Identification of New Ciliary Signaling Pathways in the Brain and Insights into Neurological Disorders

Abstract

Primary cilia are conserved sensory hubs essential for signaling transduction and embryonic development. Ciliary dysfunction causes a variety of developmental syndromes with neurological features and cognitive impairment whose basis mostly remains unknown. Despite connections to neural function, the primary cilium remains an overlooked organelle in the brain. Most neurons have a primary cilium; however, it is still unclear how this organelle modulates brain architecture and function, given the lack of any systemic dissection of neuronal ciliary signaling. Here, we present the first in vivo glance at the molecular composition of cilia in the mouse brain. We have adapted in vivo proximity-dependent biotin identification (iBioID), targeting the biotin ligase BioID2 to primary cilia in neurons of male and female mice. We identified tissue-specific signaling networks residing in neuronal cilia, including Eph/Ephrin signaling. We also uncovered a novel connection between primary cilia and gamma-aminobutyric acid signaling. Our iBioID ciliary network presents a wealth of new and neural-specific ciliary signaling proteins and yields new insights into neurological disorders. Our findings are a promising first step in defining the fundamentals of ciliary signaling and their roles in shaping neural circuits and behavior. In the future, this work can be extended to pathological conditions of the brain, with the goal of identifying ciliary signaling pathways disrupted in these disorders and the ultimate aim of finding novel therapeutic strategies.

Keywords: iBioID; mature brain; neurodevelopmental disorders; neurons; primary cilia.

Figure 1.

Targeting of the biotin ligase BirA2 to neuronal cilia in vivo. A, Experimental procedure to label the protein composition of the neuronal cilia in the brain. B, A simplified schematic of the targeting strategy for BirA2 using ciliary GPCRs. C, Immunohistochemical labeling with antibodies to HA (fused to BirA2, magenta), biotin (green), and Arl13b (red) on mouse cerebral sections transduced with AAV-BirA2 or AAV-GPCRs-BirA2. DNA was stained with DAPI. Arrows point to neuronal primary cilia. Scale bars, 20 µm. D, Streptavidin pulldowns were performed on total cell lysates from brains transduced with AAV-BirA2 or AAV-GPCRs-BirA2. WB were probed with Streptavidin-680 and Arl13b antibodies. See Extended Data Figures 1-1 and 1-2 for more details.

Figure 2.

Cilia iBioID identified neural-specific clusters in neuronal cilia in vivo. A, A volcano plot of SSTR3-proximate proteins labeled by BioID. The red dots show the proteins considered hits, delimited by the selected thresholds: log2 fold change ≥ 2 and significance ≥ 0.7. The top hit, Tnfrsf21, is highlighted in brown; GABA receptor signaling in green, axon guidance proteins in blue, and Eph/Ephrin signaling in black. The X-axis denotes the log2 fold change of SSTR3-BirA2: BirA2 control. In the Y-axis, significance displays the negative log10 transformed p value for each protein. B, The neuronal ciliary network. This is a clustergram topology of all proteins enriched in SSTR3 iBioID. It is composed of protein networks, with proteins shown as spheres and their interactions as lines. The Cytoscape software was deployed to generate this network, relying on its mining capabilities to extensively search the literature. The size of each sphere is correlated to the fold change value and the color shade to the p value. C, Clustergram topology showing overlap of SSTR3 iBioID hits with other published centrosome–cilia proteomes [Gupta et al. (2015), red, and Mick et al. (2015), green]. D–G, Clustergram topologies of hits associated with the indicated functional category (cyan). See Extended Data Figures 2-1–2-7 for more details.

Figure 3.

Selected iBioID hits localize to neuronal primary cilia by immunofluoresence. A, SIM acquisitions show that the top hit, Tnfrsf21, localizes to the ciliary tip of primary mouse cortical neurons. Two representative examples are shown with cilia by Arl13b (red), centrosomes using Pericentrin (magenta), and Tnfrsf21 (green). Individual confocal sections were analyzed with the line intensity scan function, illustrated with dashed lines, to generate fluorescence spectra graphs [Intensity (A.U.) by distance (nm)], showing the intensity profiles of Tnfrsf21 (green) and Arl13b (red). Scale bars, 500 nm. B, CSNK1G1 and (C) CSNK1G3 are labeled in primary cortical neurons and IMCD3 cells, respectively, with Arl13b (red) and Pericentrin (magenta). DNA was stained with DAPI (blue). Scale bars, 1 µm.

Figure 4.

Primary cilia harbor Eph/Ephrin signaling. A, Clustergram topology of Eph/Ephrin proteins enriched in the cilia iBioID dataset. Most of the highlighted hits are validated in the following sections of this figure. B, Representative immunofluorescence confocal images acquired with the structured illumination microscopy (SIM). Wild-type primary cortical neurons are labeled with EphA4 (green) localizing to cilia costained with the ciliary marker Arl13b (red) and centrosomal marker Pericentrin (magenta). Arrows indicate areas of colocalization of EphA4 and Arl13b or Pericentrin. Dashed lines indicate the selections used to generate fluorescence spectra graphs [a, b, c, and d; graphs showing intensity (A.U.) by distance (nm)], illustrating the intensity profiles of EphA4 (green) and Arl13b (red). Scale bars, 1 µm. C, Representative immunofluorescence images of overexpressed EphA4-GFP (green) in IMCD3 cells. Primary cilia are labeled with the ciliary marker ADCY3 (red) and centrosomal marker Pericentrin (magenta). Arrows show ciliary localization of EphA4-GFP. Scale bars, 1 µm. D, Representative immunofluorescence images of overexpressed EphB3-V5 (green) in IMCD3 cells. Primary cilia are labeled with the ciliary marker ADCY3 or Arl13b (red) and centrosomal marker Pericentrin (magenta). Arrows show ciliary localization of EphB3-V5. Scale bars, 1 µm. E, Experimental procedure to genetically ablate cilia in mature brains. F, The graph shows the mean number of primary cilia in the motor cortex from oil or tamoxifen-injected Ttbk2^fl/fl;Ubc-Cre-ERT2+ male mice. A statistical comparison was performed using the nonparametric Mann–Whitney test. The p value is shown on the graph. Fields of view of 333 µm² from the motor cortex were selected between oil (12 fields of view) and tamoxifen-injected (21 fields of view) mice. G, WB from total brain lysates from oil or tamoxifen-injected Ttbk2^fl/fl;Ubc-Cre-ERT2+ male mice. Each of the brains comes from a separate animal. Blots are probed with antibodies to Phospho (P)-EphA4 (Tyr596) and EPHA4. GAPDH was used as a loading control. Bands were quantified with the area under the curve using the Fiji software. See Extended Data Figure 4-1 for more details.

Figure 5.

Effectors of GABAergic signaling reside within neuronal primary cilia. A, Representative immunofluorescence images of GABRG2 (green) localizing to neuronal primary cilia labeled with Arl13b (red). Wild-type neurons are labeled with and identified by NeuN (magenta). Arrows indicate areas of colocalization of GABRG2 and Arl13b. Scale bars, 1 µm. B, Representative immunofluorescence images of SLC6A1 (green) in IPSC line derived from a patient carrying the SLC6A1 p.S295L mutation with its counterpart isogenic control. Primary cilia and centrosome are labeled with polyglutamylated in magenta. DNA was stained with DAPI (blue). Scale bars, 2 µm. The graph shows the mean SLC6A1 intensity in cilia. Statistical analysis was performed using the nonparametric Mann–Whitney test. The p value is shown on the graph. Quantification was done on 91 cells for the IPSC isogenic control and 88 cells for the IPSCs SLC6A1 p.S295L.