The hormone retinoic acid (RA) has been implicated in the organization of the anteroposterior (AP) body axis. In this paper, we describe the effects of RA on the activity of the RA-inducible retinoic acid receptor-2 (RAR 2) promoter. When transgenic embryos carrying a RAR 2-lacZ reporter gene were exposed to a single dose of RA between gestational days 8.5 to 10.5, lacZ expression was induced in the anterior central nervous system (CNS). Strikingly, the transgene was expressed in a segmented pattern reminiscent of that of Drosophila ‘pair-rule’ genes. RA treatment of midgastrulation embryos at day 7.5 disturbed the segmentation and pro-duced severe craniofacial defects. We discuss the possi-bility that the entire anterior CNS is segmented and that this segmentation is reflected by the RAR 2-lacZ induc-tion pattern.

Vertebrate development involves the spatial organization of embryonic tissues into repetitive metameric units or seg-ments. These segments become most conspicuous as the paraxial mesoderm condenses into somites, epithelial struc-tures that will eventually give rise to dermis, axial muscles and the axial skeleton. It is primarily the axial skeleton that elicits the segmented pattern in the adult organism.

The segmented organization of the ectoderm was first described more than 160 years ago (Baer, 1828) as tran-siently formed metameric bulges of the neural tube (termed neuromeres; Orr, 1887). However, the functional signifi-cance of neuromeres has only recently come to light. Tissue transplantation and gene expression studies, as well as cell tracing experiments demonstrated clearly that neuromeres (rhombomeres), the overlying ectoderm, and the neural crest of the branchial region are domains for cellular differ-entiation, gene expression, and cell lineage restriction (Noden, 1983; Lumsden and Keynes, 1989; Murphy et al., 1989; Couly and Le Douarin, 1990; Fraser et al., 1990; Sundin and Eichele, 1990; Hunt et al., 1991). The status of the neural tube rostral to the hindbrain is still unclear. It is often considered to be unsegmented (Kandel et al., 1991), although neuromeres have been morphologically identified (Baer, 1828; Bartelmez, 1923; Adelmann, 1925; reviewed by Vaage, 1969).

The molecular mechanisms underlying segmentation in mammals are poorly understood. Several lines of evidence, however, suggest a role for the hormone retinoic acid. Exposure of gastrulating amphibian embryos to low RA concentrations results in CNS transformation (Dursten et al., 1989; Sive et al., 1990; Ruiz i Altaba and Jessell, 1991). Congenital CNS malformations and axial defects have also been reported after maternal RA administration in rodents (Kochhar, 1967; Morriss, 1972; Shenefelt, 1972) and humans (Lammer et al., 1985). It has recently been shown that RA can disturb the segmented organization of the hind-brain (Holder and Hill, 1991; Morriss-Kay et al., 1991; Papalopulu et al., 1991) and the axial skeleton (Kessel and Gruss, 1991; Kessel, 1992).

RA modulates the transcriptional activity of its target cells via nuclear retinoic acid receptors. These receptors belong to the steroid/thyroid hormone receptor superfamily of transcription factors (Benbrook et al., 1988; Brand et al., 1988; Giguére et al., 1987; Krust et al., 1989; Petkovich et al., 1987; Zelent et al., 1989). Two subfamilies (termed RARs and RXRs) have been distinguished based on sequence-homology and ligand-affinity (Mangelsdorf et al., 1990). Each sub-family is composed of at least three unlinked genes (-α, -β, and -γ) which in turn seem to encode several receptor-isoforms as a result of differential splicing and/or promoter usage (Giguére et al., 1990; Kastner et al., 1990; Zelent et al., 1991).

We (Reynolds et al., 1991), and others (Mendelsohn et al., 1991) recently utilized RAR 2 promoter sequences to drive the expression of a lacZ reporter gene in transgenic mice. We found that 250 bp of RAR 2 promoter sequences were sufficient to direct expression of the lacZ reporter gene in a pattern that closely resembled that of the endogenous RAR 2 (Reynolds et al., 1991; Ruberte et al., 1991).

Interestingly, RAR 2 expression can be induced by RA treatment in vitro in tissue culture cells (de The et al., 1989) and also in vivo in chicken limb buds (Noji et al., 1991). This induction is dependent on a retinoic acid responsive element (RAREβ) in the RAR 2 promoter. As RARβ can bind to this element, an autoregulatory mechanism for RAR 2 expression has been suggested (de The et al., 1990; Sucov et al., 1990). Maternal administration of RA may also influence RAR 2 expression in mammalian embryos, thus contributing to its teratogenic effects.

We describe here the expression pattern of RAR 2-lacZ reporter genes in transgenic mouse embryos after maternal RA administration. Exposure of day-7.5 embryos to low RA concentrations was highly teratogenic. The same con-centration had no obvious effects when administered at day 8.5. RAR 2-lacZ induction was found in the migrating cra-nial mesoderm of day-7.5 embryos and in a segmented pat-tern in the anterior CNS after RA treatment at day 8.5 or 9.5.

Animals

Inbred FVB mice (obtained from Taconic, Germantown, NY) were used in this study. Females were adapted to a dark-light cycle (light cycle from 6 a.m. to 6 p.m.) for at least one week prior to matings. Transgenic males were housed together with three non-transgenic females. Mice housed under these conditions will usually mate around midnight. Successful matings were evaluated every morning on the presence of vaginal plugs. The morning of this day was counted as day 0.25, the noon as day 0.5 and the evening at day 0.75.

Retinoic acid treatment and RARβ2-lacZ expression analysis

All-trans RA (Sigma) was dissolved in methylsulfoxide to a con-centration of 25 mg/ml. This stock-solution was aliquoted and stored in the dark at −20oC. It was suspended prior to use in sesame oil (Sigma) to a final concentration of 1 mg/ml. Pregnant mice were fed by gavage a single dose of 10 mg RA per kg body weight (0.2-0.25 ml). Times of RA treatment and the isolation of embryos is summarized in Table 1. Embryos were isolated, stained and sectioned as described previously (Reynolds et al., 1991). Briefly, embryos were dissected in phosphate-buffered saline, fixed in 4% formaldehyde, 0.5% glutaraldehyde and stained for β-gal activity in staining solution (1 mg/ml 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside, 5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, 2 mM MgCl2). Cryosections were cut from some of the embryos at 25 μm and counterstained with 0.1% neutral red.

Table 1.

Schedule of RA treatment and isolation of embryos

Schedule of RA treatment and isolation of embryos
Schedule of RA treatment and isolation of embryos

The data presented in this paper were obtained with two independent RAR 2-lacZ transgenic lines, TG3 and TG16. These lines carry the reporter-constructs ΔRLZ79 (TG16) or RLZ79 (TG3). These constructs differ in the 5′-RAR 2 promoter-sequences utilized to drive the expression of the lacZ gene (ΔRLZ79 : −250 to +625; RLZ79: −620 to +625; Reynolds et al., 1991). We have previously analyzed the expression pattern of the lacZ reporter gene during embry-onic development in these lines (Reynolds et al., 1991): no expression was detected in day-7.5 embryos. The onset of expression was between day 8.5 (4 somites) and day 8.75 (10 somites). The expression pattern of both lines very closely resembled that of the endogenous gene (Reynolds et al., 1991). Expression in the central nervous system was restricted to the spinal cord, with an anterior boundary at the level of rhombomere 7 (rh7, Figs 1D, 2G).

Fig. 1.

Whole-mount β-gal staining of embryos after maternal RA administration at day 7.25. Embryos were isolated and stained for β-gal activity 6 hours (A), 12 hours (B), 16 hours (C), 24 hours (E), and 48 hours (F) after treatment. Embryos were partially dissected from their membranes (B-F). Day-7.75 embryos (A) revealed the rapid RAR 2-lacZ induction in a wing-shaped pattern. This pattern suggests that lacZ is expressed in mesoderm cells which ingress through the primitive streak (ps) and migrate laterally into the head process (hp). a, anterior; p, posterior. Activation of transgene is transient. No expression in mesoderm derivatives (e.g. the heart) is detected at day 8.25 (B, C) or day 8.5 (E). RAR 2-lacZ induction is, however, found in the caudal hindbrain region as compared to untreated controls (D). Embryos isolated at day 9.5 (F) exhibit the same expression boundaries as untreated embryos of the same stage (Fig. 2G). Note the delay in neural tube closure after RA exposure (arrows). The asterisk indicates the otic vesicle. All bars, 250 μm.

Fig. 1.

Whole-mount β-gal staining of embryos after maternal RA administration at day 7.25. Embryos were isolated and stained for β-gal activity 6 hours (A), 12 hours (B), 16 hours (C), 24 hours (E), and 48 hours (F) after treatment. Embryos were partially dissected from their membranes (B-F). Day-7.75 embryos (A) revealed the rapid RAR 2-lacZ induction in a wing-shaped pattern. This pattern suggests that lacZ is expressed in mesoderm cells which ingress through the primitive streak (ps) and migrate laterally into the head process (hp). a, anterior; p, posterior. Activation of transgene is transient. No expression in mesoderm derivatives (e.g. the heart) is detected at day 8.25 (B, C) or day 8.5 (E). RAR 2-lacZ induction is, however, found in the caudal hindbrain region as compared to untreated controls (D). Embryos isolated at day 9.5 (F) exhibit the same expression boundaries as untreated embryos of the same stage (Fig. 2G). Note the delay in neural tube closure after RA exposure (arrows). The asterisk indicates the otic vesicle. All bars, 250 μm.

In addition to the comprehensive analysis of RA treat-ment on RAR 2-lacZ expression in transgenic lines TG3 and TG16, we have also analyzed two other RLZ79 and ΔRLZ79 lines at day 9.5 and 10.5. The expression pattern after RA treatment was identical at all times in all trans-genic lines. Hence, the unique expression pattern described below was independent of the integration site of the trans-gene.

Effects of RA exposure at day 7.5

Gastrulating rodent embryos are particularly sensitive to low doses of RA. Administration of a single dose of all-trans-RA (10 mg/kg) by gavage at day 6.75 produced severe teratogenic effects in 50% of all litters, while RA treatment at day 7.25 and day 7.75 resulted in a delay in neural tube closure and microcephaly in 100% of all litters (group B-E in Table 1; Fig. 1F). Exposure to the same dose of RA at day 8.5 caused brain defects with only very low incidence (1 out of 9 litter; group H in Table 1). Hence, the period of strongest RA sensitivity starts at day 7 and ends at day 8.

To determine if RAR 2-lacZ expression could be pre-maturely induced by RA in gastrulating embryos, day-7.75 embryos were stained for β-gal activity 6 hours after RA treatment (group A in Table 1). We found strong lacZ expression in a characteristic, wing-shaped pattern with the origin in the primitive streak and the tips at the head process (Fig. 1A). This pattern is indicative of migrating cranial mesoderm cells (Lawson et al., 1991). No staining was detected in the neuroectoderm (Figs 1A, 3A). Hence, the RAR 2 promoter present in our constructs is selectively activated in the cranial mesoderm. This suggests that meso-derm transformation may play a role in the teratogenic effects of RA treatment at this stage.

When embryos were isolated 12 to 16 hours after RA administration (at day 7.75), we found that RAR 2-lacZ was induced in the midbrain and hindbrain. Expression levels were high in the caudal hindbrain and decreased progres-sively towards the midbrain (Fig. 1B,C). Expression in the trunk region was similar to untreated embryos (Fig. 1D), but the expression level in somites was slightly elevated. RAR 2-lacZ induction was still detectable after 24 hours (day 8.5; group C in Table 1; Fig. 1E), but embryos iso-lated at day 9.5 (group D in Table 1; Fig. 1F) or 11.5 (group E in Table 1) had the same anterior and posterior expression boundaries as untreated embryos (compare Figs 1F, 2G).

RA exposure at later stages induces segmented RARβ2-lacZ expression

To test if RAR 2-lacZ could be induced after the RA sen-sitive period, we treated them with the same dose between day 8 and day 12 (Table 1). When embryos were stained for β-gal activity 6-16 hours later, we found strong lacZ expression in a segment-like pattern in the CNS (Fig. 2).

Fig. 2.

Whole-mount staining of embryos treated with RA after the onset of CNS segmentation at day 8.25 (A-C), day 8.75 (D), day 9.25 (E, F) or day 9.75 (H, I). Embryos were isolated 6 hours (A, B, C, E, F) or 16 hours (D, H, I) after treatment. (G) RAR 2-lacZ expression in untreated embryos. The asterisk indicates the otic vesicle at the level of rh5. RAR 2-lacZ induction in a segmented pattern was first detected in day-8.75 embryos (A, B). At this stage, induction is strong in rh4 and rh6. Individual rhombomeres are identified by numbers. Weak transgene expression was found in pro-rhombomere A (RhA, open arrow), in the midbrain and the forebrain. Neural crest cells derived from rh4 (arrows) also express the transgene. Expression in the mid- and forebrain is markedly enhanced in embryos that have already turned (C). These rather diffuse RAR 2-lacZ induction domains become subsequently very distinct at day 9.5 (D-F) and delineate alternating expressing and nonexpressing segments (classified as telencephalic domains t1 and t2; diencephalic domains d1, d2, d3; and mesencephalic domains m1 and m2). Note that expression levels are higher in embryos isolated 6 hours after RA treatment (E, F). The RAR 2-lacZ induction pattern remains basically unchanged in day-10.5 embryos (H, I). rh, rhombencephalon; m, mesencephalon; t, telencephalon. Arrowheads indicate corresponding expression domains in day-9.25 (D) and day-10.25 embryos (H). All bars, 250 μm.

Fig. 2.

Whole-mount staining of embryos treated with RA after the onset of CNS segmentation at day 8.25 (A-C), day 8.75 (D), day 9.25 (E, F) or day 9.75 (H, I). Embryos were isolated 6 hours (A, B, C, E, F) or 16 hours (D, H, I) after treatment. (G) RAR 2-lacZ expression in untreated embryos. The asterisk indicates the otic vesicle at the level of rh5. RAR 2-lacZ induction in a segmented pattern was first detected in day-8.75 embryos (A, B). At this stage, induction is strong in rh4 and rh6. Individual rhombomeres are identified by numbers. Weak transgene expression was found in pro-rhombomere A (RhA, open arrow), in the midbrain and the forebrain. Neural crest cells derived from rh4 (arrows) also express the transgene. Expression in the mid- and forebrain is markedly enhanced in embryos that have already turned (C). These rather diffuse RAR 2-lacZ induction domains become subsequently very distinct at day 9.5 (D-F) and delineate alternating expressing and nonexpressing segments (classified as telencephalic domains t1 and t2; diencephalic domains d1, d2, d3; and mesencephalic domains m1 and m2). Note that expression levels are higher in embryos isolated 6 hours after RA treatment (E, F). The RAR 2-lacZ induction pattern remains basically unchanged in day-10.5 embryos (H, I). rh, rhombencephalon; m, mesencephalon; t, telencephalon. Arrowheads indicate corresponding expression domains in day-9.25 (D) and day-10.25 embryos (H). All bars, 250 μm.

Segmentation of the hindbrain involves successive sub-divisions of three pro-rhombomeres (RhA, RhB and RhC) into 8 final rhombomeres (rh1 to rh8; for review see Vaage, 1969; Lumsden, 1990). The central rhombomeres 4 and 5 are the first to establish clear boundaries. We found strong lacZ expression in day-8.75 embryos that had not yet turned (6-8 somites) 6 hours after RA exposure (Group F in Table 1) in rh4 and caudal to rh5. Rh5 and rh3 became resistant to RAR 2-lacZ induction. A dorsoventral band of express-ing cells spanned from rh4 to the second brachial arch (Fig. 2B). The analysis of serial horizontal sections revealed that these cells are migrating neural crest cells (Fig. 3B). We also found RAR 2-lacZ induction in one domain rostral to rh3. This domain corresponds to pro-rhombomere A, which subsequently divides into rh1 and rh2. RAR 2-lacZ induc-tion was again transient as embryos isolated between day 10.5 and 13.5 (Group H in Table 1) had the same expression pattern as untreated embryos.

Fig. 3.

(A) Sagittal section through a day-7.75 embryo as shown in Fig. 1A. Embryos were isolated 6 hours after RA treatment, stained for β-gal expression, sectioned, and counterstained with neutral red. The primitive streak (PS) shows high lacZ expression levels, but not the neural fold (NF) region. (B) Transverse section through a day-8.75 embryo 6 hours after RA treatment demonstrates RAR 2-lacZ induction in migrating neural crest (NC) cells originating from rh4. Picture was taken with Nomarski optics. NG, neural groove; F, foregut. All bars, 100 μm.

Fig. 3.

(A) Sagittal section through a day-7.75 embryo as shown in Fig. 1A. Embryos were isolated 6 hours after RA treatment, stained for β-gal expression, sectioned, and counterstained with neutral red. The primitive streak (PS) shows high lacZ expression levels, but not the neural fold (NF) region. (B) Transverse section through a day-8.75 embryo 6 hours after RA treatment demonstrates RAR 2-lacZ induction in migrating neural crest (NC) cells originating from rh4. Picture was taken with Nomarski optics. NG, neural groove; F, foregut. All bars, 100 μm.

Rhombomeres are fully developed at day 9.5 (Tuckett et al., 1985) and RAR 2-lacZ was markedly induced in rh1, rh4, and caudal of rh5 6 hours after treatment. Expression was also weakly induced in rh2, rh3 and rh5 (Fig. 2E, Group I in Table 1). These differential expression levels were most distinct in day-9.5 embryos that were treated at day 8.75 (group G in Table 1; Fig. 2D). No expression was detected in the midline region and at inter-rhombomeric junctions.

RAR 2-lacZ induction was also detected in the mesen-cephalon and prosencephalon at day 8.75 (Fig. 2A,C). The expression domains were initially rather diffuse, but sub-sequently formed clearly demarcated bands dividing the anteroposterior (A-P) axis of the brain into alternating expressing and non-expressing regions (Figs 2D,F,H,I, 4). We classified these regions consecutively, into two telen-cephalic (t1 and t2), three diencephalic (d1, d2, and d3) and two mesencephalic (m1 and m2) domains (Figs 2F, 4B, 6).

The anterior limit of the expressing mesencephalic domain m1 was at the mes-diencephalic boundary (Fig. 2D,F,H,I). This was particularly evident in parasagittal sec-tions through the ventral mesencephalon, where β-gal stain-ing was just rostral to the tuberculum posterius (Fig. 4A,B). In the diencephalon, we found RAR 2-lacZ induction in the lateral hypothalamus rostral to the optic stalk (d2; Figs 2F,H, 4A,B). The most anterior region of the diencephalon (d1) did not express the transgene. In the telencephalon, we found an expression domain (t2) in the caudal region of the lateral vesicles (Figs 2C,D,F,H, 4A). The posterior bound-ary of this expression domain delineated the di-telen-cephalic boundary.

Fig. 4.

(A) Parasagittal sections through a day-10.25 embryo demonstrates the segmented RAR 2-lacZ induction in the CNS. Embryos were treated with RA at day 9.75 and isolated approximately 12 hours later. T, lateral ventricle (telencoel); III, third ventricle; IV, fourth ventricle. (B) High magnification of the area indicated in A shows the alternating RAR 2-lacZ expression domains in the ventrolateral region of the diencephalon (d2), the mesencephalon (m1) and the rostral hindbrain (rh1). Arrows indicate expression boundaries. All bars, 100 μm.

Fig. 4.

(A) Parasagittal sections through a day-10.25 embryo demonstrates the segmented RAR 2-lacZ induction in the CNS. Embryos were treated with RA at day 9.75 and isolated approximately 12 hours later. T, lateral ventricle (telencoel); III, third ventricle; IV, fourth ventricle. (B) High magnification of the area indicated in A shows the alternating RAR 2-lacZ expression domains in the ventrolateral region of the diencephalon (d2), the mesencephalon (m1) and the rostral hindbrain (rh1). Arrows indicate expression boundaries. All bars, 100 μm.

High levels of induction were still detected in the mid-brain and caudal hindbrain of day-11.25 embryos (group L in Table 1), while induction in the forebrain was weak (data not shown). In day-12.25 embryos (group M in Table 1), RAR 2-lacZ expression levels were still high in the caudal hindbrain, and low in the midbrain. No induction was seen in the forebrain. lacZ expression in the spinal cord of these embryos was uninterrupted, while untreated embryos had two expression domains at this stage (data not shown).

RA exposure at day 7.5 disrupts the segment-like induction pattern

The hindbrain neuroepithelium of embryos treated with RA at day 7.75 appeared very smooth (Fig. 1E), in contrast to the rhombomeric bulges seen in untreated embryos (Fig. 2A,B). This finding supports the observation that RA expo-sure during gastrulation can lead to morphological alter-ations of the hindbrain (Holder and Hill, 1991; Morriss-Kay et al., 1991; Papalopulu et al., 1991).

To analyze the effects of RA on the organization of the neuromeres more closely, we re-examined the RAR 2-lacZ induction pattern in embryos that were treated twice (group N and O in Table 1). The first exposure at day 7.75, causes teratogenic effects and induces RAR 2-lacZ transiently. This transient induction does not last much longer than 24 hours (see Fig. 1). The RAR 2-lacZ expression pattern induced by a second RA treatment at day 8.25 or day 9.25 should therefore reveal alterations in the CNS segmentation caused by the first treatment. Embryos were isolated and stained for β-gal activity 4 hours after the second treatment. The results are shown in Fig. 5.

Fig. 5.

Whole-mount β-gal staining of embryos that were exposed to RA at day 7.75 and again at day 8.25 (A, B) or day 9.25 (C-E). All embryos were isolated 4 hours after the second treatment. RA exposure during gastrulation made embryos initially highly susceptible to RAR 2-lacZ induction and the second RA treatment at day 8.25 induced the transgene in virtually all tissues, except the most anterior region of the neuroectoderm (A, B). Subsequently, anterior regions of the embryo became resistant to RAR 2-lacZ induction and we observed three different lacZ expression patterns after a second treatment at day 9.25. In the first group (C), RAR 2-lacZ was induced uniformly in the hindbrain, in small cell patches in the midbrain and in one domain in the forebrain. The otic vesicle is indicated by an arrow. All tissues rostral to rh6 were resistant to RAR 2-lacZ induction in embryos of the second group (D). Embryos of the third group (E) showed expression of the transgene in a large domain rostral to the otic vesicle (arrow) and in very small domains in the midbrain and forebrain. All bars, 250 μm.

Fig. 5.

Whole-mount β-gal staining of embryos that were exposed to RA at day 7.75 and again at day 8.25 (A, B) or day 9.25 (C-E). All embryos were isolated 4 hours after the second treatment. RA exposure during gastrulation made embryos initially highly susceptible to RAR 2-lacZ induction and the second RA treatment at day 8.25 induced the transgene in virtually all tissues, except the most anterior region of the neuroectoderm (A, B). Subsequently, anterior regions of the embryo became resistant to RAR 2-lacZ induction and we observed three different lacZ expression patterns after a second treatment at day 9.25. In the first group (C), RAR 2-lacZ was induced uniformly in the hindbrain, in small cell patches in the midbrain and in one domain in the forebrain. The otic vesicle is indicated by an arrow. All tissues rostral to rh6 were resistant to RAR 2-lacZ induction in embryos of the second group (D). Embryos of the third group (E) showed expression of the transgene in a large domain rostral to the otic vesicle (arrow) and in very small domains in the midbrain and forebrain. All bars, 250 μm.

All embryos analyzed at day 8.75 expressed the transgene at very high levels in the heart and in the entire CNS except in the rostral tip of the headfolds after the second RA treatment (Fig. 5A,B; Group N in Table 1). We did not find any segmented expression pattern. These results indi-cate that RA treatment during gastrulation induces molec-ular transformations so that most tissues become suscepti-ble to RAR 2-lacZ induction.

The expression pattern after a second treatment at day 9.25 (Group O in Table 1) was more complex. Embryos of this group fall into three categories. The first group is shown in Fig. 5C. In those embryos, expression was strong in the entire hindbrain region and exhibited no signs of hindbrain segmentation. Expression was also found in small patches in midbrain and at a low level in one domain in the forebrain. There was no expression in the anterior CNS of embryos of the second category (Fig. 5D). How-ever, RAR 2-lacZ expression was induced in the caudal hindbrain (compare Fig. 5D with 1F). Embryos of the third category (Fig. 5E) showed expression in the entire hind-brain except for rh5. There was also expression in small domains in the prosencephalon and mesencephalon. Common among these three different categories was RAR 2-lacZ induction caudal to the otic vesicle, and a reduction in size of the midbrain and forebrain domains that are susceptible to RAR 2-lacZ induction. It is impor-tant to note that these different categories were equally rep-resented in both litters that were analyzed. We can, there-fore, exclude the possibility of simple experimental variations. These results show that the disruption of the morphological segmentation of the hindbrain is reflected by alterations in the RAR 2-lacZ expression domains. They also suggest that RA disturbs the segmented organization of anterior brain regions.

Morphological and evolutionary considerations argue against a common segmented ancestor of insects and ver-tebrates. Yet, studies in Drosophila and mice established remarkable analogies in the molecular mechanisms of pat-tern formation in these diverse species. It appears that both organisms have utilized similar segmentation strategies and recruited similar groups of genes to establish the primary body axis (Wilkinson and Krumlauf, 1990; Kessel and Gruss, 1990). The analysis of homeobox containing genes in particular led to the conclusion that these genes have (i) a similar genomic organization in gene clusters; (ii) simi-lar expression patterns in segmented structures along the AP axis and in respect to their genomic position; and (iii) similar functions, i.e. the specification of positional infor-mation.

We have analyzed the activity of the RA inducible RAR 2 promoter in transgenic RAR 2-lacZ mouse embryos after maternal RA administration. Strikingly, RAR 2-lacZ was strongly induced in a segment-like pattern in alternat-ing rhombomeres and in the anterior CNS. This pattern is very reminiscent of the expression pattern of Drosophila ‘pair-rule’ genes and suggests that segmentation is a fea-ture of the whole brain.

The segmented organization of the neural tube is easily discernible by morphological means only in the hindbrain region during a brief developmental period. Segmentation of the hindbrain is, however, supported by cell tracing experiments which demonstrated that rhombomeres are units of cell lineage restriction (Fraser et al., 1990). More-over, the expression patterns of Krox-20 and many Hox genes correlates well with segment boundaries (Wilkinson et al., 1989; Murphy et al., 1989; Sundin and Eichele, 1990).

The RAR 2-lacZ induction domains at day 8.5 and day 9.5 are summarized in Fig. 6. At day 9.5, RAR 2-lacZ induction was strong in rh4 and rh6 and low in rh3 and rh5. We also found an alternating pattern in the mesencephalon and prosencephalon. The ‘pair-rule’ type pattern was dis-turbed only in the anterior rhombomeres. Here, RAR 2-lacZ was strongly induced in rh1, and very weakly in rh2 and rh3. However, day-8.5 embryos (prior to the differentiation of rhombomeres 1 to 3) showed expression in only one domain in the rostral hindbrain, in RhA, and thus showed a perfectly alternating expression pattern (Fig. 2C). Rh1 and rh2 are the latest to differentiate (Lumsden, 1990). Hence, it is possible that the subdivision of rh1 and rh2 is a sec-ondary event to an underlying perfectly alternating two-seg-ment-repeated molecular patterning in CNS development.

Fig. 6.

Summary of the RAR 2-lacZ induction pattern in the brain at day 8.5 and 9.5. Rhombomeres 3-8 are readily detectable in day-8.5 embryos. However, the most anterior rhombomere RhA has yet to divide into rh1 and rh2. The formation of neuromeres in the mesencephalon (M) is just starting at this developmental stage (Tuckett et al., 1985). RAR 2-lacZ induction in the prosencephalon (P), mesencephalon, and in RhA was found at lower levels and without clearly demarcated boundaries (indicated by grey areas). Segment boundaries are fully established at day 9.5. Indicated are two telencephalic (t1, t2), three diencephalic (d1 to d3), two mesencephalic (m1, m2), and eight rhombencephalic (rh1 to rh8) domains. Relative lacZ expression levels in different brain regions are indicated by differential shading.

Fig. 6.

Summary of the RAR 2-lacZ induction pattern in the brain at day 8.5 and 9.5. Rhombomeres 3-8 are readily detectable in day-8.5 embryos. However, the most anterior rhombomere RhA has yet to divide into rh1 and rh2. The formation of neuromeres in the mesencephalon (M) is just starting at this developmental stage (Tuckett et al., 1985). RAR 2-lacZ induction in the prosencephalon (P), mesencephalon, and in RhA was found at lower levels and without clearly demarcated boundaries (indicated by grey areas). Segment boundaries are fully established at day 9.5. Indicated are two telencephalic (t1, t2), three diencephalic (d1 to d3), two mesencephalic (m1, m2), and eight rhombencephalic (rh1 to rh8) domains. Relative lacZ expression levels in different brain regions are indicated by differential shading.

Evidence for functional diversity of alternating rhom-bomeres came from the analysis of neurogenesis in the chick hindbrain (Lumsden and Keynes, 1989). Motor neu-rons develop first in rhombomeres 2, 4, and 6 and subse-quently in rhombomeres 3 and 5. Furthermore, motor neuron pathways are different in even-numbered and odd-numbered rhombomeres.

Several morphological studies suggested that the neural tube rostral to the hindbrain has a neuromeric organization (Baer, 1828; Bartelmez, 1923; Adelmann, 1925; Vaage, 1969; Tuckett et al., 1985). But, the number of potential neuromeres and their developmental appearence remained controversial due to the lack of supportive experimental cel-lular and molecular evidence (reviewed by Vaage, 1969). A relatively good consensus concerning the number of neu-romeres in anterior brain regions exists only for the mes-encephalon, where most authors found two neuromeres or mesomeres (Bartelmez, 1923; Adelmann, 1925; Bartelmez and Evans, 1926; Bergquist, 1952; Tuckett et al., 1985; reviewed by Bergquist and Källén, 1954; Vaage, 1969). The neuromeric organization of the mesencephalon becomes evident at the beginning of day 8.5 with the subdivision of an initially single large mesomeric sulcus (Tuckett et al., 1985). The rostral neuromere subsequently enlarges and forms the optic tectum later. RAR 2-lacZ induction in the mesencephalon was first detected in the rostral region (m1) of day-8.5 embryos. Thus, RAR 2-lacZ induction correlates well with the morphological onset of mesencephalon seg-mentation.

The RAR 2-lacZ expression pattern suggests the pres-ence of two possible segments in the telencephalon and three in the diencephalon. A subdivision of the dien-cephalon into three segments has also been suggested on the basis of the pattern of neurogenesis and cell migration (Bergquist and Källén, 1954; Puelles et al., 1987). Close examination of cell tracing experiments in the diencephalon indicates, however, the presence of four segments (Vigdor and Stern, personal communication). It is possible that one of the diencephalic segments that we suggest here, subse-quently divides to yield a total of four final segments. This situation would thus be similar to that seen in the rostral hindbrain. Interestingly, thymidine radiographic studies indicated that the fate of hypothalamic and thalamic neu-rons may be predetermined by the mosaic organization of the diencephalon (Altman and Bayer, 1979). Support for segmentation of the telencephalon was also recently pro-vided by observations of the expression of murine homologs to the Drosophila homeobox gene distalless in distinct areas of the forebrain (Price et al., 1991; Porteus et al., 1991).

The segmented RAR 2-lacZ expression pattern was dis-rupted by RA administration at day 7.5 prior to the onset of neuromere morphogenesis. RA treatment at this stage was also highly teratogenic. These results suggest that endogenous RA plays an important role in the specification of the CNS axis during early stages of neurogenesis. It has previously been shown that exposure of gastrulating mam-malian embryos to exogenous RA can result in homeotic transformations of the axial skeleton (Kessel and Gruss, 1991) and in disruption of the segmented hindbrain organization (Morriss-Kay et al., 1991). Our results indicate that RA has a similar effect on the midbrain and forebrain.

What could be the molecular basis for the segmented RAR 2-lacZ induction by RA in the CNS? When the RAREβ is placed in the context of heterologous promoters, RA-dependent transcriptional activation can be detected in virtually all cells in tissue culture (Sucov et al., 1990) as well as in mouse embryos (Rossant et al., 1991). However, the RAR 2 promoter is itself apparently very tightly regu-lated. Rossant et al. (1990) used a RARE -hsp-lacZ gene to assess RA sensitive tissues during murine development. They found strong expression of the RARE-hsp-lacZ expression throughout the CNS after exposure of embryos to RA. It is therefore likely that the segmented RAR 2-lacZ expression pattern results from the combination of elevated free RA levels and the tissue specific distribution of trans-acting factors. The absolute concentration of RA may be of little influence on the RAR 2-lacZ expression as we found the identical pattern when dams were treated on day 9.5 with a 10-fold higher dose of RA (100 mg/kg; A. Z. and A. Z., unpublished observations). This dose was sufficient to induce strong teratogenic effects.

It has recently been shown that RARβ and the TATA box binding factor TFIID act cooperatively to transactivate the RAR 2 promoter. Interestingly, this cooperativity was dependent on the promoter context and was not observed, when the RAREβ was inserted into the TK or SV40 pro-moters (Berkenstam et al., 1992). Furthermore, the cooper-ativity was dependent on an E1A-like activity found in embryonic carcinoma (EC) cells (Imperiale et al., 1984). It is conceivable that RAR 2-lacZ activation is governed by the distribution of such an E1A-like factor in the develop-ing CNS. It will be most challenging to isolate the factors that direct the segmented RAR 2-lacZ expression, as they may have functional similarity to ‘pair-rule’ genes and play a central role in the organization of the AP axis of the neural tube.

We would like to thank Kay Reynolds for assistance; Dr Mike Brownstein for support; Dr Michael Vigdor for sharing his unpub-lished results; and Drs Andrew Lumsden, Dervla Mellerick-Dressler, Michael Kessel, and Andreas Kurtz for helpful com-ments on the manuscript. Andreas Z. was supported by the Deutsche Forschungsgemeinschaft. Anne Z. was supported by the National Alliance for Research on Schizophrenia and Depression.

Adelmann
,
H. B.
(
1925
).
The development of the neural folds and cranial ganglia of the rat
.
J. Comp. Neurol
.
39
,
19
171
.
Altman
,
J.
and
Bayer
,
S. A.
(
1979
).
Development of the diencephalon in the rat. IV. Re-evaluation of the embryonic development of the thalamus on the basis of thymidine-radiographic datings
.
J. Comp. Neurol
.
188
,
501
524
.
Baer
,
K. E. von
(
1828
).
Ueber Entwicklungsgeschichte der Thiere. Bd
., pp.
1
3
.
Königsberg
:
Bornträger
.
Bartelmez
,
G. W.
(
1923
).
The subdivisions of the neural folds in man
.
J. Comp. Neurol
.
35
,
231
247
.
Bartelmez
,
G. W.
and
Evans
,
H. M.
(
1926
).
Development of the human embryo during the period of somite formation, including embryos with 2 to 16 pairs of somites
.
Contr. Embryol. Carneg. Inst
.
17
,
1
67
.
Benbrook
,
D.
,
Lernhard
,
E.
and
Pfahl
,
M.
(
1988
).
A new retinoic acid receptor identified from a hepatocellular carcinoma
.
Nature
333
,
669
672
.
Bergquist
,
H.
(
1952
).
Studies on the cerebral tube in vertebrates. The neuromeres
.
Acta Zool
.
33
,
117
187
.
Bergquist
,
H.
and
Källén
,
B.
(
1954
).
Notes on the early histogenesis and morphogenesis of the central nervous system in vertebrates
.
J. Comp. Neurol
.
100
,
627
659
.
Berkenstam
,
A.
,
Ruiz
,
M. d. M. V.
,
Barettino
,
D.
,
Horikoshi
,
M.
and
Stunnenberg
,
H. G.
(
1992
).
Cooperativity in transactivation between retinoic acid receptor and TFIID requires an activity analogous to E1A
.
Cell
69
,
401
412
.
Brand
,
N.
,
Petkovich
,
M.
,
Krust
,
A.
,
Chambon
,
P.
,
De The
,
H.
,
Marchio
,
A.
,
Tiollais
,
P.
and
Dejean
,
A.
(
1988
).
Identification of a second human retinoic acid receptor
.
Nature
332
,
850
853
.
Couly
,
G.
and
Le Douarin
,
N. M.
(
1990
).
Head morphogenesis in embryonic avian chimeras: evidence for a segmental pattern in the ectoderm corresponding to the neuromeres
.
Development
108
,
543
558
.
de The
,
H.
,
Marchio
,
A.
,
Tiollais
,
P.
and
Dejean
,
A.
(
1989
).
Differential expression and ligand regulation of the retinoic acid receptor α and β genes
.
EMBO J
.
8
,
429
433
.
de The
,
H.
,
del Mar Vivanco-Ruiz
,
M.
,
Tiollais
,
P.
,
Stunnenberg
,
H.
and
Dejean
,
A.
(
1990
).
Identification of a retinoic acid responsive element in the retinoic acid receptor β gene
.
Nature
343
,
177
180
.
Dursten
,
A. J.
,
Timmermans
,
J. P. M.
,
Hage
,
W. J.
,
Hendriks
,
H. F. J.
,
de Vries
,
N. J.
,
Heideveld
,
M.
and
Nieuwkoop
,
P. D.
(
1989
).
Retinoic acid causes an anteroposterior transformation in the developing central nervous system
.
Nature
340
,
140
144
.
Fraser
,
S.
,
Keynes
,
R.
and
Lumsden
,
A.
(
1990
).
Segmentation in the chick embryo hindbrain is defined by cell lineage restrictions
.
Nature
344
,
431
435
.
Giguére
,
V.
,
Ong
,
E. S.
,
Segui
,
P.
and
Evans
,
R. M.
(
1987
).
Identification of a receptor for the morphogen retinoic acid
.
Nature
330
,
624
629
.
Giguére
,
V.
,
Shago
,
M.
,
Zirngibl
,
R.
,
Tate
,
P.
,
Rossant
,
J.
and
Varmuza
,
S.
(
1990
).
Identification of a novel isoform of the retinoic acid receptor γ expressed in the mouse embryo
.
Mol. Cell. Biol
.
10
,
2335
2340
.
Holder
,
N.
and
Hill
,
J.
(
1991
).
Retinoic acid modifies development of the midbrain-hindbrain border and affects cranial ganglion formation in zebrafish embryos
.
Development
113
,
1159
1170
.
Hunt
,
P.
,
Gulisano
,
M.
,
Cook
,
M.
,
Sham
,
M.-H.
,
Faiella
,
A.
,
Wilkinson
,
D.
,
Boncinelli
,
E.
and
Krumlauf
,
R.
(
1991
).
A distinct Hox code for the branchial region of the vertebrate head
.
Nature
353
,
861
864
.
Imperiale
,
M. J.
,
Kao
,
H. T.
,
Feldman
,
L. T.
,
Nevins
,
J. R.
and
Strickland
,
S.
(
1984
).
Common control of the heat shock gene and early adenovirus genes: evidence for a cellular E1A-like activity
.
Mol. Cell. Biol
.
4
,
867
874
.
Kandel
,
E. R.
,
Schwartz
,
J. H.
and
Jessell
,
T. M.
(
1991
).
Principles of Neural Science
, pp.
1135
.
New York
:
Elsevier
.
Kastner
,
P.
,
Krust
,
A.
,
Mendelsohn
,
C.
,
Garnier
,
J.-M.
,
Zelent
,
A.
,
Leroy
,
P.
,
Staub
,
A.
and
Chambon
,
P.
(
1990
).
Murine isoforms of retinoic acid receptor-γ with specific patterns of expression
.
Proc. natl. Acad. Sci. USA
87
,
2700
2704
.
Kessel
,
M.
and
Gruss
,
P.
(
1990
).
Murine developmental control genes
.
Science
249
,
374
379
.
Kessel
,
M.
and
Gruss
,
P.
(
1991
).
Homeotic transformations of murine vertebrae and concomitant alteration of Hox codes induced by retinoic acid
.
Cell
67
,
89
104
.
Kessel
,
M.
(
1992
).
Respecification of vertebral identities by retinoic acid
.
Development
115
,
487
501
.
Kochhar
,
D. M.
(
1967
).
Teratogenic activity of retinoic acid
.
Acta Path. Microbiol. Scandinav
.
70
,
398
404
.
Krust
,
A.
,
Kastner
,
P.
,
Petkovich
,
M.
,
Zelent
,
A.
and
Chambon
,
P.
(
1989
).
A third human retinoic acid receptor, hRAR-γ
.
Proc. Natl. Acad. Sci. USA
86
,
5310
5314
.
Lammer
,
E. J.
,
Chen
,
D. T.
,
Hoar
,
R. M.
,
Agnish
,
N. D.
,
Benke
,
P. J.
,
Braun
,
J. T.
,
Curry
,
C. J.
,
Fernhoff
,
P. M.
,
Grix
Jr,
A. W.
,
Lott
,
I. T.
,
Richard
,
J. M.
and
Sun
,
S. C.
(
1985
).
Retinoic acid embryopathy
.
N. Eng. J. Med
.
313
,
837
841
.
Lawson
,
K. A.
,
Meneses
,.
J.
and
Pedersen
,
R. A.
(
1991
).
Clonal analysis of epiblast fate during germ layer formation in the mouse embryo
.
Development
113
,
891
911
.
Lumsden
,
A.
and
Keynes
,
R.
(
1989
).
Segmental patterns of neuronal development in the chick hindbrain
.
Nature
337
,
424
428
.
Lumsden
,
A.
(
1990
).
The cellular basis of segmentation in the developing hindbrain
.
Trends Neurosc
.
13
,
329
335
.
Mangelsdorf
,
D. J.
,
Ong
,
E. S.
,
Dyck
,
J. A.
and
Evans
,
R. M.
(
1990
).
Nuclear receptor that identifies a novel retinoic acid response pathway
.
Nature
345
,
224
229
.
Mendelsohn
,
C.
,
Ruberte
,
E.
,
LeMeur
,
M.
,
Morriss-Kay
,
G.
and
Chambon
,
P.
(
1991
).
Developmental analysis of the retinoic acid-inducible RAR-β2 promoter in transgenic animals
.
Development
113
,
723
734
.
Morriss
,
G. M.
(
1972
).
Morphogenesis of the malformations induced in rat embryos by maternal hypervitaminosis A
.
J. Anat
.
113
,
241
250
.
Morriss-Kay
,
G. M.
,
Murphy
,
P.
,
Hill
,
R. E.
and
Davidson
,
D. R.
(
1991
).
Effects of retinoic acid excess on expression of Hox-2.9 and Krox-20 and on morphological segmentation in the hindbrain of mouse embryos
.
EMBO J
.
10
,
2985
2995
.
Murphy
,
P.
,
Davidson
,
D. R.
and
Hill
,
R. E.
(
1989
).
Segment-specific expression of a homeobox-containing gene in the mouse hindbrain
.
Nature
341
,
156
159
.
Noden
,
D. M.
(
1983
).
The role of the neural crest in patterning of avian cranial skeletal, connective and muscle tissues
.
Dev. Biol
.
96
,
144
165
.
Noji
,
S.
,
Nohno
,
T.
,
Koyama
,
E.
,
Muto
,
K.
,
Ohyama
,
K.
,
Aoki
,
Y.
,
Tamura
,
K.
,
Ohsugi
,
K.
,
Ide
,
H.
,
Taniguchi
,
S.
and
Saito
,
T.
(
1991
).
Retinoic acid induces polarizing activity but is unlikely to be a morphogen in the chick limb bud
.
Nature
350
,
83
86
.
Orr
,
H.
(
1887
).
Contribution to the embryology of the lizard
.
J. Morphol
.
1
,
311
372
.
Papalopulu
,
N.
,
Clarke
,
J. D.
,
Bradley
,
L.
,
Wilkinson
,
D.
,
Krumlauf
,
R.
and
Holder
,
N.
(
1991
).
Retinoic acid causes abnormal development and segmental patterning of the anterior hindbrain in Xenopus embryos
.
Development
113
,
1145
1158
.
Petkovich
,
M.
,
Brand
,
N. J.
,
Krust
,
A.
and
Chambon
,
P.
(
1987
).
A human retinoic acid receptor which belongs to the family of nuclear receptors
.
Nature
330
,
444
450
.
Porteus
,
M. H.
,
Bulfone
,
A.
,
Ciaranello
,
R. D.
and
Rubenstein
,
J. L.
(
1991
).
Isolation and characterization of a novel cDNA clone encoding a homeodomain that is developmentally regulated in the ventral forebrain
.
Neuron
7
,
221
229
.
Price
,
M.
,
Lemaistre
,
M.
,
Pischetola
,
M.
,
Lauro
,
R. D.
and
Duboule
,
D.
(
1991
).
A mouse gene related to Distal-less shows a restricted expression in the developing forebrain
.
Nature
351
,
748
751
.
Puelles
,
L.
,
Amat
,
J. A.
and
Martinez-de-la-Torre
,
M.
(
1987
).
Segment-related, mosaic neurogenetic pattern in the forebrain and mesencephalon of early chick embryos: I. Topography of AChE-positive neuroblasts up to stage HH18
.
J. Comp. Neurol
.
266
,
247
268
.
Reynolds
,
K.
,
Mezey
,
E.
and
Zimmer
,
A.
(
1991
).
Activity of the retinoic acid receptor-β promoter in transgenic mice
.
Mech. Dev
.
36
,
15
29
.
Rossant
,
J.
,
Zirngibl
,
R.
,
Cado
,
D.
,
Shago
,
M.
and
Giguére
,
V.
(
1991
).
Expression of a retinoic acid response element-hsplacZ transgene defines specific domains of transcriptional activity during mouse embryogenesis
.
Genes Dev
.
5
,
1333
1344
.
Ruberte
,
E.
,
Dolle
,
P.
,
Chambon
,
P.
and
Morris-Kay
,
G.
(
1991
).
Retinoic acid receptors and cellular retinoid binding proteins. II Their differential pattern of transcription during early morphogenesis in mouse embryos
.
Development
111
,
45
60
.
Ruiz i Altaba
,
A.
and
Jessell
,
T.
(
1991
).
Retinoic acid modifies mesodermal patterning in early Xenopus embryos
.
Genes Dev
.
5
,
175
187
.
Shenefelt
,
R. E.
(
1972
).
Morphogenesis of malformations in hamsters caused by retinoic acid: relation to dose and stage at treatment
.
Teratology
5
,
103
118
.
Sive
,
H. L.
,
Draper
,
B. W.
,
Harland
,
R. M.
and
Weintraub
,
H.
(
1990
).
Identification of a retinoic acid-sensitive period during primary axis formation in Xenopus laevis
.
Genes Dev
.
4
,
932
942
.
Sucov
,
H. M.
,
Murakami
,
K. K.
and
Evans
,
R. M.
(
1990
).
Characterization of an autoregulated response element in the mouse retinoic acid receptor type β gene
.
Proc. Natl. Acad. Sci. USA
87
,
5392
5396
.
Sundin
,
O. H.
and
Eichele
,
G.
(
1990
).
A homeo domain protein reveals the metameric nature of the developing chick hindbrain
.
Genes Dev
.
4
,
1267
1276
.
Tuckett
,
F.
,
Lim
,
L.
and
Morriss-Kay
,
G. M.
(
1985
).
The ontogenesis of cranial neuromeres in the rat embryo I. A scanning electron microscope and kinetic study
.
J. Embryol. exp. Morph
.
87
,
215
228
.
Vaage
,
S.
(
1969
).
The segmentation of the primitive neural tube in chick embryos (Gallus domesticus). A morphological, histochemical and autoradiographical investigation
.
Ergebn. Anat. EntwGesch
.
41
,
4
87
.
Wilkinson
,
D. G.
,
Bhatt
,
S.
,
Chavrier
,
P.
,
Bravo
,
R.
and
Charnay
,
P.
(
1989
).
Segment-specific expression of a zinc-finger gene in the developing nervous system of the mouse
.
Nature
337
,
461
464
.
Wilkinson
,
D. G.
and
Krumlauf
,
R.
(
1990
).
Molecular approaches to the segmentation of the hindbrain
.
Trends Neurosc
.
13
,
335
339
.
Zelent
,
A.
,
Krust
,
A.
,
Petkovich
,
M.
,
Kastner
,
P.
and
Chambon
,
P.
(
1989
).
Cloning of murine α and β retinoic acid receptors and a novel receptor γ predominantly expressed in skin
.
Nature
339
,
714
717
.
Zelent
,
A.
,
Mendelsohn
,
C.
,
Kastner
,
P.
,
Krust
,
A.
,
Garnier
,
J.-M.
,
Ruffenach
,
F.
,
Leroy
,
P.
and
Chambon
,
P.
(
1991
).
Differentially expressed isoforms of the mouse retinoic acid receptor β are generated by usage of two promoters and alternative splicing
.
EMBO J
.
10
,
71
81
.