate map could be constructed describing the rhombomeric origin of branchial arch mesenchyme. Later in development, axons throughout the hindbrain expressed CRABP I. The results are considered in terms of the role of retinoic acid in the specification of neuronal phenotype in the hindbrain and in axon outgrowth.     

Xerl, a novel CNS-specific secretory protein, establishes the boundary between neural plate and neural crest.

A novel gene, Xerl, has been found as a CNS-specific gene encoding a secretory protein. In order to clarify a function of Xerl, we first examined Xerl-expressing areas during early development. Comparison with XlSox-2-positive neural plate and ADAM13-positive neural crest showed that Xerl expression was limited within the neural plate area. Microinjection of Xerl mRNA into 2- or 4-cell stage embryos indicated that Xerl overexpression caused the regional expansion of XlSox-2- and NCAM-positive neural plate, which was concomitant with the outer shift of ADAM13-positive region. The Xerl injection resulted in incomplete neural closure because of the local overproduction of the neuroepithelium. In contrast, loss of function analysis of Xerl indicated that Xerl inhibition caused the ectopic differentiation of neural crest cells. In the conjugation experiment using chordin-injected animal caps, Xerl promoted chordin-induced XlSox-2 expression, whereas Xerl inhibition caused ADAM13expression even in the injection with a high dose of chordin. Animal cap assays also showed that Xerl expression was induced by chordin. In the functional analysis using truncated forms of Xerl, Xerl deltaL (lacking LNS domain) worked as a dominant negative form that induced the overproduction of neural crest cells. These results suggest that Xerl is involved in the boundary formation of the neural plate through exclusion of neural crest cell differentiation.     

Apoptosis of premigratory neural crest cells in rhombomeres 3 and 5: consequences for patterning of the branchial region

In the avian hindbrain, premigratory neural crest cells undergo programmed cell death (apoptosis) in rhombomeres 3 and 5 (r3, r5). Here, we have attempted to analyze the significance of the loss of neural crest cells from these odd-numbered rhombomeres. When apoptosis is prevented in r3 and r5, r3 crest migrate into the first arch and r5 into the third arch. Interestingly, these extra neural crest cells contributed to the formation of ectopic muscle attachment sites that are also found in those species in which r3 and r5 neural crest cells do not undergo apoptosis. Thus, apoptosis in the odd-numbered rhombomeres appears to be an evolutionarily derived mechanism that is required to eliminate r3 and r5 crest migration into first and third arches and thereby remove these muscle attachment sites.     

Retinoic acid inhibits migration of cranial neural crest cells in the cultured mouse embryo

Clinical observations have demonstrated that isotretinoin (13-cis-retinoic acid; cis-RA) is a human teratogen causing primarily heart and craniofacial malformations. Isotretinoin exposure to the early postimplantation mouse embryo in culture results in specific defects in craniofacial development that may be due to an interference in the early migration of cranial neural crest (CNC) cells [Goulding and Pratt, 1986]. The present study was designed to test this hypothesis by examining the migration of these cells in whole embryo culture. Day 8 CD-1 mouse embryos were cultured for 6-48 hr in the presence or absence of cis-RA at 2 X 10(-6) to 2 X 10(-5) M. Embryos either were fixed for light microscopy using Nichols' method for localization of CNC cells or were processed for scanning and transmission electron microscopy. At the light microscopic level, CNC cells in the mid-brain region of control embryos had migrated to the region of the first and second visceral arches after 6 hr in culture. Cis-RA interfered with this migration; CNC cells in treated embryos either did not leave the neuroepithelium (NE) or were aggregated near the NE. Autoradiographic studies indicated that cis-RA did not affect the overall viability or DNA synthesis of the CNC cells. However, at the TEM level, there was a dramatic increase in the number of cellular blebs in the CNC cells. Our results demonstrate a direct effect of 13-cis-RA on the CNC cells and suggest that this effect is due to alterations in the cell surface.     

Retinoic acid inhibits cardiac neural crest migration by blocking c-Jun N-terminal kinase activation

Retinoic acid (RA), a potent teratogen, produces a characteristic set of embryonic cardiovascular malformations similar to those observed in neural crest ablated avians. While the effects of RA on neural crest are well described, the molecular mechanism(s) of RA action on these cells is less clear. The present study examines the relationship between RA and mitogen-activated protein kinase signaling in neural crest cells and demonstrates that c-Jun N-terminal kinase (JNK) activation is severely repressed by RA. RA suppressed migration and proliferation of primary cultures of mouse neural crest cells treated in vitro as well as from animals treated in vivo. On Western blots, JNK activation/phosphorylation in neural crest cultures was reduced, while neither extracellular signal-regulated kinase (ERK) nor p38 pathways were affected. Both the dose-dependent stimulation of neural crest outgrowth and JNK phosphorylation by platelet-derived growth factor AA, which promotes outgrowth but not proliferation of neural crest cultures, were completely abrogated by RA. To establish the relevance of the JNK signaling pathway to cardiac neural crest migration, dominant negative adenoviral constructs were used to inhibit upstream activation of JNK or c-Jun downstream responses. Both adenoviral constructs markedly reduced neural crest cell outgrowth, while a dominant negative inhibitor of the p38 pathway had no effect. These data demonstrate that the JNK signaling pathway and c-Jun activation are critical for cardiac neural crest outgrowth and are potential targets for the action of RA.     

Development of the neural crest.

Mutations that affect the morphogenetic behaviour and differentiation of neural crest-derived cells in mouse embryos have been shown to alter genes that code for growth factors or growth factor receptors. Identification of these and other gene products provide opportunities to understand when and how developmentally distinct embryonic cell populations arise, and how interactions between localized developmental cues and responsive cell subpopulations can be modulated during development.     

Xenopus cadherin-11 restrains cranial neural crest migration and influences neural crest specification.

Cranial neural crest (CNC) cells migrate extensively, typically in a pattern of cell streams. In Xenopus, these cells express the adhesion molecule Xcadherin-11 (Xcad-11) as they begin to emigrate from the neural fold. In order to study the function of this molecule, we have overexpressed wild-type Xcad-11 as well as Xcad-11 mutants with cytoplasmic (deltacXcad-11) or extracellular (deltaeXcad-11) deletions. Green fluorescent protein (GFP) was used to mark injected cells. We then transplanted parts of the fluorescent CNC at the premigratory stage into non-injected host embryos. This altered not only migration, but also the expression of neural crest markers. Migration of transplanted cranial neural crest cells was blocked when full-length Xcad-11 or its mutant lacking the beta-catenin-binding site (deltacXcad-11) was overexpressed. In addition, the expression of neural crest markers (AP-2, Snail and twist) diminished within the first four hours after grafting, and disappeared completely after 18 hours. Instead, these grafts expressed neural markers (2G9, nrp-I and N-Tubulin). Beta-catenin co-expression, heterotopic transplantation of CNC cells into the pharyngeal pouch area or both in combination failed to prevent neural differentiation of the grafts. By contrast, deltaeXcad-11 overexpression resulted in premature emigration of cells from the transplants. The AP-2 and Snail patterns remained unaffected in these migrating grafts, while twist expression was strongly reduced. Co-expression of deltaeXcad-11 and beta-catenin was able to rescue the loss of twist expression, indicating that Wnt/beta-catenin signalling is required to maintain twist expression during migration. These results show that migration is a prerequisite for neural crest differentiation. Endogenous Xcad-11 delays CNC migration. Xcad-11 expression must, however, be balanced, as overexpression prevents migration and leads to neural marker expression. Although Wnt/beta-catenin signalling is required to sustain twist expression during migration, it is not sufficient to block neural differentiation in non-migrating grafts.     

Migration and proliferation of cultured neural crest cells in W mutant neural crest chimeras.

Chimeric mice, generated by aggregating preimplantation embryos, have been instrumental in the study of the development of coat color patterns in mammals. This approach, however, does not allow for direct experimental manipulation of the neural crest cells, which are the precursors of melanoblasts. We have devised a system that allows assessment of the developmental potential and migration of neural crest cells in vivo following their experimental manipulation in vitro. Cultured C57Bl/6 neural crest cells were microinjected in utero into neurulating Balb/c or W embryos and shown to contribute efficiently to pigmentation in the host animal. The resulting neural crest chimeras showed, however, different coat pigmentation patterns depending on the genotype of the host embryo. Whereas Balb/c neural crest chimeras showed very limited donor cell pigment contribution, restricted largely to the head, W mutant chimeras displayed extensive pigmentation throughout, often exceeding 50% of the coat. In contrast to Balb/c chimeras, where the donor melanoblasts appeared to have migrated primarily in the characteristic dorsoventral direction, in W mutants the injected cells appeared to migrate in the longitudinal as well as the dorsoventral direction, as if the cells were spreading through an empty space. This is consistent with the absence of a functional endogenous melanoblast population in W mutants, in contrast to Balb/c mice, which contain a full complement of melanocytes. Our results suggest that the W mutation disturbs migration and/or proliferation of endogenous melanoblasts. In order to obtain information on clonal size and extent of intermingling of donor cells, two genetically marked neural crest cell populations were mixed and coinjected into W embryos. In half of the tricolored chimeras, no co-localization of donor crest cells was observed, while, in the other half, a fine intermingling of donor-derived colors had occurred. These results are consistent with the hypothesis that pigmented areas in the chimeras can be derived from extensive proliferation of a few donor clones, which were able to colonize large territories in the host embryo. We have also analyzed the development of pigmentation in neural crest cultures in vitro, and found that neural tubes explanted from embryos carrying wt or weak W alleles produced pigmented melanocytes while more severe W genotypes were associated with deficient pigment formation in vitro.     

Draxin,an axon guidance protein,affects chick trunk neural crest migration

The neural crest is a multipotent population of migratory cells that arises in the central nervous system and subsequently migrates along defined stereotypic pathways. In the present work, we analyzed the role of a repulsive axon guidance protein, draxin, in the migration of neural crest cells. Draxin is expressed in the roof plate of the chick trunk spinal cord and around the early migration pathway of neural crest cells. Draxin modulates chick neural crest cell migration in vitro by reducing the polarization of these cells. When exposed to draxin, the velocity of migrating neural crest cells was reduced, and the cells changed direction so frequently that the net migration distance was also reduced. Overexpression of draxin also caused some early migrating neural crest cells to change direction to the dorsolateral pathway in the chick trunk region, presumably due to draxin’s inhibitory activity. These results demonstrate that draxin, an axon guidance protein, can also affect trunk neural crest migration in the chick embryo.     

Retinoic acid induces CDK inhibitors and growth arrest specific (Gas) genes in neural crest cells

Retinoic acid (RA), the active metabolite of vitamin A, regulates cellular growth and differentiation during embryonic development. In excess, this vitamin is also highly teratogenic to animals and humans. The neural crest is particularly sensitive to RA, and high levels adversely affect migration, proliferation and cell death. We investigated potential gene targets of RA associated with neural crest proliferation by determining RA-mediated changes in gene expression over time, using microarrays. Statistical analysis of the top ranked RA-regulated genes identified modest changes in multiple genes previously associated with cell cycle control and proliferation including the cyclin-dependent kinase inhibitors Cdkn1a (p21), Cdkn2b (p15(INK4b)), and Gas3/PMP22. The expression of p21 and p15(INK4b) contribute to decreased proliferation by blocking cell cycle progression at G1-S. This checkpoint is pivotal to decisions regulating proliferation, apoptosis, or differentiation. We have also confirmed the overexpression of Gas3/PMP22 in RA-treated neural crests, which is associated with cytoskeletal changes and increased apoptosis. Our results suggest that increases in multiple components of diverse regulatory pathways have an overall cumulative effect on cellular decisions. This heterogeneity contributes to the pleiotropic effects of RA, specifically those affecting proliferation and cell death.     

Hox genes, neural crest cells and branchial arch patterning.

Proper craniofacial development requires the orchestrated integration of multiple specialized tissue interactions. Recent analyses suggest that craniofacial development is not dependent upon neural crest pre-programming as previously thought but is regulated by a more complex integration of cell and tissue interactions. In the absence of neural crest cells it is still possible to obtain normal arch patterning indicating that neural crest is not responsible for patterning all of arch development. The mesoderm, endoderm and surface ectoderm tissues play a role in the patterning of the branchial arches, and there is now strong evidence that Hoxa2 acts as a selector gene for the pathways that govern second arch structures.     

Cytoplasmic protein methylation is essential for neural crest migration

As they initiate migration in vertebrate embryos, neural crest cells are enriched for methylation cycle enzymes, including S-adenosylhomocysteine hydrolase (SAHH), the only known enzyme to hydrolyze the feedback inhibitor of trans-methylation reactions. The importance of methylation in neural crest migration is unknown. Here, we show that SAHH is required for emigration of polarized neural crest cells, indicating that methylation is essential for neural crest migration. Although nuclear histone methylation regulates neural crest gene expression, SAHH and lysine-methylated proteins are abundant in the cytoplasm of migratory neural crest cells. Proteomic profiling of cytoplasmic, lysine-methylated proteins from migratory neural crest cells identified 182 proteins, several of which are cytoskeleton related. A methylation-resistant form of one of these proteins, the actin-binding protein elongation factor 1 alpha 1 (EF1α1), blocks neural crest migration. Altogether, these data reveal a novel and essential role for post-translational nonhistone protein methylation during neural crest migration and define a previously unknown requirement for EF1α1 methylation in migration.     

Retinoic acid-binding protein in the axolotl: distribution in mature tissues and time of appearance during limb regeneration

Analysis of cytoplasmic protein preparations from axolotl tissues revealed the presence of a cytoplasmic retinoic acid-binding protein (CRABP), of approximate molecular weight 17K. This protein was found to be present at various concentrations in skin, muscle, and limb tissue preparations, but not in liver and serum preparations. The distribution and molecular weight of this protein agrees with that reported in mammalian studies. The level of CRABP in cone stage blastemas was found to be significantly higher than that found in nonregenerating whole limb preparations. The level falls gradually, to approach normal, towards the completion of regeneration. Such an increase, at the start of regeneration, was not altered by 4 days pretreatment with 36 mg/liter all-trans-retinoic acid, a sufficient dose to produce pattern effects. Competition experiments confirmed that the all-trans and 13-cis isomers of retinoic acid bind to CRABP with similar high efficiencies, and that the arotinoid, Ro 13-6298, exhibits only a fraction of this binding activity. Retinol, retinol palmitate, and retinol acetate were unable to compete with [3H]retinoic acid for binding to CRABP. The results presented here are discussed in terms of their possible value to understanding pattern specification in the regenerating urodele limb.