|Home | About | Journals | Submit | Contact Us | Français|
Wnt4 and β-catenin are both required for nephrogenesis, but studies using TCF-reporter mice suggest that canonical Wnt signaling is not activated in metanephric mesenchyme (MM) during its conversion to the epithelia of the nephron. To better define the role of Wnt signaling, we treated rat metanephric mesenchymal progenitors directly with recombinant Wnt proteins. These studies revealed that Wnt4 protein, which is required for nephron formation, induces tubule formation and differentiation markers Lim1 and E-cadherin in MM cells, but does not activate a TCF reporter or up regulate expression of canonical Wnt target gene Axin-2 and has little effect on the stabilization of β-catenin or phosphorylation of disheveled-2. Furthermore, Wnt4 causes membrane localization of ZO-1 and occludin in tight junctions. To directly examine the role of β-catenin/TCF-dependent transcription, we developed synthetic cell-permeable analogs of β-catenin’s helix C, which is required for transcriptional activation, in efforts to specifically inhibit canonical Wnt signaling. One inhibitor blocked TCF-dependent transcription and induced degradation of β-catenin but did not affect tubule formation and stimulated the expression of Lim1 and E-cadherin. Since a canonical mechanism appears not to be operative in tubule formation, we assessed the involvement of the non-canonical Ca2+-dependent pathway. Treatment of MM cells with Wnt4 induced an influx of Ca2+ and caused phosphorylation of CaMKII. Moreover, Ionomycin, a Ca2+-dependent pathway activator, stimulated tubule formation. These results demonstrate that the canonical Wnt pathway is not responsible for mesenchymal-epithelial transition (MET) in nephron formation and suggest that the non-canonical calcium/Wnt pathway mediates Wnt4-induced tubulogenesis in the kidney.
Reciprocal interactions between the metanephric mesenchyme (MM) and ureteric bud (UB) are required for nephron formation in kidney development. The UB induces MM to undergo epithelial conversion to form the nephron through a series of morphogenetic transformations, including condensation, pretubular aggregation, and mesenchymal-epithelia transition (MET) (Dressler, 2006). Multiple inductive signaling ligands have been elucidated in this process, including Wnts (Kispert et al., 1998), fibroblast growth factors (FGFs) (Perantoni et al., 1995), transforming growth factor-β2 (TGF-β2) (Barasch et al., 1999; Plisov et al., 2001) and leukemia inhibitory factor (LIF) (Barasch et al., 1999; Plisov et al., 2001), using an explant culture system with isolated MMs. These studies imply the existence of progenitor cells in MM rudiments with the retained capacity to produce the epithelia that comprise the various segments of the nephron (Karavanova et al., 1996; Osafune et al., 2006; Perantoni et al., 1995).
Wnt/β-catenin signaling is essential for both normal development and tumorigenesis (Koesters and von Knebel Doeberitz, 2003; Logan and Nusse, 2004; Smalley and Dale, 1999). In the canonical Wnt/β-catenin signaling pathway, engagement of a Wnt receptor Frizzled leads to stabilization of β-catenin, which is then translocated into the nucleus to initiate gene expression through interaction with a member of the TCF transcription factor family. Alternatively, two major TCF-independent Wnt signaling mechanisms have also been described: a Ca2+-releasing pathway (Kohn and Moon, 2005; Kuhl et al., 2000b) that is activated by Wnt-stimulated G proteins and the planar cell polarity (PCP) pathway that involves Wnt-activated c-Jun N-terminal kinase (Katoh, 2005; Yamanaka et al., 2002).
During kidney development, the requirement for Wnt9b and Wnt4 in nephrogenesis has been demonstrated using genetically modified mice. Wnt4 is expressed along with Fgf8 (Perantoni et al., 2005) in pretubular aggregates and is essential for normal conversion of MM to the epithelia of the nephron (Carroll et al., 2005; Stark et al., 1994). Wnt9b, which is expressed in the UB, functions upstream of Wnt4, and its loss could be rescued with Wnt1, a putative canonical Wnt signaling activator (Carroll et al., 2005). In addition, removal of β-catenin in MM showed reduced nephron formation (Park et al., 2007). These results suggest that canonical Wnt/β-catenin is involved in tubule formation in MM; however, constitutive expression of stabilized β-catenin cannot induce MET in cultured MMs or transgenic mice (Park et al., 2007; Schmidt-Ott et al., 2007). Furthermore, studies with β-catenin-activated transgenic reporter (BAT-gal) mice or TCF-driven β-gal reporter mice revealed that β-catenin-dependent transactivation occurs in the branching UB but not in the surrounding MM or its derivatives (Iglesias et al., 2007; Maretto et al., 2003), suggesting that β-catenin/TCF-dependent transcriptional activation may not be required for mesenchymal cell differentiation. Although the Wnt/β-catenin pathway has been implicated as a regulator of epithelial formation, the precise mechanism(s) by which Wnt4 mediates its cellular effects in nephrogenesis remains unclear.
In this study, we confirm that BATlacZ transgenic mice have strong β-gal activity in the UB but not in MM cells. We also report that Wnt4 does not activate canonical signaling in MM but rather induces MET through a non-canonical Ca2+-dependent pathway.
Ionomycin, A23187 and BAPTA-AM were purchased from BIOMOL (Plymouth Meeting, PA, USA). Wnt4, Wnt9b and Wnt3a recombinant proteins were from R&D systems (Minneapolis, MN, USA). Antibodies were obtained as follows: for β-catenin (#9562 ), Dvl2 (#3224), phospho-CaMKII (#3361) and pan-CaMKII (#4436) - from Cell Signaling Technology (Beverly, MA, USA); for ZO-1 (#339100), Occludin (#711500), rabbit Alexa Fluor 488 and mouse Alexa Fluor 555-conjugated secondary antibodies - from Invitrogen (Carlsbad, CA, USA); for β-Actin (A5441) - from Sigma (St.Louis, MO, USA); for E-cadherin (#610181) - from BD Biosciences (San Jose, CA, USA).
Generation and confirmation of BATlacZ transgenic mice as well as X-gal staining were previously described (Nakaya et al., 2005). All animal procedures were performed following the guidelines from NCI-Frederick Animal Care and Use Committee. NCI-Frederick is accredited by AAALAC International and follows the Public Health Service Policy for the Care and Use of Laboratory Animals. Animal care was provided in accordance with the procedures outlined in the “Guide for Care and Use of Laboratory Animals” (National Research Council; 1996; National Academy Press; Washington, D.C.). Metanephric mesenchymes dissected away from T-shaped ureteric buds were prepared from E13.5 rat embryos and cultured as described previously (Perantoni et al., 1991). Cultured mesenchymes were treated with 50 ng/ml human FGF2 and 10 ng/ml human TGF-α (R&D systems). Wnt4-soaked beads were prepared with Affi-gel Blue Gel (BioRad, Hercules, CA, USA). Beads were rinsed in PBS several times and soaked in 100 ng/ml mouse Wnt4 recombinant protein (R&D system), or 100 ng/ml BSA-PBS (for control) solution overnight at 4°C. Prepared beads were implanted into the MM rudiments. Wnt3a conditioned medium (Wnt3aCM) was prepared as described (Willert et al., 2003).
Metanephric mesenchymes were enzymatically separated from the ureteric buds and cultured on Matrigel-coated BD-Biocoat 10-cm dishes (BD Biosciences) in serum-free Ham’s F12:DMEM 1:1 (Gibco, BRL) medium with supplements as described (Perantoni et al., 1991) and with the addition 10 ng/ml human TGF-α and 50 ng/ml human FGF2. Tissues were grown for 10 days, resulting in a 50–100 fold increase in cell numbers. Cells were removed from plates with trypsin-EDTA solution (Gibco-Invitrogen).
WISH was performed essentially as described previously (Pizard et al., 2004), except that BM purple (Roche Diagnostics GmbH, Mannheim, Germany) was used as the chromogenic substrate. In short, plasmids with cDNAs encoding the Lim1 (Karavanova et al., 1996) gene were linearized at the 5′-end of the inserts and used to generate antisense probes corresponding to the coding sequences with a transcription kit and DIG-labeled dNTPs (Roche). Probes were precipitated and tested on agarose gels for integrity. Prior to the in situ hybridization procedure, the tissues or explants were fixed overnight with 4% paraformaldehyde in PBS after 3–6 days culture, and then tissues were pre-hybridized for 1 h at 70°C in hybridization buffer (50% formamide, 5 × SSC, 50 μg/ml heparin, 500 μg/ml tRNA, 0.1% Tween20, 9.0 mM citric acid) and incubated with 50ng of probe overnight at 70°C in hybridization buffer. The samples were washed with optimized wash buffer. The samples were placed for 3 hours in blocking solution at room temperature and then incubated overnight at 4°C with an alkaline phosphatase-coupled anti-digoxigenin antibody (Roche) at a 1/1000 dilution in blocking buffer. Finally, the samples were washed, and stained with BM purple (Roche). The reaction was stopped in 4% PFA in PBST.
Total RNA was purified using TRIzol reagent (Invitrogen), and RNase-Free DNase Set (Qiagen, Valencia, CA, USA) was used to digest genomic DNA during RNA purification. For RT-PCR, random hexamer primers were used to generate cDNA from total RNA. Gene-specific primers were then used to amplify PCR fragments. RT reactions were performed with a Superscript III RT kit according to manufacturer’s instructions (Invitrogen). PCR reactions were performed using a PCR enzyme mix (Roche) with primers. Incubation conditions consisted of 35 cycles (30 seconds at 94°C, 30 seconds at 55–60°C (depending upon primer optimal annealing temperatures) and 30 seconds at 72°C). Primer sequences, annealing temperatures, and cycle parameters are shown in supplementary Table 1.
After treatments, cells were washed with PBS, pH7.4, scraped with Laemmli reducing sample buffer, and resolved in 4–20 % Tris-glycine gels (Invitrogen). Proteins were electrotransfered to PVDF membranes (Bio-Rad, Hercules, CA, USA). Membranes were blocked with 5% nonfat dry milk in TBST and incubated overnight at 4°C with primary antibodies. Membranes were washed with TBST, followed with species-specific HRP-conjugated secondary antibodies. Bound antibodies were visualized using the SuperSignal West Pico Chemiluminescent Substrate system (ThermoFisher Scientific, Rockford, IL) according to manufacturer’s instructions.
Primary MM cells were transfected with the Amaxa Nucleofector 96-well Shuttle System (Lonza Group Ltd, Basel, Switzerland). Briefly, 1×106 cells per transfection were resusupended in 20 μl of 96- well Nucleofector solutions. TOPflash luciferase plasmid (1μg) and 0.1 μg pRL-TK-renilla plasmids (Promega, Madison, WI, USA) were mixed with cell suspensions and transferred to the well of a 96-well Nucleocuvette module. Nucleofections were performed according to manufacturer’s instructions using program DN-100. Twenty-four hrs after transfection, cells were treated with various factors for an additional 24h incubation. Cells were lysed and assayed for reporter activity using a Dual luciferase assay system (Promega) and a Monolight 3010 Luminometer (Pharmagen, San Diego, CA, USA).
Primary MM cells were cultured on 3-cm cover slip dishes and treated as indicated. Cells were washed with PBS, fixed in ice-cold methanol for 10 min, and washed with PBS for 10 min. The cells were blocked for 1 hr at room temperature with 10 % goat serum in PBS, and incubated with ZO-1 (1:50) or occludin (1:50) antibodies for 1 hr. The cells were then washed with 1 % goat serum in PBS and incubated with Alexa Fluor 488-conjugated secondary antibody for 1 hr. The cells were treated with DAPI, mounted on slides with Vectashield (Vector laboratories, Inc., Burlingame, CA, USA), and visualized using a Zeiss LSM510 META confocal microscope (Carl Zeiss, Jena, Germany) at 400× magnification. Confocal images were captured and analyzed by the LSM image Browzer program (Zeiss).
Calcium indicator Fluo-4 NW (Invitrogen) was used for intracellular calcium influx detection following manufacturer’s instructions. Primary MM cells were cultured at 50,000 cells per well in black, clear-bottom poly-Lysine coated 96 well plate (BD Biosciences). Calcium assays were performed in a NOVOstar multifunctional plate leader (BMG LABTECH Gmbh, Offenburg, Germany). The cells were treated with different concentrations of Wnt4 by using a NOVOstar pipettor. Samples were added at 10 sec and readings were taken up to 4 min at 1 sec intervals. The fluorescence intensity mode used included a 485/10 excitation filter and a 520/10 emission filter. Background fluorescence for each well was measured every 1 sec for 4 min without addition of ligand, and the average background was subtracted from each value. Data are representative of 3–5 independent experiments.
The peptides were synthesized on a 433A Peptide Synthesizer (Applied Biosystems) using Fmoc chemistry. The peptides were cleaved from the resin and deprotected with a mixture of 90.0 % (v/v) trifluoroacetic acid (TFA) with 2.5 % water, 2.5 % triisopropyl-silane, and 5 % thioanisol. The resin and deprotection mixture were pre-chilled to −5ºC and reacted for 15 minutes at −5ºC with stirring. The reaction was allowed to continue at room temperature for 1.75 hr. The resin was filtered off and the product was precipitated with cold diethyl ether. The resin was washed with neat TFA. Peptide suspended in diethyl ether was centrifuged at −20ºC and the precipitate was washed with diethyl ether four more times and left to dry in a vacuum overnight. The dried crude peptide was dissolved in DMSO and purified on a preparative (25 mm × 250mm) Atlantis C18 reverse phase column (Agilent Technologies) in a 90 minute gradient of 0.1 % (v/v) trifluoroacetic acid in water and 0.1 % trifluoroacetic acid in acetonitrile, with a 10 mL/min flow rate. The fractions containing peptides were analyzed on Agilent 1100 LC/MS spectrometer with the use of a Zorbax 300SB-C3 Poroshell column and a gradient of 5 % acetic acid in water and acetonitrile. Fractions that were more than 95 % pure were combined and freeze dried. Retro-inverso peptide made of all-D amino acids was synthesized using essentially the same protocol, except that palmitic acid had to be introduced in the side chain. Resin preloaded with α-Fmoc-ε-palmytoil-D-Lys was prepared as described (Remsberg et al., 2007). Purity (>95%) and structure of all peptides were confirmed by mass spectrometry using Agilent C18 reversed-phase columns (4.6×150 mm) and an Agilent 1100 LC/MC system.
Data were evaluated for statistical significance by analysis using student t-test. Statistical differences were considered significant if p<0.05* or p<0.01**. All experiments were performed independently at least three times, and the data presented are from a representative experiment. The results are presented as mean±S. D. (error bars)
To assess the role of Wnt/β-catenin signaling in nephron formation, we examined the LacZ expression pattern in the kidneys of BATLacZ reporter transgenic mice (Nakaya et al., 2005). As shown in Fig. 1A, strong β-gal staining was observed in the UB but not in MM. To confirm that the reporter is inducible in MM, we assessed expression in isolated MM from a BATLacZ reporter transgenic mouse treated with the GSK3 inhibitor BIO or LiCl. As shown, both established activators of canonical Wnt signaling can induce β-galactosidase activity in explanted MMs from BATLacZ reporter mice (Suppl. Fig. 1). These results suggest that β-catenin/TCF-dependent transactivation is not required for conversion of MM to nephronic epithelia, although it may still be involved in progenitor proliferation or maintenance (Schmidt-Ott et al, 2007).
Since Wnt4 is required for nephronic differentiation in MM (Kispert et al., 1998; Stark et al., 1994), we first addressed whether purified recombinant protein can induce tubule formation in explant culture. For this, primary MMs were treated with Wnt4-soaked beads (100 ng/ml). As shown in Fig. 1B, tubule formation and expression of tubular epithelial marker Lim1 were observed in cultures treated with Wnt4 but not in control cultures. These results confirm the biological activity of the commercially produced recombinant protein and further demonstrate that soluble Wnt4 protein can induce MET.
Studies on signal transduction in metanephric mesenchymal cells have been significantly hindered by the limited number of mesenchymal cells that can be obtained from each embryonic metanephros. To overcome this handicap, we cultured primary metanephric mesenchymes on Matrigel in serum-free medium with TGFα and Fgf2 for 10 days. This allowed us to expand 30–40 isolated MMs into 4–8 million cells without obvious loss of responsiveness to tubule induction. In order to assess the stability of expression of several developmentally relevant genes in cultured cells, we isolated RNA from uncultured MMs and 10-day cultures of MM cells explanted to Matrigel-coated dishes and performed semi-quantitative RT-PCR for expression. All of the genes examined were expressed in both uncultured (Fig. 2A, lane 1) and cultured (Fig. 2A, lane 2) cells. The genes included Six1, Six2, Cited1, Wt1, Osr1, Eya1, Sall1, Gdnf, Pax2, Hoxa11, Lim-1 and E-Cadherin. Only epithelial markers Lim1 and E-cadherin were not expressed and Sall1 and Pax2 were somewhat reduced in cultured cells. These results suggest that the expansion of these cells in culture does not dramatically alter the expression of genes associated with progenitor differentiation, nor is there an obvious selection of a sub population within the tissue. Next, the suitability for use in differentiation studies was confirmed by cocultivation with embryonic spinal cord (SC) on type IV collagen-coated filter. As shown in Fig. 2B, cultured cells, which were co-cultivated with SC, formed tubules which expressed epithelial marker Lim1. Furthermore, treatment of cells with Wnt4 protein induced the expression of epithelial markers Lim1 and E-cadherin, the condensation markers Pax2 and Pax8, and the tight-junction markers ZO-1 and occludin (Fig. 2C). In addition, co-localization of junctional adhesion protein ZO-1 and transmembrane protein occludin, which is indicative of epithelial cell tight-junction formation, were observed with Wnt4 treatment (Fig. 2D). These results confirm that cultured primary MM cells have retained the ability to undergo epithelial morphogenesis and form tubules.
To investigate the effect of Wnt4 protein on canonical Wnt signaling, a TOPflash reporter construct was transfected into MM cells. As shown in Fig. 3A, Wnt4 treatment (100 ng/ml) did not affect TCF/LEF dependent transactivation, whereas Wnt3a induced a 4–6 fold increase in reporter activity in these cells. Moreover, the canonical Wnt target gene Axin2 was induced by Wnt3a treatment but not by Wnt4 protein (Fig. 3B), indicating that Wnt4 did not stimulate TCF/LEF transactivation in this system. Wnt9b showed a modest induction of reporter activation, suggesting it may also function through a canonical Wnt mechanism, although we did not pursue this further. We then examined the effect of Wnt4 on β-catenin protein stabilization and disheveled-2 (Dvl2) phosphorylation prior to differentiation. Dvl2, which undergoes phosphorylation in response to Wnt signals that stabilize β-catenin (Rothbacher et al., 2000), and β-catenin were analyzed by immunoblotting (Fig. 3C and D). Wnt4 recombinant protein dramatically enhanced the level of E-cadherin but did not alter the levels of total β-catenin at 24 hrs or Dvl2 phosphorylation (Fig. 3C), while Wnt3a induced the accumulation of both β-catenin and Dvl2 phosphorylation within 5 min of treatment (Fig. 3D). A detectable increase in β-catenin at 48 and 72 hrs post-Wnt4 treatment was concomitant with the increased expression of E-cadherin and may result from a sequestration of β-catenin to membrane adherent complexes. Taken together, these results suggest that Wnt4 induces differentiation in a β-catenin/TCF-independent manner.
Recent reports documenting the crystal structure of β-catenin have revealed novel structural details that may permit the separation and differential regulation of adhesion activity from transcription functions (Xing et al., 2008). A C-terminal region called helix C, which represents a predicted 13th Armadillo repeat and encompasses amino acid residues 667–683, has now been identified as an essential motif for transcription but not adhesion (Xing et al., 2008). We hypothesized therefore that synthetic analogs of the helix could function as dominant-negative inhibitors of β-catenin transcriptional activity. Based on the X-ray structure of full-length β-catenin, we designed lipopeptide analogs of the C-terminal helix (Fig. 4A and B). Lipidation of the peptides was used to enhance cell penetration and structure stabilization of the antagonist, as it was shown previously to dramatically improve the biological activity of protein fragments (Remsberg et al., 2007; Timofeeva et al., 2007). Helix C is also believed to interact with β-catenin-binding proteins Chibby and ICAT, which are involved in inhibiting transcriptional activation of Wnt-responsive genes. Structural analysis of β-catenin suggests that helix C forms extensive interactions with the last armadillo repeat. Leu674, Leu678 and Leu682 fit into a hydrophobic groove formed by the second and third helices of armadillo repeat 12 (Xing et al., 2008). Lys681 is likely to form an ion pair with Glu664, while Arg684 forms a hydrogen bond with Ser646. We used these interactions for generation of peptides that are likely to compete with the native helix C in the assembly of the protein. Since negative charges of the peptides impede cell membrane penetration, we replaced Glu677 with Gln since, according to structural data, this residue is not involved in any intra-molecular interactions (CT-HC-3, Fig. 4B). To generate metabolically stable versions of peptides, we prepared retro-inverso derivatives constructed from all-D amino acids (CT-HC-5). An analysis of the three CT-HC peptides revealed significant cytotoxicity on primary MM cells following treatments with CT-HC-3 and -5 (Fig. 4C). We therefore selected CT-HC-1 for subsequent experiments. Concurrently, we designed peptides for targeting the intracellular domain of EpCAM (Fig. 4B), which interacts with β-catenin and Lef1 to form a nuclear complex that regulates β-catenin-dependent transcription (Maetzel et al., 2009). These peptides failed to inhibit both BIO- and Wnt3aCM-induced transcriptional activity (Suppl. Fig. 2) but also showed no cell toxicity (Fig. 4C and G), so they were used as negative controls in subsequent studies. On the other hand, CT-HC-1 completely blocked BIO- and Wnt3aCM-induced transcription in cultured MM cells (Fig. 4D and E). Similar results were obtained with incubations for 1, 24, and 48 hrs (only 24-hr period shown) following the single addition of CT-HC-1. Cells were also evaluated for levels of total β-catenin. In these studies, β-catenin was degraded within 3 hrs of treatment with CT-HC-1 (Fig. 4F and G), and levels had not recovered after 24 hrs (Fig. 4C). Since this suggested that the effect on transcriptional activation may be mediated through degradation of β-catenin, cells were also treated with proteasomal inhibitor MG132. As observed, β-catenin levels were stabilized even in CT-HC-1-treated cells, demonstrating that peptide activity is likely due to the accelerated degradation of its target (Fig. 4F).
To determine if the loss of β-catenin affects differentiation, MM cells were co-cultivated with inductor, embryonic spinal cord, and treated with CT-HC-1 (Fig. 4H). In these cultures, tubule formation was unaffected. This finding was supported by semi-quantitative RT-PCR analysis of BIO-induced MM cells in which levels of epithelial markers Lim-1 and E-cadherin were unaltered with CT-HC-1 treatment (Fig. 4I). Furthermore, in cells exposed only to CT-HC-1 and not an inductive ligand, peptide treatment caused a decrease in Axin-2 expression, as expected, but it also enhanced expression of Lim-1 and E-cadherin (Fig. 4J). This observation is consistent with recent reports describing an antagonistic relationship between canonical and non-canonical Wnt mechanisms (Park et al., 2007; Schmidt-Ott et al., 2007).
Since our markers of differentiation were elevated in expression following CT-HC-1 treatment, we asked whether inhibition of β-catenin is sufficient to induce tubulogenesis in MMs. Despite repeated attempts, there was no consistent evidence to support this hypothesis. MMs showed condensation with minimal evidence of tubule formation (data not shown). Since we have previously demonstrated that FGF2 (Karavanov et al., 1998), which is required to maintain MM cells, is sufficient to weakly induce tubules, we were unable to distinguish between the effects of the growth factor and our inhibitor. Regardless, the response did not resemble the robust production of tubules elicited with Wnt4 treatment.
Since Wnt4 had little effect on the canonical Wnt pathway in MM cells, we asked whether Wnt4-induced differentiation was due to the activation of a non-canonical Wnt mechanism. To address the possibility of a non-canonical Wnt mechanism, we evaluated the activity of CaMKII, a target of intracellular Ca2+ flux, by immunoblotting with an anti-phospho CaMKII antibody. Like calcium ionophores, Ionomycin and A23187, Wnt4 induced a rapid (60 sec) phosphorylation of CaMKII (Fig. 5A and B). When these cells were instead pretreated with Ca2+ chelator BAPTA-AM, Wnt4 failed to induce phosphorylation of CaMKII (Fig. 5C), indicating that Wnt4 activates CaMKII by stimulating Ca2+ influx in MM cells. To confirm whether Wnt4 directly induces Ca2+ influx in MM cells, Wnt4-mediated Ca2+ influx was examined using the Ca2+ indicator Fluo-4NW. As shown in Fig. 5D, Ca2+ influx occurred in 1 min following Wnt4 treatment and in a dose-dependent manner. A significant difference between control and Wnt4 treatment was observed at a range of Wnt4 treatments (100–500 ng/ml; Fig. 5E). These results indicate that Wnt4 stimulates a rapid Ca2+ influx in MM cells at levels that induce tubule formation.
To confirm that Ca2+ influx is sufficient to induce tubule formation, MMs were treated with 10 nM Ionomycin and cultured for 3 days. Tubular morphogenesis was readily apparent in Ionomycin-treated MMs by microscopy (Fig. 6A). Ionomycin-induced MET was also demonstrated by up regulation of differentiation markers Lim1, E-cadherin, ZO-1, occludin, Pax8, and pax2 (Fig. 6B). These results suggest that intracellular Ca2+ influx is sufficient to induce tubule formation. We also examined the effect of Ca2+ influx on stabilization of β-catenin in primary MM cells treated with Ionomycin for 24 hrs. As shown in supplementary Fig. 3, there was no significant increase in levels of β-catenin following Iononmycin treatment, whereas Wnt3a stabilized β-catenin under these conditions. These data support the hypothesis that Wnt4, through the stimulation of calcium influx, induces MM progenitors to undergo MET independent of canonical Wnt signaling.
Genetic studies have revealed an essential role for Wnt signaling in mesenchymal-epithelial transition (MET) during nephron formation. There is evidence that β-catenin is necessary for epithelial differentiation although the precise molecular mechanism is not yet clear. In the current study, we have found that Wnt4-mediated tubulogenesis in metanephric mesenchyme occurs by a non-canonical Wnt/calcium-dependent mechanism and not by a canonical β-catenin/TCF-transcription dependent mechanism. An analysis of a BATlacZ reporter mouse suggested that TCF-dependent transcription is activated in the branching UB, but not in MM. However, recombinant Wnt4 protein could induce tubulogenesis without activation of the canonical Wnt pathway. This was demonstrated by an inability of recombinant Wnt4 protein to activate a TCF reporter, induce the expression of canonical Wnt pathway target gene Axin-2, or stabilize β-catenin in MM cells. Moreover, the novel dominant-negative inhibitor for β-catenin CT-HC-1 abrogated canonical Wnt signaling yet failed to inhibit tubulogenesis. These results by exclusion then suggest that Wnt4 induces nephrogenesis through a non-canonical Wnt mechanism. In our studies, we have found that a calcium-dependent pathway is activated by Wnt4 and is involved in tubulogenesis. Calcium-dependent activities, including calcium influx and CaMKII phosphorylation, were observed in MM cells in response to Wnt4 treatment. Interestingly, the calcium ionophore Ionomycin induced tubule formation, suggesting that calcium influx in MM is sufficient to cause tubule formation.
A number of laboratories have found little reporter activity in MMs and renal vesicles in transgenic mice expressing a TCF-driven β-galactosidase reporter (Iglesias et al., 2007; Maretto et al., 2003; Schmidt-Ott et al., 2007; Schwab et al., 2007). It has been suggested that the absence of activity may be due to the insensitivity of the reporter transgene (Barolo, 2006). It is also conceivable that the TCF-driven constructs are not expressed in MMs because of the lack of certain components required for expression in specific cellular environments. We show in the present study that β-galactosidase activity can be detected in MMs from BATlacZ transgenic mice when exposed to BIO or WA, indicating that the construct is active in MMs (Suppl. Fig. 1). Interestingly, two of these articles report that β-gal activity can also be detected in S-shaped bodies, i.e., polarized epithelial structures (Iglesias et al., 2007; Schmidt-Ott et al., 2007), suggesting that TCF-dependent transcription is active after the establishment of epithelial polarity. Unfortunately, no effort was made to rule out possible staining from the fused tip of the UB in tissue sections. Park et al. (2007) showed that in β-catenin GOF mutants crossed with a conditional β-galactosidase reporter allele (R26R), β-gal positive cells condense but are E-cadherin negative and do not form tubules, indicating that MET is blocked in MMs when stabilized β-catenin is present continuously (Park et al., 2007). Similarly, MMs with stabilized β-catenin proliferate, aggregate, and escape apoptosis, but cannot complete epithelial conversion (Park et al., 2007; Schmidt-Ott et al., 2007). In light of this, our data support the notion that a β-catenin/TCF-independent mechanism, such as Wnt-induced Ca2+ influx, is required for MET and tubule formation in MM progenitors, and that the role of transcription-dependent Wnt/β-catenin signaling may be limited to the survival and proliferation of cells in the MM. Furthermore, they suggest that β-catenin signaling must become attenuated before MET can occur.
In efforts to regulate Wnt transcriptional activation, we designed and evaluated peptides that target different canonical Wnt regulators: one involving the intracellular domain of EpCAM, which interacts with β-catenin to activate transcription, and the other helix C in β-catenin, a hydrophobic region that caps the 12th armadillo repeat in β-catenin and is essential for transcriptional activation but not cell adhesion. Peptides designed to target EpCAM yielded no reduction in transcriptional activation, so these peptides were used as negative controls in subsequent studies. On the other hand, helix C peptides exhibited a striking inhibition of transcription. One of these peptides CT-HC-1 showed little toxicity and also apparently facilitated differentiation as reflected in the increased expression of tubular markers. This peptide appeared to function by inducing proteasomal degradation of β-catenin (Fig. 4F), since β-catenin levels were restored in the presence of proteasomal inhibitor MG132. This observation is consistent with a previous report that the C-terminal domain of β-catenin (CTD) interferes with its binding to Axin and APC (Choi et al., 2006) and that the CTD facilitates the accumulation of β-catenin by shielding the armadillo repeat domain from the Axin-scaffold phospho-destruction complex (Mo et al., 2009). Therefore, CT-HC-1 may cause a conformational change in β-catenin, thus allowing the formation of a destruction complex, including Axin and APC, on the armadillo repeat domain of β-catenin.
There is general agreement that the development of reagents that target the Wnt pathway is a worthwhile endeavor, considering its established role in human neoplasms, such as the pediatric renal cancer called Wilms tumor, and its involvement and interactions with a wide variety of signaling pathways. Ideally, such small molecules would permit the investigation of those interactions as well as potentially provide more effective chemotherapies for tumors that depend upon canonical Wnt signaling. A number of small molecules that target canonical Wnt/β-catenin signaling components have been identified (Chen et al., 2010). Most affect components of the canonical Wnt/β-catenin pathway, including Axin (Chen et al., 2009; Huang et al., 2009), CK2 (Gao and Wang, 2006), or the TCF/β-catenin complex (Gandhirajan et al., 2010). Recently identified XAV939 antagonizes the canonical Wnt pathway by inducing β-catenin degradation in SW480 cells; however, this is accomplished by first stabilizing axin through inhibition of tankyrase, which, in turn, stimulates β-catenin degadation (Huang et al., 2009). In a cell-based screen of some 63,000 small molecule inhibitors, Ewan et al., identified a compound, which blocked TCF-dependent transcription without affecting the degradation of β-catenin in some cell lines, suggesting that the compound acts independently of β-catenin levels (Ewan et al., 2010). In another large screen for Wnt antagonists an evaluation of ~200,000 compounds in a library yielded nine compounds that were distributed in two functional groups - those that blocked ligand production and those that blocked ligand signaling (Chen et al., 2009). The former include compounds that inhibit porcupine, an O-acyltransferase that adds a palmitoyl group to Wnts for secretion and signaling. The latter are thought to inhibit canonical Wnt signaling by direct interaction with Axin-2 to stimulate β-catenin degradation. Thus far, inhibitors discovered in screening do not directly target β-catenin, perhaps due to its critical membrane interaction with E-cadherin complexes in adherens junctions. Our C helix inhibitor appears not to disrupt junctional complexes, since tubulogenesis occurs normally, yet it completely blocked canonical Wnt signaling at levels which were not toxic to these progenitors. This peptide then may serve as an effective tool for dissecting canonical from noncanonical mechanisms and may have efficacy against tumor cells that depend upon canonical Wnt signaling, but further study is needed. Furthermore, the design of such structure-based cell permeable peptides for β-catenin may provide a useful alternative approach for the discovery of molecules that target specific Wnt-dependent processes, in lieu of costly high-throughput screening methods.
Previous studies have revealed a role for the Wnt/calcium pathway in zebrafish and Xenopus embryos. Overexpression of Wnt5a can trigger intracellular Ca2+ influx, leading to the activation of Ca2+-dependent effector molecules such as CaMKII (Kuhl et al., 2000a; Moon and Shah, 2002). It has been proposed that active CaMKII protein can then initiate a protein kinase signaling cascade, resulting in NLK-mediated phosphorylation of TCF/LEF transcription factors. This phosphorylation of TCF/LEF prevents the β-catenin TCF/LEF transcriptional complex from binding to DNA (Ishitani et al., 2003; Ishitani et al., 1999). In this model, inhibition of Wnt/β-catenin signaling due to Wnt5a-stimulated Ca2+ influx occurs downstream of β-catenin stabilization and at the level of TCF-mediated transcription. Based on these findings, our data suggest that Wnt4 might induce a Ca2+-mediated phosphorylation of CaMKII, and, thus, in turn, terminate canonical Wnt signaling in nephrogenesis. Additionally, since snail is a target of canonical Wnt signaling and since it inhibits E-cadherin expression (Batlle et al., 2000; Cano et al., 2000; Nieto, 2002), a Wnt4-induced attenuation of the canonical Wnt pathway may be sufficient to induce E-cadherin expression and facilitate MM morphogenesis.
Alternatively, an increase of intracellular Ca2+ leads to activation of a calcineurin-dependent pathway including the transcription factor Nuclear Factor of Activated T cells (NFAT), which is activated in most immune system cells to stimulate cytokine transcription and is involved in the development and morphogenesis of a variety of vertebrate tissues (Medyouf and Ghysdael, 2008). In unstimulated cells, NFATc transcription factors are highly phosphorylated and sequestered to the cytoplasm. When a stimulus results in an increase in intracellular Ca2+, the heterodimeric serine/threonine phosphatase calcineurin dephosphorylates NFATc, resulting in an allosteric switch that exposes a nuclear localization sequence and conceals a nuclear export sequence. NFATc then translocates to the nucleus and binds to specific regions in the promoters of target genes. The activity of NFAT proteins is tightly regulated by the calcium/calmodulin-dependent phosphatase calcineurin, a primary target for inhibition by cyclosporine A (CSA) and FK506 (Beals et al., 1997a). The nuclear import and activation of NFATc is opposed by rephosphorylation of NFATc by GSK3 (Beals et al., 1997b) or CKI (Okamura et al., 2004). The role of a calcineurin-dependent pathway in developing kidney has been suggested using CSA in mouse embryonic kidney. Treatment of mouse embryonic kidneys with CSA inhibits proliferation in embryonic kidneys in culture (Alcalay et al., 2007). Furthermore, calcineurin A knockout mice display renal hypoplasia and postnatal lethality (Gooch et al., 2004). Inhibition of calcineurin with CSA treatment in rat embryonic kidneys reduced nephron formation, suggesting that a calcineurin-dependent pathway is involved in nephrogenesis (Merlet-Bénichou et al., 1996). In addition, dominant-negative NFAT inhibits canonical Wnt signaling by suppressing β-catenin expression in Xenopus (Saneyoshi et al., 2002). In our study, we have noted that Wnt4 activates an NFAT-dependent promoter as does Ionomycin treatment in primary MM cells (Suppl. Fig. 4). Burn and colleagues have also found that a non-canonical calcium/NFAT pathway functions downstream of Wnt4 in nephrogenesis (Burn et al., 2011). In these studies, they demonstrate that all five NFAT family members are expressed in the developing mouse metanephros and that inhibtion of calcium/NFAT signaling reduces nephron formation. Furthermore, they were able to rescue nephron formation in Wnt4−/− metanephroi with the calcium ionophore Ionomycin. Therefore, a Wnt4-induced Ca2+ influx may trigger the calcineurin/NFAT pathway; however, the role of this pathway on tubule formation is not yet clear. Calcineurin dephosphorylates not only NFAT, but also Cux-1, which is highly expressed in the nephrogenic zone of the developing kidney (Vanden Heuvel et al., 1996) and functions as a transcriptional repressor of the cyclin kinase inhibitors p21 and p27 (Coqueret et al., 1998; Ledford et al., 2002). Ectopic expression of Cux-1 in transgenic mice causes renal hyperplasia (Ledford et al., 2002). Therefore, the calcineurin-dependent pathway may function as a general regulator of cell proliferation during tubule formation. In our study, we have revealed that Ca2+ influx in MM is sufficient to induce tubule formation. However, the precise downstream targets of calcium-dependent signaling, including CaMKII and calcineurin still remain undefined and future studies are required to elucidate their roles.
In summary, our work provides a better understanding of the mechanism of MET in MM, implicating the non-canonical Wnt/calcium pathway. Particularly, Wnt4-induced Ca2+ influx plays a crucial role in tubulogenesis, and such observations contribute to our understanding of the mechanism and potential connection between canonical Wnt and non-canonical Wnt signaling during early kidney development.
Isolated MMs from BATLacZ reporter transgenic mouse were treated with the GSK3 inhibitor BIO or LiCl, and β-gal activity was detected.
TOPflash reporter constructs were transfected into primary MM cells, using an Amaxa nucleofector system. Twenty-four hrs after transfection, MM cells were pre-treated with EpCAM peptide inhibitors (1, 5 and 10 μM) for 30 min, and then exposed to BIO (500 nM) (D) or Wnt3a CM (50%) (E) for 24 hr. CT-HC-1 (1–10 μM) was used as a positive control. Luciferase activities were measured by a Dual reporter assay system and normalized by renilla activity. Each value represents the mean ± SD of three or four separate experiments. **p < 0.01 vs Wnt3aCM or BIO-induced luciferase activity.
Ca2+ influx does not affect β-catenin stabilization. Primary MM cells were treated with Ionomycin (100 nM) for 24 hr and were then incubated with Wnt3a (100 ng/ml) for 12 hr. Cells were harvested with SDS sample buffer and analyzed by immunoblotting with β-catenin antibodies.
pGL4.30-NFAT luciferase plasmid and pGL4.74-TK-renilla (Promega Corp., Madison, WI) were transfected into primary MM cells using an Amaxa Nucleofector system. Twenty-four hrs after transfection, cells were treated with Wnt4 recombinant protein (10–100 ng/ml) or Ionomycin (1μM) and cultured for an additional 18 hr. Luciferase activities were measured by a Dual reporter assay system and normalized to renilla activity.
We are grateful to Dr. Joost Oppenheim, Dr. O.M. Zack Howard, and Nathan Kadan (Laboratory of Molecular Immunoregulation, NCI -Frederick) for Calcium influx experiments. This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research. This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.