Sox7 plays crucial roles in parietal endoderm differentiation in F9 embryonal carcinoma cells through regulating Gata-4 and Gata-6 expression

Abstract

During early rodent development, the parietal endoderm appears from an inner cell mass and produces large amounts of basement membrane components, such as laminin-1 and collagen IV. To elucidate the regulatory network for gene expression during these procedures, we constructed a series of short interfering RNA expression vectors targeted to various transcription factors, transfected them into F9 embryonal carcinoma cells, and evaluated the effects of the gene silencing on the induction of parietal endoderm differentiation and basement membrane component production by treating F9 cells with all trans-retinoic acid and dibutyryl cyclic AMP. Among the transcription factors tested, silencing of Sox7 or combined silencing of Gata-4 and Gata-6 resulted in suppression of cell shape changes and laminin-1 production, which are the hallmarks of parietal endoderm differentiation. In cells silenced for Sox7, induction of Gata-4 and Gata-6 by retinoic acid and cyclic AMP treatment was inhibited, while induction of Sox7 was not affected in cells silenced for Gata-4 and Gata-6, indicating that Sox7 is an upstream regulatory factor for these Gata factors. Nevertheless, silencing of Sox7 did not totally cancel the action of retinoic acid, since upregulation of coup-tf2, keratin 19, and retinoic acid receptor beta2 was not abolished in Sox7-silenced F9 cells. Although overexpression of Sox7 alone was insufficient to induce parietal endoderm differentiation, overexpression of Gata-4 or Gata-6 in Sox7-silenced F9 cells restored the differentiation into parietal endoderm. Sox7 is therefore required for the induction of Gata-4 and Gata-6, and the interplay among these transcription factors plays a crucial role in parietal endoderm differentiation.

FIG. 1.

Structure of the pH1RNAi vector and efficiency of gene silencing. (A) Structure of the pH1RNAi vector. The 19-mer siRNA target sequence is shown as N19. The sequence of the linker (ttcaagaga) and five thymidines (t) signaling termination are also shown. The transcribed RNA molecules self anneal to form hairpin-shaped short interfering RNA. (B) Silencing effect of the pH1Lama1 vector on laminin α1 expression at the mRNA level. The expression level of laminin α1 mRNA estimated by qPCR in cells transfected with the pH1Lama1 vector (pH1Lama1) is shown as a percentage relative to that in control cells transfected with the pH1S vector (control). The means ± SD of results from triplicate transfections are shown. (C) Silencing effect of the pH1Lama1 vector on laminin α1 expression at the protein level. Western blotting for laminin-1 proteins in conditioned medium from control cells or cells transfected with pH1Lama1 is shown. The solid arrowhead indicates laminin-1 (heterotrimer of α1, β1, and γ1 subunits), and the open arrowhead indicates either the α1 monomer or a heterodimer of β1 and γ1.

FIG. 2.

Laminin α1 expression in cells silenced for various transcription factors. The siRNA target genes are indicated at the bottom. In Gata4/6 and Cited1/2, both Gata-4 and Gata-6 and both Cited-1 and Cited-2 were silenced, respectively. The laminin α1 mRNA levels estimated by qPCR are presented relative to that in control cells transfected with pH1S, a control vector without a hairpin-forming insert, and treated with RA/Bt₂cAMP for 96 h after transfection. The means ± SD of results from triplicate experiments are shown.

FIG. 3.

Effects of Gata-4, Gata-6, or Sox7 silencing on the morphological changes and gene expression induced by RA/Bt₂cAMP treatment in F9 cells. (A) Morphology of cells treated with RA/Bt₂cAMP for 96 h after transfection. Note that the cells transfected with pH1S (control) are round with a scattered distribution, whereas those silenced for both Gata-4 and Gata-6 (Gata4/6) or Sox7 show a flat epithelium-like shape. (B) Expression levels of the mRNAs for basement membrane components (Lamb1, laminin β1; Lamc1, laminin γ1; Col4a1, type IV collagen α1) in F9 cells silenced for Gata-4, Gata-6, Gata-4/6, or Sox7 after RA/Bt2cAMP treatment. (C) Secretion of laminin-1 into conditioned media. The solid arrowhead indicates laminin-1 (heterotrimer of α1, β1, and γ1 subunits), and the open arrowhead indicates either the α1 monomer or a heterodimer of β1 and γ1. Conditioned media from F9 cells silenced for the indicated genes and treated with RA/Bt₂cAMP for 96 h were subjected to Western blotting under nonreducing conditions. Control cells were transfected with the pH1S vector and treated with RA/Bt₂cAMP. Conditioned media from undifferentiated F9 cells and F9 cells treated with RA/Bt₂cAMP for 96 h without transfection of exogenous plasmids (F9-S and F9-PE, respectively) were also subjected to the analysis for reference. (D) Expression levels of the mRNAs for the transcription factors (Sox17, Hnf1b, and Foxa2/Hnf3b) upregulated during parietal endoderm differentiation. (E to G) Expression levels of the mRNAs for Coup-tf2 (E), keratin 19 (F), and RARb2 (G). Cells were treated with RA/Bt₂cAMP for 96 h after transfection with pH1RNAi vectors targeting the indicated genes. Control indicates F9 cells transfected with the pH1S vector and treated with RA/Bt₂cAMP. The expression levels relative to those in the control cells are shown as the means ± SD of results from triplicate experiments.

FIG. 4.

Regulatory network among Gata-4, Gata-6, and Sox7. (A) Expression of Gata-4, Gata-6, and Sox7 mRNAs. F9 cells were transfected with pH1RNAi vectors targeting the genes indicated at the bottom and treated with RA/Bt₂cAMP for 96 h. Control indicates F9 cells transfected with the pH1S vector and treated with RA/Bt₂cAMP. The expression levels relative to those in the control cells are shown as the means ± SD of triplicate experiments. (B) Expression of Gata-4 and Gata-6 proteins. F9 cells were transfected with pH1RNAi vectors targeting the genes indicated at the top and treated with RA/Bt₂cAMP for 96 h. Nuclear extracts were subjected to Western blotting using polyclonal antibodies against Gata-4 and Gata-6. Nuclear extracts from undifferentiated and RA/Bt₂cAMP-treated F9 cells without transfection (F9-S and F9-PE, respectively) were also analyzed. The lower molecular mass band indicated by the open arrowhead in the Gata-6 blot is considered to be a nonspecific signal, and the solid arrowhead indicates the Gata-6 protein.

FIG. 5.

Effects of Gata-4, Gata-6, or Sox7 overexpression on undifferentiated F9 cells. F9 cells were transfected with the expression vectors indicated at the bottom of each panel and cultured for 96 h without RA/Bt₂cAMP treatment. The expression levels of laminin α1 (Lama1) (A), Hnf1b (B), endogenous Gata-4 (C), endogenous Gata-6 (D), and endogenous Sox7 (E) were estimated by qPCR. The primers used were designed to selectively amplify the endogenous, but not the transfected, gene products. The mRNA expression levels in undifferentiated and RA/Bt₂cAMP-treated F9 cells without transfection (F9-S and F9-PE, respectively) are also shown for reference. The expression levels relative to those in F9-PE cells are shown as the means ± SD of results from triplicate experiments.

FIG. 6.

Rescue of F9 cells stably silenced for Sox7 by exogenous expression of Sox7 or the Gata factors. Three independent clones of the F9 cells stably transfected with the pH1RNAi vector targeting Sox7 (F9^H1Sox7; solid bars) and wild-type F9 (wt; open bars) were analyzed for their mRNA expression levels by qPCR. Cells were transfected with expression vectors and cultured for 96 h with or without RA/Bt₂cAMP treatment (+ or −, respectively). The panels show the endogenous gene expression levels of Sox7 (A), laminin α1 (B), Gata-4 (C), Gata-6 (D), Hnf1b (E), Col4a1 (F), and Sox17 (G). The overexpressed genes are indicated at the bottom of each panel. The expression levels are presented relative to those in the nontransfected wild-type F9 cells (indicated as “no”) treated with RA/Bt₂cAMP. The means ± SD of results from triplicate experiments for three clones of F9^H1Sox7 cells and wild-type F9 cells are shown. Note that some bars (indicated by asterisks) are hardly visible due to the extremely low levels of expression (less than 2%).

FIG. 7.

Time course analyses of gene expression patterns during F9 differentiation. Cells were harvested every 12 h until 48 h and then every 24 h until 96 h after RA/Bt₂cAMP treatment, and the mRNA expression levels were estimated by qPCR. The analyzed genes are indicated at the top of each panel. The mRNA expression levels are expressed as the percentages relative to those at 96 h, and the means ± SD of results from triplicate experiments are shown. Lama1, laminin α1.

FIG. 8.

Schematic model of the regulatory network operating during the parietal endoderm differentiation of F9 cells. Arrows indicate positive regulation of the gene expression. Sox7 and an unidentified factor, factor X, function synergistically to induce Gata-4 and Gata-6 expression. Gata-4 and Gata-6 constitute a functional unit with mutual positive regulation and functional redundancy. The curved line with the flattened arrowhead indicates suppressive regulation.