Autotaxin is required for the cranial neural tube closure and establishment of the midbrain–hindbrain boundary during mouse development
Abstract
Autotaxin (ATX) is a lysophospholipid‐generating exoenzyme expressed in embryonic and adult neural tissues. We previously showed that ATX is expressed in the neural organizing centers, anterior head process, and midbrain‐hindbrain boundary (MHB). To elucidate the role of ATX during neural development, here we examined the neural phenotypes of ATX‐deficient mice. Expression analysis of neural marker genes revealed that lateral expansion of the rostral forebrain is reduced and establishment of the MHB is compromised as early as the late headfold stage in ATX mutant embryos. Moreover, ATX mutant embryos fail to complete cranial neural tube closure. These results indicate that ATX is essential for cranial neurulation and MHB establishment. Developmental Dynamics 240:413–421, 2011. © 2011 Wiley‐Liss, Inc.
INTRODUCTION
The central nervous system (CNS) emerges through a dynamic morphogenetic process called neurulation (see Colas and Schoenwolf,2001, for review), in which a portion of the dorsal ectoderm is specified to become the neural ectoderm, bends to form neural folds, and eventually closes as a neural tube. As development proceeds, three primary brain vesicles, forebrain, midbrain, and hindbrain, are further subdivided along the anterior‐posterior axis in parallel with the dorsal‐ventral axis specification. These processes are orchestrated by local organizing centers of the developing brain and spinal cord, such as the anterior neural ridge (ANR) in the rostral forebrain, the midbrain‐hindbrain boundary (MHB), and the floor plate in the ventral midline of the neural tube. These organizing centers direct intricate patterning events via precise control of local signaling molecules, transcription factors, and ultimately gene expression changes (see Kiecker and Lumsden,2005, for review). Therefore, to fully understand how the CNS develops, it is crucial to elucidate the roles of genes specifically expressed in these neural organizing centers.
Here we focus on autotaxin (ATX), because it is expressed in the head process, MHB, and floor plate, thereby well positioned to play an instructive role during early neural development (Bachner et al.,1999; Ohuchi et al.,2007; Koike et al.,2009). ATX is a secreted protein originally identified as a cell‐motility factor from a human melanoma cell line (Stracke et al.,1992). It is also known as ectonucleotide pyrophosphatase‐phosphodiesterase 2 (Enpp2), a member of the nucleotide pyrophosphatase (NPP) family (see Stefan et al.,2005, for review). In addition to its ability to hydrolyze nucleotides in vitro, ATX was found to be identical to lysophospholipase D (lysoPLD), a secreted enzyme present in the plasma that converts lysophosphatidylcholine into bioactive lysophosphatidic acid (LPA) (Tokumura et al.,2002; Umezu‐Goto et al.,2002). LPA stimulates cell proliferation, migration, and survival by acting on G protein–coupled receptors LPA1 through LPA6 as well as some as yet uncharacterized receptors (see van Meeteren and Moolenaar,2007; Chun et al.,2010, for review).
To elucidate its physiological functions, ATX‐deficient mice have been generated using several strategies. Deletion of exons 6 and 7, which encode the active catalytic domain of ATX (van Meeteren et al.,2006; Koike et al.,2009), and deletion of exons 1 and 2, which include the initiation codon and first 45 amino acids (Tanaka et al.,2006; Fotopoulou et al.,2010), abolished ATX protein and activity. Moreover, mice in which threonine 210 in the catalytic domain is replaced with alanine have also been generated (Ferry et al.,2007). All of these ATX‐deficient homozygous mice die during early development and exhibit angiogenic defects in the yolk sac, allantois malformation, and severe brain/neural tube defects. These results indicate that ATX and its product LPA are crucial for early mouse development.
In this study, to elucidate the role of ATX in brain patterning, we examined the neural phenotypes of ATX‐null mice in early developmental stages.
RESULTS AND DISCUSSION
ATX (Enpp2, Mouse Genome Informatics) is reportedly expressed in the head process and yolk sac at embryonic day 7.5 (E7.5) (Koike et al.,2009). Because ATX‐deficient embryos exhibit obvious morphological defects, such as neural tube malformation and asymmetric headfolds, by E8.5 (van Meeteren et al.,2006; Fotopoulou et al.,2010), we re‐examined the embryonic ATX expression pattern between E7.0 and E7.5. At E7.0, ATX was diffusely expressed in the extraembryonic region (Fig. 1A). By E7.5, ATX became expressed in the yolk sac and anterior head process (AHP) (Fig. 1B–F). Later, at E8.5 and before embryonic turning, ATX was expressed in the rostral headfold anterior to the prospective hindbrain (not shown; Koike et al.,2009). After embryonic turning (at E8.75), expression in the developing brain became restricted to the MHB (not shown; Ohuchi et al.,2007). The MHB, referred to as the isthmus, is the tissue that can induce the surrounding cells to form an ectopic midbrain and cerebellum with precise patterns when transplanted into the posterior part of the diencephalon and rhombomeres of the hindbrain, respectively (see Joyner et al.,2000, for review). Although Fibroblast Growth Factor (FGF8) secreted from the MHB was shown to mimic such isthmic activity, it is still a challenge to determine how the molecular cascade that arises from the MHB patterns the midbrain–hindbrain region.
In situ hybridization for ATX in E7.0 and E7.5 embryos. A–C: Lateral views, posterior is to the right. Arrowheads indicate the constriction delineating embryonic and extraembryonic regions. A–C are shown at the same magnification. B′, C′: Anterior upward views of B and C, respectively. By E7.5 ATX is expressed in the yolk sac (ys) and AHP notochord (ahp). D–F: Transverse sections of the embryo. After whole mount in situ hybridization, the embryo was sectioned at the levels shown in C. E, F: ATX is expressed in the visceral endoderm of the yolk sac and AHP notochord (E), but not in the posterior notochordal plate (D) or the future prechordal plate (F). al, allantois; epc, ectoplacental cone; ne, neuroepithelium; ng, neural groove; pg, primitive groove. Scale bar = 100 μm (D); D–F are shown at the same magnification.
The early expression pattern of ATX in the rostral axial mesoderm prompted us to examine the expression domain of forebrain marker genes orthodenticle homolog 2 (Otx2) and sine oculis‐related homeobox 3 homolog (Six3). Otx2 was expressed in the wild‐type prospective forebrain and midbrain during brain regionalization (E8.0; Fig. 2A,C; see Boyl et al.,2001, for review). In ATX‐deficient embryos, intense expression of Otx2 was observed in the rostral region of the embryo, although a large effusion in the head caused Otx2 expression to be asymmetric (Fig. 2B,D; n = 3/3). Six3 was also expressed in the anterior neural fold, which is located in the rostral part of the Otx2‐expression domain, including the ANR (Fig. 2E,G; Oliver et al.,1995). In the ATX‐deficient embryo, Six3 was expressed in the rostral part of the forebrain, though asymmetrically due to effusion (E8.5; Fig. 2F,H; n = 2/2). To visualize the notochord and posterior primitive streak region, we examined the expression of Brachyury (T) (Herrmann,1991). T was expressed in the mutant notochord and primitive streak region, but its expression was disrupted in places (E8.0, n = 3/3; Fig. 2J,L; compare with Fig. 2I,K) and the notochord became kinky and wavy in the rostral part of the embryo by around E8.5 (n = 7/7) (Fig. 2N,P; compare with Fig. 2M,O), suggesting a defect in anterior notochord formation. Simultaneous in situ hybridization with T and Wnt1 probes showed that such abnormal expression of T was observed from the midbrain to hindbrain levels where the neural tube closure defect was obvious (see Fig 6 and Supp. Fig. S1, which is available online). Cross‐sections of the hybridized embryos showed that expression domain of T appeared to either expand or reduce in ATX‐deficient mice depending on the axial level (Fig. 2R,S; compare with Fig. 2Q).
Expression patterns of Otx2 (A–D), Six3 (E–H), and T (Brachyury) (I–S) in control and ATX‐deficient embryos at E8.0 (A–L) and E8.5 (M–S). A, B, E, F, I, J, M, N: Lateral views. C, D, G, H, K, L, O, P: Frontal views. Q–S: Cross‐section of the hybridized embryos after in situ hybridization. D, H: Large effusion in the head repressed the expression of Otx2 and Six3 on the right side. J, L: At E8.0, the linear expression of T in the notochord was discontinuous in the rostral portion (arrowheads). N, P: At E8.5, rostral expression of T was kinky and wavy (arrowheads). Q–S: Expression of T in the notochord (arrow) at the level of hindbrain in the ATX‐heterozygote (Q) and ATX−/− mouse (R, S). Mid‐hindbrain level (Q), upper hindbrain level (R), and lower hindbrain level (S). Scale bar = 100 μm (S); Q–S are shown at the same magnification.
FGF8 which is secreted from the ANR and the MHB, is involved in the patterning of the anterior forebrain and midbrain/hindbrain, respectively (see Joyner et al.,2000; Rubenstein et al.,1998, for review). We thus examined the expression of Fgf8 and other marker genes to determine whether abnormalities in the ANR and MHB were present in ATX‐deficient mice. We found that Fgf8 expression in the MHB was decreased in some mutants (Fig. 3D,E; n = 6/10) and unchanged in others (Fig. 3G,H; n = 4/10). Fgf8 expression in the ANR by around E8.5 was very weak in some mutants (Fig. 3I; n = 2/8) and more restricted to the medial portion in others (Fig. 3F; n = 6/8). Since it was reported that Fgf8 hypomorphic mutations in mouse results in a small telencephalon correlated with cell death and reduced proliferation at E9 (Storm et al.,2006), it would be intriguing to know whether alterations in Fgf8 expression of the ATX‐deficient mouse may also cause a similar prosencephalic phenotype at later stages, which should be clarified by conditional knockout mice.
Expression patterns of Fgf8 in wild‐type (A–C) and ATX−/− (D–I) embryos at E8.5. A, D, G: Lateral views. B, E, H: Dorsal views of midbrain and hindbrain regions. C, F, I: Frontal views. Asterisks indicate the MHB. Arrowheads indicate the ANR. A–C: Fgf8 was normally expressed in the ANR, MHB, and primitive streak region. D, E: In this ATX−/− embryo, Fgf8 was hardly detectable in the MHB (asterisk). The large central region of Fgf8 expression is that in the primitive streak region seen from this angle. F: Fgf8 expression domain in this mutant ANR did not expand laterally. G–I: In this ATX−/− embryo, Fgf8 was expressed in the MHB (G, H) but not in the ANR.
We next examined the expression of Tcf4 (transcription factor 7‐like 2), homeobox expressed in ES cells 1 (Hesx1), and NK2 homeobox 1 (Nkx2.1) to further examine the rostral forebrain patterning of mutant embryos. In the wild‐type embryos, Tcf4 is expressed in the rostral forebrain, which represents the prospective ventral telencephalon (Fig. 4A,C; Ishibashi and McMahon,2002). Tcf4 was expressed in the mutant forebrain neural fold, but lateral expansion of the Tcf4‐expressing domain was reduced and the mutant had a V‐shaped rostral neural fold (Fig. 4B,D; n = 2/2). At E8.5, Hesx1 is normally expressed in the anterior neural fold (Fig. 4E,G; Hermesz et al.,1996), whereas its expression was also diminished and confined medially in ATX mutants (Fig. 4F,H; n = 2/2). At around E8.0, Nkx2.1 is normally expressed in the rostral midline, which represents the prospective rostral diencephalon, hypothalamus (Fig. 4I; Shimamura and Rubenstein,1997; Puelles and Rubenstein,2003). Nkx2.1 was detected in the rostral midline of ATX‐deficient embryos, though affected by effusion (Fig. 4J; n = 3/3). These marker expression patterns indicate that the rostral forebrain neural folds do not expand laterally and exhibit a V‐shaped morphology, which seems to be related to cranial neural tube defects (mentioned below) in ATX‐deficient mice.
Expression patterns of Tcf4, Hesx1, and Nkx2.1 at E8.0 and E8.5. A, B: Fronto‐ventral views of the wild‐type (A) and ATX−/− (B) embryo. Tcf4 was expressed in the ANR. C, D: Frontal views of the wild‐type (C) and ATX−/− (D) embryo. In the mutant, Tcf4 expression did not expand laterally and exhibited a V‐shaped morphology in the midline region of the ANR. E–H: Hesx1 was expressed in the rostral forebrain in the ATX+/− embryo (E, G), whereas its expression domain was restricted to medial and rostral areas in the ATX‐deficient forebrain (F, H). Lateral views (E, F) and frontal views (G, H) are shown. I, J: Nkx2.1 was normally expressed in the medial anterior neural plate (I), whereas it was asymmetrically expressed in the mutant embryo due to a large effusion on the right side in this case (J).
Previous studies have shown that ATX expression overlaps with the Wnt1 expression domain in the MHB, which is anterior to the Gbx2 and Pax2 expression domains (Ohuchi et al.,2007). The expression pattern of Wnt1 in the developing MHB changed as the development of the boundary proceeded. Initially, Wnt1 was broadly expressed and the boundaries of its expression were not sharply defined (Fig. 5A′). Subsequently, cell‐sorting processes resulted in sharpening of the Wnt1 expression borders (Kiecker and Lumsden,2005) and restricting Wnt1 expression to the posterior end of the midbrain abutting the future MHB (Fig. 5A). We found that some ATX‐deficient mice expressed Wnt1 in a wild‐type pattern (n = 3/5; Fig. 5D; Fig. 5A as control), whereas other ATX‐deficient embryos had a broader Wnt1 expression domain than the wild‐type embryos at E8.0 (n = 2/5; Fig. 5D′). Similarly, Gbx2 and Pax2 were more broadly expressed, and at lower levels, in the mutant MHB region (n = 2/2 for Gbx2; n = 4/4 for Pax2; Fig. 5B,C, E, F). Thus, in the absence of functional ATX, the MHB is initially formed but its establishment is compromised in some mutant embryos.
Expression patterns of Wnt1, Gbx2, and Pax2 in wild‐type (A–C) and ATX−/− (D–F) embryos at E8.0 and E8.5. Anterior is at the top. A, D: In both wild‐type and ATX‐deficient embryos, Wnt1 was expressed in the posterior end of the midbrain abutting the MHB and future dorsal margin of the midbrain and hindbrain neural fold. A′: At E8.0, Wnt1 was expressed broadly in the posterior end of the midbrain abutting the future MHB. D′: The Wnt1 expression domain in the posterior midbrain was broader in mutants than in wild‐type. B, E: Gbx2 was distinctly expressed in the anterior end of the hindbrain in wild‐type embryos (B), whereas it was expressed weakly in the ATX−/− neural fold (E). C, F: Pax2 is distinctly expressed in the MHB of the wild‐type embryo (C), whereas it was weakly and broadly expressed in the ATX−/− neural fold (F).
ATX‐deficient embryos exhibit defects in neural tube closure. In the mouse, primary neural tube closure is initiated at the hindbrain‐cervical boundary at E8.5 (closure 1) (Copp,2005). Following closure 1, two additional closures occur: one at the forebrain‐midbrain boundary (closure 2) and the other at the extreme rostral end of the forebrain (closure 3). Bidirectional spread of neurulation between these closures leads to the completion of neural tube formation. Wnt1 expression in the future roof of the neural tube revealed that in normal embryos closure 2 extends posteriorly to the anterior hindbrain by E8.75, and closure 1 extends anteriorly to the level of the otic vesicles (Fig. 6A). By contrast, neurulation between closures 1 and 2 did not proceed in the mutant embryos (n = 2/2; Fig. 6B,D). These findings demonstrate that ATX is required for the completion of the cranial closure.
Expression patterns of Wnt1 in the wild‐type (A, C, E) and ATX−/− (B, D, F) embryos at E8.5 and E9.0. A, B: Dorsal views of the MHB and hindbrain region. C, D: Frontal views. E, F: Lateral views. Asterisks show the MHB. Yellow arrowheads show the forebrain‐midbrain boundary, where closure 2 of the neural folds occurs and anterior dorsal Wnt1 expression tapers at this stage. In the ATX‐deficient embryo, defects in neural tube closure were seen from closure 2 toward the hindbrain. The arrow in F indicates a large effusion in the head. ov, otic vesicle.
It has been known that the AHP, where ATX is expressed in early stages, is formed from dispersed notochord progenitors found anterior to the forming node in the gastrula, which converge directly onto the midline (Yamanaka et al.,2007). The AHP is a structure that forms between the prechordal plate and trunk notochord, under the midbrain and rostral hindbrain (Rowan et al.,1999). Given that the activity of the AHP is required for the development of the forebrain (Camus et al.,2000), our findings suggest that ATX secreted from the AHP and/or proper development of the AHP by virtue of ATX may be important for the lateral expansion of rostral forebrain neural folds in association with a cranial neural tube closure at the midbrain to hindbrain levels.
ATX and LPA are known to promote cellular proliferation (Umezu‐Goto et al.,2002; Ferry et al.,2003). To examine whether mitotic defects underlie the mutant brain abnormalities, immunostaining for the mitotic marker phosphorylated histone H3 (ph H3) was performed. At E7.5, ph‐H3‐positive cells were observed in the neuroectoderm as well as in the head mesenchyme of both wild‐type and ATX‐deficient embryos (Fig. 7B,D). At E8.5, ph‐H3‐positive cells were still observed in mutant embryos as well as wild‐type embryos (Fig. 7E–J, K–P). Normally, mitotic cells are located on the apical side of the neuroepithelium. In both wild‐type and ATX mutant embryos, ph‐H3‐positive cells in the neuroectoderm were located on the apical side, indicating that the apico‐basal polarity of the mutant neuroepithelium formed normally. However, α‐tubulin staining of the neuroepithelium at E8.5 revealed a slightly irregular structure of the epithelium, particularly on the basal side (Fig. 7K,M,O; E,G, I as control). To better visualize the basement membrane separating epithelial and mesencehymal tissues, we examined laminin localization. In the E8.5 ATX‐deficient embryo, localization of laminin appeared almost normal, but laminin underlying the prospective hindbrain neural fold, which abuts a large effusion, was disrupted and rough (Fig. 8G, H; C, D as control). The mesenchymal cells abutting the disrupted basal membrane were compressed to the overlying epithelium, lost the loose appearance of mesenchymal cells, and exhibited an epithelial‐like packed appearance (Fig. 8H).
Localization of α‐tubulin (green) and phospho‐histone H3 (ph‐H3) (magenta) in wild‐type (top panels) and ATX−/− (bottom panels) embryos at E7.5 (A–D) and E8.5 (E–P). Transverse sections are shown. A′, C′, E′, K′: Nuclei were stained with 4′, 6‐diamidino‐2‐phenylindole (DAPI). High‐magnification pictures of neural folds in E, F, K, L are shown in G, I; H, J; M, O; N, P, respectively. Asterisks in K′ show a large effusion in the mutant head region. Comparable amounts of mitotic cells were seen in the neural folds and head mesenchyme in the wild‐type and ATX−/− embryos at both stages. At E8.5, localization of α‐tubulin on the basal side of the mutant neural folds appeared to be wavy (K; arrowheads in M, O). a, amnion; c, coelomic cavity; cn, caudal neural plate; fd, foregut diverticulum; hb, hindbrain neural fold; m, head mesenchyme; ne, neural ectoderm.
Localization of laminin (green) in wild‐type (top panels) and ATX−/− (bottom panels) embryos at E8.5. Transverse sections are shown. Nuclei were stained with DAPI. C, G: Merged images of A, B and E, F, respectively. D, H: High‐magnification pictures of the hindbrain neural fold in the boxed areas in C and G are shown. Laminin staining revealed a smooth basement membrane between the neuroepithelium and the underlying mesenchyme in the wild‐type embryo (D), whereas the mutant basement membrane splits finely and the mesenchymal cells abut the overlying neuroepithelium (H). Asterisks in F, H indicate a large effusion in the mutant head region. cn, caudal neural plate; da, dorsal aorta; fb, forebrain neural fold; fd, foregut diverticulum; hb, hindbrain neural fold; hd, hindgut diverticulum.
To examine whether programmed cell death is correlated with the early neural abnormalities in ATX mutants, we further performed terminal UTP nick end labeling (TUNEL) staining. In the E8.5 wild‐type embryos, no TUNEL‐positive cells were found in the forebrain or hindbrain neuroepithelia (Fig. 9A,D). By contrast, in mutant embryos, TUNEL‐positive cells were distinctly observed in some restricted areas. A population of apoptotic cells was observed in the prospective hindbrain neuroepithelium abutting a large effusion (Fig. 9B,C, C′), where the basement membrane was disrupted (Fig. 8G,H). The finding that apoptosis occurs near the effusions is consistent with that previously reported by van Meeteren et al. (2006). However, TUNEL‐positive cells were not observed in the forebrain neural fold at E8.5. Thus, in ATX‐deficient mice, destruction of the basement membrane and apoptosis were observed in the brain neural fold abutting the effusion at the late headfold stage (E8.5). As reported recently, ATX‐deficient mice have effusions (head cavity formation) in almost all embryos at E8.5 due to deficiency in the LPA‐receptor‐Rho‐ROCK pathway (Koike et al.,2010). Laterality of the head cavities (effusions) in the ATX‐deficient mice was not obvious in E7.5 to E8.5 embryos (right, 28.5%; left, 21.4%, bilateral, 42.8%; n = 28). It is difficult to attribute all the neural phenotypes in ATX‐deficient embryos to the direct consequence of ATX disruption, because the head effusions observed in almost all ATX‐deficient embryos may influence gene expression changes described here and neural tube defects. Given that ATX has pleiotropic roles during early mouse development, future studies on the role for ATX in brain development using conditional knockout mice would be helpful to discriminate the effect of ATX deficiency in different regions.
TUNEL staining of wild‐type (A, D) and ATX−/− (B, C, C′, E, F, F′) embryos at E8.5. Transverse sections are shown. Nuclei in A–C, F were stained with DAPI. D: In the wild‐type embryo, no TUNEL‐positive cells were detected. A, B, C, F: Merged images of TUNEL and nuclear staining are shown. C, C′, F, F′: High‐magnification pictures of the hindbrain (C, C′) and foregut/forebrain region (F, F′) shown in B, E. In the ATX‐deficient embryo, TUNEL‐positive cells were seen in the neuroepithelium and the underlying mesenchyme abutting a large effusion (C, C′), and in a few cells near the forgut diverticulum and forebrain neuroepithelium (F, F′). Asterisks in B, C indicate a large effusion in the mutant head region. cn, caudal neural plate; da, dorsal aorta; fb, forebrain neural fold; fd, foregut diverticulum; hb, hindbrain neural fold.
In summary, we have shown that the LPA‐producing enzyme ATX, which is secreted from the anterior notochord and MHB, is an important player in the morphogenesis of the rostral forebrain, the establishment of the MHB, and neural tube closure at the midbrain‐hindbrain region. Although a decrease in cell proliferation and massive cell death were observed at later stages in the absence of ATX (Fotopoulou et al.,2010), our findings suggest the possibility that perturbation of mitosis or programmed cell death in the neuroepithelium is not the primary cause of the abnormalities in neural development observed in ATX‐deficient mice. Nevertheless, variability in gene expression pattern of the ATX‐deficient mice shows that decreased cell proliferation and cell death more or less influence the early phenotype in brain development.
EXPERIMENTAL PROCEDURES
Mice, Genotyping, and Staging
All the experiments using animals were approved by the Animal Care and Use Committee of the University of Tsukuba and performed under its guidelines. Noon of the day on which a vaginal plug was observed was taken as embryonic day 0.5 (E0.5). The generation and genotyping of ATX (Enpp2)‐deficient mice was described previously (Koike et al.,2009). Embryos were staged according to Kaufman (1995). Because the ATX‐deficient mice may be developmentally delayed, careful examination was performed on their morphology of the head folds and whole bodies for staging. Highly deteriorated mutant embryos were excluded from the analysis in this study.
In Situ Hybridization
Whole mount in situ hybridization (ISH) was either performed as described previously (Koike et al.,2009), or using a semi‐automatic ISH apparatus (HS‐5100; Aloka, Co. Ltd., Japan) (Ohuchi et al.,2007). In situ probes used were described previously or cloned in this study as listed in Supp. Table S1. For selected embryos, frozen sections (10 μm) were prepared after in situ hybridization to further examine gene expression domains.
Immunofluorescence and TUNEL Staining
Embryos fixed with 4% paraformaldehyde in phosphate buffered saline (PBS) were paraffin‐embedded and sectioned at 5 μm thickness. Pretreatment of the sections were done as previously described (Mizugishi et al.,2005; Gersdorff et al.,2005) with minor modifications. For anti‐laminin staining, sections were treated with proteinase K (10 μg/ml) for 5 min at room temperature. Primary antibodies used were mouse monoclonal anti‐α‐tubulin (1:250; Sigma T9026, St. Louis, MO), rabbit polyclonal anti‐phosphohistone H3 (Ser10) (1:2,000; Upstate Biotechnology, East Syracuse, NY), and rabbit polyclonal anti‐laminin A1/B1 (1:200; Abcam ab11575, Cambridge, MA). Secondary antibodies used include AlexaFluor 488‐conjugated anti‐mouse IgG (1:500; Invitrogen, San Diego, CA), Cy3‐conjugated anti‐rabbit IgG (1:500; Jackson ImmunoResearch, West Grove, PA), and AlexaFluor 488‐conjugated anti‐rabbit IgG (1:500; Invitrogen). TUNEL staining was performed using a Click iT TUNEL AlexaFluor 488 imaging assay kit (Invitrogen). Fluorescent images were obtained using a laser scanning confocal microscope (Eclipse C1si Confocal System; Nikon, Melville, NY). Immunofluorescence and TUNEL analysis were performed on at least two embryos for each genotype.
Acknowledgements
Probes were generously provided through Shinichi Aizawa (Riken Kobe, Japan), Yasuhide Furuta (MD Anderson, Texas), and Shinji Takada (NIBB, Japan). We thank Takumi Kawaue, Eri Taguchi, and Takuya Okada for their technical assistance.
