The polycystin complex mediates Wnt/Ca(2+) signalling

Abstract

WNT ligands induce Ca(2+) signalling on target cells. PKD1 (polycystin 1) is considered an orphan, atypical G-protein-coupled receptor complexed with TRPP2 (polycystin 2 or PKD2), a Ca(2+)-permeable ion channel. Inactivating mutations in their genes cause autosomal dominant polycystic kidney disease (ADPKD), one of the most common genetic diseases. Here, we show that WNTs bind to the extracellular domain of PKD1 and induce whole-cell currents and Ca(2+) influx dependent on TRPP2. Pathogenic PKD1 or PKD2 mutations that abrogate complex formation, compromise cell surface expression of PKD1, or reduce TRPP2 channel activity suppress activation by WNTs. Pkd2(-/-) fibroblasts lack WNT-induced Ca(2+) currents and are unable to polarize during directed cell migration. In Xenopus embryos, pkd1, Dishevelled 2 (dvl2) and wnt9a act within the same pathway to preserve normal tubulogenesis. These data define PKD1 as a WNT (co)receptor and implicate defective WNT/Ca(2+) signalling as one of the causes of ADPKD.

Figure 1. WNT9B binds to the extracellular domain of PKD1

(a) Membrane topology of full length PKD1, PKD1L1, and TRPC1. (b) Interaction of Flag-tagged PKD1 (F-PKD1), but not bacterial alkaline phosphatase (F-BAP), F-PKD1L1, or F-TRPC1 and WNT9B in lysates of transfected HEK293T cells. (c) Interaction of WNT9B and the LRR-WSC domain of PKD1 fused to Fc in conditioned media of co-transfected HEK293T cells. Mouse FZD8-CRD-Fc was used as positive control. (d) Interaction of WNT9B with the LRR-WSC domain of PKD1 fused to Fc in conditioned media in trans using co-cultures of HEK293T cells singly transfected with plasmids carrying Fc fusions or WNT9B. Red or blue arrow heads indicate Fc fusions or WNT9B, respectively. In upper panel, Fc fusions were non-specifically labeled (indicated by red arrow heads) by the secondary bovine α-goat used to detect α-WNT9B. (e) Interaction of WNT9B with a minimal domain in PKD1 containing the C-terminal cysteine rich domain of the LRR and WSC domain (LRR-CT+WSC) in conditioned media of co-transfected HEK293T cells. (f-h) Interaction of WNT9B with the LRR-WSC domain of PKD1 using purified proteins. (f) Coomassie blue staining showing the purification of LRR-WSC-Fc or Fc from the conditioned media of stably transfected HEK293 cells. (g) Western blotting of purified Fc fusions using α-Fc. (h) 390 ng of purified LRR-WSC-Fc or Fc were incubated with 500 ng/ml purified WNT9B. Fc fusions were pulled down with protein G and immunocomplexes were blotted with α-WNT9B. 87 ng of WNT9B bound to 390 ng of LRR-WSC-Fc (lane 1). 120, 60, or 30 ng of purified WNT9B were used as reference points. Experiments were successfully repeated 2 (b,d,f,g,h) or 5 (c,e) times. Unprocessed blots are shown in Supplementary Fig. 8.

Figure 2. WNT9B activates PKD1/TRPP2

Time course of WNT9B-induced Δ[Ca²⁺]_i (shown as fluorescence ratio 340/380) in untransfected (a, red, n=52 cells pooled from 3 independent experiments) or PKD1/TRPP2 co-transfected CHO-K1 cells (b, red, n=30 cells pooled from 5 independent experiments) in 1.8 mM extracellular Ca²⁺. Ionomycin (I, 2 μM) was added at time point 35 min. (c) Time course of WNT9B-induced Δ[Ca²⁺]_i in PKD1/TRPP2 co-transfected CHO-K1 cells (red, n=65 cells pooled from 5 independent experiments) in Ca²⁺ free extracellular solution (10 μM EGTA). Δ[Ca²⁺]_i in PBS-treated cells in all three conditions is shown in black. (d-l) Time course (d-f), step currents (g-i), or I-V curves of WNT9B (500 ng/ml)-induced whole cell currents in untransfected (d,g,j), PKD1/TRPP2 co-transfected (e,h,k) or PKD1^S99I/TRPP2-co-transfected CHO-K1 cells (f,i,l) in 50 nM intracellular Ca²⁺ concentration. La³⁺ (100 μM) was added at the indicated times. I-V curves were taken before (black) and 5 min after (red) the addition of WNT9B in the bath solution of untransfected (j, n=6 cells pooled from 2 independent experiments), PKD1/TRPP2- (k, n=12 cells pooled from 4 independent experiments), or PKD1^S99I/TRPP2-co-transfected CHO-K1 cells (l, n=6 cells pooled from 4 independent experiments). Representative time courses (d-f) and step currents (g-i) from 6 (untransfected), 12 (PKD1+TRPP2), or 6 (PKD1^S99I+TRPP2) cells. All statistical tests were performed using paired Student's t-test, **P<0.01. Data are shown as mean ± SEM.

Figure 3. Pathogenic mutations in PKD1 or TRPP2 suppress activation by WNT9B

(a-h) I-V curves taken before (black) and 5 min after the addition of WNT9B (500 ng/ml) in the bath solution (red) of untransfected (a, n=5 cells pooled from 2 independent experiments), PKD1/TRPP2- (b, n=8 cells from 3 independent experiments), or PKD1^S99I/TRPP2- (c, n=7 cells from 3 independent experiments), PKD1/TRPP2^D511V- (d, n=11 from 4 independent experiments), PKD1/TRPP2^R872X- (e, n=7 cells from 3 independent experiments), PKD1/TRPP2^Kv1.3- (f, n=12 cells from 5 independent transfections), PKD1L1/TRPP2- (g, n=9 cells from 5 independent experiments), or PKD1/TRPP2/ZNRF3- (h, n=11 cells pooled from 3 independent transfections) co-transfected CHO-K1 cells in zero intracellular Ca²⁺. (i) Summary data of inward (−100 mV) or outward currents (+100 mV) from all groups. Differences between current densities before and after WNT9B for any given membrane potential in I-V curves were determined using paired Student's t-test, *P<0.05, **P<0.01. Data are shown as mean ± SEM.

Figure 4. TRPP2 mediates WNT9B-induced whole cell currents in MEFs

(a-c) Time course (a), step currents (b), and I-V curves of WNT9B-induced whole cell currents in 50 nM intracellular Ca²⁺ (c, n=7 cells pooled from 3 independent experiments), before (Cont-Pkd2^+/+) or after WNT9B (500 ng/ml, WNT9B-Pkd2^+/+) in wild type Pkd2^+/+ MEFs. Time course of WNT9B-induced currents in Pkd2^−/− MEFs is shown as open blue circles in (a), whereas I-V curves are shown as open black (Cont-Pkd2^−/−) and open blue circles (WNT9B-Pkd2^−/−) in (c) (n=6 cells from 3 independent experiments). (d-l) Time course, step currents, and I-V curves of WNT9B-induced whole cell currents in zero intracellular Ca²⁺ in mock- transfected (just CD8α) (d-f, n=6 cells from 3 independent experiments), wild type TRPP2- transfected (g-i, n=12 cells from 5 experiments), or TRPP2^Kv1.3- transfected Pkd2^−/− cells (j-l, n=5 cells pooled from 3 independent experiments). Time course and I-V curves of WNT9B-induced currents in zero intracellular Ca²⁺ in Pkd2^−/− cells is shown as open blue circles in (d) and (f, n=6 cells pooled from 3 independent experiments). (m-o) Time course, step currents, and I-V curves of WNT9B-induced whole cell currents in zero intracellular Ca²⁺ in ZNRF3-transfected Pkd2^+/+ MEFs (n=12 cells pooled from 4 independent experiments). Step currents are shown 5 min after the addition of WNT9B (500 ng/ml) in the bath solution. I-V curves were taken before (black) and 5 min after the addition of WNT9B (red). Differences between current densities before and after WNT9B for any given membrane potential in I-V curves were determined using paired Student's t-test, *P<0.05, **P<0.01. Data are shown as mean ± SEM. Representative images of time courses (a,d,g,j,m) and step currents (b,e,h,k,n) were obtained from 7, 6, 12, 5, 12 cells, respectively.

Figure 5. TRPP2 mediates WNT3A-induced whole cell currents in MEFs

Time course, step currents, and I-V curves of WNT3A-induced whole cell currents in zero intracellular Ca²⁺ in mock- (a-c, n=16 cells pooled from 6 independent experiments), or ZNRF3-transfected Pkd2^+/+ (d-f, n=20 cells pooled from 7 independent experiments) or Pkd2^−/− MEFs (g-i, n=9 cells from 4 independent experiments) before (black) and after the addition of WNT3A (500 ng/ml) in the bath solution (red). Step currents are shown at 5 min after the addition of WNT9B in the bat solution. Statistical analysis was performed using paired Student's t-test, *P<0.05, **P<0.01. Data are shown as mean ± SEM. Representative images of time courses (a,d,g) and step currents (b,e,h) were obtained from 16, 20, 9 cells, respectively.

Figure 6. Interaction of DVL1 and DVL2 with PKD1 in MEFs and transiently transfecte HEK293T cells

(a) Wild type MEF cell lysates were incubated with mouse IgG1 control antibody or mouse monoclonal α-PKD1 (E4). Immunocomplexes were immunoblotted with rabbit polyclonal α-DVL1 (upper left panel), α-DVL2 (middle left panel), or mouse monoclonal α-PKD1 (7e12, lower left panel). Expression levels of DVL1, DVL2, or PKD1 in lysates incubated with mouse IgG1 control or α-PKD1 (E4) are shown in right panels (input). Experiments were successfully repeated 3 times. (b) Cell lysates from Pkd1^−/− or Pkd2^−/− MEFs were incubated with mouse IgG1 control antibody or mouse monoclonal α-PKD1 (E4) and DVL2 was detected in the immunocomplexes, as described in (a). Experiments were successfully repeated twice. (c) HEK293T cells were transiently transfected with indicated plasmids and PKD1 or TRPP2 was pulled down with α-HA. Flag-tagged wild type or mutant DVL2 was detected using α-Flag. DIX (DIshevelled and aXin), PDZ, or DEP (Dishevelled, Egl-10 and Plecstrin) domains are shown red in DVL2 diagram. Position of E499G mutation and PY motif are shown by a black arrow head. Experiments were successfully repeated 3 times. (d) Time course, step currents and I-V curves in zero intracellular Ca²⁺ of wild type MEFs transiently transfected with a DVL2-specific siRNA (n=18 cells pooled from 5 independent experiments) or a mixture of DVL1 and DVL2 siRNAs (n=13 cells pooled from 4 independent experiments). Statistical tests were performed using paired Student's t-test, **P<0.01. Data are shown as mean ± SEM. (e) Knockdown efficiency of DVL1 and DVL2 siRNAs. Wild type MEFs were transiently transfected with indicated siRNAs and cell lysates were immunoblotted with α-DVL1, α-DVL2, or α-β-actin. Experiment was done once. Unprocessed original scans of blots are shown in Supplementary Fig. 8).

Figure 7. TRPP2 mediates WNT9B-induced cell migration

(a) Pkd2^+/+ cells stained with phalloidin (green, to visualize F-actin), Myosin IIB (red), or DAPI (blue) at 0 (Control) or 30 min after the addition of WNT9B (500 ng/ml). Arrow heads indicate accumulation of F-actin and Myosin IIB in the trailing edge of a migrating cell. Scale bar, 50 μm. (b) Percentage of Pkd2^+/+ cells with trailing edge per field (total number of fields of cells pooled from 3 independent experiments shown in graph) incubated with WNT9B or PBS at different time points. (c) Percentage of cells with trailing edge per field (total number of fields of cells pooled from 3 independent experiments shown in graph) in Pkd2^+/+ MEFs incubated with WNT9B in the presence or absence (DMSO) of 10 μM BAPTA/AM at different time points. (d) Representative images of GFP-, GFP+TRPP2^WT- or GFP+TRPP2^Kv1.3-transfected Pkd2^−/− cells stained for F-actin (yellow) and Myosin IIB (red). Scale bar, 50 μm. (e) Percent of migrating GFP-positive cells in three groups treated for 1 h with PBS or WNT9B (500 ng/ml) from one representative out of three independent transfections (total number of fields of cells from a single experiment, that was independently repeated three times are shown in graph). Statistics source data for all three experiments are available in Suppl. Table 3. (f) Cells migrated through a trans-well filter in the presence of WNT9B (1 μg/ml) or PBS for 6 h. Scale bar, 500 μm. (g) Number of migrating Pkd2^+/+ or Pkd2^−/− cells through a trans-well filter in the presence of PBS or WNT9B. n= 18 fields per group pooled from 3 independent experiments scoring a total of 2462 (Pkd2^+/+-PBS), 5665 (Pkd2^+/+-WNT9B), 3091 (Pkd2^−/−-PBS), or 2796 (Pkd2^−/−-WNT9B) cells. (h) Number of migrating cells in the presence of PBS or PDGF-BB (25 ng/ml). Data were pooled from 3 independent experiments in quadruplicates (n= 12 fields per group scoring a total of 3625 (Pkd2^+/+-PBS), 8927 (Pkd2^+/+-PDGF-BB), 3904 (Pkd2^−/−-PBS), or 9048 (Pkd2^−/−-PDGF-BB) cells. N.S means non-significant, *P<0.05, ***P<0.001 in one-way ANOVA followed by Neuman-Keuls post hoc test. Data are shown as mean ± SEM.

Figure 8. Cooperativity between PKD1 and DVL2 in Xenopus

(a-l) Xenopus embryos were injected with the indicated amounts of Pkd1-sMO1, Dvl2-MO and Wnt9A-MO either alone or in combination. Embryos were analyzed at stage 43 by morphology for edema formation or at stage 40 by immunofluorescence with the proximal tubular marker 3G8 (green) and the distal tubule/nephric duct marker 4A6 (yellow). Representative images of an uninjected control (a,e) and embryos injected with suboptimal amounts of Pkd1-sMO1 (b,f), Dvl2-MO (c,g), a mixture of Pkd1-sMO1 and Dvl2-MO (d,h), as well as with a Wnt9A-sMO alone (j) or in combination with suboptimal amounts of Pkd1-sMO1 (k) are shown. Bar diagrams showing the percentage of embryos exhibiting edema or dysplastic kidneys from 4 or 3 independent experiments are shown in (i) or (l), respectively. (m) Rescue experiment of Dvl2 morphants with a unilateral injection of either 4 ng wild type human DVL2 mRNA or 4 ng mutant DVL2-E499G mRNA. Embryos were analyzed at stage 40 by 3G8/4A6 immunofluorescence comparing the mRNA injected side with the contralateral side. Bar diagram depicts the summary of 3 independent experiments. (Inset) Expression levels of Flag-tagged human DVL2 and DVL2-E499G mutant in stage 25 embryos injected with 1 ng mRNA into a single blastomere at the 4-8 cell stage. Experiment was done twice. Total number of scored embryos is shown on top of each bar.