Suppression of Cyclic GMP-dependent Protein Kinase Is Essential to the Wnt/cGMP/Ca2+ Pathway*

Novel downstream effectors sensing changes in intracellular concentrations of Ca2+ and cyclic GMP in response to activation of the Wnt/Frizzled-2 pathway were sought. Activation of Frizzled-2 suppressed protein kinase G activity while activating NF-AT-dependent transcription. Each of these responses was abolished by pertussis toxin and by knock-down of the expression of either Gαt2 or Gαo. Activation of NF-AT-dependent transcription in response to Wnt5a stimulation was suppressed by activation of protein kinase G and by buffering intracellular Ca2+. Elevation of intracellular cyclic GMP either by inhibition of cyclic GMP phosphodiesterase or by addition of 8-bromocyclic GMP was shown to activate protein kinase G, to block Ca2+ mobilization, as well as to markedly attenuate activation of NF-AT-dependent transcription in response to Wnt5a stimulation. Chemical inhibition of protein kinase G by Rp-8-pCPT-cGMP, conversely, was shown to provoke increased NF-AT gene transcription and Ca2+ mobilization in the absence of Wnt stimulation. Protein kinase G is shown to be a critical downstream effector of the noncanonical Wnt-Frizzled-2/cGMP/Ca2+ pathway.


MATERIALS AND METHODS
Cell Culture-Mouse F9 teratocarcinoma cells were obtained from the ATCC collection (Manassas, VA). The cells were propagated and maintained in Dulbecco's modified Eagle's medium supplemented with 15% heat-inactivated fetal bovine serum (Hyclone, South Logan, UT) at 37°C in a 5% CO 2 incubator. Clones stably transfected with pcDNA3 harboring either Rfz2 or ␤ 2 AR/Rfz2 or co-transfected with NF-AT-sensitive luciferase reporter gene (NFAT-Luc) (Stratagene, La Jolla, CA) were selected in medium containing neomycin analogue, G418 (0.4 mg/ml; Invitrogen, Carlsbad, CA). At least three independent clones were propagated for each transfection and maintained in Dulbecco's modified Eagle's medium supplemented with 15% fetal bovine serum, 100 units/ml penicillin, 0.1 mg/ml streptomycin, and 0.1 mg/ml G418. The H7 clone of human embryonic stem cells (hESC) was obtained from the WiCell Research Institute. H7 cells were cultured on mytomycin C-treated mouse embryonic fibroblasts in Dulbecco's modified Eagle's medium/Ham's F-12 medium supplemented with 20% knock-out serum replacement, 1% nonessential amino acids, 1 mM L-glutamine, 0.1 mM ␤-mercaptoethanol, and 4 ng/ml basic * This work was supported by Grant GM 069375 from the NIGMS, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 To whom correspondence should be addressed. Tel.: 631-444-6317; Fax: 631-444-3432; E-mail: wangh@pharm.stonybrook.edu.
fibroblast growth factor (all from Invitrogen). Purified mouse recombinant Wnt5a (645-WN) and anti-Wnt5a antibody (AF645) were purchased from R & D Systems (Minneapolis, MN). Assay of PKG Activity-PKG activity was determined as described previously with modifications (15). Briefly, supernatants (20 l) from the crude whole cell fractions were employed with 200 mg/ml PKG-specific heptapeptide substrate (RKRSRAE; Bachem, San Siego, CA), 50 M ATP, and 2.5 Ci of [␥-32 P]ATP in TMG buffer containing 20 mM Tris-HCl, pH 7.4, 20 mM magnesium acetate, 10 mM glycerophosphate, 100 nM okadaic acid, and a protease inhibitor mixture (10 g/ml leupeptin, 10 g/ml aprotinin, and 100 M phenylmethylsulfonyl fluoride). The reaction mixtures were incubated at 37°C for 20 min, and the reaction was terminated by the addition of 4 l of 2 N HCl. Preboiled or HCl-treated supernatants were used as blank samples. Twenty microliters of the resultant mixture were spotted onto P-81 phosphocellulose filters. Air-dried P-81 filters were washed with H 3 PO 4 (75 mM) three times. Incorporated ␥-32 P in substrates were measured by liquid scintillation spectrometry. The samples were assayed in triplicate. Kinase activity is normalized with protein concentration and presented in percentages of PKG activity of treatment group to that of the control group.
Cytoplasmic Calcium Measurement-Mouse F9 cells were plated on collagen-coated coverslips inserted in 35-mm dishes (MatTek Corporation, Ashland, MA) and cultured overnight. hESC free of contaminating mouse embryonic fibroblasts were cultured on Matrigel-coated (BD Biosciences) coverslips in medium conditioned by mouse embryonic fibroblasts and supplemented with basic fibroblast growth factor. The intracellular Ca 2ϩ was measured as described (16). Briefly, coverslip-attached cells were loaded with 2 M Fura-2 acetoxymethyl ester (Molecular Probes, Inc., Eugene, OR) for 40 min at 37°C in the dark. The cells were then washed three times with Krebs buffer composed of 128 mM NaCl, 5 mM KCl, 77.5 mM NaH 2 PO 4 , 1.3 mM MgSO 4 , and 1.3 mM CaCl 2 and kept in the same buffer at room temperature for 10 min in the dark. Fura-2-loaded cells were treated with isoproterenol (10 M) or with Wnt5a (25 ng/ml) or buffer alone. The cells were immediately monitored via a Hamamatsu OREA ER-AG digital CCD camera coupled to a Nikon inverted microscope. Excitation at 340 and 380 nm were generated using a 175-watt ozone-free xenon lamp and a Fura-2 fluorescence filter set coupled to a Sutter Lambda DG-4 high speed filter switcher. The emitted light was channeled into the CCD camera via an ultraviolet dichroic mirror and an interference filter of 510 mm. Measurement of fluorescence intensity was performed at a rate of 200 points/s, and the ratio of absorbance at 340/381 nm (A 340 /A 380 ) was computed using Dynamic Intensity Analysis (Compix Inc., Cranberry Township, PA).
Treatment of Cells with Morpholino Antisense Oligonucleotides-Morpholino phosphorodiamidate antisense oligonucleotides targeting the translational initiation sites of G␣ t2 , G␣ t1 , and G␣ o were purchased from Gene Tools (Corvallis, OR). The sequences of morpholinos for G␣ t2 , G␣ t1 , and G␣ o are as follows: CACTCCCCATTTCTGCTGTCTCCTC is for G␣ t2 , CTCCCCGGCCTCCTCAGACGACCTT is for G␣ t1 , and CCTCTGCGCTCAGCGTACATCCCAT is for G␣ o . F9 cells (wild-type or stably transfected with the desired genes according to the requirements of the experiments) were treated with morpholino antisense oligonucleotides according to the manufacturer's Special Delivery protocol. Briefly, the cells were seeded onto 60-mm culture dishes. The cells were 80% confluence on the day morpholino treatment was applied. Growth medium was replaced by the serum-free medium containing 1.4 M morpholino antisense oligonucleotides and 0.6 M ethoxylated polyethylenimine delivery reagent. After 3 h of incubation at 37°C, the cells were grown in standard culture medium with serum, but free of morpholinos, for an additional 72 h before the expression levels of target proteins were assessed by immunoblotting. For cultures that were subsequently subjected to further PKG or NF-AT activity assays, the cells were treated either with isoproterenol or Wnt5a or not (control).
Immunoblotting-Whole cell lysates were used for detecting PKG and G protein subunits. The samples (30 g of protein/ lane) were subjected to SDS-PAGE for separation. The resolved proteins were transferred onto nitrocellulose membranes. The blots were probed with antibodies against PKGI␤, PKGI␣/␤ (Stressgen Biotechnologies, Victoria, Canada), or the indicated G protein subunits (anti-G␣ t1 and anti-G␣ t2 from Santa Cruz Biotechnology, Santa Cruz, CA; anti-G␣ o from Chemicon, Temecula, CA) and followed by incubation with a corresponding peroxidase-conjugated secondary antibody (anti-mouse from Santa Cruz Biotechnology; anti-rabbit from Kirkegaard and Perry Laboratories, Gaithersburg, MD). Immune complexes were detected by enhanced chemiluminescence method, as per the manufacturer's instructions. Equal loading was verified by probing with antibodies to ␤-actin (Sigma-Aldrich).
NF-AT-sensitive Luciferase Assay-F9 clones stably co-transfected with NF-AT reporter gene and with expression vector harboring either Rfz2 or ␤ 2 AR/Rfz1 chimera were cultured on 12-well plates. The cells were starved from serum for overnight and then were stimulated with Wnt5a (25 ng/ml) or isoprotenerol (10 M) for 5 h or indicated periods. The cells were lysed with 1ϫ luciferase cell culture lysis reagent (Promega, Madison, WI). Supernatants from cell lysates were subjected to luciferase assay according to the manufacturer's instruction (Stratagene, La Jolla, CA). Briefly, 20 l of supernatant were mixed with 100 l of luciferase assay buffer (40 mM Tricine, pH 7.8, 0.5 mM ATP, 10 mM MgSO 4 , 0.5 mM EDTA, 10 mM 1,4-dithiothreitol, 0.5 mM coenzyme A, and 0.5 mM luciferin), and the intensity of luminenscence was immediately measured using a luminometer (Lumat LB 9507; Berthold Technologies, Oak Ridge, TN). Samples were assayed in triplicate, and the luciferase activity was normalized based on protein concentrations of the cell lysate. The results are presented as ratios of relative light units of treatment groups to control groups.
Statistical Analysis-The experiments were conducted at least in triplicate. All of the data are expressed as the means Ϯ S.E. from at least three separate experiments. Comparisons of data among groups were performed with one-way analysis of variance followed by the Newman-Keuls test. Statistical significance (p value of less than 0.05) is denoted with asterisks or pound symbols.

Ca 2ϩ Mobilization in Response to Frizzled-2 Activation in
Mouse Embryonic F9 Cells-The mouse F9 embryonic teratocarcinoma (F9) cells are totipotent cells that were shown earlier to be capable of being transfected with ␤ 2 -adrenergic receptor/ Frizzled chimeras or authentic Frizzleds that, when activated, promote these embryonic cells in culture to form a primitive endoderm-like phenotype characteristic of early stages in mouse development (2,13). These cells express a modest number of ␤ 2 -adrenergic receptors. Frizzled-2 expression by wildtype F9 cells, if any, is insufficient to enable Wnt5a-stimulated activation of the cyclic GMP/Ca 2ϩ pathway.
Mouse F9 cells stably expressing the ␤ 2 -adrenergic/rat Frizzled-2 chimera (␤ 2 AR/Rfz2), a chimera that binds ␤-adrenergic agonist but stimulates the Frizzled-2 pathway (17), were loaded with the Ca 2ϩ -sensitive dye Fura-2 and treated with the isoproterenol (10 M; Fig. 1A). In these embryonic cells, activation of the Frizzled-2 chimera provokes an increase in intracellular concentrations of Ca 2ϩ within 60 s that can be visualized by fluorescence microscopy and monitored in real time as the change in the ratio of absorbance of light at 340 nm versus that at 380 nm, that is characteristic for Ca 2ϩ -activated Fura-2. The kinetics of the Ca 2ϩ response agrees well with earlier measurements assayed by spectroscopy (17). The Frizzled-2 stimulation of Ca 2ϩ mobilization plateaus within 3 min and is maintained for up to 10 min in the continued presence of agonist stimulation. Stimulating zebrafish embryos expressing ␤ 2 AR/Rfz2 chimera with isoproterenol leads to Ca 2ϩ transients (13), identical in character to those stimulated by Wnt5a alone (12). Wildtype (WT) and F9 cells stably transfected with an "empty" expression vector display no Ca 2ϩ response to activation of the Frizzled-2 pathway (Fig. 1A, inset).
The pharmacological properties of the Ca 2ϩ response of F9 cells expressing the Frizzled-2 chimera to ␤-adrenergic agonist was probed further. The increase in intracellular Ca 2ϩ stimulated by activation of the Frizzled-2 chimera with agonist could be abolished by simultaneous treatment with the ␤-adrenergic antagonist propranolol (ϩPro, 10 M) (Fig. 1B). Expression of this Frizzled-2 chimera in zebrafish embryos by-passed the need for purified Wnt5a, allowing the use of isoproterenol to activate the noncanonical pathway, a response that also was sensitive to antagonism by propranolol (13). The Frizzled-2 activation of Ca 2ϩ mobilization was sharply attenuated by the pretreatment of the F9 cells with pertussis toxin (Fig. 1B), an agent that catalyzes the ADP-ribosylation and inactivation of G protein ␣-subunits of the G␣ i family, including G␣ i1,2,3 , G␣ o , and G␣ t (18 -20).
We compared the Ca 2ϩ response stimulated by purified Wnt5a using F9 cells expressing the authentic rat Frizzled-2 ( Fig. 1C) to the response stimulated by the activation of the Frizzled-2 chimera (Fig. 1B). Wnt5a stimulates a mobilization of intracellular Ca 2ϩ in the F9 clones expressing Frizzled-2 ( Fig.  1C). Application of Wnt3a, which can activate Frizzled-1 but not Frizzled-2 receptors, in contrast, did not stimulate the Ca 2ϩ response (data not shown). Differences in the kinetics of the stimulated Ca 2ϩ mobilization between the clones expressing Frizzled-2 versus the ␤ 2 AR/Rfz2 chimera are apparent (Fig. 1, B and C). First, the kinetics of the Ca 2ϩ response to Wnt5a stimulation was biphasic, whereas the response to stimulation of the chimeric receptor was not. Second, the kinetics of Ca 2ϩ response to Wnt5a stimulation was not as rapid as that of F9 cells expressing the chimera and stimulated by ␤-adrenergic agonist. The slower and atypical kinetics of the Wnt5a-stimulated response are likely to be a reflection of the palmitoylated, glycoprotein nature of the Wnt5a, which tends to form micelles in slow equilibrium with the bulk solution and Frizzled receptors (21). Pertussis toxin treatment nearly abolished Wnt5astimulated Ca 2ϩ mobilization for F9 cells expressing the Frizzled-2, as it did for clones expressing the ␤ 2 AR/Rfz2 chimera stimulated by isoproterenol (Fig. 1, B and C). These experiments are the first to demonstrate a bona fide, pertussis toxinsensitive Ca 2ϩ mobilization of mammalian embryonic cells in culture, following activation of the Frizzled-2 pathway. With this model system, detailed biochemical analysis of the downstream signaling of the Wnt/Ca 2ϩ /cGMP that is not possible in zebrafish embryos was enabled in F9 cells. Finally, we tested in another embryonic cell line, H7 hESC in culture, whether purified Wnt5a regulated the intracellular Ca 2ϩ concentration, using the same approach adopted for F9 cells. Although not fully characterized, hESC do express Frizzled-2 receptor (data not shown). Treating H7 hESC with Wnt5a stimulated an increase in intracellular Ca 2ϩ levels ( Fig. 1D), which agrees with the response observed in F9 cells (Fig. 1C). The addition of excess anti-Wnt5a antibody to the Wnt5a mixture neutralized the ligand, blocking the Wnt5a-stimulated Ca 2ϩ response (Fig. 1D).
Frizzled-2 Regulates NF-AT-dependent Transcriptional Activation-Whereas the canonical Wnt/␤-catenin pathway signals to the level of Lef/Tcf-sensitive transcriptional activation (5), the noncanonical Wnt/Ca 2ϩ pathway has been reported in Xenopus embryos to activate a group of developmentally important genes, including those dependent upon the nuclear transcription factor of activated T-cells (NF-AT) (11). Using a luciferase-based reporter gene construct to assay NF-AT-dependent gene transcription, we set out to test whether the Frizzled-2 signaling downstream of Ca 2ϩ mobilization and cyclic GMP extended in mammalian cells to the level of NF-AT-sensitive transcription. NF-AT-dependent transcription indeed was activated in F9 clones expressing the Frizzled-2 chimera treated with ␤-adrenergic agonist ( Fig. 2A). The transcriptional response to Frizzled-2 activation was first observed at 2 h poststimulation, with a peak response noted within 5 h. The NF-AT transcriptional response remained elevated for more than 5 h, thereafter declining to near basal levels within 20 h. For F9 cells expressing the ␤ 2 AR/Rfz2 chimera, treatment with ␤-adrenergic antagonist propranolol effectively blocks the Frizzled-2-mediated activation of the NF-AT-dependent transcription in response to isoproterenol, as it does the Ca 2ϩ mobilization (Fig. 1B).
If the results obtained from cells expressing the ␤ 2 AR/Rfz2 chimera accurately reflected Frizzled-2 action, then treatment of F9 clones stably expressing authentic rat Frizzled-2 (Rfz2) withpurifiedWnt5awouldbeexpectedalsotostimulateNF-ATdependent transcription (Fig. 2B). Wnt5a stimulates a tran-scriptional response in the Rfz2-expressing clones, but not in those clones expressing the empty expression vector. The NF-AT-dependent transcriptional response to Wnt5a stimulation displays kinetics (Fig. 2B) in good agreement with those obtained in F9 cells expressing the ␤ 2 AR/Rfz2 chimera ( Fig.  2A).
The NF-AT-dependent transcriptional response was tested for specificity. Wnt5a antibody, which neutralizes the ligand, FIGURE 1. Activation of Frizzled-2 promotes intracellular Ca 2؉ mobilization via G proteins. A, WT F9 cells, F9 clones transfected with pCDNA 3 empty vector (EV), and F9 clones expressing the ␤ 2 AR/Rfz2 chimera were treated with the ␤-adrenergic agonist isoproterenol (Iso, 10 M). The concentration of intracellular free Ca 2ϩ measurements was measured as described under "Materials and Methods." The upper panel shows artificial color images of free Ca 2ϩ in WT and ␤ 2 AR/Rfz2 chimera-expressing clones. The green is designed as the fluorescence emitted from excitation at 380 nm, and red is designed as it emitted from excitation at 340 nm. The data were analyzed and presented as ratios of absorbance intensity of 340 over 380 nm, tracked in real time (lower panel). The data are representative of multiple imaging experiments performed on at least three separate clones. B, WT and ␤ 2 AR/Rfz2 chimera-expressing clones were pretreated with either buffer alone or pertussis toxin (10 ng/ml, ϩPTX) for 2 h, and the cells were loaded with Fura-2 during the last 45-min incubation. The cells were washed and stimulated with isoproterenol (10 M) in the presence or absence of the ␤-adrenergic antagonist propranolol (10 M, ϩPro). The intracellular free Ca 2ϩ was measured immediately. C, WT and rat Frizzled-2 (Rfz2)-expressing clones were pretreated with either buffer alone or pertussis toxin (10 ng/ml, ϩPTX) for 2 h, and the cells were loaded with Fura-2 during the last 45 min of incubation. The cells were washed and stimulated with Wnt5a (25 ng/ml). The intracellular free Ca 2ϩ was measured immediately. D, H7 hESC were treated with Wnt5a (25 ng/ml) or Wnt5a in a mixture with excess anti-Wnt5a antibody (AbϩWnt5a; 0.2 g/ml purified antibody). Intracellular Ca 2ϩ concentration was monitored by Fura-2 imaging. The data are representative of multiple experiments performed on at least three separate clones.
abolishes Wnt5a-stimulated activation of NF-AT-dependent transcription in rat Frizzled-2-expressing cells (Fig. 2C). F9 clones expressing the ␤ 2 AR/Rfz1 chimera were compared with the ␤ 2 AR/Rfz2 chimera-expressing clones for NF-AT transcriptional activation in response to stimulation with ␤-agonist (Fig. 2D). Treatment with the ␤-adrenergic agonist isoproterenol stimulates activation of the NF-AT-dependent transcription in clones expressing the ␤ 2 AR/Rfz2 chimera. The F9 clones expressing the ␤ 2 AR/Rfz1 chimera, in sharp contrast, displayed little NF-AT-sensitive transcriptional activation, i.e. the same level of NF-AT-dependent transcription observed in the WT or empty vector transfected F9 cells (data not shown). These data agree with the observations that Wnt5a, but not the Frizzled-1 agonist ligand Wnt3a, stimulates the intracellular Ca 2ϩ response in the F9 clones expressing authentic Rfz2 (data not shown). A, mouse F9 cells were stably co-transfected with ␤ 2 AR/Rfz2 and NF-AT luciferase reporter gene (NF-AT-Luc). Stable clones were treated with isoproterenol (ϩIso, 10 M) in the presence or absence of the ␤-antagonist propranolol (ϩPro, 10 M) for periods indicated, and the activity of NF-AT-dependent luciferase reporter was measured. *, p Ͻ 0.001 versus control (ϩIso at time 0); #, p Ͻ 0.001 versus the corresponding ϩIso group. B, F9 cells were stably co-transfected with NF-AT-Luc and either rat Frizzled-2 (Rfz2) or its empty vector (EV). Stable clones were treated with Wnt5a for periods indicated, and the activity of NF-AT-dependent luciferase reporter was measured. C, mouse F9 clones stably expressing rat Frizzled-2 were transfected with NF-AT-Luc. The cells were treated with (ϩWnt5a, 25 ng/ml) or without Wnt5a (ϪWnt5a) for 6 h. Parallel experiments were performed with Wnt5a in a mixture of vehicle (Ϫ) alone or a mixture with anti-Wnt5a antibody at an antibody concentration of 0.1 and 0.2 g/ml. ***, p Ͻ 0.001, versus control ϪWnt5a, Ϫanti-Wnt5a group; ###, p Ͻ 0.001, versus ϩWnt5a, Ϫanti-Wnt5a antibody group. D, mouse F9 cells were transfected with NF-AT-Luc alone or co-transfected with expression vectors harboring either the ␤ 2 AR/Rfz1 or ␤ 2 AR/Rfz2 chimera. The expressions of ␤ 2 AR/Rfz1 or ␤ 2 AR/Rfz2 chimeric receptors analyzed by radioligand binding assay were comparable (␤ 2 AR/Rfz1, 12.3 pmol/mg membrane protein; ␤ 2 AR/Rfz2, 11.7 pmol/mg membrane protein) (30). The cells were stimulated with isoproterenol (ϩIso, 10 M) for 6 h, and the luciferase activity assay was performed using a whole cell lysate to determine NF-AT-dependent transcriptional activity. ***, p Ͻ 0.001. E, F9 clones expressing ␤ 2 AR/Rfz2 chimera were treated with pertussis toxin (PTX, 10 ng/ml for 2 h) or with morpholino antisense oligonucleotides (Antisense) for 72 h to knock down G␣ t1 , G␣ t2 , or G␣ o individually. The cells were then stimulated by isoproterenol (Iso, 10 M) for 5 h, and the luciferase activity assay was performed using a whole cell lysate to determine NF-AT-dependent transcriptional activity (upper panel). The protein levels of the designated G protein ␣ subunits were assessed by immunoblotting (IB) to test the efficiency of the antisense morpholino-induced knock-down, and ␤-actin was probed by specific antibody to ensure the equal loading of protein samples (lower panel). ***, p Ͻ 0.001 versus the control group; ###, p Ͻ 0.001 versus the ϩIso alone group. F, F9 cells were co-transfected with ␤ 2 AR/Rfz2 chimera and NF-AT luciferase reporter gene. The clones were treated either with buffer (Ϫ) or with BAPTA for 30 min prior stimulation by isoproterenol (ϩIso). The cells were lysed after 6 h of stimulation and assayed for NF-AT-dependent luciferase activity. The results are presented as the mean values Ϯ S.E. from at least three separate experiments. ***, p Ͻ 0.001.
We explored whether or not G proteins were mediating the Frizzled-2 signaling to the level of NF-AT transcriptional activation. As was observed for the Wnt5a/Ca 2ϩ response, pertussis toxin pretreatment largely abolished the NF-AT-dependent transcriptional response (Fig. 2E). We used morpholino antisense oligonucleotides to suppress the expression of the two pertussis toxin G protein ␣-subunits, G␣ t2 and G␣ o , previously shown to mediate Wnt regulation of Ca 2ϩ mobilization and activation of cyclic GMP PDE (13). The antisense morpholinos were able to knock down the expression of the targeted G protein subunit by more than 75% (Fig. 2E, insets, and Table 1). Knock-down of either G␣ t2 or G␣ o , but not G␣ t1 (control), with antisense morpholinos attenuates the Frizzled-2-mediated activation of NF-AT-dependent transcription (Fig. 2E). Taken together these data demonstrate that NF-AT transcriptional activation by Frizzled-2 is a G protein-mediated downstream readout of the Wnt/Ca 2ϩ /cGMP pathway.
We performed similar diagnostic tests on the role of intracellular Ca 2ϩ on Frizzled-2 regulation of NF-AT-sensitive transcriptional activation. We made use of the membrane-permeable form of a Ca 2ϩ -buffering agent, 1,2-bis(o-aminophenoxy)ethane-N,N,NЈ,NЈ,-tetraacetic acid tetra(acetoxymethyl) ester (BAPTA), to neutralize Frizzled-2-stimulated changes in intracellular Ca 2ϩ levels. Preloading the cells with the acetoxymethyl ester of BAPTA buffers intracellular Ca 2ϩ and was found to abolish the Frizzled-2 activation of the NF-AT transcription (Fig. 2F). The levels of "basal" NF-AT transcription, in contrast, were largely unaffected by buffering intracellular Ca 2ϩ with BAPTA. Thus, Ca 2ϩ mobilization in response to activation of Frizzled-2 is upstream to and obligate for the expression of NF-AT transcriptional regulation by the Wnt/Ca 2ϩ /cGMP pathway.

Noncanonical Wnt/Ca 2ϩ /cGMP PDE Signaling Regulates
Protein Kinase G-Earlier it was shown that the Wnt/Frizzled-2 pathway includes not only Ca 2ϩ mobilization but also activation of cyclic GMP-sensitive PDE6 (14), catalyzing a sharp decline in intracellular concentrations of cyclic GMP (13). The intracellular sensor(s) for cyclic GMP in the Wnt/Ca 2ϩ /cGMP pathway is (are) not known. Likewise, whether or not the regulation of the Ca 2ϩ mobilization and of the intracellular concentrations of cyclic GMP reflects common or distinct signaling pathways downstream of Wnt5a/Frizzled-2 is unknown. A prime candidate effector worthy of analysis is the cyclic GMPdependent serine/threonine protein kinase, PKG. PKG was tested as a possible downstream effector by measuring its activity directly. F9 clones stably expressing the Frizzled-2 chimera were stimulated with ␤-adrenergic agonist, and PKG activity was assayed. If PKG is an effector for the Frizzled-2 pathway, activation of Frizzled-2/PDE signaling would be expected to reduce PKG activity. PKG activity was found to be reduced in response to activation of the Frizzled-2 pathway (Fig. 3A). The decline in PKG activity mimics the decline of the intracellular concentration of cyclic GMP provoked by activation of the Wnt/Ca 2ϩ /cGMP pathway (13). The decline in PKG activity in response to Frizzled-2 activation was not reflected at the level of PKG mRNA, because reverse transcription-PCR amplification of the cellular RNA revealed no change in PKG mRNA abundance over this time frame (data not shown). At the protein level, the expression of I␣ and I␤ subunits of PKG measured by immunoblotting of whole cell lysates of F9 cells also revealed no change in PKG subunit abundance (Fig. 3A). Thus, activation of Frizzled-2 stimulates a decline in PKG activity, but not in its abundance. We examined whether the regulation of PKG activ-

Efficiency of knockdown of G-protein subunits targeted by antisense morpholinos
The abundance of each subunit in the absence of antisense morpholino treatment is set to a value of 1.00. The results displayed are the mean values Ϯ S.E. from four or more separate experiments. An additional control, measurement of ␤-actin, provided an independent measure of the loading uniformity in the immunoblotting of whole-cell lysates from F9 cells. See the text for details. *, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001. OCTOBER 13, 2006 • VOLUME 281 • NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 30995 ity by Frizzled-2 activation was readily reversible, making use of the ␤ 2 AR/Rfz2 chimera-expressing cells, in which isoproterenol can activate and propranolol antagonize the Frizzled-2 pathway (2,13,17). The activation of Frizzled-2 chimera by isoproterenol was maintained for 30 min, with a corresponding decline in PKG activity (Fig. 3B). At that point, propranolol was added to half of the tubes to block ␤-adrenergic agonist stimulation of Frizzled-2 chimera, and the incubation continued for 90 additional min. The addition of propranolol antagonizes and fully reverses within 90 min the decline in PKG activity caused by the stimulation of the Frizzled-2 pathway (Fig. 3B).

Wnt Signaling via Protein Kinase G
As noted above, inhibition of PKG activity for F9 clones expressing the Frizzled-2 chimera in response to the ␤-agonist isoproterenol is sensitive to blockade by the ␤-adrenergic antagonist propranolol (Fig. 3C). The role of G proteins in development was first detected by the ability of pertussis toxin, which inactivates members of the G i family of heterotrimeric G proteins, to block key pathways. We tested whether pertussis toxin treatment also blocks the ability of Frizzled-2 chimera to activate the downstream events that culminate in the inhibition of PKG activity (Fig. 3C). Pertussis toxin abolishes the ability of Frizzled-2 to reduced PKG activity, much like it did for the Ca 2ϩ mobilization response stimulated by the activation of Frizzled-2 (Fig. 1B).
We tested these primary observations in F9 clones expressing authentic rat Frizzled-2 and stimulated with purified Wnt5a. Wnt5a treatment stimulates a frank decrease in PKG activity (Fig. 3D), in good agreement with the data derived from cells expressing the Frizzled-2 chimera (Fig. 3, A and C). Furthermore, Frizzled-2-mediated suppression of PKG activity in response to Wnt5a was blocked by pretreatment of the cells with pertussis toxin (Fig. 3D). These observations suggest that members of the G i family mediate Wnt5a signaling downstream to PKG. If this premise is correct, then knock-down of the G proteins implicated in Frizzled-2 signaling (as shown in Fig. 2E) should also abolish the ability of Frizzled-2 signaling to regulate PKG activity. Knock-down studies of G proteins involved in the Wnt/Frizzled-2/Ca 2ϩ /cGMP pathway revealed that G␣ o and G␣ t2 , but not G␣ t1 (control), were essential for the reduction of PKG activity in response to activation of Frizzled-2 (Fig. 3E).
To further test these observations, we investigated the ability of knock-down of selective G protein ␣ subunits to block Wnt5a-induced reduction in PKG activity in F9 cells stably expressing the authentic Rfz2 (Fig. 3F). Knock-down with antisense morpholinos specific for the endogenous G␣ t2 blocks the ability of Wnt5a to reduce PKG activity. Having attenuated the ability of Wnt5a to regulate PKG by knock-down of endogenous G␣ t2 , we performed an experiment designed to try to rescue the Wnt5a-stimulated response by expressing G␣ t2 exogenously in the knock-down cells. The G␣ t2 mRNA expressed by the expression vector lacks the 5Ј-untranslated region targeted by the antisense morpholinos. Expression of exogenous G␣ t2 rescued the Wnt5a-stimulated regulation of PKG (Fig. 3F). Expression of exogenous G␣ t1 , in sharp contrast, does not rescue the PKG response lost in the G␣ t2 -deficient cells. Parallel rescue experiments were performed on F9 cells made deficient in G␣ o (Fig. 3G). In F9 cells stably expressing Rfz2 and treated with morpholinos antisense to G␣ o , the ability of Wnt5a to regulate PKG is lost. Expression of exogenous G␣ o rescued the Wnt5a-stimulated regulation of PKG in these cells (Fig. 3G). Expression of exogenous G␣ t1 , however, does not rescue the PKG response lost in the G␣ o -deficient cells.
If the decline of the intracellular concentration of cyclic GMP is primarily responsible for the Frizzled-2 stimulated decline in PKG activity, then increasing the intracellular concentration of cyclic GMP would be expected to block Wnt-stimulated action on PKG activity. Treating F9 clones either with the PDE6-selective inhibitor Zaprinast or with a cell-permeable, nonhydrolyzable cyclic GMP analogue, 8-bromo-cyclic GMP (8-Br-cGMP), elevates intracellular concentrations of this cyclic nucleotide (22). Inhibition of cyclic GMP PDE with Zaprinast elevates basal PKG activity and abolished Frizzled-2-stimulated reduction of PKG activity (Fig. 3H). Increasing intracellular cyclic GMP levels with cyclic GMP analogue 8-Br-cGMP also elevates PKG activity and, like Zaprinast treatment, blocks Frizzled-2-stimulated reduction of FIGURE 3. Protein kinase G activity is regulated by the noncanonical Wnt/cGMP/Ca 2؉ pathway. A, F9 clones expressing the ␤ 2 AR/Rfz2 chimera were treated with isoproterenol (10 M), and the activity of PKG was measured in whole cell lysates prepared at the time period indicated (upper panel). Major isoforms of PKG, PKG1␣ and ␤ as well as ␤-actin in each sample were probed by immunoblotting (IB) with specific antibodies (lower panel). *, p Ͻ 0.05. B, F9 cells stably expressing ␤ 2 AR/Rfz2 chimeric receptor were stimulated by isoproterenol alone (ϩIso) or in the presence of the ␤-antagonist propranolol after 30 min of incubation of isoproterenol (ϩIso ϩPro). PKG activity was measured as described under "Materials and Methods." **, p Ͻ 0.01; ***, p Ͻ 0.001 versus the ϩIso group at 0 min; #, p Ͻ 0.05; ##, p Ͻ 0.01 versus the corresponding ϩIso groups. C, clones expressing the ␤ 2 AR/Rfz2 were stimulated by isoproterenol alone (ϩIso) or in the presence of propranolol (ϩIso ϩPro), or these clones were pretreated with pertussis toxin for 2 h prior to the stimulation by isoproterenol (ϩIso ϩPTX). PKG activity was measured. *, p Ͻ 0.05 versus the ϩIso at 0 min; #, p Ͻ 0.05, versus the corresponding ϩIso groups. D, F9 clones stably expressing rat Frizzled-2 were stimulated by Wnt5a and PKG activity in whole cell lysates collected at the indicated time was determined. PKG activity from Frizzled-2-expressing cells pretreated with pertussis toxin (ϩPTX) prior Wnt5a stimulation for 1 h was compared. **, p Ͻ 0.01 versus the ϩWnt5a at 0 min; ##, p Ͻ 0.01 versus the ϩWnt5a at 60 min. E, F9 cells expressing ␤ 2 AR/Rfz2 chimera were treated with morpholino antisense oligonucleotides (Antisense) for 72 h to knock down G␣ t1 , G␣ t2 , or G␣ o individually prior to addition of isoproterenol (ϩIso, 10 M). The cells were lysed after 1 h of incubation of isoproterenol, and PKG activity was measured. The protein levels of G␣ t1 , G␣ t2 , and G␣ o were assessed by immunoblotting (Fig. 2C). *, p Ͻ 0.05 versus control. F, expression of G␣ t2 , but not G␣ t1 , rescued the effect of knock-down of G␣ t2 on PKG activity. F9 cells expressing rat Frizzled-2 receptor were treated with morpholino antisense oligonucleotides to G␣ t2 (Antisense-G␣ t2 ) for 72 h to knock down G␣ t2 . Twenty-four hours after morpholino treatment, the cells were transfected with expression vectors harboring the cDNA of either G␣ t2 or G␣ t1 (versions not sensitive to the antisense morpholinos directed to the 5Ј-untranslated region of the native G protein counterparts) for 2 days. The cells were stimulated by Wnt5a (ϩWnt5a) or without stimulation (ϪWnt5a) for 20 min, and PKG activity in whole cell lysates was measured. ***, p Ͻ 0.001. G, expression of G␣ o , but not G␣ t1 , rescued the effect of knock-down of G␣ o on PKG activity. F9 cells expressing rat Frizzled-2 were treated with morpholino antisense oligonucleotides to G␣ o (Antisense-G␣ o ) for 72 h to knock down G␣ o . Twenty-four hours after morpholino treatment, the cells were transfected with G␣ o or G␣ t1 for 2 days. The cells were stimulated by Wnt5a (ϩWnt5a) or without stimulation (ϪWnt5a) for 30 min, and PKG activity in whole cell lysates was measured. **, p Ͻ 0.01. H, F9 clones stably expressing ␤ 2 AR/Rfz2 chimera were treated with Zaprinast (1 M), 8-Br-cGMP (100 M), or Rp-8-pCPT-cGMP (2 M) for 30 min. The F9 clones were then challenged by isoproterenol (ϩIso, 10 M) for 1 h, and the activity of PKG was measured in whole cell lysates. PKG activity is presented as a percentage by setting the activity at time 0 or without stimulation as 100%. The results are presented as the mean values Ϯ S.E. of three or more separate experiments. **, p Ͻ 0.01; ***, p Ͻ 0.001. OCTOBER 13, 2006 • VOLUME 281 • NUMBER 41 PKG activity (Fig. 3H). Thus, Frizzled-2-stimulated inhibition of the cellular activity of PKG is sensitive to the inactivation of G␣ i protein family members by pertussis toxin as well as to suppression of either G␣ o or G␣ t2 . Conversely, elevation of intracellular cyclic GMP levels can both activate PKG activity and abolish Frizzled-2stimulated inhibition of PKG activity.

Wnt Signaling via Protein Kinase G
PKG and Ca 2ϩ Converge in Frizzled-2 Signaling at NF-AT-dependent Gene Transcription-What, then, is the relationship between the Ca 2ϩ mobilization and the changes in intracellular concentration of cyclic GMP downstream of Frizzled-2 signaling? The effect of increasing intracellular concentration of cyclic GMP on the ability of Frizzled-2 activation to mobilize intracellular Ca 2ϩ was investigated. F9 clones were treated with either the PDE inhibitor Zaprinast or with the 8-Br-cGMP (Fig. 4A). Treatment with either Zaprinast or with 8-Br-cyclic GMP attenuates the ability of Frizzled-2 to stimulate Ca 2ϩ mobilization, i.e. increasing PKG activity can suppress the ability of Frizzled-2 to provoke Ca 2ϩ mobilization. Based upon these observations, PKG appears to function upstream of Ca 2ϩ mobilization in response to Frizzled-2 activation. This hypothesis was tested further by making use F9 clones expressing authentic rat Frizzled-2 and stimulated with purified Wnt5a (Fig. 4B). The Wnt5a-stimulated mobilization of intracellular Ca 2ϩ was nearly abolished by treating the clones with 8-Br-cGMP and Zaprinast, in good agreement with the observations derived with the F9 clones expressing the Frizzled-2 chimera.
We next tested our hypothesis by comparing the effects of activation versus inhibition of protein kinase G on the Ca 2ϩ mobilization response. F9 cells were treated with 8-Br-cGMP to activate and with Rp-8-(4chlorophenylthio)-guanosine 3Ј:5Јcyclic monophosphorothioate (Rp-8-pCPT-cyclic GMP) to inhibit PKG (23), and Ca 2ϩ mobilization was monitored (Fig. 4C). Elevation of intracellular cyclic GMP activates  Fig. 3H) but has little effect on the intracellular Ca 2ϩ levels, as imaged with Fura-2 (Fig. 4C). Direct inhibition of PKG activity with Rp-8-pCPT-cyclic GMP (Fig. 3H), in sharp contrast, provokes an increase in intracellular concentration of Ca 2ϩ (Fig. 4C). A 20-min lag in the time course of intracellular concentrations of Ca 2ϩ was observed between the addition of Rp-8-pCPT-cyclic GMP and a mobilization of Ca 2ϩ . The observed time lag likely reflects the time required for diffusion of the inhibitor to the intracellular compartment responsible for PKG-mediated Ca 2ϩ mobilization.

PKG (
If a relationship between cyclic GMP and Ca 2ϩ impacts the regulation of NF-AT-dependent transcription, one would expect that 1) the activation of PKG, like the inhibition of cyclic GMP-sensitive PDE, would attenuate the effect of Frizzled-2 activation and 2) the inhibition of PKG, like the activation of cyclic GMP-sensitive PDE, would mimic the effects of Frizzled-2 activation. NF-AT-dependent transcription was measured in F9 clones expressing the Frizzled-2 chimera and then stimulated with isoproterenol (Fig. 4D). Activation of Frizzled-2 chimera leads to increased cyclic GMP PDE activity, a decrease in intracellular cyclic GMP levels, yet increased Ca 2ϩ mobilization (13). Zaprinast inhibits PDE6, elevates cyclic GMP levels (13), and increases PKG activity (Fig. 3H). If the hypothesis is correct, Zaprinast treatment should provoke an attenuation of Frizzled-2 activation of NF-AT-dependent transcription. Zaprinast indeed attenuates Frizzled-2 activation of NF-AT-dependent transcriptional response (Fig. 4D), as it does the Ca 2ϩ mobilization (Fig. 4A). Treating cells with the 8-Br-cGMP should increase PKG activity and, if the hypothesis is correct, should likewise attenuate transcription of the NF-ATdependent transcription in response to Frizzled-2 activation. As shown for Ca 2ϩ mobilization (Fig. 4A), treating F9 clones with 8-Br-cyclic GMP (and thereby directly activating PKG) does attenuate the ability of Frizzled-2 activation to stimulate NF-AT-dependent transcription (Fig. 4D). Finally, the hypothesis suggests that treating the F9 clones with the PKG inhibitor Rp-8-pCPT-cyclic GMP would be expected to mimic the activation of Frizzled-2. Treating cells with Rp-8-pCPT-cyclic GMP indeed stimulates NF-AT-dependent transcription in the absence of Frizzled-2 activation (Fig. 4D). Surprisingly, this level of inhibition of PKG activity by Rp-8-pCPT-cyclic GMP not only elevated basal transcription of the NF-AT gene but still also enabled an activation of NF-AT-dependent transcription by Frizzled-2 (Fig. 4D). Higher concentrations of the PKG inhibitor, which would likely more strongly activate PKG activity and preclude further activation in the presence of Frizzled-2 activation, were not tolerated by the cells, so this point remains speculative. Cyclic GMP PDE, cyclic GMP, and PKG appear to function upstream of the signaling from Frizzled-2 to the level of Ca 2ϩ mobilization.
Although we show that Ca 2ϩ mobilization and NF-AT-sensitive transcription in response to Frizzled-2 activation requires inhibition of PKG, we sought to test whether the inhibition of PKG by Frizzled-2 activation required Ca 2ϩ mobilization. We preloaded the cells with the acetoxymethyl ester of BAPTA to buffer intracellular Ca 2ϩ . Buffering of intracellular Ca 2ϩ with BAPTA in F9 cells stably expressing rat Frizzled-2 chimera indeed blocks activation of NF-AT-dependent gene transcription in response to isoproterenol (Fig. 2F). We tested the effects of BAPTA loading on Frizzled-2 activation of PKG activity in F9 clones stably expressing Frizzled-2 and stimulated with purified Wnt5a. Wnt5a stimulates inhibition of PKG activity, as shown earlier (Fig. 3D). If the mobilization of Ca 2ϩ is obligate for PKG activation by Wnt5a, then buffering intracellular Ca 2ϩ with BAPTA might be predicted to block Wnt5a-stimulated inhibition of PKG. Loading F9 clones with BAPTA, however, did not alter the character of the suppression of PKG activity in response to Wnt5a (Fig. 4E).

DISCUSSION
The current studies illuminate several novel facets of the Wnt/cGMP/Ca 2ϩ pathway mediated by Frizzled-2 (Fig. 5). In the Xenopus embryo, activation of the Wnt/Ca 2ϩ pathway promotes ventral cell fate by interfering with the canonical Wnt/ ␤-catenin pathway (24). In the zebrafish embryo, Wnt5a stimulates phosphatidylinositol signaling and Ca 2ϩ transients essential to normal development (9,12). The mouse embryonic F9 cells were exploited in the current work to elucidate the signal linkage map from a proximal step (i.e. activation of Frizzled-2) downstream to the activation of the developmentally regulated, luciferase reporter gene sensitive to NF-AT. A key, cyclic GMP sensor downstream of Frizzled-2 was identified, and the interplay between the two dominant cellular signaling responses from the noncanonical pathway, Ca 2ϩ mobilization and decreased cellular concentrations of cyclic GMP, was uncovered.
Protein kinase G is shown to be a key sensor for cyclic GMP in the Wnt/cGMP/Ca 2ϩ pathway (Fig. 5), although additional cyclic GMP sensors may be participating to a lesser extent. PKG activity is reduced, and NF-AT-sensitive transcription is increased in response to stimulation of Frizzled-2 by Wnt5a. Decreasing PKG activity with Rp-8-pCPT-cyclic GMP mimics the effect of Wnt5a stimulation on PKG activity, induces Ca 2ϩ transients, and increases NF-AT-sensitive transcription, even in the absence of Wnt. Furthermore, elevating the intracellular concentration of cyclic GMP either by inhibition of cyclic GMP phosphodiesterase (PDE6) or by addition of the 8-bromo analogue of cyclic GMP increases PKG activity and effectively blocks Wnt5a-stimulated NF-AT transcriptional activation. Thus, the Wnt/Ca 2ϩ /NF-AT signaling pathway is mediated downstream of cyclic GMP, in part, by protein kinase G. Recently, a genome-wide Drosophila RNA interference screen also indicated PKG as a potential regulator of NF-AT (25).
The immediate upstream regulator of cyclic GMP (and thereby PKG) is a cyclic GMP phosphodiesterase that is sensitive to inhibition by either dipyridamole or Zaprinast (7,13). Zaprinast is a selective inhibitor of the phosphodiesterase PDE6 (26), identified in mouse F9 cells and a well known effector of the heterotrimeric G protein G t2 (27). We show that each aspect of the signal linkage map shown in Fig. 5 is sensitive to inactivation of G proteins by pertussis toxin, a toxin that catalyzes the ADP-ribosylation of members of the Gi protein family, including G␣ t2 and G␣ o . The G protein G␣ o has been shown to function in Wnt signaling in both the canonical and noncanonical Wnt pathways in a variety of systems, from mouse F9 cells (2,7,13,17) to Drosophila embryos (28). Finally, as was observed for zebrafish embryos injected with RNA encoding a unique chimera composed of the exofacial and transmembrane domains of the human ␤ 2 -adrenergic with the cytoplasmic domains of rat Frizzled-2 (i.e. ␤ 2 AR/Rfz2 chimera), isoproterenol stimulates the noncanonical pathway mimicking the action of Wnt5a binding to Rfz-2, again reflecting upon the G proteincoupled nature of Frizzleds (13,29).
The interplay between Ca 2ϩ mobilization and cyclic GMP regulation at a macroscopic level is revealed in the current study. Wnt5a stimulates a decline in intracellular cyclic GMP, attenuates PKG activity, and enables robust Ca 2ϩ mobilization (Fig. 5). The increase in Ca 2ϩ mobilization appears to be dependent upon a decline in intracellular cyclic GMP and a corresponding reduction in the activity of PKG. Rp-8-pCPTcyclic GMP inhibits PKG activity, mimicking Wnt5a action, and increases Ca 2ϩ mobilization, even in the absence of Wnt5a. Elevation of intracellular cyclic GMP either by inhibition of the cyclic GMP phosphodiesterase or by addition of cell-permeable cyclic GMP analogue, 8-Br-cyclic GMP, effectively blocks the Ca 2ϩ mobilization in response to Wnt5a. These observations provide compelling evidence that Wnt5a-stimulated Ca 2ϩ mobilization is regulated downstream of cyclic GMP and its sensor protein kinase G (Fig. 5). This is not to imply that some initial component of the Ca 2ϩ mobilized in response to Wnt5a cannot escape or be linked to Wnt signaling outside of the schematic shown here. Changes in Ca 2ϩ levels, which can be readily measured directly by Fura-2 imaging, appear to occur in advance of the measured changes in PKG activity. Lacking information on the "gain" possible in Wnt signaling pathway, we cannot rule out a Ca 2ϩ mobilization in response to a rapid activation of phosphatidylinositol signaling by Wnt, as observed elsewhere (9). The bulk of Ca 2ϩ mobilized, however, appears to require signaling from Wnt via cyclic GMP PDE/ cyclic GMP/PKG (Fig. 5).
Interestingly, blocking the Wnt-stimulated Ca 2ϩ mobilization by buffering intracellular Ca 2ϩ with BAPTA precludes Wnt5a-stimulated activation of NF-AT-sensitive transcription, but not the ability of Wnt to inhibit PKG activity. Because basal NF-AT transcriptional activity is observed in the cells treated with BAPTA under these conditions, the buffering of intracellular Ca 2ϩ does not appear so severe as to preclude basal activity of the transcriptional response. These data help to clarify earlier results in which PDE inhibitors (either Zaprinast or dipyridamole) were found to block Wnt5a stimulation of Ca 2ϩ mobilization in zebrafish embryos (13). Blockade of the cyclic GMP PDE activity not only attenuated the Ca 2ϩ mobilization in response to Wnt5a, but also stunted normal development of the zebrafish embryos (13). The Wnt5a-stimulated increase in intracellular Ca 2ϩ regulates the activity of downstream Ca 2ϩsensitive enzymes, such as the phosphoprotein phosphatase calcineurin and Ca 2ϩ /calmodulin-dependent protein kinase II. It is the activation of calcineurin that largely is responsible for the activation of NF-AT-dependent transcription. Thus, signals from the Wnt/cGMP/Ca 2ϩ pathway are required for the regulation of NF-AT-sensitive gene transcription, a major element in the Wnt5a response mediated by Frizzled-2. The signals downstream of cyclic GMP and PKG enable the Wnt-stimulated Ca 2ϩ mobilization, although molecular description of this interplay and identification of PKG substrates essential for signaling of the Wnt/cGMP/Ca 2ϩ pathway are not known.