Lim1 is a homeobox gene expressed in the extraembryonic anterior visceral endoderm and in primitive streak-derived tissues of early mouse embryos. Mice homozygous for a targeted mutation of Lim1 lack head structures anterior to rhombomere 3 in the hindbrain. To determine in which tissues Lim1 is required for head formation and its mode of action, we have generated chimeric mouse embryos and performed tissue layer recombination explant assays. In chimeric embryos in which the visceral endoderm was composed of predominantly wild-type cells, we found that Lim1−/− cells were able to contribute to the anterior mesendoderm of embryonic day 7.5 chimeric embryos but that embryonic day 9.5 chimeric embryos displayed a range of head defects. In addition, early somite stage chimeras generated by injecting Lim1−/− embryonic stem cells into wild-type tetraploid blastocysts lacked forebrain and midbrain neural tissue. Furthermore, in explant recombination assays, anterior mesendoderm from Lim1−/− embryos was unable to maintain the expression of the anterior neural marker gene Otx2 in wild-type ectoderm. In complementary experiments, embryonic day 9.5 chimeric embryos in which the visceral endoderm was composed of predominantly Lim1−/− cells and the embryo proper of largely wild-type cells, also phenocopied the Lim1−/− headless phenotype. These results indicate that Lim1 is required in both primitive streak-derived tissues and visceral endoderm for head formation and that its inactivation in these tissues produces cell non-autonomous defects. We discuss a double assurance model in which Lim1 regulates sequential signaling events required for head formation in the mouse.

The organizer experiment of Spemann and Mangold demonstrated that the dorsal lip region of an amphibian gastrula stage embryo could direct the formation of a secondary embryonic axis and induce neural tissue from ectoderm when transplanted to a host amphibian embryo (Spemann and Mangold, 1924). Later, Spemann also observed that the inductive properties of the dorsal lip region changed during the course of gastrulation. Transplantation of dorsal lip regions from early gastrula stage embryos resulted in the formation of secondary axes that contained head structures whereas transplantation of dorsal lip regions from late gastrula stage embryos resulted in the formation of secondary axes that contained trunk and tail structures. These findings led Spemann to propose the existence of a distinct head and trunk organizer in vertebrate embryos (Spemann, 1931).

In the mouse, the node is a morphologically distinct structure located at the rostral end of the primitive streak of embryonic day (E) 7.5 embryos. The node from late streak stage embryos is able to induce a secondary embryonic axis when transplanted to the lateral aspect of a similarly staged mouse host embryo; however, the induced secondary axis in these embryos lack any discernable anterior neural tissue (Beddington, 1994). One possible explanation for the failure of the node to induce anterior neural tissue is that at the late streak stage, the node may be equivalent to the amphibian late gastrula dorsal lip or trunk organizer. The inductive properties of the presumptive node have been tested at an earlier stage, when it may be analogous to the early dorsal lip, by transplanting tissue from the posterior epiblast region of early primitive streak stage embryos (Tam et al., 1997). Cells from this region express several node-specific genes and fate mapping studies indicate that they contribute descendants to node derivatives including the definitive endoderm and notochord (reviewed by Tam and Behringer, 1997). When transplanted to a late streak stage host embryo, the posterior epiblast tissue is able to induce a secondary embryonic axis but none of the induced structures displayed any anterior characteristics (Tam et al., 1997). These experiments suggest that the mouse node, unlike the amphibian dorsal lip region, is not sufficient to induce ectoderm to become anterior neural tissue.

Recent experiments have implicated the visceral endoderm in anterior neural induction in mouse. The visceral endoderm is an extraembryonic tissue that surrounds the epiblast of the egg cylinder stage embryo (Rossant, 1986). During gastrulation the visceral endoderm is replaced by definitive endoderm that derives from the anterior portion of the primitive streak (Lawson and Pedersen, 1987). Although no morphological asymmetries are apparent in the visceral endoderm, molecular studies have shown that a distinct anterior-posterior pattern exists in the visceral endoderm prior to the formation of the primitive streak (Rosenquist and Martin, 1995; Hermesz et al., 1996; Thomas and Beddington, 1996; Thomas et al., 1998). These studies revealed that the VE-1 antigen and the homeobox genes Hesx1/Rpx and Hex are expressed in the anterior visceral endoderm that underlies the ectoderm fated to form the anterior portion of the neural plate. Ablation experiments have shown that, if the anterior visceral endoderm is removed at the early streak stage, expression of Hesx1/Rpx in the anterior neuroectoderm in late streak/headfold stage embryos is absent or greatly reduced (Thomas and Beddington, 1996).

Chimera experiments have provided genetic evidence that extraembryonic tissues play an important role in establishing anterior identity in the mouse embryo. Mice homozygous for a retroviral insertion in the nodal gene fail to gastrulate and die during embryogenesis (Conlon et al., 1991). In chimeric embryos in which the visceral endoderm was composed of primarily nodal mutant cells and the embryonic portion of primarily wild-type cells, the gastrulation defect was rescued but the resulting embryos lacked anterior head structures (Varlet et al., 1997). Similarly, Otx2, a homeobox gene expressed in the visceral endoderm and the ectoderm, has been shown to be required in the visceral endoderm and subsequently in the neuroectoderm for development of the forebrain and midbrain (Rhinn et al., 1998). Furthermore, the expression of the anterior neural marker Bf1 and the midbrain/hindbrain neural marker Engrailed1 (En1) can occur in the absence of embryonic mesendoderm as seen in embryos homozygous for a mutation in Cripto, which encodes a putative signaling molecule (Deng et al., 1998). These results suggest that the anterior visceral endoderm is necessary for mammalian head formation (Bouwmeester and Leyns, 1997).

The mouse Lim1 homeobox gene (also known as Lhx1) is expressed in both the anterior visceral endoderm and in primitive streak-derived tissues including the node, the mesodermal wings and the anterior mesendoderm that underlies the presumptive anterior neural plate (Barnes et al., 1994; Shawlot and Behringer, 1995; Belo et al., 1997, Shimono and Behringer, 1999). Lim1 expression in the anterior primitive streak-derived tissues is transient and is downregulated after embryonic day (E) 7.5 (Barnes et al., 1994; W. S., unpublished data). To study the function of Lim1 during mouse embryogenesis, we previously generated a mutation in Lim1 by homologous recombination in embryonic stem (ES) cells (Shawlot and Behringer, 1995). Homozygous mutant embryos have defects in node and axial mesendoderm development and do not form head structures anterior to rhombomere 3 in the hindbrain (Shawlot and Behringer, 1995). These studies however, have not resolved the cellular site of Lim1 action. To determine in which tissues Lim1 is required for head formation and its mode of action, we have performed chimera studies and tissue layer recombination assays. Surprisingly, we found that Lim1 is required in both primitive streak-derived tissues and extraembryonic tissues for head formation and that its inactivation in these tissues produces cell non-autonomous defects. Thus the anterior visceral endoderm is necessary for head formation but this process also requires streak-derived tissues. We discuss a double assurance model in which Lim1 regulates sequential signaling events from the anterior visceral endoderm and the anterior definitive mesendoderm that are necessary for head formation in the mouse.

Generation of ES cell lines

ES cell lines were isolated from blastocysts obtained from Lim1+/− females crossed with Lim1+/−, ROSA26 males using standard methods (Robertson, 1987). Twelve independent ES cell lines were derived and genotyped by Southern hybridization using a 5′ Lim1 probe (Shawlot and Behringer, 1995).

Generation and X-gal staining of chimeras

Chimeras were generated by blastocyst injection (Bradley, 1987). To assess the role of Lim1 in epiblast tissues, ES cells were injected into blastocysts obtained from Swiss Webster mice (Taconic, Germantown, NY). To assess the role of Lim1 in the visceral endoderm, Lim1+/+ ES cells (ES31) were injected into blastocysts obtained by crossing Lim1+/− mice. Injected blastocysts were transferred to pseudopregnant Swiss Webster females and recovered at E7.5 and E9.5. Embryos were fixed in 0.2% glutaraldehyde and processed for X-gal staining. Embryos were postfixed in 4% paraformaldehyde and paraffin embedded (Hogan et al., 1994).

Lim1−/−↔tetraploid +/+ chimeras were generated by injecting Lim1−/− ES cells into wild-type tetraploid blastocysts (Swiss Webster strain). 2-cell stage wild-type embryos were isolated and electrofused using a CF-150 Cell Fusion Instrument (Biochemical Laboratory Services, Budapest, Hungary). Embryos were fused in M2 medium with 3 pulses of 90 volts and 94 microseconds. Using these parameters, 80-90% of the embryos fused within 1 hour. The fused embryos were cultured to the blastocyst stage in M16 media. Blastocyst injections were performed on the morning of E4. Approximately one third of the injected tetraploid embryos were recovered at E8.

Whole-mount in situ hybridization

Whole-mount in situ hybridization was performed by standard methods (Wilkinson, 1992). The Otx2 probe used was pBSotd 9 (Ang et al., 1994). Hybridization and washes were performed at 65°C.

Tissue layer recombination explant cultures

Ectoderm from the distal tip of the egg cylinder was isolated from E6.5 early streak stage embryos using pulled glass capillary needles. Anterior mesendoderm from E7.5 embryos was isolated using tungsten needles after incubating the embryos in 0.5% trypsin, 0.25% pancreatin in PBS at 4°C for 10 minutes (Ang and Rossant, 1993). Tissue fragments were cultured for 2 days in drops of Dulbecco’s modified Eagles (DME) media containing 15% fetal calf serum, 2 mM glutamine and 0.1 mM 2-mercaptoethanol.

Generation of Lim1-τ-lacZ mice

The 5′ 1.2 kb HindIII-HindIII DNA fragment and the 3′ 4.3 kb BamHI-BamHI DNA fragment from the Lim1 locus (Shawlot and Behringer, 1995) were inserted into the IRES-τ-lacZ-loxPGKneo-lox vector of Mombaerts et al. (1996). Correct targeting results in the replacement of the entire Lim1-coding region with the lacZ-neo cassette, thereby creating a null allele. The targeting vector was electroporated into W95 129/Sv ES cells and 96 G418-resistant clones were isolated. The G418-resisitant ES cell clones were analyzed by Southern hybridization using the 5′ and 3′ DNA probes described by Shawlot and Behringer (1995). One correctly targeted ES cell clone was identified that produced a 5 kb wild-type band and a 1.8 kb targeted band in genomic DNA digested with KpnI and hybridized with the 5′ probe. Correct targeting was confirmed by digestion of the genomic DNA with EcoRI and hybridization with the 3′ probe, which produced a 12 kb wild-type band and a 10 kb targeted band. Blastocyst injections were performed as described previously (Mombaerts et al., 1996).

Function of Lim1 in primitive streak-derived tissues

Previous studies have demonstrated that ES cells injected into blastocysts do not contribute extensively to the extraembryonic endoderm (Beddington and Robertson, 1989). We have used this developmental bias of ES cells to generate highly polarized Lim1 chimeras in which the embryo proper is predominantly of one genotype and the extraembryonic endoderm is of another.

Twelve independent Lim1 ES lines were generated from blastocysts obtained from Lim1+/− females mated to Lim1+/− males homozygous for the ROSA26 lacZ transgene (data not shown) which is widely expressed during development (Friedrich and Soriano, 1991). To determine if Lim1−/− cells can contribute normally to primitive streak-derived tissues, we injected separately two Lim1−/− ES cell lines, 3-21 and ES27, into wild-type blastocysts. These chimeric embryos are referred to as Lim1−/−↔+/+ chimeras. As controls, we also injected two Lim1+/+ ES cell lines, ES10 and ES31, into wild-type blastocysts. These chimeric embryos are referred to as Lim1+/+↔+/+ chimeras. The injected blastocysts were transferred to pseudopregnant recipients and potential chimeric embryos were isolated at E7.5 and stained with X-gal to determine the location of the ES-derived cells.

Approximately 150 E7.5 Lim1−/−↔+/+ chimeras were recovered. The chimeric embryos resulting from the injection of Lim1−/− ES cell lines 3-21 and ES27 were similar. When compared with Lim1+/+↔+/+ control chimeras (data not shown), the lacZ-marked Lim1−/− cells were randomly distributed in Lim1−/−↔+/+ chimeras (Fig. 1A). Lim1−/− cells did not appear to be excluded from or concentrate in any region of the embryo proper. Unlike E7.5 Lim1−/− mutant embryos, very high percentage Lim1−/−↔+/+ chimeras (greater than 95% Lim1−/− cells) did not have a constriction between the embryonic and extraembryonic region (Fig. 1B). As the extraembryonic tissues in these chimeric embryos are composed of predominantly wild-type cells, the constriction present in Lim1−/− embryos is likely to be caused by a defect in extraembryonic tissues.

Fig. 1.

E7.5 Lim1−/−↔+/+ chimeras. Distribution of Lim1−/− cells (3-21) in Lim1−/−↔+/+ chimeras. Lim1−/− cells are randomly distributed in the embryo proper. Phenotypic appearance of a E7.5 Lim1−/−↔+/+ chimera (3-21) with a contribution of Lim1−/− cells estimated to be greater than 95%. Unlike E7.5 Lim1−/− embryos, no constriction is present between the embryonic and extraembryonic region of Lim1−/−↔+/+ chimeric embryos. (C) Transverse section of a E7.5 Lim1−/−↔+/+ chimera (3-21) at the level of the presumptive forebrain region. The contribution of Lim1−/− cells (blue) is estimated to be greater than 95%. Lim1−/− cells are present in the definitive endoderm (de), the mesodermal wings (mw), the midline mesendoderm (mme) and the ectoderm (ect).

Fig. 1.

E7.5 Lim1−/−↔+/+ chimeras. Distribution of Lim1−/− cells (3-21) in Lim1−/−↔+/+ chimeras. Lim1−/− cells are randomly distributed in the embryo proper. Phenotypic appearance of a E7.5 Lim1−/−↔+/+ chimera (3-21) with a contribution of Lim1−/− cells estimated to be greater than 95%. Unlike E7.5 Lim1−/− embryos, no constriction is present between the embryonic and extraembryonic region of Lim1−/−↔+/+ chimeric embryos. (C) Transverse section of a E7.5 Lim1−/−↔+/+ chimera (3-21) at the level of the presumptive forebrain region. The contribution of Lim1−/− cells (blue) is estimated to be greater than 95%. Lim1−/− cells are present in the definitive endoderm (de), the mesodermal wings (mw), the midline mesendoderm (mme) and the ectoderm (ect).

To analyze the contribution of Lim1−/− cells to the germ layers, transverse sections of E7.5 neural plate stage Lim1−/−??↔+/+ chimeras were made. In moderate and high percentage Lim1−/−↔+/+ chimeras, lacZ-tagged Lim1−/− cells were able to contribute to the anterior midline mesendoderm, the mesodermal wings and the definitive anterior endoderm even in chimeric embryos in which the contribution of mutant cells was estimated to be greater than 95% (Fig. 1C). Thus it is unlikely that the mutant cells are rescued by the presence of wild-type cells. These results indicate that Lim1−/− cells can contribute normally to the primitive streak-derived tissues of E7.5 Lim1−/−↔+/+ chimeras.

To determine whether Lim1 is required in primitive streak-derived cells for the development of anterior structures, we analyzed the head phenotype of E9.5 Lim1−/−↔+/+ chimeric embryos. The visceral endoderm of these chimeric embryos is predominantly composed of wild-type cells. The chimeric embryos generated with Lim1−/− ES cell lines 3-21 and ES27 were phenotypically similar and the results from these two lines were combined. Non-chimeric and low percentage chimeric embryos had normal anterior head development (Fig. 2A). In contrast, 16 of 23 chimeric embryos with an estimated contribution of Lim1−/− cells greater than 50% had defects in head development. Moderate percentage chimeric embryos (50-75% Lim1−/− cells) were microcephalic with the forebrain region either reduced or absent (Fig. 2B-D). Very high percentage chimeric embryos (>95% Lim1−/− cells) lacked head structures just anterior to the otic vesicle and were identical to E9.5 Lim1−/−embryos (Fig. 2E, F). Lim1+/+↔+/+ control chimeras (n=20) that were generated using ES cell lines ES10 and ES31 had normal head development (data not shown). As Lim1 is not expressed in the central nervous system until E10 (Barnes et al., 1994; Fujii et al., 1994), it is unlikely that the anterior defects observed in E9.5 Lim1−/−↔+/+ chimeras are due to a requirement for Lim1 in the anterior neural tissue itself. Despite the absence of a constriction between the embryonic and extraembryonic region of E7.5 chimeric embryos with a very high contribution of mutant cells, a headless phenotype identical to that seen in Lim1−/− embryos was observed in E9.5 chimeric embryos with a very high contribution of mutant cells. This suggests that the constriction and any defects in cell movement caused by the constriction are not the primary cause of the headless phenotype.

Fig. 2.

Head phenotype of E9.5 Lim1−/−↔+/+ chimeras. (A) Non-chimeric embryo with normal anterior neural development displayingforebrain (fb), midbrain (mb) and hindbrain (hb). (B) Moderate percentage chimera in which the forebrain region is reduced. An optic placode was not present. (C) Moderate percentage chimera having a small anterior projection bearing a single optic placode (arrowhead). (D) A moderate to high percentage chimera in which the forebrain region is absent. (E) A very high percentage chimera that lacks head structures just anterior to the otic vesicle. (F) Lim1−/− embryo displaying a loss of head structures just anterior to the otic vesicle. All chimeras shown here were generated using Lim1−/− ES cell line 3-21. The arrows in A-F mark the otic vesicle. Expression of Otx2 in an E8.0 wild-type embryo. (H) Expression of Otx2 in a E8.0 Lim1−/−↔tetraploid (4n) +/+ chimera. Expression of Otx2 in an E8.0 Lim1−/− embryo. The arrows in G-I mark the first branchial arch.

Fig. 2.

Head phenotype of E9.5 Lim1−/−↔+/+ chimeras. (A) Non-chimeric embryo with normal anterior neural development displayingforebrain (fb), midbrain (mb) and hindbrain (hb). (B) Moderate percentage chimera in which the forebrain region is reduced. An optic placode was not present. (C) Moderate percentage chimera having a small anterior projection bearing a single optic placode (arrowhead). (D) A moderate to high percentage chimera in which the forebrain region is absent. (E) A very high percentage chimera that lacks head structures just anterior to the otic vesicle. (F) Lim1−/− embryo displaying a loss of head structures just anterior to the otic vesicle. All chimeras shown here were generated using Lim1−/− ES cell line 3-21. The arrows in A-F mark the otic vesicle. Expression of Otx2 in an E8.0 wild-type embryo. (H) Expression of Otx2 in a E8.0 Lim1−/−↔tetraploid (4n) +/+ chimera. Expression of Otx2 in an E8.0 Lim1−/− embryo. The arrows in G-I mark the first branchial arch.

To analyze anterior neural development at an earlier stage, we next examined Otx2 expression in early somite stage E8.0 Lim1−/−↔+/+ tetraploid chimeras. When ES cells are introduced into tetraploid blastocysts, the embryo proper is almost completely ES cell-derived while the extraembryonic endoderm is composed of mainly tetraploid cells (Nagy et al., 1990). To generate these chimeras, we injected Lim1−/− ES cells (3-21) into wild-type tetraploid blastocysts. In E8.0 wild-type embryos, Otx2 is expressed in the forebrain and midbrain and in the first branchial arch (Fig. 2G) (Ang et al., 1994). In E8.0 Lim1−/−↔+/+ tetraploid chimeras (n=2) and E8.0 Lim1−/− embryos (n=2), Otx2 was expressed only in the first branchial arch indicating that anterior neural tissue was not present (Fig. 2H,I). These results indicate that Lim1 is required in primitive streak-derived tissues for the development of anterior head structures.

Lim1 tissue layer recombination explants

Previous studies have shown that, in explant cultures, E7.5 anterior mesendoderm can maintain the expression of Otx2 in E6.5 ectoderm (Ang et al., 1994). As Lim1−/− cells were able to contribute normally to the anterior mesendoderm of high percentage E7.5 Lim1−/−↔+/+ chimeras but E8.0 Lim1−/−↔+/+ tetraploid chimeras lacked anterior neural structures, we hypothesized that Lim1 regulated the production of a factor that acted cell non-autonomously. To test this hypothesis, we performed tissue layer explant recombination experiments using anterior mesendoderm from E7.5 Lim1−/− embryos. To confirm that anterior mesendoderm is present in E7.5 Lim1−/− embryos, we examined the expression of the goosecoid (gsc) gene, an anterior mesendoderm marker expressed between E7.0 and E7.75 (Faust et al., 1995). To visualize gsc expression, we used a mouse strain containing a lacZ insertion in the gsc locus. E7.5 embryos heterozygous for this insertion expressed lacZ in the node and the anterior mesendoderm (M. W. and R. B., unpublished data) (Fig. 3A). In E7.5 Lim1−/− embryos carrying the gsc-lacZ marker, lacZ expression was observed in the distal portion of the embryo corresponding to the anterior portion of the primitive streak and the anterior mesendoderm (Fig. 3B). The location of the anterior aspect of the primitive streak in Lim1−/− embryos, as revealed by the gsc- lacZ marker, is in agreement with previous molecular marker studies that examined the expression of the primitive streak and node markers Brachyury, nodal and HNF3β (Shawlot and Behringer, 1995). These gsc-lacZ marker results indicate that anterior type mesendoderm is present in Lim1−/− embryos and that it is located in the distal portion of the embryo.

Fig. 3.

Expression of Otx2 in tissue layer recombination explants. lacZ expression in a E7.5 wild-type embryo that carries a lacZ transgene integrated into the gsc locus. The lacZ transgene is expressed in the anterior mesendoderm and the node. (B) lacZ expression in a E7.5 Lim1−/− embryo that carries a lacZ transgene targeted to the gsc locus. The lacZ transgene is expressed in the distal portion of embryo in the rostral portion of the primitive streak and in the anterior mesendoderm region. The lines through the embryos in A and B indicate the cuts made to isolate anterior mesendoderm. The line outside the embryo in B indicates the extent of the primitive streak. am, anterior mesendoderm; n, node; ps, primitive streak. (C) Expression of Otx2 in control E7.5 wild-type anterior mesendoderm + E6.5 wild-type ectoderm explants. (D) Lack of Otx2 expression above background in E7.5 Lim1−/− anterior mesendoderm + E6.5 wild-type ectoderm explants. (E) Absence of Otx2 expression in E7.5 wild-type mesendoderm explants cultured alone. (F) Absence of Otx2 expression in E6.5 wild-type ectoderm explants cultured alone.

Fig. 3.

Expression of Otx2 in tissue layer recombination explants. lacZ expression in a E7.5 wild-type embryo that carries a lacZ transgene integrated into the gsc locus. The lacZ transgene is expressed in the anterior mesendoderm and the node. (B) lacZ expression in a E7.5 Lim1−/− embryo that carries a lacZ transgene targeted to the gsc locus. The lacZ transgene is expressed in the distal portion of embryo in the rostral portion of the primitive streak and in the anterior mesendoderm region. The lines through the embryos in A and B indicate the cuts made to isolate anterior mesendoderm. The line outside the embryo in B indicates the extent of the primitive streak. am, anterior mesendoderm; n, node; ps, primitive streak. (C) Expression of Otx2 in control E7.5 wild-type anterior mesendoderm + E6.5 wild-type ectoderm explants. (D) Lack of Otx2 expression above background in E7.5 Lim1−/− anterior mesendoderm + E6.5 wild-type ectoderm explants. (E) Absence of Otx2 expression in E7.5 wild-type mesendoderm explants cultured alone. (F) Absence of Otx2 expression in E6.5 wild-type ectoderm explants cultured alone.

In control explant experiments, wild-type E7.5 (late streak- neural plate stage) anterior mesendoderm was recombined with E6.5 (early streak stage) distal tip ectoderm, cultured for 2 days and then assayed for Otx2 expression by RNA whole-mount in situ hybridization. 11 of 13 wild-type recombinant explants expressed Otx2 (Fig. 3C). When cultured alone, 0/7 mesendoderm explants expressed Otx2 (Fig. 3E). Two of 12 ectoderm explants expressed Otx2 when cultured alone (Fig. 3F). The two ectoderm explants that expressed Otx2 may have contained some mesendoderm tissue, or alternatively, the ectoderm may have already received inductive signals from the mesendoderm prior to its isolation. When E7.5 Lim1−/− anterior mesendoderm from the distal tip region, which excludes the posterior region of the embryo, was recombined with E6.5 wild-type ectoderm, 0/11 explants expressed Otx2 above background (Fig. 3D). The results of these explant recombination experiments indicate that Lim1−/− anterior mesendoderm is defective in anterior neural signaling.

Function of Lim1 in extraembryonic tissues

To determine if Lim1 is required in extraembryonic tissues for anterior neural development, we injected Lim1+/+ ES cells (ES31) into blastocysts obtained from Lim1+/− intercrosses. Approximately one quarter of the injected blastocysts should be of the genotype Lim1−/− and give rise to chimeras in which the extraembryonic tissues are composed almost exclusively of Lim1−/− cells. These embryos are referred to as +/+↔Lim1−/−chimeras. Of the thirty embryos recovered at E9.5, eight were phenotypically indistinguishable from Lim1−/− embryos, which lack head structures just anterior to the otic vesicle (Fig. 4A). We did not observe any chimeric embryos with microcephaly. After staining with X-gal, four of the headless embryos were estimated to contain greater than 50% wild-type cells with one chimeric embryo containing greater than 95% wild-type cells in the embryo proper (Fig. 4B). Histological analysis of this chimeric embryo demonstrated that the wild-type cells had contributed extensively to the anterior portion of the embryo including the neural tube, notochord and foregut tissues (Fig. 4C,D). A similar distribution of wild-type cells was seen in the three moderate percentage +/+↔Lim1−/− headless chimeras (data not shown). We were not able to genotype the extraembryonic tissues from the chimeric embryos by PCR because the yolk sac endoderm layer that we isolated for genotyping was presumably contaminated with mesoderm cells that derive from the embryo proper. Because of this, we cannot exclude the possibility that the headless phenotype was rescued in some chimeric embryos. However, the observation that four chimeric embryos with a substantial contribution of wild-type cells had a complete headless phenotype indicates that Lim1 is required in extraembryonic tissues for head formation.

Fig. 4.

Analysis of +/+↔Lim1−/− chimeras at E9.5. (A) Phenotype of a E9.5 Lim1−/− embryo showing the absence of head structures just anterior to the otic vesicle (ov). Taken from Shawlot and Behringer (1995). (B) Phenotype of a E9.5 +/+↔Lim1−/− chimera. The embryo proper contains a high percentage of wild-type cells (blue) but the embryo lacks head structures anterior to the otic vesicle. The lines through the embryo indicate the level of the transverse section in C and D. (C) Transverse section of a +/+↔Lim1−/− chimera at the level of the otic vesicle. (D) Transverse section of same embryo at the level of the heart. In both sections, wild-type cells (blue) have contributed extensively to the neural tube (nt), notochord (nc) and foregut endoderm (fg) but do not rescue head development

Fig. 4.

Analysis of +/+↔Lim1−/− chimeras at E9.5. (A) Phenotype of a E9.5 Lim1−/− embryo showing the absence of head structures just anterior to the otic vesicle (ov). Taken from Shawlot and Behringer (1995). (B) Phenotype of a E9.5 +/+↔Lim1−/− chimera. The embryo proper contains a high percentage of wild-type cells (blue) but the embryo lacks head structures anterior to the otic vesicle. The lines through the embryo indicate the level of the transverse section in C and D. (C) Transverse section of a +/+↔Lim1−/− chimera at the level of the otic vesicle. (D) Transverse section of same embryo at the level of the heart. In both sections, wild-type cells (blue) have contributed extensively to the neural tube (nt), notochord (nc) and foregut endoderm (fg) but do not rescue head development

Lim1 is not required in the visceral endoderm for Lim1 expression in primitive streak-derived tissues

Next, we determined if the lack of anterior neural development in +/+↔Lim1−/− chimeras was due to an inability of the mutant visceral endoderm to induce Lim1 expression in the primitive streak-derived tissues. We analyzed expression directed by the endogenous Lim1 promoter in Lim1 mutant embryos that carry a τ-lacZ transgene targeted to the Lim1 locus (A. K. and T. J., unpublished data). This Lim1lacZ allele is functionally equivalent to our original Lim1-null allele. The details of these mice will be published elsewhere. Lim1lacZ/+ males were mated with Lim1+/− females and embryos were collected at E7.5 (neural plate stage) and stained with X-gal. In Lim1lacZ/+ embryos, lacZ expression was observed in the node and the anterior mesendoderm, which is identical to the endogenous Lim1 expression pattern (Fig. 5A). Analysis of transverse histological sections showed that lacZ was expressed in the mesodermal wings, the endoderm and the anterior midline mesendoderm tissue (Fig. 5C). In Lim1lacZ/− mutant embryos, lacZ expression was also observed in the node region and in the anterior mesendoderm (Fig. 5B). Analysis of transverse sections showed that lacZ was expressed in the mesodermal wings and the endoderm of Lim1 mutant embryos (Fig. 5D). These results indicate that Lim1 is not required in the visceral endoderm to induce Lim1 expression in the primitive streak-derived tissues. Thus, the +/+↔Lim1−/− chimera results indicate that Lim1 is required in extraembryonic tissues to directly induce anterior neural tissue.

Fig. 5.

Analysis of Lim1 expression in primitive streak-derived tissues. (A) Expression of lacZ in an E7.5 Lim1lacZ/+ embryo.Expression of lacZ in an E7.5 Lim1lacZ/− mutant embryo. Expression of lacZ is present in the node region and anterior mesendoderm in both the wild-type and mutant embryos.Transverse section of an E7.5 Lim1lacZ/+ embryo. (D) Transverse section of Lim1lacZ/− mutant embryo. In both embryos, lacZ is expressed in the mesodermal wings and the endoderm. ect, ectoderm; end, endoderm; mme, midline mesendoderm; mw, mesodermal wing.

Fig. 5.

Analysis of Lim1 expression in primitive streak-derived tissues. (A) Expression of lacZ in an E7.5 Lim1lacZ/+ embryo.Expression of lacZ in an E7.5 Lim1lacZ/− mutant embryo. Expression of lacZ is present in the node region and anterior mesendoderm in both the wild-type and mutant embryos.Transverse section of an E7.5 Lim1lacZ/+ embryo. (D) Transverse section of Lim1lacZ/− mutant embryo. In both embryos, lacZ is expressed in the mesodermal wings and the endoderm. ect, ectoderm; end, endoderm; mme, midline mesendoderm; mw, mesodermal wing.

Our results demonstrate that Lim1 is required in both primitive streak-derived tissues and extraembryonic tissues for head formation and that its inactivation in these tissues leads to cell non-autonomous defects. Anterior visceral endoderm therefore is necessary for head formation but additional signals from the streak-derived definitive anterior mesendoderm are also essential for head development. These findings are consistent with previous recombination explant experiments in mice showing that mid- to late streak stage anterior mesendoderm can induce anterior neural fates in naïve epiblast (Ang and Rossant, 1993). We have also demonstrated that Lim1 is not required in the anterior visceral endoderm to initiate or maintain Lim1 expression in primitive streak-derived tissues. Furthermore, Lim1 is not required in the primitive streak- derived tissues to maintain Lim1 expression in the anterior visceral endoderm because Lim1 expression is observed in the anterior visceral endoderm of E7 Wnt3 mutant embryos, which lack a primitive streak and associated tissues (Liu et al., 1999). Based on our chimera findings we hypothesize that Lim1 regulates sequential signaling events, first in the anterior visceral endoderm and then subsequently in the anterior definitive mesendoderm that underlies the future anterior neural plate.

A double assurance model for head development

Spemann (1927) described a two-step induction or double assurance model for neural plate determination in amphibian embryos. Spemann hypothesized that the first step occurred by forward spreading of a neuralizing agent in the surface layer from the dorsal blastopore lip to the adjacent ectoderm. The second step occurred when the involuting dorsal mesoderm came to underlie the presumptive neural plate. Spemann wrote, “It would be entirely conceivable that [its] induction which occurs after its exposure to the subjacent mesoderm is merely the continuation of another induction which was initiated when the [prospective mesodermal and ectodermal] materials were still lying side by side on the surface” (Spemann, 1927; Hamburger, 1988).

A two-step model for anterior neural induction in the mouse has recently been proposed (Thomas and Beddington, 1996). This model suggests that the visceral endoderm is responsible for inducing anterior identity in the embryo and that this identity is subsequently reinforced and maintained by primitive streak-derived anterior mesendoderm. Our results provide genetic support for this model. We hypothesize that Lim1 regulates the production of a secreted factor or factors from the anterior visceral endoderm at the pre- to early streak stage that causes the overlying ectoderm to differentiate as anterior neural tissue. This anterior neural fate is labile, however, and a second Lim1-regulated signal from the primitive streak-derived definitive mesendoderm, which replaces the visceral endoderm beginning at the mid- to late streak stage, is required to maintain or complete the differentiation of the anterior ectoderm as anterior neural tissue.

We interpret our chimera results with respect to this two-step induction model (Fig. 6). In Lim1−/−↔+/+ chimeras, the first signaling event occurs normally because the visceral endoderm is composed of wild-type cells. The second signaling event, however, is altered because the definitive mesendoderm is a mixture of both wild-type cells and Lim1−/− cells. Consequently, there is a reduction in the amount and or the distribution of the second Lim1-regulated signal, which results in the development of chimeric embryos with microcephaly.

Fig. 6.

Model for Lim1 function in establishing anterior identity. Lim1 is expressed in the anterior visceral endoderm (AVE) in pre- and early streak stage embryos. Lim1 is also expressed in the primitive streak-derived definitive mesendoderm (Def Mesend) that replaces the visceral endoderm at the mid- to late primitive streak stage. Lim1 chimera experiments indicate that two sequential signaling events are required for head development. The first Lim1-regulated signaling event from the anterior visceral endoderm induces the anterior ectoderm towards an anterior neural fate. This anterior neural fate is labile and is stabilized by a second Lim1-regulated signal from the definitive mesendoderm. In Lim1−/−↔+/+ chimeras, the first signaling event occurs normally but the second signaling event is quantitatively reduced because fewer wild-type cells are present in the definitive mesendoderm. This results in chimeras with microcephaly. In +/+↔Lim1−/− chimeras, the first signaling event from the anterior visceral endoderm does not occur and no subsequent anterior development takes place despite the presence of a high percentage of wild-type cells in the definitive mesendoderm. The size of the genotype labels next to the mesendoderm boxes approximates the contribution of wild-type and mutant cells in the mesendoderm.

Fig. 6.

Model for Lim1 function in establishing anterior identity. Lim1 is expressed in the anterior visceral endoderm (AVE) in pre- and early streak stage embryos. Lim1 is also expressed in the primitive streak-derived definitive mesendoderm (Def Mesend) that replaces the visceral endoderm at the mid- to late primitive streak stage. Lim1 chimera experiments indicate that two sequential signaling events are required for head development. The first Lim1-regulated signaling event from the anterior visceral endoderm induces the anterior ectoderm towards an anterior neural fate. This anterior neural fate is labile and is stabilized by a second Lim1-regulated signal from the definitive mesendoderm. In Lim1−/−↔+/+ chimeras, the first signaling event occurs normally but the second signaling event is quantitatively reduced because fewer wild-type cells are present in the definitive mesendoderm. This results in chimeras with microcephaly. In +/+↔Lim1−/− chimeras, the first signaling event from the anterior visceral endoderm does not occur and no subsequent anterior development takes place despite the presence of a high percentage of wild-type cells in the definitive mesendoderm. The size of the genotype labels next to the mesendoderm boxes approximates the contribution of wild-type and mutant cells in the mesendoderm.

In +/+↔Lim1−/− chimeras, the visceral endoderm is predominantly composed of Lim1−/− cells so the first signaling event does not occur and the ectoderm remains uninduced. In the absence of the first signal, the second signal from the wild-type cells in the definitive mesendoderm is unable to direct anterior neural development despite the expression of Lim1 by the definitive anterior mesendoderm. Consequently no microcephalic embryos are present and the chimeras completely lack anterior head structures.

Our Lim1 chimera studies are consistent with this two-step model for anterior neural induction. However, the phenotype of Cripto mutant mice does not appear to fit with this model (Deng et al., 1998). Cripto mutant mice form primitive streak- derived extraembryonic mesoderm but not other streak-derived tissues. Still, Bf1 and En1, forebrain and midbrain/hindbrain marker genes respectively, are expressed in E8.5 Cripto mutant embryos. Interestingly, anterior visceral endoderm markers are expressed but are located in the distal region of the mutant embryo. These findings suggest that anterior neural fates can be expressed in the absence of embryonic streak-derived tissues. One possible explanation to reconcile the Cripto findings with the results reported here and by others (Liu et al., 1999) is that Cripto functions to render the initial signal from the visceral endoderm labile (Thomas and Beddington, 1996). Thus, in the absence of Cripto, the initial visceral endoderm- derived signal is stable and can cause the expression of anterior neural fates in the epiblast in the absence of embryonic streak- derived tissues.

One candidate Lim1 downstream gene is the cerberus- related gene, Cerr1. In Xenopus, cerberus encodes a secreted molecule that can induce ectopic head structures when its mRNA is injected into Xenopus embryos (Bouwmeester et al., 1996). In mouse, Cerr1 is expressed in both the anterior visceral endoderm and the primitive streak-derived anterior mesendoderm (Belo et al., 1997; Thomas et al., 1997; Biben et al., 1998; Shawlot et al., 1998) and its expression is downregulated in Lim1−/− embryos (Shawlot et al., 1998). The partitioning of the same signaling molecule to both the visceral endoderm and the anterior definitive mesendoderm would allow the anterior ectoderm to be continuously in contact with the inducing signal(s) for an extended period of time despite the visceral endoderm being displaced during the course of gastrulation. Surprisingly, Cerr1-null mice created by gene targeting have normal head development, indicating that additional or alternative factors are involved in mouse head formation (W. S. and R. B., unpublished data).

In summary, we found that Lim1 is required in both primitive streak-derived tissues and extraembryonic tissues for head formation and that its inactivation in these tissues produces cell non-autonomous defects. Our results imply that sequential Lim1-regulated signals from distinct and separate tissue layers are required for the establishment of anterior identity in the mouse. The results of our study will aid in the identification of Lim1 downstream genes involved in vertebrate head formation.

We would like to thank Jenny Deng for tissue culture assistance and Barbara Han and Monica Mendelsohn for ES cell work and blastocysts injections in the generation of the Lim1-lacZ mice. We also thank Patrick Tam for comments on an earlier draft of this manuscript. This work was supported by a fellowship from the Theodore N. Law Endowment (W. S.) and a grant from the Human Frontiers Science Program (R. R. B.); T. M. J. is an investigator of the Howard Hughes Medical Institute.

Ang
,
S. L.
and
Rossant
,
J.
(
1993
).
Anterior mesendoderm induces mouse Engrailed genes in explant cultures
.
Development
118
,
139
149
.
Ang
,
S. L.
,
Conlon
,
R. A.
,
Jin
,
O.
and
Rossant
,
J.
(
1994
).
Positive and negative signals from the mesoderm regulates the expression of mouse Otx2 in ectoderm explants
.
Development
120
,
2979
2989
.
Barnes
,
J. D.
,
Crosby
,
J. L.
,
Jones
,
C. M.
,
Wright
,
C. V. E.
and
Hogan
,
B. L.M.
(
1994
).
Embryonic expression of Lim-1, the mouse homolog of Xlim-1, suggests a role in lateral mesoderm differentiation and neurogenesis
.
Dev. Biol
.
161
,
168
178
.
Beddington
,
R. S. P.
and
Robertson
,
E. J.
(
1989
).
An assessment of the developmental potential of embryonic stem cells in the midgestation embryo
.
Development
105
,
733
737
.
Beddington
,
R. S. P.
(
1994
).
Induction of a second neural axis by the mouse node
.
Development
120
,
613
620
.
Belo
,
J. A.
,
Bouwmeester
,
T.
,
Leyns
,
L.
,
Kertesz
,
N.
,
Gallo
,
M.
,
Follettie
,
M.
and
De Robertis
,
E. M.
(
1997
).
Cerberus-like is a secreted factor with neuralizing activity expressed in the anterior primitive endoderm of the mouse gastrula
.
Mech. Dev
.
68
,
45
57
.
Biben
,
C.
,
Stanley
,
E.
,
Fabri
,
L.
,
Kotecha
,
S.
,
Rhinn
,
M.
,
Drinkwater
,
C.
,
Lah
,
M.
,
Wang
,
C.-C.
,
Nash
,
A.
,
Hilton
,
D.
,
Ang
,
S. L.
,
Mohun
,
T.
and
Harvey
,.
P
. (
1998
).
Murine cerberus homologue mCer-1: a candidate anterior patterning molecule
.
Dev. Biol
.
194
,
135
151
.
Bouwmeester
,
T.
,
Kim
,
S.
,
Sasai
,
Y.
,
Lu
,
B.
and
De Robertis
,
E. M.
(
1996
).
Cerberus is a head-inducing secreted factor expressed in the anterior endoderm of Spemann’s organizer
.
Nature
382
,
595
601
.
Bouwmeester
,
T.
and
Leyns
,
L.
(
1997
).
Vertebrate head induction by anterior primitive endoderm
.
BioEssays
19
,
855
863
.
Bradley
,
A.
(
1987
).
Production and analysis of chimeric mice
.
In Teratocarcinomas and Embryonic Stem Cells: A Practical Approach
(ed.
E. J.
Robertson
), pp.
113
151
.
Oxford
:
IRL Press
.
Conlon
,
F. L.
,
Barth
,
K. S.
and
Robertson
,
E. J.
(
1991
).
A novel retrovirally-induced embryonic lethal mutation in the mouse; assessment of the developmental fate of ES cells homozygous for the 413d proviral integration
.
Development
111
,
969
981
.
Deng
,
J.
,
Yang
,
L.
,
Yan
,
Y-T.
,
Chen
,
A.
,
Desai
,
N.
,
Wnyshaw-Boris
,
A.
and
Shen
,
M. M.
(
1998
).
Cripto is required for correct orientation of the anterior-posterior axis in the mouse embryo
.
Nature
395
,
702
707
.
Faust
,
C.
,
Schumacher
,
A.
,
Holdener
,
B.
and
Magnuson
,
T.
(
1995
).
The eed mutation disrupts anterior mesoderm production in mice
.
Development
121
,
273
285
.
Friedrich
,
G.
and
Soriano
,
P.
(
1991
).
Promoter traps in embryonic stem cells: A genetic screen to identify and mutate developmental genes in mice
.
Genes Dev
.
5
,
1513
1523
.
Fujii
,
T.
,
Pichel
,
J. G.
,
Taira
,
M.
,
Toyama
,
R.
,
Dawid
,
I. B.
and
Westphal
,
H.
(
1994
).
Expression patterns of the murine LIM class homeobox gene lim1 in the developing brain and excretory system
.
Dev. Dyn
.
199
,
73
83
.
Hamburger
,
V.
(
1988
).
The Heritage of Experimental Embryology: Hans Spemann and the Organizer
.
New York
:
Oxford University Press
.
Hermesz
,
E.
,
Mackem
,
S.
and
Mahon
,
K. A.
(
1996
).
Rpx: a novel anterior-restricted homeobox gene progressively activated in the prechordal plate, anterior neural plate and Rathke’s pouch of the mouse embryo
.
Development
122
,
41
52
.
Hogan
,
B.
,
Beddington
,
R.
,
Costantini
,
F.
and
Lacy
,
E.
(
1994
).
Manipulating the Mouse Embryo: A Laboratory Manual
.
Cold Spring Harbor
:
Cold Spring Harbor Press
.
Lawson
,
K. A.
and
Pedersen
,
R. P.
(
1987
).
Cell fate, morphogenetic movement and population kinetics of embryonic endoderm at the time of germ layer formation in the mouse
.
Development
101
,
627
652
.
Liu
,
P.
,
Wakamiya
,
M.
,
Shea
,
M. J.
,
Albrecht
,
U.
,
Behringer
,
R. R.
and
Bradley
,
A.
(
1999
).
Requirement for Wnt3 in vertebrate axis formation
.
Nature Genetics
22
,
361
365
.
Mombaerts
,
P.
,
Wang
,
F.
,
Dulac
,
C.
,
Chao
,
A. K.
,
Nemes
,
A.
,
Mendelsohn
,
M.
,
Edmondson
,
J.
and
Axel
,
R.
(
1996
).
Visualizing an olfactory sensory map
.
Cell
87
,
675
686
.
Nagy
,
A.
,
Gocza
,
E.
,
Merentes Diaz
,
E.
,
Prideaux
,
V. R.
,
Ivanyi
,
E.
,
Markkula
,
M.
and
Rossant
,
J.
(
1990
).
Embryonic stem cells alone are able to support fetal development in the mouse
.
Development
110
,
815
821
.
Rhinn
,
M.
,
Dierich
,
A.
,
Shawlot
,
W.
,
Behringer
,
R. R.
,
Le Meur
,
M.
and
Ang
,
S. L
. (
1998
).
Sequential roles for Otx2 in visceral endoderm and neuroectoderm for forebrain and midbrain induction and specification
.
Development
125
,
845
856
.
Robertson
,
E. J.
(
1987
).
Embryo-derived stem cell lines
.
In Teratocarcinomas and Embryonic Stem Cells: a Practical Approach
(ed
E. J.
Robertson
), pp.
71
112
.
Oxford
:
IRL Press
.
Rosenquist
,
T. A.
and
Martin
,
G. R.
(
1995
).
Visceral endoderm-1 (VE-1): an antigen marker that distinguishes anterior from posterior embryonic visceral endoderm in the early post-implantation mouse embryo
.
Mech. Dev
.
49
,
117
121
.
Rossant
,
J.
(
1986
).
Development of extraembryonic cell lineages in the mouse embryo
.
In Experimental Approaches to Mammalian Development
(ed.
J.
Rossant
and
R. A.
Pedersen
), pp.
97
120
.
Cambridge
:
Cambridge University Press
.
Shawlot
,
W.
and
Behringer
,
R. R.
(
1995
).
Requirement for Lim1 in head organizer function
.
Nature
374
,
425
430
.
Shawlot
,
W.
,
Deng
,
J. M.
and
Behringer
,
R. R.
(
1998
).
Expression of the mouse cerberus-related gene, Cerr1, suggests a role in anterior neural induction and somitogenesis
.
Proc. Natl. Acad. Sci. USA
95
,
6198
6203
.
Shimono
,
A.
and
Behringer
,
R. R.
(
1999
).
Isolation of novel cDNAs by subtractive hybridizations between the anterior mesendoderm of single gastrula stage embryos
.
Dev. Biol
.
209
,
369
380
.
Spemann
,
H.
and
Mangold
,
H.
(
1924
).
Über induktion von embryonalanlagen durch implantation artfremder organisatoren
.
Roux’s Arch. Dev. Biol
.
100
,
599
638
.
Spemann
,
H.
(
1927
).
Neue arbeiten über organisatoren in der tierischen entwicklung
.
Naturwissenschaften
15
,
946
951
.
Spemann
,
H.
(
1931
).
Über den anteil von implantat und wirtskem an der orientierung und beschaffenheit der induzierten embryonalanlage
.
Wilhelm Roux Arch. EntwMech. Org
.
123
,
389
517
.
Tam
,
P. P. L.
,
Steiner
,
K. A.
and
Quinlan
,
G. A.
(
1997
).
Lineage and functional analyses of the mouse organizer
.
In Cold Spring Harbor Symposia on Quantitative Biology LXII
, pp.
135
144
,
Cold Spring Harbor
:
Cold Spring Harbor Laboratory Press
.
Tam
,
P. P. L.
and
Behringer
,
R. R.
(
1997
).
Mouse gastrulation: the formation of the mammalian body plan
.
Mech. Dev
.
68
,
3
25
.
Thomas
,
P.
and
Beddington
,
R. S. P.
(
1996
).
Anterior primitive endoderm may be responsible for patterning the anterior neural plate in the mouse embryo
.
Curr. Biol
.
6
,
1487
1496
.
Thomas
,
P.
,
Brickman
,
J. M.
,
Popperl
,
H.
,
Krumlauf
,
R.
and
Beddington
,
R.S. P.
(
1997
).
Axis duplication and anterior identity in the mouse embryo
.
In Cold Spring Harbor Symposia on Quantitative Biology LXII
, pp.
115
125
.
Cold Spring Harbor
:
Cold Spring Harbor Laboratory Press
.
Thomas
,
P. Q.
,
Brown
,
A.
and
Beddington
,
R. S.
(
1998
).
Hex: a homeobox gene revealing peri-implantation asymmetry in the mouse embryo and an early transient marker of endothelial cell precursors
.
Development
125
,
85
94
.
Varlet
,
I.
,
Collignon
,
J.
and
Robertson
,
E. J.
(
1997
).
nodal expression in the primitive endoderm is required for specification of the anterior axis during mouse gastrulation
.
Development
124
,
1033
1044
.
Wilkinson
,
D. G.
(
1992
).
Whole mount in situ hybridization of vertebrate embryos
.
In In Situ Hybridization: A Practical Approach
(ed.
D. G.
Wilkinson
), pp.
75
83
.
Oxford
:
Oxford University Press
.