Region-specific regulation of cell proliferation by FGF receptor signaling during the Wolffian duct development.

The Wolffian duct (WD) is a primordium of the male reproductive tract and kidney collecting duct system. Fibroblast growth factor receptors (FGFRs), members of the receptor tyrosine kinase (RTK) family, are essential for kidney development. Although the functions of FGFR signaling in kidney morphogenesis have been analyzed, their function in WD development has not been comprehensively investigated. Here, we demonstrate that Fgfr2 is the major Fgfr gene expressed throughout the WD epithelia and that it is essential for the maintenance of the WD, specifically in the caudal part of the WD. Hoxb7-Cre mediated inactivation of Fgfr2 in the mouse WD epithelia resulted in the regression of the caudal part of the WD and abnormal male reproductive tract development. Cell proliferation and expression of the downstream target genes of RTK signaling (Etv4 and Etv5) were decreased in the caudal part of the WD epithelia in the mutant embryos. Cranial (rostral) WD formation and ureteric budding were not affected. Ret, Etv4, and Etv5 expression were sustained in the ureteric bud of the mutant embryos. Taken together, these data suggest region-specific requirements for FGFR2 signaling in the developing caudal WD epithelia.

Introduction

The Wolffian duct (WD) is a primordium of the male reproductive tract, which develops into the epididymis, vas deferens, and seminal vesicles. In addition, the WD plays a crucial role in kidney development as the primordium for the ureter and kidney collecting duct system (Bouchard et al., 2002; Saxen, 1987). In mouse embryos, the WD is formed in the anterior intermediate mesoderm at embryonic (E) day 8.5 (E8.5), elongating caudally to reach the cloaca by E9.5 (Atsuta et al., 2013; Davidson, 2009). WD insertion into the cloaca has been reported to be essential for ureter positioning (Chia et al., 2011; Weiss et al., 2014). Three successive kidneys develop along the WD in an anterior–posterior manner: the pronephros, mesonephros, and metanephros (Capel, 2000). The pronephros is formed at the most anterior level of the WD at E8.5, and degenerates by E9.5 (Dressler, 2006). The mesonephros is subsequently formed posterior to the pronephros. The cranial part of the mesonephric mesenchyme differentiates into the mesonephric tubules (MTs) (Dressler, 2006). Degeneration of the mesonephric mesenchyme begins from caudal at E10.5 and extends cranially (Hoshi et al., 2012). At the same time, in the caudal end of the mesonephros, the WD develops a multilayered, “pseudostratified epithelia”, forming the ureteric bud (UB) (Chi et al., 2009). The metanephros, which is the definitive kidney in mammals, begins to form from the UB and adjacent metanephric mesenchyme (MM) (Yu et al., 2004). Malformation of the WD in the caudal mesonephros, in addition to the UB, often results in the congenital anomalies in human, such as duplex ureter and absence of the vas deferens. The characterization of the nephrogenic regions, especially the caudal mesonephros, has just started to be understood for their susceptibility to such developmental abnormality (Hoshi et al., 2012; Murashima et al., 2014; Nishita et al., 2014).

The reciprocal interaction between the WD epithelia and adjacent nephrogenic mesenchyme is essential for the morphogenesis of all kidney structures (Batourina et al., 2001). In such interactions, growth factor signaling, including Glial cell line-derived neurotrophic factor (Gdnf), Wnt, and Fibroblast growth factor (Fgf), are well characterized signaling pathways. Dysregulation of these signaling pathways often causes ectopic UB formation, mostly in the caudal part of the mesonephros (Basson et al., 2005; Brophy et al., 2001; Grieshammer et al., 2004; Kume et al., 2000; Mendelsohn, 2009; Michos et al., 2004; Nishita et al., 2014; Shakya et al., 2005).

Receptor tyrosine kinase (RTK) signaling is a key requirement for the development of many organs. It regulates cellular processes including proliferation, survival, differentiation, and migration (Schlessinger, 2000). Binding of ligands to their RTKs activates several intracellular pathways such as phosphoinositide 3-kinase (PI3K)/AKT, mitogen-activated protein kinase (MAPK), and phospholipase C-gamma (PLC-gamma) (Song et al., 2011). Several RTKs such as Rearranged during transfection (RET) and Fgf receptor (FGFR) play an essential role in vertebrae embryonic development, especially in mammalian kidney development (Bates, 2011; Benazeraf and Pourquie, 2013; Costantini and Shakya, 2006; Moore et al., 1996; Pichel et al., 1996; Sainio et al., 1997; Sanchez et al., 1996; Schuchardt et al., 1994).

There are four FGFRs that interact with 18 known FGF ligands in mammals (Beenken and Mohammadi, 2009; Ornitz and Itoh, 2001). The expression of Fgfr1 and Fgfr2 is detected in the metanephros and in the cranial part of the mesonephros (Cancilla et al., 1999; Dudley et al., 1999; Kitagaki et al., 2011; Orr-Urtreger et al., 1991; Peters et al., 1992; Poladia et al., 2006; Zhao, 2004). In contrast, Fgfr3 and Fgfr4 are not detected in the mesonephros (Kitagaki et al., 2011). Conditional knockout (KO) approaches in mice have revealed that Fgfr1 and Fgfr2 in the kidney mesenchyme are crucial for early MM formation and UB induction (Hains et al., 2008; Poladia et al., 2006; Walker et al., 2013). On the other hand, epithelial FGFR signaling regulates the branching morphogenesis of the metanephric kidney, possessing minor functions on initial UB formation (Zhao, 2004). Recent studies indicate that FGFR signaling may compensate loss of GDNF/RET signaling, which is the most essential RTK signaling for the ureteric budding (Maeshima et al., 2007; Michos et al., 2010; Tee et al., 2013). Although the functions of FGF signaling in metanephric kidney morphogenesis have been analyzed, its function in WD development has not been well described. In addition, the possible compensation between different RTKs in the WD has not been investigated.

In this study, we examined the roles of FGFR signaling in the WD using mouse genetic models. WD-specific inactivation of Fgfr2 revealed a developmental region-specific function for FGFR2 on regulation of cell proliferation.

Materials and methods

Mice

The mouse strains used herein were Fgfr1 (Zhao et al., 2007), Fgfr2 (Yu et al., 2003), Hoxb7-Cre (Yu et al., 2002), ROSA26-LacZ (Soriano, 1999), ROSA26-YFP (Srinivas et al., 2001), Ret-GFP (Enomoto et al., 2001) and ICR (CLEA, Tokyo, Japan). Hoxb7-Cre mice and ROSA26-YFP mice were obtained from Jackson Laboratory. Fgfr2-null alleles were generated by crossing Fgfr2 mice to CAGGS-Cre mice (Araki et al., 1997), which express Cre recombinase in the germline. All experimental procedures and protocols were approved by the Animal Research Committee of the Wakayama Medical University and Kumamoto University. Embryos for each experiment were collected from at least three pregnant females. Noon on the day when a vaginal plug was detected was designated as E0.5. Embryos used for analyses at E10.5 were not identified their gender. Male embryos were used for the analyses at E12.5 and later.

Histology and immunohistochemistry

Tissues were fixed in 4% paraformaldehyde, dehydrated, embedded in paraffin, and sectioned. Hematoxylin and eosin (HE) staining was performed by standard procedures as previously described (Haraguchi et al., 2000). For immunohistochemistry, deparaffinized sections were treated for antigen retrieval (autoclave 121 °C for 1 min in 10 mM citrate buffer at pH 6.0) and incubated with 3% H2O2 in methanol for 10 min to inactivate endogenous peroxidase before 1 h incubation with a blocking solution (1.5% fetal bovine serum (FBS) in PBS). Anti-cleaved caspase-3 polyclonal antibody (Cell Signaling Technology, Danvers, MA, 1:1000), anti-ZO-1 polyclonal antibody (Invitrogen, Carlsbad, CA, 1:200), anti-laminin polyclonal antibody (Sigma-Aldrich, St. Louis, MO, 1:100), or anti-E-cadherin monoclonal antibody (BD Biosciences, San Jose, CA, 1:400) in PBS solution was added to the slides and incubated for 1 h at room temperature. Negative controls were obtained by excluding the primary antibodies. After washing with PBS, the sections were stained with the Vectastain ABC Kit (Vector Laboratories, Burlingame, CA) or incubated with Alexa 546/488-labeled goat anti-rabbit/mouse IgG (Life Technologies, Grand Island, NY) or horseradish peroxidase-labeled rabbit anti-goat IgG (Zymed, South San Francisco, CA) for 1 h at room temperature. Where indicated, the sections were subsequently incubated with 3, 3-diaminobenzidine tetrahydrochloride (DAB) containing H2O2. Sections were counterstained by Hoechst 33342 (Sigma-Aldrich), Hematoxylin (Wako, Osaka, Japan), or Methyl Green (Sigma-Aldrich).

Cell proliferation assay

Pregnant females were injected intraperitoneally with 25 mg 5-ethynyl-2′-deoxyuridine (EdU; Invitrogen) per 1 kg body weight. One hour after injection, embryos were collected. Embryos were fixed and processed for paraffin-embedding as described above. Serial horizontal sections (6 μm/section) were made from the cranial end to the caudal end of the WD. EdU-labeled cells were stained using the Click-iT EdU Alexa Fluor 555 Imaging Kit (Invitrogen) according to the supplier’s instructions. The total cell number (stained by Hoechst) and the total number of EdU-positive cells were counted for each epithelial portion. The percentage proliferation was calculated as the fraction of EdU-positive cells over the total number of nuclei. Statistical analysis was performed using the Mann–Whitney test (P<0.05 was considered significant).

Whole-mount and section RNA in situ hybridization for gene expression

Whole-mount and section in situ hybridization was performed using digoxigenin-labeled probes as previously described (Haraguchi et al., 2000). Probes for the following genes were used: Pax2 (Dressler et al., 1990), Fgfr1 (kindly provided by Dr. J. Partanen), Fgfr2 (Ozawa et al., 1996), Etv4 (provided by Dr. D. M. Ornitz), Etv5 (provided by Dr. D. M. Ornitz), Ret (Nishinakamura et al., 2001), Gdnf (Nishinakamura et al., 2001), and Fgf10 (kindly provided by Dr. H. Ohuchi and Dr. N. Itoh). For the confirmation of the Fgfr2 mutation, a probe that does not recognize the mutated allele was used (Yu et al., 2003).

Results

Fgfr2 is the major Fgfr gene expressed in the WD epithelia

Although Fgfr expression has been partially reported, their expression in the mesonephros has been poorly described. To identify patterns of Fgfr expression in the WD, whole-mount RNA in situ hybridization was performed at E10.5. Fgfr1 and Fgfr2 were expressed throughout the mesonephros of wild-type embryos at E10.5 (Fig. 1A and E, arrow). Fgfr1 expression was more prominent in the MTs (Fig. 1A and B, arrowheads) and mesenchyme (Fig. 1B–D) than that in the epithelia of the WD (Fig. 1B–D, dashed line, indicated in Fig. 1I). In contrast, Fgfr2 expression was more prominent in the epithelia of the WD (Fig. 1F–H, dashed line, indicated in Fig. 1I) compared to the MTs (Fig. 1E and F, arrowheads) and mesenchyme (Fig. 1F–H). Fgfr2 expression was also observed in the WD epithelia locating near the cloaca (Fig. 1J, dashed line).

Fig. 1

Fgfr2 is the major Fgfr gene expressed in the WD epithelia. Whole-mount RNA in situ hybridization to detect Fgfr1 (A) and Fgfr2 (E) expression in the mesonephros of wild-type embryos was performed at E10.5. Section RNA in situ hybridization to detect Fgfr1 (B–D) and Fgfr2 (F–H, and J) expression in the mesonephros of wild-type embryos was performed at E10.5. For the histological analyses of horizontal sections at E10.5, the WD is divided into three parts along the axis by the following definition: (1) cranial part, the WD where MTs are present; (2) caudal part, the WD of the caudal mesonephros (the WD between the cranial part and the UB); and (3) UB part, the WD where the pseudostratified epithelia is observed. A schematic representation of the cranial (B and F), caudal (C and G), and UB (D and H) parts in the WD at E10.5 are presented in I. Fgfr2 expression in the portion that lies near the cloaca is shown in J. Scale bars: A and E, 500 μm; B–D, F–H, and J, 50 μm. Arrow, WD; arrowheads, MTs; dashed line, WD epithelia; Cl, cloaca.

WD-specific inactivation of Fgfr2 results in WD regression in the caudal mesonephros

Cre activity driven by the Hoxb7 promoter was analyzed by crossing Hoxb7-Cre mice with the ROSA26-LacZ reporter strain and embryos were analyzed for β-galactosidase (βgal) enzyme activity. βgal-positive cells were located throughout the WD epithelia in Hoxb7-Cre; ROSA26 embryos at E10.5 and E9.5 (Supplemental Fig. S1A–C) (Kobayashi et al., 2005; Mugford et al., 2008; Yu et al., 2002). Most of the epithelial cells were βgal-positive in the WD, and a few βgal-negative cells were observed in the cranial part (Supplemental Fig. S1A, arrow). A small number of βgal-positive cells were observed in MTs (Supplemental Fig. S1A, arrowhead). The percentage of βgal-positive cells in each region of the WD epithelia of Hoxb7-Cre; ROSA26 embryos at E10.5 was shown in Supplemental Fig. S1 (n=3).

To examine the roles of FGFR2 signaling during WD development, Hoxb7-Cre; Fgfr2+/− male mice were crossed with Fgfr1; Fgfr2 female mice to get Hoxb7-Cre; Fgfr1; Fgfr2 embryos. Hereafter, we refer to Hoxb7-Cre; Fgfr1; Fgfr2 mice as WD-specific Fgfr2-inactivation mice (Fgfr2). Control mice are either heterozygous for the Fgfr2 floxed allele (Hoxb7-Cre; Fgfr1; Fgfr2) or lack the Hoxb7-Cre transgenic allele.

Inactivation of Fgfr2 in the WD of Fgfr2 embryos was confirmed by whole-mount RNA in situ hybridization. Fgfr2 was expressed in the WD and MTs in the control embryos at E10.5, while its expression was reduced in Fgfr2 embryos compared to the controls at E10.5 (Fig. 2A and A′, black arrow and arrowheads). To examine the development of WD structure in Fgfr2 mice, whole-mount RNA in situ hybridization for Pax2 was performed. Pax2 expression was observed in the WD, MTs, and condensed mesenchyme in both control and Fgfr2 embryos at E10.5 (Fig. 2B and B′, black arrow, black arrowheads, and white arrowheads). The MTs were formed along the cranial WD (Fig. 2B and B′, black arrowheads) and the UB invaginated dorsally (Fig. 2B and B′, yellow arrow), showing no significant differences between the control and Fgfr2 embryos. Histological analysis showed a well-established epithelial structure in the cranial part of the WD in both control and Fgfr2 embryos at E10.5 (Fig. 2E and E′, black arrow, indicated in Fig. 2S). In the caudal part, no significant differences were observed in the size of the WD between the control and some Fgfr2 embryos (3 out of total 7 samples), whereas a hypoplastic WD of the caudal mesonephros was observed in other Fgfr2 embryos (4 out of total 7 samples; Fig. 2F and F′, black arrow, also indicated in Fig. 2S). The ductal epithelia reached the cloacal epithelia in both control and Fgfr2 embryos at E10.5 (7 out of total 7 samples; Fig. 2L and L′, black arrowheads). The well-differentiated pseudostratified epithelia of the UB were detected even in Fgfr2 embryos with hypoplastic WD of the caudal mesonephros (Fig. 2G and G′, also indicated in Fig. 2S). These data indicate that the WD is formed in the mutant embryos by E10.5. At E12.5, the WD marked by Pax2 expression was observed from cranial to caudal of the control embryos (Fig. 2C). In contrast, such Pax2 expression was disrupted in the WD of the caudal mesonephros of Fgfr2 embryos (Fig. 2C′, red arrow). Histological analysis revealed that the cranial epithelial structure was indistinguishable between the control and Fgfr2 WD (Fig. 2H and H′, indicated in Fig. 2T). In contrast, the WD of the caudal mesonephros showed regressing epithelia in Fgfr2 embryos (a red arrow in Fig. 2I′, also indicated in Fig. 2T; 8 out of total 10 samples in Table 1). Taken together, these results suggest that epithelial FGFR signaling is essential for the maintenance of the WD, especially in the caudal mesonephros. At E16.5, the control embryos possessed well-differentiated WD derivatives, the epididymis and vas deferens, connected to the urogenital sinus (Fig. 2D). Fgfr2 embryos, on the other hand, showed a dilation of the duct in the cranial WD (Fig. 2D, D′, J, and J′, indicated in Fig. 2U; 4 out of total 6 samples in Table 1), and the epithelial structure was not observed in the vas deferens derived from the WD of the caudal mesonephros (Fig. 2K and K′, also indicated in Fig. 2U).

Fig. 2

WD-specific inactivation of Fgfr2 results in WD regression in the caudal mesonephros. Whole-mount RNA in situ hybridization to detect Fgfr2 expression in the WD and MTs of the control (A) and Fgfr2 (A′) embryos was performed at E10.5. Insets show high magnification views of the caudal mesonephros (A and A′). The WD and MTs are marked by Pax2 expression in the control (B and C) and Fgfr2 (B′ and C′) embryos at E10.5 (B and B′) and E12.5 (C and C′). Insets show high magnification views of the caudal mesonephros (B, B′, Cand C′). The gross morphology of the WD and testis at E16.5 is shown inD (control) and D′ (Fgfr2). HE staining of the WD in the control (E–L) and Fgfr2 (E′–L′) embryos was performed at E10.5 (E–G, L, E′–G′, and L′), E12.5 (H, I, H′, and I′), and E16.5 (J, K, J′, and K′). Schematic representations indicating the axial levels of the WD at E10.5, E12.5, and E16.5 are presented in S, T, and U. The WD that lies near the cloaca is shown in L (Control) and L′ (Fgfr2). HE staining at E12.5 and the gross morphology of the WD and testis at E16.5 in Fgfr1/2 and Fgfr1 embryos are shown in M–O and P–R. Scale bars: A–C and A′–C′, 100 μm; D, D′, O, and R, 500 μm; E–N, P, Q, and E′–L′, 50 μm. Black arrow, WD; black arrowheads, MTs; white arrowheads, condensed mesenchyme; yellow arrow, UB; red arrow, regression of WD epithelia; Cl, cloaca; t, testis.

Table 1

Phenotypic incidence of Fgfr2, Fgfr1/2 and Fgfr1 embryos.

Individual allele with genotype	E12.5	E16.5
Hoxb7-Cre; Fgfr1flox/+; Fgfr2flox/− (Fgfr2WD flox/−)	8/10	4/6
Hoxb7-Cre; Fgfr1flox/−; Fgfr2flox− (Fgfr1/2WD flox/−)	3/3	2/3
Hoxb7-Cre; Fgfr1flox/−; Fgfr2flox/+ (Fgfr1WD flox/−)	0/8	0/4

Functional redundancy of Fgfr1 and Fgfr2 has been reported in several developmental systems (Hebert, 2003; Paek et al., 2009; White et al., 2006). To examine the function of Fgfr1 upon inactivation of Fgfr2 in the WD epithelia, Hoxb7-Cre; Fgfr1; Fgfr2 (Fgfr1/2) mice were investigated. Fgfr1/2 embryos showed regressing epithelia in the WD of the caudal mesonephros at E12.5 (Fig. 2N, red arrow; 3 out of total 3 samples in Table 1), as observed in Fgfr2 embryos (Fig. 2I′; 8 out of total 10 samples in Table 1). In contrast, the epithelial structure was indistinguishable in the cranial part between the control and Fgfr1/2 WDs (Fig. 2H and M). At E16.5, the gross morphology of Fgfr1/2 embryos showed dilated duct formation in the cranial part, with an obstructed duct in the caudal part at E16.5 (Fig. 2O; 2 out of total 3 samples in Table 1), as observed in Fgfr2WD flox/− embryos (Fig. 2D′; 4 out of total 6 samples in Table 1). By the histological analysis of the Fgfr1/2 WD, the presence of epithelia in the cranial part was observed, while the epithelial structure was not evident in the caudal part (data not shown), as observed in Fgfr2 embryos (Fig. 2J′ and K′). Hoxb7-Cre; Fgfr1; Fgfr2ox/−) embryos showed no obvious phenotypes; well-differentiated WDs reached the urogenital sinus as well as the control WD at E16.5 (Fig. 2R, n = 4 in Table 1) and at E12.5 (Fig. 2P and Q; n=8 in Table 1). This is consistent with the absence of prominent Fgfr1 expression in the WD epithelia during WD development (Fig. 1B–D). These results indicate that Fgfr2 is the dominant FGF receptor for the maintenance of the WD in the caudal mesonephros, and Fgfr1 may contribute with a small amount of redundancy.

Decreased expression of the downstream target genes of RTK signaling in the WD epithelia of Fgfr2WD flox/− embryos

Etv4 and Etv5 are downstream targets of RTK signaling including FGFR (Brent and Tabin, 2004; Firnberg and Neubüser, 2002; Liu et al., 2003). The expression pattern of Etv4 and Etv5 in the WD was examined at the WD of the cranial and caudal mesonephros, and UB (Fig. 3N). Both Etv4 and Etv5 were expressed in the WD epithelia, MT, and condensed nephrogenic mesenchyme of E10.5 control embryos (dashed line, black arrowhead, and white arrowhead in Fig. 3A–C and D–F). Expression of these genes was reduced in the WD epithelia of the cranial and caudal mesonephros of Fgfr2 embryos, compared to control embryos (Fig. 3A, A′, B, B′, D, D′, E and E′, dashed line). The expression of these genes was not significantly different in the MTs (Fig. 3A, A′, D, and D′, black arrowhead) and the condensed mesenchyme (Fig. 3B, B′, E, and E′, white arrowhead) between the control and Fgfr2 embryos. Notably, the expression of Etv4 and Etv5 in the epithelia adjacent to the MM (MM side) was retained at the UB level, whereas it was decreased in the epithelia at the opposite side of the MM (coelom side) in Fgfr2 embryos (Fig. 3C, C′, F, and F′, red arrowheads, indicated in Fig. 3O).

Fig. 3

Decreased expression of the downstream target genes of RTK signaling in the WD epithelia of Fgfr2 embryos. Section RNA in situ hybridization for Etv4 (A–C and A′–C′), Etv5 (D–F and D′–F′), and Ret (G–I and G′–I′) in E10.5 embryos in the cranial (A, A′, D, D′, G, and G′), caudal (B, B′, E, E′, H, and H′), and UB (C, C′, F, F′, I, and I′) parts was performed in the control (A–I) and Fgfr2 (A′–I′) embryos. Whole-mount RNA in situ hybridization to detect Ret expression was performed in the control (J) and Fgfr2 (J′) embryos at E10.5. Section RNA in situ hybridization for Etv5 in wild-type (K–M) and Ret KO (K′–M′) embryos for the cranial (K and K′), caudal (L and L′), and UB (M and M′) parts was performed at E10.5. A schematic representation of the cranial, caudal, and UB parts in the WD at E10.5 is shown in N. A schematic representation of the MM and coelom sides of the UB at E10.5 is shown in O. Scale bars: A–I, K–M, A′–I′, and K′–M′, 50 μm; J and J′, 100 μm. Dashed line, WD epithelia; black arrowhead, MT; white arrowhead, condensed mesenchyme; MM, metanephric mesenchyme; red arrows, coelom side of the UB; yellow arrow, UB.

In addition to FGFR signaling, GDNF/RET signaling is also essential RTK signaling for UB formation (Cacalano et al., 1998; Enomoto et al., 1998; Moore et al., 1996; Pichel et al., 1996; Sanchez et al., 1996; Schuchardt et al., 1994). Failure of UB formation is the most frequent phenotype associated with Gdnf or Ret deficiency (Costantini and Shakya, 2006). In control embryos, Ret was expressed throughout the WD epithelia (Fig. 3G–I, dashed line; Fig. 3J, black arrow), with the highest level of expression in the UB at E10.5 (Fig. 3I, dashed line; Fig. 3J, yellow arrow). In contrast, Ret expression was reduced in the WD of the cranial and caudal mesonephros, and in the coelom side of the UB of Fgfr2 embryos (Fig. 3G–I and G′–H′, dashed line, red arrowheads; Fig. 3J and J′, black arrow). Ret expression in the MM side of the UB was not affected in Fgfr2 embryos at E10.5 (Fig. 3I′, dashed line; Fig. 3J′, yellow arrow). These results suggest that Ret expression may be regulated by Fgfr2 in the WD epithelia, except in the MM side of the UB. This observation is consistent with the finding that UB outgrowth was unaffected in Fgfr2 embryos until E13.5 (data not shown) (Zhao, 2004).

In the WD, it is reported that chimeric expression of Ret in the UB results in epithelial cell rearrangement to exclude the Ret-negative cells from the MM side of the UB (Chi et al., 2009). To trace the lineage of cells in which Cre was activated in the Fgfr2 background, we crossed Hoxb7-Cre; Fgfr2+/− mice with Fgfr2; ROSA26 mice to obtain Hoxb7-Cre; ROSA26; Fgfr2 embryos. Recombined cells, marked by GFP, were distributed throughout the WD and displayed no significant differences when compared with Hoxb7-Cre; ROSA26 embryos at E10.5 (Supplemental Figs. S1 and S2). Uniform distribution of the recombined cells in the UB was observed in Fgfr2 embryos (Supplemental Fig. S2). These results suggest that Hoxb7-Cre-mediated Fgfr2 gene recombination in the WD was introduced ubiquitously and recombined cell-specific epithelial cell rearrangement may not occur in the Fgfr2 WDs. Taken together, these results suggest that FGFR2 signaling is the major RTK signaling pathway in the WD epithelia, but not in the MM side of the UB, at E10.5.

To examine the involvement of GDNF/RET signaling in Etv expression in the WD, we analyzed Ret KO mice at E10.5. Ret gene mutations lead to renal agenesis and hypoplasia, and ectopic ureter termination (de Graaff et al., 2001; Enomoto et al., 2001; Jain et al., 2006; Schuchardt et al., 1996). Sustained expression of Etv4 and Etv5 was observed in all regions of the WD epithelia of Ret KO embryos (dashed line in Fig. 3K–M and K′–M′, data not shown). However, reduced phospho-Erk expression was observed in the MM side of the UB of Ret KO embryos at E10.5 (data not shown), which agrees with the previous finding (Chi et al., 2009). These results suggest that the reduced expression of Etv4 and Etv5 observed in Fgfr2 embryos is independent of RET-mediated RTK signaling.

FGFR signaling regulates cell proliferation in the WD of the caudal mesonephros and the coelom side of the UB

To identify the mechanism leading to the regression of the WD of the caudal mesonephros in Fgfr2 embryos, cell proliferation was assessed by EdU incorporation at the WD of the cranial and caudal mesonephros, and UB at E10.5 (Fig. 4A–C and A′–C′; indicated in Fig. 4N). In control embryos, the WD of the caudal mesonephros and the UB displayed higher rate of cell proliferation than the cranial part (Fig. 4A–C). The WD of the caudal mesonephros in the Fgfr2 embryos demonstrated significantly reduced epithelial cell proliferation, compared to the control embryos (Fig. 4B, B′, and M). There were no significant differences in cell proliferation in the cranial part and in the MM side of the UB between the control and Fgfr2 WDs (Fig. 4A, A′, C, C′ and M). Fewer proliferating cell nuclei were observed in the coelom side of the UB in Fgfr2 embryos than in the control embryos (Fig. 4C, C′, and M). Apoptosis, indicated by cleaved caspase-3 expression, was also analyzed in control and Fgfr2 embryos. No significant differences in apoptosis were observed between the control and Fgfr2 embryos at E10.5 (Fig. 4D–F and D′–F′, red arrowheads). These results indicate that FGFR signaling regulates cell proliferation in the WD of the caudal mesonephros and in the coelom side of the UB.

Fig. 4

FGFR signaling regulates cell proliferation in the WD of the caudal mesonephros and the coelom side of the UB. Proliferative cells were labeled with EdU (red) and co-stained with E-cadherin (green) at the cranial (A and A′), caudal (B and B′), and UB (C and C′) parts in the control (A–C) and Fgfr2 (A′–C′) embryos. Nuclei were stained with Hoechst 33342. Immunohistochemistry for cleaved caspase-3 demonstrated the presence of apoptotic cells at the axial levels of the cranial (D and D′), caudal (E and E′), and UB (F and F′) parts in the control (D–F) and Fgfr2 embryos (D′–F′). Sections were counterstained with hematoxylin. Immunohistochemistry for ZO-1 was performed at the axial levels of the cranial (G and G′), caudal (H and H′), and UB (I and I′) parts in the control (G–I) and Fgfr2 (G′–I′) embryos. Sections were counterstained with methyl green. Immunofluorescence staining of laminin (red) and E-cadherin (green) in sections of E10.5 WD was performed at the axial levels of the cranial (J and J′), caudal (K and K′), and UB (L and L′) parts in the control (J–L) and Fgfr2 (J′–L′) embryos. Nuclei were stained with Hoechst 33342. The graph (M) demonstrated the rate of EdU-positive cells per total cells in the WD epithelia in the control and Fgfr2 embryos. A schematic representation of the cranial, caudal, and UB parts in the WD at E10.5 is presented in N. The UB is divided into two domains, the MM side and coelom side, as indicated in O. Scale bars: 50 μm. Red arrowheads, signals of cleaved caspase-3; dashed line, WD epithelia; MM, metanephric mesenchyme.

**P< 0.01; n = 4 per genotype (cranial and caudal), n = 6 per genotype (UB); values are mean + SD.

To investigate whether loss of epithelial cell polarity contributes to the epithelial regression observed in Fgfr2 WDs, expression patterns of ZO-1 (Fig. 4G–I and G′–I′), E-cadherin, and laminin (Fig. 4J–L and J′–L′) was analyzed. The core tight junction protein, ZO-1, was detected at the apical side of epithelial cells in both control and Fgfr2 WDs (Fig. 4G–I and G′–I′). E-cadherin, an adherence junction protein, and laminin, a basement membrane component, were detected at the apical and lateral side of the cells and at the basement membrane, respectively, in Fgfr2 WDs as well as in control embryos (Fig. 4J–L and J′–L′). These results suggest that polarized epithelia are established and maintained in Fgfr2 WDs.

Region-specific expression of RTK ligands in the mesonephros and metanephros

FGF7 and FGF10 are ligands with high specificity for the FGFR2-IIIb isoform (Ornitz et al., 1996; Zhang et al., 2006). Previous studies have demonstrated that Fgf7 is not detected in the developing kidney before E14.5 (Finch et al., 1995; Mason et al., 1994). On the other hand, FGF10 is considered the main ligand for FGFR in the WD (Donjacour et al., 2003).

Fgf10 expression was observed in both the MM and the mesenchyme of the coelom side with a pattern showing a higher expression at the UB level that gradually decreases as it proceeds cranially (Fig. 5A and B, Supplemental Fig. S4). Gdnf, a ligand for RET signaling, was expressed mainly in the MM of the control embryos (Fig. 5C and D). Localization of Fgf10 and Gdnf expression in the Fgfr2 embryos was essentially similar to that in the control embryos (Fig. 5A–D and A′–D′).

Fig. 5

Region-specific distribution of RTK ligands in the mesonephros and metanephros. Section RNA in situ hybridization for Fgf10 (A, A′, B, and B′) and Gdnf (C, C′, D, and D′) in the caudal mesonephros (A, A′, C, and C′) and UB (B, B′, D, and D′) was performed in the control (A–D) and Fgfr2 (A′–D′) embryos at E10.5. Scale bars: 50 μm. Dashed line, WD epithelia; MM, metanephric mesenchyme.

Discussion

WD as a model for ductal formation and its abnormalities

The current study demonstrates that FGFR2 signaling prevents WD regression in the caudal mesonephros, and that loss of FGFR2 in the WD may lead to dilation of the cranial part of the WD. A previous study using FGF10 null mice reported cystic dilation in the epididymis and degeneration of the vas deferens (Donjacour et al., 2003). However, the corresponding mechanisms were not identified. In the current study, we demonstrated that reduced FGF signaling in the WD epithelia results in WD regression in the caudal mesonephros. Luminal fluid from the testis to the WD is essential for further development of the WD into male reproductive tract (Hinton et al., 2000; Shum et al., 2011; Tong et al., 1996). Since fluid secretion commences at approximately E13–14 in mice (Joseph et al., 2009), the obstruction of the WD in the caudal mesonephros in Fgfr2 embryos may cause accumulation of the fluid, leading to dilation of the cranial part of the WD at E16.5.

Formation of the common nephric duct (CND), the WD which locates caudally to the UB, was observed in both the control and Fgfr2 embryos at E12.5 (Supplemental Fig. S3A and A′). At E14.5, the CND was eliminated, and the WD and ureter were separated in the control and Fgfr2 embryos (Supplemental Fig. S3B, B′, C, and C′; the level of B and B′ is more anterior than that of C and C′). Our observation that the WD and the ureter were inserted into the urogenital sinus separately at E14.5 suggests that the dilation of the cranial epididymal duct was a consequence of the obstruction of the duct, not of back-flow of urine. Another potential mechanism underlying such dilation might be the disruption of oriented cell division, as described previously in the kidney of Fgfr1 and Fgfr2 double mutant mice (Sims-Lucas et al., 2012).

The WD also grows craniocaudally after reaching the cloaca. Immunohistochemical analysis for ZO-1, E-cadherin, and laminin expressions indicated that the epithelial cells maintained their characteristics in Fgfr2 WDs at E10.5. Cell proliferation is considered to be the major contributor for WD elongation, and the WD of the caudal mesonephros is the most mitotically active in the WD (Fig. 4A–C) (Dyche, 1979; Joseph et al., 2009; Michael and Davies, 2004). Considering that apoptosis was not increased in Fgfr2 WDs compared to control WDs at E10.5, it is possible that decreased cell proliferation in the WD of the caudal mesonephros causes the ductal thinning observed in Fgfr2 embryos at E12.5 and later. The current study suggests that active cell proliferation regulated by FGFR signaling in the WD of the caudal mesonephros is essential to maintain the ductal size and epithelial integrity in the growing duct at a later stage.

RTK signaling via FGFR is required for WD development

The current study demonstrates that WD epithelial cell proliferation is regulated by distinct region-specific FGFR signaling.

The frequency of βgal-negative cells in the cranial WD epithelia of Hoxb7-Cre; ROSA26 embryos was higher than that in the WD of the caudal mesonephros and UB. It is possible that the sustained epithelial cell proliferation in the cranial part of mutant WDs might be due to lower activity of Hoxb7-Cre in the cranial WD epithelia. However, significant reduction of active RTK signaling, indicated by Etv4 and Etv5 expression, was observed in the cranial part of Fgfr2 WDs compared to control WDs. These results suggest that RTK signaling is sufficiently reduced in the cranial part of Fgfr2 WDs. Taken together, RTK signaling through Etv4 and Etv5 may possess less contribution to epithelial cell proliferation in the cranial WD. Fgf8 is expressed in MTs at E10.5 and its mesoderm-specific conditional KO mice display degeneration in the cranial part of the WD and missing cranial MTs (Kitagaki et al., 2011). Moreover, mesoderm-specific conditional KO mice for Fgfr1 and Fgfr2 show loss of the WD and MTs (Kitagaki et al., 2011). These observations may suggest that FGFR signaling in MTs or in the nephrogenic mesenchyme is essential for the maintenance of the cranial part of the WD.

In Fgfr2 embryos, epithelial cell proliferation in the MM side of the UB did not show any significant reduction. The expression pattern of Ret, Etv4, and Etv5 in Fgfr2 embryos indicated that RTK signaling via RET was retained in the MM side of the UB. These results are consistent with the previous findings that UB outgrowth is regulated mainly by GDNF/RET signaling (Moore et al., 1996; Pichel et al., 1996; Sainio et al., 1997; Sanchez et al., 1996; Schuchardt et al., 1994). In Ret KO mice, almost 40% of embryos show ureteric budding, and additional inhibition of FGFR signaling completely abolishes UB formation (Michos et al., 2010; Schuchardt et al., 1994). These previous observations indicate partial redundancy between RET and FGFR signaling in the UB.

The current work demonstrated that Ret expression was reduced in the WD epithelia, except in the MM side of the UB in Fgfr2 embryos, suggesting that Fgfr2 probably regulates GDNF/RET signaling in the WD of the caudal mesonephros and the coelom side of the UB. However, analysis of Ret KO embryos suggests that GDNF/RET signaling is dispensable for Etv expression and also for cell proliferation (data not shown) in these parts. These results indicate that RTK signaling via FGFR is required for WD development.

To our knowledge, Fgf10 is the only RTK ligand that is distinctly expressed in the mesenchyme of the caudal mesonephros and of the coelom side of the UB at E10.5. In contrast, Gdnf is prominently expressed in the metanephric mesenchyme. This region-specific expression pattern of RTK ligands may explain the region-specific FGFR2 dependency of RTK signaling in the WD epithelia, i.e., the possible redundancy of RTK signaling between FGF and GDNF in the MM side of the UB but not in the coelom side and the WD of the caudal mesonephros.

In conclusion, the current study demonstrates that WD epithelial cell proliferation is regulated by distinct region-specific mechanisms during its development. The results suggest that FGFR signaling is essential for cell proliferation of the WD in the caudal mesonephros and the coelom side of the UB. The WD is fundamental not only for the development of the transient or definitive kidneys but also for the development of the reproductive tract. The WD is also essential for the formation of the Mullerian duct (MD), female reproductive tract anlarge. The MD extends from cranial to caudal under the guidance of the WD around E12.5 (Mullen and Behringer, 2014). Therefore, it is possible that the abnormality observed in the current mutant embryos may also consequently affect the MD development and female reproductive tract development. Further analyses regarding female embryos may provide broad function of FGF signaling in reproductive tract formation. The current study provides novel insights into the regulation of urogenital tract development, including male and female reproductive tract, and kidney.