Wnt Secretion Is Regulated by the Tetraspan Protein HIC-1 through Its Interaction with Neurabin/NAB-1

Abstract

The aberrant regulation of Wnt secretion is implicated in various neurological diseases. However, the mechanisms of Wnt release are still largely unknown. Here we describe the role of a C. elegans tetraspan protein, HIC-1, in maintaining normal Wnt release. We show that HIC-1 is expressed in cholinergic synapses and that mutants in hic-1 show increased levels of the acetylcholine receptor AChR/ACR-16. Our results suggest that HIC-1 maintains normal AChR/ACR-16 levels by regulating normal Wnt release from presynaptic neurons, as hic-1 mutants show an increase in secreted Wnt from cholinergic neurons. We further show that HIC-1 affects Wnt secretion by modulating the actin cytoskeleton through its interaction with the actin-binding protein NAB-1. In summary, we describe a protein, HIC-1, that functions as a neuromodulator by affecting postsynaptic AChR/ACR-16 levels by regulating presynaptic Wnt release from cholinergic motor neurons.

Keywords: C. elegans; F-actin; Wnt; claudin.

Figure 1. Mutants in hic-1 Are Hypersensitive to Aldicarb

(A) Illustration of the genomic region of hic-1 introns and exons. The red bar indicates the hic-1 (ok3475) deletion. Also see Figure S1. (B) Percentage paralysis of C. elegans at the 60 min time point. Attempted rescue of the Aldicarb phenotype using the following promoters; Prab-3 (pan-neuronal), Pmyo-3 (body-wall muscles), Plet-413 (epithelial cells), Punc-17 (cholinergic neurons), and Punc-25 (GABAergic neurons). In all Aldicarb bar graphs, N is the number of trials and n is the total number of animals tested per genotype (~20 animals/trial). (C) Expression of phic-1::mCherry in the ventral nerve cord (VNC) cholinergic neurons that are tagged with GFP. n > 10. Also see Figure S2. (D) The punctate expression of the HIC-1::mCherry overlaps with the SNB-1::GFP at the cholinergic synapses of the dorsal nerve cord (DNC). Percentage co-localization of HIC-1 was calculated using the following formula: (number of HIC-1 puncta co-localized with SNB-1/total number of HIC-1 puncta) × 100 in 100 μm. n = 10. Also see Figure S2. (E) Representative images and quantitation of Gelsolin (GSLN-1)::GFP expressed in a subset of cholinergic neurons. WT (n = 35), hic-1 (n = 28), and hic-1; PACh::HIC-1 (n = 25). (F) Representative images and quantitation of fluorescence intensity of the C. elegans NMJ expressing ACR-16::GFP transgene in the body-wall muscles in WT, hic-1, and hic-1; PACh::HIC-1 animals. n = 35 (WT), n = 40 (hic-1), and n = 25 (hic-1; PACh::HIC-1). Also see Figure S3. p values were calculated using one-way ANOVA and Bonferroni’s multiple-comparison test. ***p < 0.001. ns, not significant. Data are represented as mean ± SEM.

Figure 2. Muscle Responsiveness Is Aberrant in hic-1-Mutant Animals

(A) Representative time-lapse images at multiple time points of single punctum of ACR-16::GFP; −1 m refers to the time before bleaching. The punctum was photobleached at the 0 min time point. The fluorescence recovery and the percentage of recovery were calculated till the fluorescence intensity reached a plateau (at the 15 min time point). (B) Recovery rate of a single ACR-16::GFP puncta after photobleaching in WT, hic-1, and hic-1; PACh::HIC-1 animals. (C) Percentage of recovery at the 15 min time point. The number of puncta analyzed and the genotypes tested were n = 15 (WT), n = 18 (hic-1), and n = 13 (hic-1; PACh::HIC-1). (D) Whole-cell recordings on the muscles were performed to record endogenous acetylcholine (ACh) release (mEPSCs) from WT, hic-1, and hic-1; PACh::HIC-1 animals. The mEPSC frequency is greater in hic-1-mutant C. elegans and is rescued by expressing HIC-1 in cholinergic neurons. The mEPSC amplitude was not significantly different across genotypes. The animals tested were n = 17 (WT), n = 12 (hic-1), and n = 9 (hic-1; PACh::HIC-1). Also see Figure S4A. (E) GCaMP is expressed in the body-wall muscles using a muscle-specific promoter. Representative time-lapse fluorescence images and data analysis of calcium transients in the C. elegans muscles are shown here. The hic-1 mutants show increased calcium transients, which are rescued by expressing HIC-1 in cholinergic neurons. The dot-plot graphs represent rise time (t1s), dwell time (t2s), and fall time (t3s) constants for calcium transients in different genotypes. A representation of the time constants is indicated in Figure S4B. Animals tested: n = 15 (WT), n = 12 (hic-1), and n = 11 (hic-1;PACh::HIC-1). Also see Figures S4B and S4C. p values were calculated using one-way ANOVA and Bonferroni’s multiple-comparison test. ***p < 0.001; ns, not significant. Data are represented as mean ± SEM.

Figure 3. Mutants in hic-1 Show Increased Wnt Release

(A) Aldicarb assay for double mutants containing wntless/mig-14 and hic-1 along with control animals. (B) Aldicarb assay for frizzled/lin-17; hic-1 mutants along with control strains. (C) Illustration of the coelomocyte uptake assay for Wnts. (D) Wnt/CWN-2 fluorescence intensity along the DNC from WT (n = 20), hic-1 (n = 18), HIC-1 rescue in ACh neurons (n = 15), mig-14 (n = 15), and mig-14; hic-1 (n = 15) animals. (E) Representative images and dot plot of coelomocyte fluorescence intensity of Wnt/CWN-2::mCherry in cholinergic neurons from WT (n = 57), hic-1 (n = 44), PACh::HIC-1 (n = 25), mig-14 (n = 25), and mig-14; hic-1 (n = 25) animals. Also see Figure S5. (F) Representative images and quantitation of coelomocyte fluorescence intensity in animals expressing Wnt/LIN-44::mCherry in cholinergic neurons. WT (n = 38), hic-1 (n = 42), and hic-1; PACh::HIC-1 (n = 23). Also see Figure S5. p values were calculated using one-way ANOVA and Bonferroni’s multiple-comparison test. ***p < 0.001; ns, not significant. Data are represented as mean ± SEM.

Figure 4. HIC-1 Is Required to Maintain a Normal Actin Cytoskeleton at Cholinergic Synapses

(A) Representative images and quantitation of PACh::GFP-UtrCH fluorescence intensity along the DNC of C. elegans. WT (n = 30), hic-1 (n = 27), PACh::HIC-1 (n = 23). (B) Representative images of the DNC and quantitation for PGABA::GFP-UtrCH in WT (n = 15) and hic-1 (n = 17) mutant animals. (C) Representative images and quantitation of the DNC of animals expressing PACh::GFP-UtrCH. WT (n = 19) and WT animals injected with DMSO (n = 14) or DMSO and latrunculin-A (LAT-A) (n = 15) were imaged for this experiment. (D) Representative images and quantitation of coelomocyte fluorescence in WT and mig-14 strains that express PACh::Wnt/CWN-2::mCherry. WT injected with DMSO (n = 20), WT injected with LAT-A (n = 15), mig-14 injected with DMSO (n = 20), and mig-14 injected with LAT-A (n = 16). In (A), (C), and (D), p values were calculated using one-way ANOVA and Bonferroni’s multiple-comparison test. In (B), p values were calculated using two-tailed unpaired Student’s t test. ***p < 0.001; ns, not significant. Data are represented as mean ± SEM.

Figure 5. HIC-1 Interacts with Neurabin through Its PDZ(bm)

(A) Percentage paralysis of C. elegans at 60 min after Aldicarb exposure. HIC-1ΔC(4aa) indicates a deletion of the last four amino acids from C terminus of HIC-1. Also see Figure S6A. (B) Percentage paralysis of C. elegans at 60 min after Aldicarb exposure indicating genetic interaction between hic-1 and nab-1. Also see Figure S6B. (C) Schematic indicating possible results of the BiFC assay between HIC-1 and NAB-1. HIC-1 (pink) is tagged with the C-terminal half of YFP (green) via a linker sequence (blue), the PDZ(bm) is indicated as circles. The C terminus of NAB-1 is tagged to the N-terminal half of YFP (green) using a linker sequence (blue). The interaction between NAB-1 and HIC-1 leads to reconstitution of YFP fluorescence (yellow glow), while no fluorescence is detected in the absence of the PDZ(bm) of HIC-1. (D) Representative images and quantification of the DNC of WT animals expressing either HIC-1::SpYFP and NAB-1::SpYFP together or HIC-1ΔC(4aa) and NAB-1::SpYFP together in the cholinergic neurons. The cholinergic synapses are labeled with RAB-3::mCherry. Right: quantification of the YFP reconstitution between HIC-1 and NAB-1 along with multiple controls. Also see Figures S6C and S6D. (E) Representative image of the DNC of C. elegans expressing HIC-1::mCherry and NAB-1:GFP. Partial co-localization was seen for HIC-1 and NAB-1 (n > 10). (F) Representative images and quantitation of NAB-1::GFP fluorescence intensity along the DNC in WT (n = 21), hic-1 (n = 23), and hic-1; PACh::HIC-1 (n = 13) animals. Also see Figure S6E. p values were calculated using one-way ANOVA and Bonferroni’s multiple-comparison test. ***p < 0.001; ns, not significant. Data are represented as mean ± SEM.

Figure 6. HIC-1 and Neurabin Function Together for Normal Wnt Release

(A) Representative images and quantitation of fluorescence intensity along the DNC in animals expressing Pmyo-3::ACR-16::GFP.WT (n = 45), hic-1 (n = 48), nab-1 (n = 25), nab-1; PACh::NAB-1 (n = 18), and nab-1; hic-1 (n = 21). (B) Representative images and quantitation of coelomocyte fluorescence intensity from PACh::Wnt/CWN-2::mCherry. WT (n = 57), hic-1 (n = 44), nab-1 (n = 35), nab-1; PACh::NAB-1 (n = 28), and nab-1; hic-1 (n = 32). Also see Figure S7A. (C) Representative images and quantitation of DNC fluorescence from PACh::GFP-UtrCH expressing animals. WT (n = 25), hic-1 (n = 20), nab-1 (n = 20), nab-1; PACh::NAB-1 (n = 21), and nab-1; hic-1 (n = 25). p values were calculated using one-way ANOVA and Bonferroni’s multiple-comparison test. ***p < 0.001; ns, not significant. Data are represented as mean ± SEM.

Figure 7. The ABD of Neurabin Linked to HIC-1 Is Sufficient to Rescue the NMJ Defects of the nab-1; hic-1 Double Mutants

(A) Schematic of the fusion construct HIC-1(ΔC(4aa)+NAB-1(ABD)). The C-terminal four amino acids were deleted from HIC-1, and the ABD of NAB-1 was added in frame with the above HIC-1 construct. The domains of NAB-1 are shown (Chia et al., 2012). (B) Time course paralysis on Aldicarb for the following strains: WT, nab-1; hic-1, nab-1; hic-1; PACh::HIC-1(ΔC(4aa)), and nab-1; hic-1; PACh::HIC-1(ΔC (4aa)+NAB-1(ABD)). The experiment is performed six times, and the total number of animals used in each experiment was 120 (20 animals/trial) for each genotype. (C) Representative images and quantitation of coelomocyte fluorescence intensity from PACh::Wnt/CWN-2::mCherry. WT (n = 15), nab-1;hic-1 (n = 18), nab-1; hic-1; PACh::HIC-1(ΔC(4aa)) (n = 17), and nab-1; hic-1; PACh::HIC-1(ΔC(4aa)+NAB-1(ABD)) (n = 22). (D) Representative images and quantitation of the DNC of animals expressing GFP-UtrCH as a transgene in cholinergic neurons. The genotypes used in this experiment were WT (n = 16), nab-1; hic-1 (n = 15), nab-1; hic-1; PACh::HIC-1(ΔC(4aa)) (n = 15), and nab-1; hic-1; PACh::HIC-1(ΔC(4aa)+NAB-1(ABD)) (n = 19). In (C) and (D), p values were calculated using one-way ANOVA and Bonferroni’s multiple-comparison test. ***p < 0.001; ns, not significant. Data are represented as mean ± SEM.