Neural tube defects and impaired neural progenitor cell proliferation in Gβ1-deficient mice

Authors

  • Hiroaki Okae,

    1. Center for Experimental Medicine and Systems Biology, Institute of Medical Science, University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Yoichiro Iwakura

    Corresponding author
    1. Center for Experimental Medicine and Systems Biology, Institute of Medical Science, University of Tokyo, Tokyo, Japan
    2. Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, Saitama, Japan
    • Center for Experimental Medicine and Systems Biology, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo 108-8639, Japan
    Search for more papers by this author

Abstract

Heterotrimeric G proteins are well known for their roles in signal transduction downstream of G protein–coupled receptors (GPCRs), and both Gα subunits and tightly associated Gβγ subunits regulate downstream effector molecules. Compared to Gα subunits, the physiological roles of individual Gβ and Gγ subunits are poorly understood. In this study, we generated mice deficient in the Gβ1 gene and found that Gβ1 is required for neural tube closure, neural progenitor cell proliferation, and neonatal development. About 40% Gβ1−/− embryos developed neural tube defects (NTDs) and abnormal actin organization was observed in the basal side of neuroepithelium. In addition, Gβ1−/− embryos without NTDs showed microencephaly and died within 2 days after birth. GPCR agonist-induced ERK phosphorylation, cell proliferation, and cell spreading, which were all found to be regulated by Gαi and Gβγ signaling, were abnormal in Gβ1−/− neural progenitor cells. These data indicate that Gβ1 is required for normal embryonic neurogenesis. Developmental Dynamics 239:1089–1101, 2010. © 2010 Wiley-Liss, Inc.

INTRODUCTION

Heterotrimeric G proteins consist of three subunits, Gα, Gβ, and Gγ, and are best known for their roles in signal transduction downstream of G protein–coupled receptors (GPCRs). GPCRs activated by ligand binding catalyze exchange of GDP for GTP on the Gα subunit, and binding of GTP results in dissociation of Gα from tightly associated Gβγ dimer. Gα and Gβγ subunits both interact with downstream effector molecules such as phospholipases, adenylate cyclases, and ion channels, and initiate intracellular signal transduction pathways (Oldham and Hamm, 2008; Smrcka, 2008).

Sixteen Gα genes have been identified in the human and mouse genomes, and are divided into four classes, Gαs, αi, αq, and α12/13. Almost all Gα genes have now been knocked out in mice and the function of each Gα subunits is relatively well known. However, the roles of individual Gβ and Gγ subunits are poorly understood (Chen et al., 2003; Schwindinger and Robishaw, 2001; Schwindinger et al., 2003, 2004). Five different Gβ genes and twelve Gγ genes have been identified in the human and mouse genomes. While the amino acid identity is low among Gγ subunits, Gβ1–4 subunits share 79–90% amino acid identity and about 50% identity for Gβ5 subunit. Importantly, most of individual Gβ and Gγ subunits share nearly 100% amino acid identity between mammalian species, suggesting distinct conserved functions for each Gβ and Gγ subunits.

It is known that high concentrations of heterotrimeric G proteins are present in developing neural tissues (Morishita et al., 1999). Several GPCR agonists, GPCRs, and heterotrimeric G proteins are reported to regulate various steps in embryonic neurogenesis: neural tube closure, neuronal proliferation, and neural migration. Sphingosine-1-phosphate (S1P), lysophosphatidic acid (LPA), and Gpr161 are involved in the regulation of neural tube closure (Contos et al., 2000; Matteson et al., 2008; Mizugishi et al., 2005; van Meeteren et al., 2006). Many GPCR agonists, such as S1P, LPA, endothelin-1 (ET1), and cannabinoids, are reported to enhance proliferation of neural progenitor cells through extracellular signal-regulated kinase (ERK) phosphorylation (Estivill-Torrús et al., 2008; Harada et al., 2004; Jiang et al., 2005; Mizugishi et al., 2005; Morishita et al., 2007; Palazuelos et al., 2006; Shinohara et al., 2004). Loss of Gα12 and Gα13 or Gpr56 results in overmigration of neurons, and Cxcr4 and Sdf1 are required for normal cerebellar neuron migration (Li et al., 2008; Ma et al., 1998; Moers et al., 2008; Zou et al., 1998).

As few Gα knockout mice are reported to show defects in embryonic neurogenesis, it is possible that Gβγ signaling plays central roles in the GPCR-dependent embryonic neurogenesis. In particular, regulation of neural progenitor cell proliferation is frequently mediated by ERK in a pertussis toxin (PTX)-sensitive manner. PTX stimulates ADP-ribosylation of Gαi and prevents functional interaction between GPCRs and heterotrimeric G proteins containing Gαi. Many of the PTX-sensitive effector activations are mediated by Gβγsubunits rather than Gα subunits (Schwindinger and Robishaw, 2001; Smrcka, 2008). Thus, Gβγ subunits may play important roles in neural progenitor cell proliferation.

Gβγ subunits are also reported to control mitotic orientation, proliferation, and differentiation of cortical progenitors in a GPCR-independent manner (Sanada and Tsai, 2005). Cortical progenitors in the ventricular zone (VZ) self-renew and generate neurons. The cleavage plane angle of cortical progenitors is important for the daughter cell fates (Doe, 2008; Gotz and Huttner, 2005; Konno et al., 2008). Sequestration of Gβγ subunits randomizes cleavage plane orientation and results in hyperdifferentiation of progenitors into neurons (Sanada and Tsai, 2005). Several Gβ and γ subunits are expressed in the VZ and the localization of Gβ1 in dividing progenitors correlates with the cleavage plane (Morishita et al., 1999; Sanada and Tsai, 2005). However, no Gβ or Gγ knockout mice are reported to show defects in embryonic neurogenesis.

In this study, we generated mice deficient in Gβ1 and found that Gβ1 is required for embryonic neural development. About 40% Gβ1−/− embryos developed neural tube defects (NTDs) and Gβ1−/− embryos without NTDs showed microencephaly. GPCR agonist-induced ERK phosphorylation, cell proliferation, and cell spreading, which were all found to be regulated by Gαi and Gβγ signaling, were abnormal in Gβ1−/− neural progenitor cells. On the other hand, Gαq-dependent JNK phosphorylation and Gα12/13-dependent cell contraction were not significantly affected in Gβ1−/− neural progenitor cells. These data suggest that Gβγ signaling plays crucial roles in embryonic neurogenesis.

RESULTS

Disruption of the Gβ1 Gene by a Gene Trap Insertion

A gene trap screening for membrane-bound proteins expressed in stem cells was performed to identify novel genes required for stem cell proliferation and differentiation. We found that a gene trap line 4B8 expressed β-gal at high levels in an undifferentiated state (H.O. and Y.I., unpublished data). Cloning of the genomic DNA sequence flanking the gene trap insertion site of this trap line revealed that the insertion event occurred in the first intron of the Gβ1 gene (Fig. 1A). The insertion site was also confirmed by Southern blot analysis and genomic PCR (Fig. 1B,C). Gβ1 mutant mice were generated from the Gβ1 gene trap clone. Gβ1 transcripts were not detected in Gβ1−/− embryos (Fig. 1D). By Western blot analysis, the level of Gβ1 protein in Gβ1+/− retina was about half of that in Gβ1+/+ retina, and Gβ1 protein was absent in Gβ1−/− retina (Fig. 1E). These data indicate that the Gβ1 gene is successfully inactivated in these Gβ1−/− mice.

Figure 1.

Disruption of the Gβ1 gene by a gene trap insertion. A: The location of the gene trap element in the Gβ1 gene. The gene trap element contains the splice acceptor site (SA), β-galactosidase (β-gal), and hygromycin phosphotransferase gene (hygR). Arrows indicate the sites of primers used for genotyping. The probe used for Southern blot analysis is also shown. The first exon is non-coding. B: Southern blot analysis of Gβ1+/− ES cells. Genomic DNA was digested with EcoRI and hybridization was carried out using the 32P-labeled DNA probe. C: Genotyping of the Gβ1 gene. D: RT-PCR analysis of RNA extracts from E10.5 embryos. E: Western blot analysis of P0 retinal extracts. Relative band intensity is presented.

Expression Pattern of Gβ1

First, we examined the expression pattern of Gβ1 at the stage of neural tube closure and at birth, because Gβ1−/− embryos exhibited NTDs and died perinatally as described below. X-gal staining and immunostaining with anti-Gβ1 antibody revealed that Gβ1 was highly expressed in the cranial neural plate and neural tube, especially in basal and apical areas (Fig. 2A–E). Gβ1 was also strongly expressed in the spinal cord and limb buds (Fig. 2B,C). Next, to investigate tissue distribution of G protein β subunits, real-time RT-PCR was performed on WT mice at P0. The expression of Gβ1 was detected in many tissues, but high levels of expression were observed in the brain and lung (Fig. 2F). Strong expression of Gβ1 was also observed in the peripheral nervous system (see Supp. Fig. S1, which is available online). Gβ2 and Gβ4 were also highly expressed in the brain and lung (Fig. 2G,I). Gβ3 was specifically expressed in the eye (Fig. 2H). Gβ5 showed relatively ubiquitous expression (Fig. 2J). As Gβ1 was highly expressed in the neural plate, which is mainly composed of undifferentiated neural cells, the expression of G protein α and β subunits was analyzed in in vitro cultured neural progenitor cells. By quantitative real-time RT-PCR, it was shown that Gβ1 was most strongly expressed in neural progenitor cells among five G protein β subunits (Fig. 2K). The expression levels of other G protein α and β subunits were unaffected by the Gβ1 mutation (Fig. 2L).

Figure 2.

Expression pattern of the Gβ1 gene. A: X-gal staining of a coronal section of E8.75 Gβ1+/− embryo. MB, midbrain; SC, spinal cord. B: Whole mount X-gal staining of E10.5 Gβ1+/− embryo. FB, forebrain; HB, hindbrain; LB, limb bud. C: X-gal staining of a coronal section of E10.5 Gβ1+/− embryo. D,E: Immunostaining with anti-Gβ1 antibody of coronal sections of the midbrain at E9.5 (green). Nuclei were stained with Hoechst (blue). F–J: Real-time RT-PCR analysis of G protein β subunits in WT pups at P0. The expression of G protein β subunits was normalized by Gapdh gene expression. K: Quantification of G protein β subunits in WT neural progenitor cells. Messenger RNA copy numbers of G protein β subunits were determined by real-time RT-PCR and are expressed relative to the copy number of Gβ1 mRNA. Standard curves were generated using cloned cDNA fragments of G protein β subunits. L: Real-time RT-PCR analysis of G protein α and β subunits in Gβ1−/− neural progenitor cells. Messenger RNA levels were normalized to those observed in WT neural progenitor cells. Bars = 100 μm in A, E; 200 μm in C. The scale bar in E also applies to D. Data are means ± SD from two to three replicates.

NTDs in Gβ1−/− Embryos

About 40% of Gβ1−/− embryos exhibited midbrain-hindbrain exencephaly at E10.5, while spinal bifida was not observed (Fig. 3A,B and Table 1). About 30% of Gβ1−/− embryos exhibited exencephaly at E18.5 and died perinatally (Fig. 3C and Table 1). No Gβ1+/+ or Gβ1+/− mice examined had NTDs (Table 1). Histological analysis of embryos revealed that dorsolateral bending of cranial neural plate was abnormal in Gβ1−/− embryos (7 of 7 embryos) at the stage just before neural tube closure (Fig. 3D,E). In addition, Gβ1−/− neuroepithelial cells and nuclei were elongated along the apical-basal axis (Fig. 3F–J).

Table 1. Genotyping of Offspring From Gβ1+/− Intercrossesa
Age+/++/−−/−
  • a

    Genotypes of offspring at various stages were determined.

  • The number of mice with NTDs is shown in parentheses.

P2141 (0)87 (0)0 (0)
P217 (0)31 (0)0 (0)
P012 (0)24 (0)5 (0)
E15.5 - E18.522 (0)35 (0)13 (4)
E10.527 (0)59 (0)30 (13)
Figure 3.

Neural tube closure defects in Gβ1−/− embryos. A,B: Lateral views of E10.5 embryos. About 40% of Gβ1−/− embryos exhibited NTDs (arrow). C: Lateral views of E18.5 embryos. About 30% of Gβ1−/− embryos exhibited exencephaly (arrow). D–G: H&E stained coronal sections of the midbrain at E8.75. Abnormal bending of the neural plate is observed in Gβ1−/− embryos (E). F and G are higher magnification views of the boxed areas in D and E. The neuroepithelial cells of Gβ1−/− embryos are relatively elongated (compare F and G). H–J: Analysis of nuclear shape. Coronal sections of the midbrain at E8.75 were stained with Hoechst (blue). The cell membrane was visualized using anti-β-catenin (β-cat) antibody (green). The length-to-width (L/W) ratio of the nuclei except mitotic nuclei at the apical surface was measured as described in J. Nuclei (n = 151) from seven WT embryos and those (n = 130) from six Gβ1−/− embryos were examined. Data are expressed as means ± SEM. *P < 0.05; Student's t-test. Bar = 100 μm in E. The scale bar also applies to D.

As NTDs are associated with abnormal cell proliferation or death in many other mutant lines, cell proliferation and death were analyzed using phospho-histone H3 (PH3, M-phase marker) and TUNEL staining, respectively (Fig. 4A–H). The number of PH3- and TUNEL-positive cells was counted in the dorsal neural folds of the midbrain at E8.75 and it was shown that cell proliferation and apoptotic cell death were normal in Gβ1−/− embryos (Fig. 4C,F). Furthermore, PH3-positive mitotic cells were located at the apical surface of the neuroepithelium in WT and Gβ1−/− embryos (Fig. 4A,B,G,H) and BrdU-positive S-phase nuclei were located at the basal side (Fig. 4G,H). It is well known that DNA synthesis occurs in the basal side of the neuroepithelium and cells divide at the apical surface, and cell cycle–dependent nuclear positioning was unaffected in Gβ1−/− embryos. Shh, which inhibits dorsolateral bending of the neural plate, was also normally expressed in the ventral neural tube of Gβ1−/− embryos at E9.5 (Fig. 4I,J).

Figure 4.

Normal cell proliferation and death in Gβ1−/− embryos at E8.75. A–C: Immunostaining with anti-phospho-histone H3 (PH3) antibody (green) of coronal sections of E8.75 embryos. Embryos were sectioned in a plane passing through the midbrain and optic vesicle (OV). The number of PH3-positive cells in the dorsal neural folds (DNFs) per section was counted (C). Five sections from five embryos were examined in each genotype. D–F: TUNEL staining of coronal sections of E8.75 embryos (green). The number of TUNEL-positive cells in the DNFs of the midbrain per section was counted (F). Five sections from five embryos were examined in each genotype. G,H: Immunostaining with anti-BrdU antibody for S-Phase labeling (green) and anti-PH3 antibody for M-Phase labeling (red) of coronal sections of the midbrain at E9.5. I,J: In situ hybridization with Shh antisense probe of coronal sections of the hindbrain at E9.5. Bars = 100 μm. Data are means ± SD.

S1P and LPA are involved in the regulation of neural tube closure (Contos et al., 2000; Mizugishi et al., 2005; van Meeteren et al., 2006) and are known to induce morphological changes of neural progenitor cells through actin remodeling (Contos et al., 2000; Fukushima et al., 2000; Harada et al., 2004; Hurst et al., 2008). As the actin cytoskeleton is known to play a crucial role in the dorsolateral bending of neural folds (Copp, 2005; Copp et al., 2003), F-actin in the neural tube was visualized with rhodamine-phalloidin (Fig. 5). Double staining with a neural progenitor cell marker, Nestin, revealed that strong F-actin staining was observed in the apical and basal sides of the neural tube (Fig. 5B–D). Around the dorsolateral hinge points (DLHPs), F-actin staining in the basal side of Gβ1−/− neuroepithelium was reduced irrespective of neural tube closure (Fig. 5E,K, and data not shown). In contrast, F-actin staining appeared normal in the apical side of Gβ1−/− neuroepithelium (Fig. 5B,H).

Figure 5.

Reduced F-actin staining in the basal side of Gβ1−/− neuroepithelium. Phalloidin staining of coronal sections of the midbrain at E9.5 (B and H, red). Neural tube was stained with anti-Nestin antibody (C and I, green) and nuclei were counterstained with Hoechst (A and G, blue). D and J are merged images of A-C and G-I, respectively. E, F, K, L: Higher magnification views of the boxed areas in D and J, respectively. F-actin staining in the basal side of Gβ1−/− neuroepithelium is reduced around the DLHPs (compare E and K). At least three embryos were examined in each genotype. Bars = 100 μm.

Abnormal Morphological Change Induced by S1P in Gβ1−/− Neural Progenitor Cells

S1P and LPA induce contraction of neural progenitor cells through actin remodeling in a Rho-dependent manner (Contos et al., 2000; Fukushima et al., 2000; Harada et al., 2004; Hurst et al., 2008). As reported, S1P induced contraction of WT neural progenitor cells at 1 μM or higher concentrations (Fig. 6A,C, and data not shown), and cell contraction was inhibited by the Rho inhibitor Y27632 (Fig. 6D,H). Interestingly, 0.1 μM S1P induced contraction of Gβ1−/− neural progenitor cells but not of WT cells (Fig. 6A,B,E,F). Similar effects were also observed when cell contraction was induced by LPA, but the differences between WT and Gβ1−/− cells were smaller than those of S1P (Supp. Fig. S3). It is known that S1P and LPA inhibit cell contraction through Gαi-dependent Rac activation and induce it through Gα12/13-dependent Rho activation in several cell types (Hurst et al., 2008; van Leeuwen et al., 2003; Watterson et al., 2007). We found that 0.1 μM S1P could induce contraction of WT cells in the presence of PTX (Gαi inhibitor), NSC23766 (Rac inhibitor), or Gallein (Gβγ inhibitor) (Fig. 6I-L). This means that Gαi, Rac, and Gβγ signaling inhibit S1P-induced cell contraction. In addition, staining with rhodamine-phalloidin revealed that actin remodeling occurred when Gβ1−/− neural progenitor cells were stimulated with 0.1 μM S1P but not when WT cells were stimulated (Fig. 6M–P). Actin-rich protrusions were lost in S1P-stimulated Gβ1−/− neural progenitor cells (Fig. 6P).

Figure 6.

Effects of S1P on the morphology of neural progenitor cells. A–H: S1P-induced cell contraction. Low-dose S1P (0.1 μM) induced contraction of Gβ1−/− neural progenitor cells but not of WT cells (compare B and F). Y27632 (Rho inhibitor) prevented cell contraction induced by 1 μM S1P (D, H). I–L: Effects of PTX (Gαi inhibitor), NSC23766 (Rac inhibitor), and Gallein (Gβγ inhibitor) on morphological changes of WT neural progenitor cells. In sum, 20 ng/ml PTX, 50 μM NSC23766, and 10 μM Gallein were added 16, 1, and 2 hr before stimulation, respectively. Low-dose S1P induced cell contraction in the presence of PTX, NSC23766, or Gallein. M–P: Phalloidin staining of neural progenitor cells stimulated with 0.1 μM S1P. Similar results were obtained from at least three independent experiments. Bars = 100 μm in D, H, L; 50 μm in P. The scale bars in D, H, L, and P also apply to A-C, E-G, I-K, and M-O, respectively.

Reduced Cortical Thickness in Gβ1−/− Mice

Gβ1−/− mice without NTDs died perinatally (Table 1). Most of Gβ1−/− mice died a few hours after birth with respiratory defects (Supp. Fig. S1). Some Gβ1−/− mice survived beyond this stage, but they did not ingest milk and died within 2 days after birth (Supp. Fig. S1). Since Gβ1 was strongly expressed in the fetal brain, we analyzed brain morphology of Gβ1−/− mice at P0. While gross morphology of the brain and anterior-posterior length of cerebral cortex appeared normal in Gβ1−/− mice, cortical thickness was reduced upon close examination (Fig. 7A–C). Gβ1−/− embryos with NTDs also had smaller brain volumes at E18.5, associated with severe brain malformations (Supp. Fig. S2). To examine whether reduced cortical thickness of Gβ1−/− brains was caused by a defect in neural progenitor cell proliferation, mitotic cells in the VZ were stained with anti-phospho-histone H3 (PH3) antibody and quantified (Fig. 7D–F). The PH3-positive cells were reduced in the cerebral cortex of Gβ1−/− embryos at E16.5 but not at E10.5, suggesting that reduced proliferation of neural progenitor cells in late gestation is the cause of thinning of cerebral cortex. The organization of cerebral cortex layers was normal in Gβ1−/− mice (Fig. 7G,H). Oct6 and Tbr1, which are highly expressed in the layers II–IV and VI, respectively, were normally expressed, suggesting that the inside-out formation of the cortical plate was not altered in Gβ1−/− mice (Fig. 7I–L).

Figure 7.

Reduced cortical thickness and mitotic cells in Gβ1−/− mouse brain. A,B: H&E staining of saggital sections of the telencephalon at P0. C: Anterior-posterior length (L) and thickness (T) of the cerebral cortex were measured and normalized to littermate controls. D,E: E16.5 sagittal sections stained with an antibody against the mitotic marker phospho-histone H3 (PH3). F: The number of PH3-positive cells in the ventriclar zone (VZ) was counted in a unit section (250 μm width for E10.5 embryos; 400 μm width for E16.5 embryos) and normalized to littermate controls. G,H: Immunostaining with anti-MAP2 (green) and -Tuj1 (red) antibodies of coronal sections of the telencephalon at E18.5. Sections were counterstained with Hoechst (blue). MZ, marginal zone; CP, cortical plate; SP, subplate; IZ, intermediate zone; VZ, ventricular zone. I,J: In situ hybridization with Oct6 (layer II–IV) antisense probe of coronal sections of the telencephalon at P0. K,L: In situ hybridization with Tbr1 (layer VI) antisense probe of coronal sections of the telencephalon at P0. Bars = 1 mm in B; 100 μm in E, H; 500 μm in J, L. Data are means ± SD (n = 5 in C; n = 4 in F). *P <0.005; Student's t-test.

Reduced ERK Phosphorylation and Cell Proliferation Induced by GPCR Agonists in Gβ1−/− Neural Progenitor Cells

Some GPCR agonists including S1P, LPA, and endothelin-1 (ET1) enhance proliferation of neural progenitor cells through ERK phosphorylation (Harada et al., 2004; Morishita et al., 2007; Shinohara et al., 2004). It is also known that ET1 induces JNK phosphorylation in a Gαq-dependent manner (Mizuno et al., 2005). ERK phosphorylation induced by S1P and ET1 was reduced in Gβ1−/− neural progenitor cells (Fig. 8A,B). However, JNK phosphorylation induced by LPA and ET1 was not affected in Gβ1−/− neural progenitor cells (Fig. 8A,C). LPA and S1P had little effect on ERK and JNK phosphorylation, respectively (Fig. 8A–C). Using inhibitors, it was shown that ERK phosphorylation induced by S1P was inhibited by the Gαi inhibitor PTX and the Gβγ inhibitor Gallein (Fig. 8D,E). ET1-induced ERK phosphorylation was also partially dependent on Gαi and Gβγ signaling (Fig. 8D,E).

Figure 8.

Defective ERK phosphorylation induced by GPCR agonists in Gβ1−/− neural progenitor cells. A–C: LPA (1 μM)-, S1P (1 μM)-, and ET1 (10 nM)-induced ERK and JNK phosphorylation in neural progenitor cells. Neural progenitor cells were stimulated for 15 min. The band intensity of phosphorylated ERK and JNK was quantified. D,E: Effects of PTX (Gαi inhibitor) and Gallein (Gβγ inhibitor) on ERK phosphorylation. Neural progenitor cells were stimulated with LPA, S1P, or ET1 for 15 min. Then, 20 ng/ml PTX and 10 μM Gallein were added 16 and 2 hr before stimulation, respectively. The band intensity of phosphorylated ERK was quantified. Data are means ± SD (n = 3 in B, E). **P < 0.005; *P < 0.05; Student's t-test.

Next, we examined GPCR agonist-induced cell proliferation. Neural progenitor cells were stimulated with EGF, S1P, LPA, or ET1 for 2 days and cell proliferation was measured as described in the Experimental Procedures section. While EGF-induced cell proliferation was not affected, ET1-induced proliferation was reduced in Gβ1−/− neural progenitor cells (Fig. 9A). LPA and S1P had less effect on the proliferation of neural progenitor cells. ET1-induced cell proliferation was partially dependent on Gαi and βγ signaling just as ET1-induced ERK phosphorylation was (Fig. 9B). The Gβγ inhibitor Gallein was not toxic to neural progenitor cells (Supp. Fig. S4).

Figure 9.

Reduced cell proliferation induced by GPCR agonists in Gβ1−/− neural progenitor cells. A: GPCR agonist-induced neural progenitor cell proliferation. Neural progenitor cells were stimulated with 1 μM S1P, 1 μM LPA, 10 nM ET1, or 20 ng/ml EGF for two days. B: Effects of PTX and Gallein on ET1-induced cell proliferation. WT neural progenitor cells were stimulated with 10 nM ET1. Then, 20 ng/ml PTX and 10 μM Gallein were added 16 and 2 hr before stimulation, respectively. Cell proliferation was measured as described in the Experimental Procedures section and is expressed as fold increase compared with untreated WT cells. Data are means ± SD. *P < 0.05; Student's t-test.

Normal Mitotic Orientation of Cortical Progenitors in Gβ1−/− Mice

It was reported that G protein βγ subunits and AGS3, which is a nonreceptor activator of Gβγ, regulate mitotic orientation of cortical progenitors and abnormal mitotic orientation results in impaired proliferation of neural progenitor cells (Sanada and Tsai, 2005). Since reduced proliferation of neural progenitor cells was observed in Gβ1−/− embryos at E16.5, we examined the mitotic orientation of Gβ1−/− cortical progenitors. Cortical VZ of embryos (without NTDs) was stained with hematoxylin and the mitotic orientation of cortical progenitors was visualized. At all embryonic stages examined, the mitotic orientation of Gβ1−/− cortical progenitors was unaffected compared with control (Supp. Fig. S5).

DISCUSSION

In this study, we demonstrated that Gβ1−/− mice were neonatal lethal and showed defects in neural tube closure and neural progenitor cell proliferation. ERK phosphorylation, cell proliferation, and cell spreading induced by GPCR agonists were abnormal in Gβ1−/− neural progenitor cells, and these were found to be dependent on Gαi and Gβγ signaling. On the other hand, Gαq-dependent JNK phosphorylation was not affected and Gα12/13-dependent cell contraction was induced in Gβ1−/− neural progenitor cells. These data suggest that Gβ1 is required for Gβγ-dependent neuronal development.

Impaired Neural Tube Development and Neural Progenitor Cell Proliferation in Gβ1−/− Embryos

Gβ1 was highly expressed in the cranial neural tube and about 40% Gβ1−/− embryos developed NTDs. Neural cell proliferation, death, and cell cycle–dependent nuclear positioning in the neural plate were normal in Gβ1−/− embryos at E8.75. We found, however, that dorsolateral bending of the cranial neural plate was abnormal in Gβ1−/− embryos and neuroepithelial cells were elongated along the apical-basal axis in Gβ1−/− neural plate at the stage of neural tube closure. In addition, F-actin staining was reduced in the basal side of Gβ1−/− neuroepithelium around the DLHPs. The apical actin microfilaments are thought to regulate neural tube closure through the apical constriction of neuroepithelial cells (Copp et al., 2003), but F-actin staining appeared normal in the apical side of Gβ1−/− neuroepithelium. While less attention has been paid to actin accumulation in the basal side of the neuroepithelium, basal actin accumulation does occur and is proposed to be important for neural tube closure (Moephuli et al., 1997; Sadler et al., 1982; Zolessi and Arruti, 2001).

S1P and LPA are involved in the regulation of neural tube closure and are also known to induce contraction of neural progenitor cells through actin remodeling (Contos et al., 2000; Fukushima et al., 2000; Harada et al., 2004; Mizugishi et al., 2005; van Meeteren et al., 2006). Both loss of S1P and inhibition of actin polymerization are known to cause cranial NTDs but not spinal bifida (Mizugishi et al., 2005; Copp et al., 2003), just as loss of Gβ1 did. We showed that abnormal cell contraction and actin reorganization were induced by low concentrations of S1P in Gβ1−/− neural progenitor cells. Consistent with these in vitro data, the morphology of neuroepithelial cells was abnormal in Gβ1−/− neural plate and F-actin staining was reduced in Gβ1−/− neuroepithelium, suggesting that defects in S1P responses may be responsible for NTDs, although involvement of other GPCR agonists is not excluded. It is known that S1P and LPA induce cell contraction through Gα12/13-dependent Rho activation and inhibit it through Gαi-dependent Rac activation in several cell types (Hurst et al., 2008; van Leeuwen et al., 2003; Watterson et al., 2007). It is also reported that Gα12/13-dependent contraction of neural progenitor cells is enhanced by PTX (Yanagida et al., 2007). At low concentrations, S1P induced contraction of Gβ1−/− neural progenitor cells but not of WT cells. Importantly, low concentrations of S1P could also induce contraction of WT neural progenitor cells in the presence of PTX (Gαi inhibitor), NSC23766 (Rac inhibitor), or Gallein (Gβγ inhibitor). As Gαi-coupled receptors are thought to activate Rac through Gβγ subunits (Niu et al., 2003; Vogt et al., 2007), it is expected that Gβγ-dependent Rac activation is suppressed in Gβ1−/− cells.

As described above, low concentrations of S1P induced contraction of Gβ1−/− neural progenitor cells but not of WT cells and abnormal morphology of neuroepithelial cells was observed in Gβ1−/− neural plate. Furthermore, F-actin staining was reduced in the basal side of Gβ1−/− neuroepithelium but not in the apical side. The apical actin microfilaments are thought to regulate neural tube closure through the apical constriction of neuroepithelial cells (Copp et al., 2003). From these data, we propose that excess cell contraction occurs in the basal side of Gβ1−/− neural plate and inhibits normal bending of the neural plate by antagonizing with the apical constriction of neuroepithelial cells.

While the organization of cerebral cortex layers was normal, reduced neural progenitor cell proliferation was observed in Gβ1−/− embryos without NTDs at late gestation. Some GPCR agonists including ET1, S1P, and cannabinoids stimulate neural progenitor cell proliferation through ERK phosphorylation (Morishita et al., 2007; Palazuelos et al., 2006; Shinohara et al., 2004). It was reported that S1P induces ERK phosphorylation in a Gαi-dependent manner and ET1 induces it in a Gαq- and Gαi-dependent manner (Harada et al., 2004; Morishita et al., 2007). As many of PTX-sensitive effector activations are mediated by Gβγ subunits rather than Gαi (Schwindinger and Robishaw, 2001; Smrcka, 2008), it is possible that GPCR agonists induce neural progenitor cell proliferation through Gβγ subunits. We found that Gβ1 was most highly expressed among G protein β subunits in neural progenitor cells, and S1P- and ET1-induced ERK phosphorylation and cell proliferation were reduced in Gβ1−/− neural progenitor cells. ERK phosphorylation and cell proliferation were also reduced in the presence of PTX or Gallein in WT neural progenitor cells. These data suggest that Gβ1 is required for Gβγ-dependent ERK phosphorylation and neural progenitor cell proliferation.

ET1 is also known to induce JNK phosphorylation in a Gαq-dependent manner (Mizuno et al., 2005), and JNK phosphorylation was not affected in Gβ1−/− neural progenitor cells. ET1-induced ERK phosphorylation, which is mediated by Gαq and Gαi subunits (Morishita et al., 2007), was reduced but not completely eliminated in Gβ1−/− neural progenitor cells and in WT cells treated with PTX or Gallein. In addition, contraction of neural progenitor cells, which is thought to be dependent on Gα12/13 subunits (Yanagida et al., 2007), was induced by LPA and S1P in Gβ1−/− neural progenitor cells. From these data, it is expected that Gαq- and Gα12/13-dependent downstream effector activation is not significantly affected in Gβ1−/− neural progenitor cells.

Reduced cortical thickness could be caused by abnormal cleavage plane orientation of cortical progenitors. Sequestration of Gβγ subunits randomizes cleavage plane orientation of cortical progenitors and results in impaired proliferation of neural progenitor cells (Sanada and Tsai, 2005). Gβ1 was most strongly expressed in neural progenitor cells among five G protein β subunits, suggesting the involvement of Gβ1 in the regulation of cleavage plane orientation. However, contrary to the expectation, the mitotic orientation of Gβ1−/− cortical progenitors was unaffected. It is possible that other Gβ subunits such as Gβ2, which is also expressed in neural progenitor cells, would substitute for the loss of Gβ1. In contrast to mitotic orientation, ERK activation was abnormal in Gβ1−/− neural progenitor cells. It may mean that the low level of Gβ protein can support normal mitotic orientation of progenitors but not ERK activation. Alternatively, it may reflect functional differences among Gβ subunits rather than the amount of Gβ protein.

Additional Defects

Gβ1−/− mice without NTDs showed abnormal suckling behavior and died in two days after birth. Gβ1−/− mice also frequently had respiratory defects. Reduced cortical thickness and impaired cell proliferation associated with reduced ERK phosphorylation were observed in Gβ1−/− neural progenitor cells or mice. However, these defects may not be the cause of the perinatal lethality, because reduced cortical thickness caused by a loss of ERK2 in the neural progenitor cells does not result in early death of the pups (Samuels et al., 2008). It is known that one of the major causes of abnormal suckling behavior and respiratory defects is craniofacial skeletal defects, the most common being cleft palate (Turgeon and Meloche, 2009). However, cleft palate was not observed in Gβ1−/− mice (Supp. Fig. S1). Furthermore, while high expression of Gβ1 was observed in the fetal lung, gross morphology of the lung was normal in Gβ1−/− embryos at E17.5 (Supp. Fig. S1).

We did not identify the mechanisms underlying perinatal lethality of Gβ1−/− mice, but abnormal neurotransmitter release could cause the inability to feed and respiratory defects (Turgeon and Meloche, 2009). We revealed that Gβγ signaling was impaired in Gβ1−/− neural progenitor cells and Gβ1 was highly expressed in the peripheral nervous system. Gβγ subunits are known to regulate neurotransmitter release through direct interactions with voltage-dependent Ca2+ channels and soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complexes (Blackmer et al., 2005; De Waard et al., 1997; Gerachshenko et al., 2005). Many mutant mice, in which neurotransmission is impaired, have respiratory or behavioral defects. Thus, loss of Gβ1 could impair normal neurotransmitter release and result in neonatal death. Alternatively, other unidentified defects may be responsible.

It was reported that homozygous inheritance of Rd4 chromosome, in which Gβ1 is disrupted by a large inversion spanning nearly all of mouse chromosome 4, is lethal, although detailed analysis had not been carried out (Kitamura et al., 2006; Roderick et al., 1997). This observation is consistent with our results. They also reported that Rd4/+ mice develop retinal degeneration and proposed that Gβ1 is responsible for the Rd4 retinal disease. Though the level of Gβ1 protein in Gβ1+/− retina was about half of that in Gβ1+/+ retina, Gβ1+/− mice did not show retinal degeneration (Supp. Fig. S6). Therefore, genes other than Gβ1 may also be damaged by the inversion in Rd4/+ mice, or the genetic backgrounds of the two mutant lines may influence the phenotypes.

EXPERIMENTAL PROCEDURES

Generation of Gβ1 Gene Trap Mutant Mice

ES (E14.1) cells were infected with the prvSStrap retroviral vector (Morita et al., 2000; Shirasawa et al., 2005), and trap clones were selected with hygromycin. Genomic DNA sequences flanking gene-trap insertion sites were determined by the anchoring PCR as described (Chen and Soriano, 2003). Chimeric mice were generated from a Gβ1 gene trap clone using a modified aggregation method as described previously (Asano et al., 1997; Nagy et al., 1993). The chimeric mice were mated with C57BL/6J mice to generate heterozygous mutant mice, and then the progenitors on a mixed background of 129/SvJ and C57BL/6J were intercrossed. Midday of the day the vaginal plug was found was termed embryonic day 0.5 (E0.5) and the day of birth was termed postnatal day 0 (P0). All mice were kept under specific pathogen-free conditions in an environmentally controlled clean room at the Center for Experimental Medicine and Systems Biology (The Institute of Medical Science, University of Tokyo). The experiments were conducted according to the institutional ethical guidelines for animal experiments and the safety guidelines for gene manipulation experiments.

Genotyping of Gβ1-Deficient Mice

Genomic DNA was prepared from the tail and used as the template. The following primers were used: (5′ CACTGGCTCTCTTTCTACCCGAAT 3′), (5′ ACTCAGGAGGCAGAAG CAGA 3′), and (5′ ACCTGAAAT GACCCTGTGCCTTAT 3′). After denaturing at 95°C for 5 min, the PCR reactions were cycled 40 times at 95°C for 30 sec, 68°C for 30 sec, and 72°C for 30 sec.

Southern Blotting

Genomic DNA was digested with EcoRI, electrophoresed, and transferred to a nylon membrane. Hybridization was carried out according to standard methods (Sambrook et al., 1989) using 32P-labeled DNA probes.

Real-Time RT-PCR

Total RNA was prepared with Sepasol RNA I Super reagent (Nacalai Tesque, Kyoto, Japan). First-strand cDNA was synthesized from total RNA with SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA). Real-time PCR reaction was done with the SYBR Green qPCR kit (Invitrogen) using the iCycler system (Bio-Rad, Gaithersburg, MD). The amount of target mRNA was determined from the appropriate standard curve and normalized to the amount of Gapdh mRNA. The primer sets are shown in Supp. Table S1.

Histology and Immunostaining

Embryos were fixed overnight in phosphate buffered saline (PBS) containing 4% paraformaldehyde at 4°C followed by cryoprotection in 30% sucrose or embedding in paraffin. Cryostat sections were sectioned at 12–16 μm, and paraffin blocks were sectioned at 5-μm thickness. Sections were stained with hematoxylin-eosin (H&E) or processed for immunostaining as follows. Sections were blocked in 2% skim milk in TBST for 1 hr, followed by incubation with primary antibodies at 4°C overnight. After the incubation, sections were washed with TBST and then incubated for 1 hr with secondary antibodies. The following antibodies were used as the primary antibodies: anti-Tuj1 (1:2,000, ab18207, Abcam, Cambridge, MA), anti-phospho-Histone H3 (1:300, ab5176, Abcam), anti-MAP2 (1:100, AP-20, Neo Markers), anti-Gβ1 (1:100, 10247-2-AP, Proteintech, Chicago, IL), anti-neurofilament (1:50, 2H3, Developmental Studies Hybridoma Bank), anti-Nestin (1:50, Rat-401, Developmental Studies Hybridoma Bank), and anti-β-catenin (1:300, C19220, BD Biosciences, Franklin Lakes, NJ). The following antibodies were used as the secondary antibodies: Cy3-conjugated goat anti-rabbit IgG (Jackson Immuno Research, West Grove, PA), Alexa488-conjugated goat anti-rabbit IgG (Molecular Probes, Eugene, OR), Alexa488-conjugated goat anti-mouse IgG (Molecular Probes), biotin-conjugated goat anti-rabbit IgG (Abcam), and HRP-conjugated goat anti-mouse IgG, IgA, and IgM (MP Biomedicals, Solon, OH). The HRP-conjugated antibody was detected with diaminobenzidine (DAB). The biotin-conjugated antibody was detected with streptavidin-HRP and DAB. F-actin filaments were visualized using rhodamine phalloidin (Cytoskeleton, Denver, CO) according to the manufacturer's instructions. Nuclei were stained with Hoechst 33342 (Sigma, St. Louis, MO).

In Situ Hybridization

Digoxigenin-labeled probes were synthesized using the Diagnostics RNA labeling kit (Roche, Nutley, NJ). Template DNA was amplified using the following primers: Shh, (5′ CGCCAA GAAGGTCTTCTACGTGAT 3′) and (5′ AGGTTGCTTGGCCTCACAAAA TAA 3′); Oct6, (5′ AGCCACTTTCT CAAGTGTCCCAAG 3′) and (5′ GAA GAGGCGCGGAAAGAATAAAGT 3′); Tbr1, (5′ TCATCCCATTATCTCGAC CACTGA 3′) and (5′ TGTTGTTTGAT GCTCCCTTGTTGT 3′). Embryos were fixed overnight in PBS containing 4% paraformaldehyde at 4°C and embedded in paraffin. Paraffin blocks were sectioned at 10-μm thickness and in situ hybridization was performed as described (Wilkinson, 1992).

BrdU Labeling

Pregnant mice were injected with BrdU at 50 μg/g of body weight. The animals were sacrificed 1 hr after BrdU injection and prepared for paraffin sections. Sections were pretreated for 20 min in 2 N HCl at 37°C. BrdU-positive cells were detected by immunostaining with mouse monoclonal antibody to BrdU (Roche).

X-gal Staining

Embryos were fixed with 0.2% glutaraldehyde for 30 min and stained overnight with PBS containing 5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, 2 mM MgCl2, 0.02% NP-40, and 1 mg/ml X-gal. Embryos were refixed with PBS containing 4% paraformaldehyde, embedded in paraffin, and sectioned at 10-μm thickness.

TUNEL Assay

TUNEL analysis was performed with in situ Cell Death Detection Kit (Roche) according to the manufacturer's instructions. TUNEL-positive cells were visualized with TUNEL POD (Roche) and DAB.

Neural Progenitor Cell Culture

Neural progenitor cells were cultured as described previously (Harada et al., 2004) with modifications. Telencephalons were freed from E15.5 to E18.5 embryos and trypsinized into single-cell suspension. Single cells were cultured in Neurobasal medium containing 2% B27 supplement (Life Technologies, Norwalk, CT), 0.2 mM L-glutamine and 20 ng/ml EGF (PeproTech, Inc., Rocky Hill, NJ) to generate neurosphreres. Neurospheres were passaged every 5–7 days and cells were used for experiments after two to five passages.

Western Blotting

Neurospheres were trypsinized into single-cell suspension and cells were plated on poly-D-lysine and laminin coated 48 well plates at 2 × 104 cells/well. Neural progenitor cells were cultured with EGF-free medium for 3 hr and stimulated with 1 μM S1P (Cayman Chemical Co., Ann Arbor, MI), 1 μM LPA (Biomol, Plymouth Meeting, PA), or 10 nM ET1 (Peptide, Osaka, Japan) for 15 min. Then, 20 ng/ml PTX (Kaketsuken, Kumamoto, Japan) and 10 μM Gallein (Merck, Rahway, NJ) were added 16 and 2 hr before stimulation, respectively. Immunoblot analysis was performed using antibodies against ERK, phospho-ERK, phospho-JNK (Cell Signaling, Danvers, MA), Gβ1 (Santa Cruz Biotechnology, Santa Cruz, CA), and β-tubulin (Abcam). HRP-conjugated goat anti-rabbit IgG (Cell Signaling) was used as the secondary antibody. Blots were developed with ECL Plus Western Blotting Detection System (Amersham Biosciences, Pittsburgh, PA).

Proliferation Assay

Neural progenitor cells were plated on poly-D-lysine- and laminin-coated 96 well plates at 1 × 104 cells/well. Neural progenitor cells were cultured with EGF-free medium for 3 hr and stimulated with 20 ng/ml EGF, 1 μM S1P, 1 μM LPA, or 10 nM ET1 for 2 days. 20 ng/ml PTX and 10 μM Gallein were added 16 and 2 hr before stimulation, respectively. Cell proliferation was measured by absorbance at 450 nm using Cell Count Reagent SF (Nacalai Tesque, Kyoto, Japan).

Morphological Analysis

Neural progenitor cells were plated on poly-D-lysine coated 48-well plates at 2 × 104 cells/well and stimulated with S1P for 1 hr. Subsequently, 20 ng/ml PTX and 10 μM Gallein were added 16 and 2 hr before stimulation, respectively. Then, 10 nM Y27632 (Nacalai Tesque) and 50 μM NSC23766 (Calbiochem, San Diego, CA) were added 1 hr before stimulation. Cell morphology was observed under phase-contrast microscopy.

Acknowledgements

We thank Dr. T. Shinomura for providing the prvSStrap gene trap vector and Dr. T. Kitamura for Plat-E cells. The monoclonal antibodies 2H3 developed by T. M. Jessell and J. Dodd, and Rat-401 developed by S. Hockfield, were obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by the University of Iowa, Department of Biological Sciences, Iowa City, IA 52242. We thank all the members of the laboratory for excellent animal care. This work was supported in part by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Ancillary