For tandem affinity purification (TAP), CIB2 isoform 1 (O75838-1, aa 1-187) was Strep II-FLAG (SF)-tagged at the N terminus. Plasmids used for pulldowns and immunoprecipitation coded for Strep-II-FLAG (SF)-tagged ADGRV1a (Uni-Prot ID Q8WXG9-1, aa 4340-6306), (HA-tagged ADGRV1_CTF (Uni-Prot ID Q8WXG9-1, aa 5891-6306), HA-tagged ADGRV1_ICD (Q8WXG9-1, aa 6155-6306), FLAG-myc-tagged CCT3 (P49368-1) and mRFP-tagged BBS6 (Q9NPJ1-1, aa 2-570), BBS10 (Q8TAM1-1, aa 2-723) and BBS12 (Uni-Prot ID Q6ZW61-1, aa 2-710).

hTERT-RPE1 cells and HEK293T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% heat-inactivated fetal calf serum (FCS). Cells were transfected with GeneJuice® (Merck Millipore) according to the manufacturer’s instructions.

Three TAPs were performed for CIB2 as described (Gloeckner et al., 2007; Knapp et al., 2019). In brief, SF-CIB2 was overexpressed in HEK293T cells for 48 h. Mock-treated cells were used as a control. The cells were lysed, and the lysate was cleared by centrifugation. The supernatant was then subjected to a two-step purification on Strep-Tactin®Superflow® beads (IBA) and anti-FLAG M2 agarose beads (Sigma-Aldrich). Competitive elution was achieved by Desbiothin (IBA) in the first step and FLAG® peptide (Sigma-Aldrich) in the second step. The eluate was precipitated by methanol-chloroform and then subjected to mass spectrometric analysis.

Mass spectrometry was performed as previously described (Boldt et al., 2016). SF-TAP-purified protein complexes were solubilized before subjecting to trypsin cleavage. Resulting peptides were desalted and purified using stage tips before separation on a Dionex RSLC system. Eluted peptides were ionized by Nano spray ionization and detected by an LTQ Orbitrap Velos mass spectrometer (Thermo Fisher Scientific). Raw spectra were searched against the human SwissProt database using Mascot and the results were verified by Scaffold (Version 4.02.01, Proteome Software Inc.) to validate MS/MS-based peptide and protein identifications. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (Perez-Riverol et al., 2022) partner repository with the dataset identifier PXD042629.

Mass spectrometry data of SF-tagged CIB2 were compared to data for mock-transfected cells. Proteins that occurred in the mock dataset were not considered for subsequent analysis of CIB2 data. The identified prey in CIB2-TAPs were compared with the data for ADGRV1a from Knapp et al. (2022). Gene names (according to HGNC) of ADGRV1 and CIB2 prey were used as input for the Cytoscape (http://www.cytoscape.org/) plugins STRING (http://apps.cytoscape.org/apps/stringapp) and ClueGO (Bindea et al., 2009). The parameter confidence (score) cutout was set to 0.40 and the parameter maximum number of interactors was set to 0 for STRING analysis. ClueGO v2.3.3 was used for Gene Ontology (GO) term enrichment analysis. Network specificity was set to default (medium). Only GO terms that are based on experimental data (setting: All_Experimental (EXP, IDA, IPI, IMP, IGI, IEP)) were included for the enrichment analysis and only pathways with a pV ≤ 0,05 were considered.

RFP-fused proteins were immobilized on RFP-Trap® agarose beads (ChromoTek) and used for co-precipitation assays according to the manufacturer’s protocol. Briefly, cell lysates from co-transfected HEK293T cells (RFP-tagged proteins or RFP alone together with HA- or SF-tagged proteins, respectively) were suspended in lysis buffer (10 mM Tris-Cl, pH 7.5, 150 mM NaCl, 0.5 mM EDTA, 0.5% NP-40), spun and the supernatant was diluted to 1 mL in dilution buffer (10 mM Tris–Cl, pH 7.5, 150 mM NaCl, 0.5 mM EDTA). Fifty microliters were separated as input (total cell lysate) and samples were added to equilibrated beads for 2 h at 4°C under constant shaking. After washing, precipitated protein complexes were eluted with SDS-sample buffer and subjected to SDS–PAGE and western blotting.

For co-IP FLAG-myc-CCT3, and HA-ADGRV1_CTF (HA-ADGRV1_ICD, HA-CIB2, HA-centrin) were expressed in HEK293T cells and lysed in TAP lysis buffer. Co-IP was performed using anti-FLAG M2 beads from Sigma according to the manufacturer’s protocol. Briefly, cell lysates were incubated with anti-FLAG M2 beads for 1 h at 4°C. Reciprocal Co-IPs were performed with anti-HA agarose beads from Biotool. After three washing steps with TAP washing buffer, samples were eluted with SDS-sample buffer and subjected to SDS-page and Western blot, using antibodies against the FLAG- and HA-tag.

Glutathione S-transferase (GST), GST-tagged BBS6 and 7xHis-tagged ADGRV1-ICD expressed in E. coli BL21 AI following the manufacturer’s instructions. Equal amounts of GST or GST fusion protein were mixed with lysates of His-tagged ADGRV1-ICD and a protease inhibitor mix (Sigma-Aldrich). Samples were incubated overnight at 4°C followed by incubation with 50 μL glutathione sepharose beads 4B (Amersham Biosciences) for 45 min with gentle agitation. Beads were centrifugated and washed 4 times with 50 mM Tris-HCl, 150 mM NaCl, 5 mM MgCl2, 1 mM EDTA, 10% glycerol, 0.01% polyoxyethylene-10-lauryl ether, pH 7.5. Subsequently, bound proteins were eluted with SDS sample buffer and subjected to SDS-PAGE and Western blotting.

The following antibodies were used: mouse anti-CCT3 (Proteintech 60264-1-Ig), mouse anti-CCT2 (Proteintech 68214-1-Ig), mouse anti-CIB2 (Abnova H00010518-A01), rabbit anti-ADGRV1 (Maerker et al., 2008), mouse anti-FLAG M2 (Sigma F3165), rabbit anti-HA antibody (Roche 11867423001), anti-RFP (Chromotek 5F8), goat-anti-GST (Sigma-Aldrich, SAB 2501414), anti-His antibody (Sigma-Aldrich, SAB1306082), goat anti-centrin 2 antibody (Giessl et al., 2004), anti-paxillin (rabbit polyclonal, Abcam, cat no ab32115; mouse monoclonal, BD Transduction Laboratories, cat no 610052) anti-pericentrin 2 (PCNT2) (Santa Cruz, C-16), mouse anti-Arl13b (Abcam, ab136648). Secondary antibodies conjugated to Alexa 488, Alexa 568, or Alexa 647 were purchased from Molecular Probes (Life Technologies) or from Rockland Inc.

All experiments described herein are conforming to the statement by the Association for Research in Vision and Ophthalmology as to care and use of animals in research. C57BL/6J mice and eGFP-Centrin2 mice (Higginbotham et al., 2004) were maintained under a 12 h light-dark cycle, with food and water ad libitum. After sacrificing the animals in CO2 and decapitation, appropriate tissues were dissected. The use of mice in research was approved by District administration Mainz-Bingen, 41a/177-5865-§11 ZVTE, 30.04.2014.

The eyes of mice were cryofixed in melting isopentane and cryosectioned as previously described (Wolfrum 1991). Cryosections were placed on poly-L-lysine-precoated coverslips, incubated with 0.01% Tween 20 in PBS, washed several times, covered with blocking solution, and incubated for a minimum of 30 min followed by overnight incubation at 4°C with primary antibodies. Washed cryosections were incubated with secondary antibodies in a blocking solution containing DAPI (1 mg/mL) (Sigma) for 1.5 h at room temperature. After washing, sections were mounted in Mowiol (Roth).

hTERT-RPE1 cells were processed for immunohistochemistry as previously described (Krzysko et al., 2022).

Specimen were analyzed on a Leica DM6000B microscope and 3D deconvoluted with Leica imaging software (three iteration steps). Images were processed with Leica imaging software and Adobe Photoshop CS. Fiji/ImageJ software (NIH, Bethesda) was used for image processing and quantifications.

To identify novel potential proteins interacting with CIB2 we applied affinity proteomics using tandem affinity purification (TAP) (Boldt et al., 2016). We fused the tandem Strep II-FLAG (SF)-tag to the N-terminus of the CIB2 isoform 1 (Figure 1). The SF-tagged CIB2 was expressed in HEK293T cells, and then subjected to TAP as described previously (Gloeckner et al., 2007; Knapp et al., 2019). Recovered protein complexes were separated by liquid chromatography and the peptide content was determined with tandem mass spectrometry (LC-MS/MS). To identify interacting proteins the raw spectra were searched against SwissProt databases and the results were verified by Scaffold. By applying these strategies we identified 386 potential novel interactors for CIB2 (Supplementary Table S1).

We compared the CIB2 TAP data with the data set which we previously described for ADGRV1a (Knapp et al., 2022). This comparison revealed a high degree of overlap (Figure 1C; Supplementary Table S1), which we did not observe for ADGRV1 and TAP results from other USH proteins (e.g., for SANS (USH1G) and harmonin (USH1C), not shown). We found 270 identical prey proteins in the data sets of CIB2 and ADGRV1 (Figure 1C). Analysis by the Cytoscape plugin STRING revealed that of the 270 shared binding partners of ADGRV1 and CIB2, 244 are highly interconnected (https://string-db.org/, confidence view) and form a common protein network (Figure 1D).

The high overlap of the ADGRV1 and CIB2 TAP interactome prompted us to test whether both proteins physically interact. To address this question, we performed RFP-Trap® pulldown experiments (Figure 2A). For this we co-expressed the 3xHA-tagged C-terminal fragment of ADGRV1 (3xHA-ADGRV1_CTF) with either RFP-tagged CIB2 (mRFP-CIB2) or RFP alone in HEK293T cells. Cell lysates were then incubated with RFP-Trap® beads to immobilize RFP-CIB2 and RFP, respectively. Subsequent Western blot analysis of the recovered proteins revealed that 3xHA-ADGRV1_CTF was pulled down by RFP-CIB2, but not by RFP alone. We next performed reciprocal immunoprecipitations with anti-HA agarose beads to immobilize 3xHA-ADGRV1_CTF and the appropriate controls (Supplementary Figure S1). In this case, Western blot analysis of the recovered proteins revealed that RFP-CIB2, but not RFP alone, co-immunoprecipitated with 3xHA-ADGRV1_CTF, but not with the HA beads alone. Taken together, these findings demonstrated the interaction between ADGRV1 and CIB2.

Next, we used indirect immunofluorescence to examine whether ADGRV1 and CIB2 are co-distributed in the mouse retina. Immunostaining of both proteins in longitudinal cryosections through the murine retinas showed that ADGRV1 and CIB2 were most prominently localized in the synaptic and ciliary region of photoreceptor cells (Figure 2B). Triple-labeling with antibodies against centrin 2, a marker for the connecting cilium, the basal body and the adjacent daughter centriole of photoreceptor cells (Trojan et al., 2008), further highlighted the ciliary association of CIB2 and ADGRV1 in photoreceptor cells (Figures 2C, D). Immunostaining of pericentrin and Arl13b, common markers for the cilia base and shaft/axoneme of primary cilia, respectively (Mühlhans et al., 2011), as well as CIB2 and ADGRV1 confirmed the localization of CIB2 and ADGRV1 at the cilia base of primary cilia in hTERT-RPE1 cells (Supplementary Figure S2). Co-staining of CIB2 and paxillin, a focal adhesion component, in hTERT-RPE1 cells did not show CIB2 staining in focal adhesions (Supplementary Figure S2A), where ADGRV1 is also localized (Supplementary Figure S2B) (Kusuluri et al., 2021; Güler et al., 2023).

The physical interaction and co-localization of ADGRV1 with CIB2, and the high degree of overlap of their interactomes, indicate their functional relation. To investigate the context of these connections, we further analyzed the shared ADGRV1 and CIB2 network. For this purpose, we used the Cytoscape plugin ClueGO (accessed 10 September 2017) and STRING data (https://string-db.org/) (accessed 20 March 2022) (Bindea et al., 2009), which allows protein enrichment analysis based on Gene Ontology (GO) terms that indicate cellular pathways and processes.

We searched within the three categories Biological Process, Cellular Component, and Molecular Function. GO terms based on experimental data and a significance pV ≤ 0.05 were considered (Figure 3; Supplementary Tables S2–S4). In the Biological Process category, seven groups showed enriched GO terms with the leading terms positive regulation of DNA biosynthetic process, protein localization to nuclear body, establishment of protein localization to organelle, acidic amino acid transport, substantia nigra development, protein refolding and mitochondrial membrane organization (Figure 3A). In the Cellular Component category, we identified eight GO term groups with the leading terms extracellular matrix, chaperone complex, chaperonin-containing T-complex, mitochondrion, mitochondrial nucleoid, mitochondrial inner membrane presequence translocase complex, organelle envelope lumen and endoplasmic reticulum (Figure 3B). In the Molecular Function category, we found eleven GO term groups with the leading terms protein binding involved in protein folding, L-glutamate transmembrane transporter activity, cation-transporting ATPase activity, RNA binding, ubiquitin protein ligase binding, NADH dehydrogenase activity, heat shock protein binding, MHC class II protein complex binding, cadherin binding and protein domain specific binding (Figure 3C).

The TAP prey, that was related to most of the GO terms in our enrichment analysis, were the eight components of the chaperonin-containing T (CCT)-complex, also known as the TCP1 ring complex (TRiC) (Ghozlan et al., 2022). In addition, STRING analysis demonstrated that the eight CCT subunits are directly connected to numerous additional prey in the ADGRV1-CIB2 interactome (Figure 3D). The CCT complex subunits form a double ring with inter- and intra-ring contact sites, which build up a cylindrical structure with a central cavity, where polypeptides are inserted and folded (Figure 3E).

Since there is increasing evidence for the participation of CCTs in retinal function (Sinha et al., 2014; Minegishi et al., 2016), we further dissected the interaction of ADGRV1 and CIB2 with the complex subunit CCT3. For this, we co-expressed FLAG-myc-tagged CCT3 and 3xHA-ADGRV1_CTF in HEK293T cells and incubated the cell lysate with anti-FLAG® M2 beads. After anti-FLAG pull-downs we subjected the recovered proteins to Western blots and observed binding of CCT3 to ADGRV1_CTF, but not to the intracellular domain of ADGRV1 (ADGRV1_ICD) alone or to the negative control centrin 1, an EF-hand motif-containing Ca²⁺-binding protein (Trojan et al., 2008) (Figure 4A). In contrast, ADGRV1_CTF was not recovered control precipitations with anti-FLAG beads only (Supplementary Figures S3A–C) and CIB2 was not co-precipitated with CCT3 in this experimental setting (Figure 4B).

Next, we performed reciprocal co-immunoprecipitations and incubated the cell lysate with anti-HA beads (Figure 4C). Western blots of anti-HA-pull-downs also showed the binding of CCT3 to ADGRV1_CTF (Figure 4C, right blot, 1st lane) and no binding to the intracellular domain of ADGRV1 (ADGRV1_ICD) or to centrin 1 (Figure 4C), confirming the findings in anti-FLAG-pull-downs (Figures 4A, B). In addition, however, a small substantial portion of CCT3 was precipitated by CIB2 in the anti-HA-immunoprecipitation settings (Figure 4C, fade band indicated by white arrow). In contrast, in mock-transfected and single transfected cells expressing FLAG-myc-CCT3, respectively no bands were detectable. Taken together, these findings indicated that CCT3 interacts with the cytoplasmic face of the 7-transmembrane part of ADGRV1, most probably with one of the three intracellular loops and that CIB2 also loosely binds to CCT3.

It has been previously shown that components of the CCT complex are localized at the base of primary cilia (Seixas et al., 2010; Seo et al., 2010; Kypri et al., 2014). Given that the connecting cilium and photosensitive outer segment of photoreceptor cells represent a modified primary cilium (Roepman and Wolfrum, 2007; May-Simera et al., 2017), we aimed to examine the expression and spatial distribution of CCT3 in the murine retina (Figure 5). In Western blots with antibodies against CCT3 we detected in protein lysate of the murine retina a prominent band at a proximal molecular weight of ∼ 60 kDa, which is in accordance with the predicted size of CCT3 (Figure 5A). Immunohistochemistry in longitudinal cryosections through murine retina cryosections revealed puncta-like staining of CCT3 (Figure 5B). Triple immunostaining of CCT3, ADGRV1, and the ciliary marker protein centrin 2 in retinal sections demonstrated the localization of CCT3 in the outer plexiform layer, the outer nuclear layer, the inner segments of photoreceptors, and indicated co-localization with ADGRV1 in the ciliary region of photoreceptor cells (Figures 5B, C). Higher magnification of the ciliary region of the triple-stained photoreceptor cells confirmed the localization of ADGRV1 at the connecting cilium and the localization of CCT3 in the basal body and proximal daughter centriole with a slight co-localization of both proteins at the junction of the basal body with the connecting cilium (arrowhead) (Figure 5D). A comparison of this staining pattern with the triple staining for ADGRV1, CIB2, and centrin (Figure 5E) indicated the co-localization of CIB2 with CCT3 in the basal body but only a slight co-localization of both proteins with ADGRV1 (Figure 5F). To determine whether the expression profile of CCT3 in the retina can be extended to other molecules of the TRiC/CCT-chaperonin complex, we also examined the expression of CCT2 in the mouse retina (Supplementary Figure S4). Anti-CCT2 Western of retinal lysates revealed bands of the expected size of CCT2 molecules, but also bands of high molecular weight, possibly representing dimers and oligomers (Supplementary Figure S4A), which were previously suspected (Collier et al. (2021). Immunohistochemistry also revealed punctate labeling of CCT2 in the retinal layers (Supplementary Figure S4B), especially in the inner segments and ciliary region of the photoreceptors (Supplementary Figure S4C).

In primary cilia, the interaction of the cytoplasmic TRiC/CCT chaperonin complex with the three chaperonin-like BBS proteins BBS6, BBS10, and BBS12 mediates the assembly of the BBSome (Figure 6A) (Seo et al., 2010). This interaction of CCTs with the chaperonin-like BBS proteins raised the possibility that these may also interact with ADGRV1 and CIB2. To investigate this, we performed RFP-Traps® with ADGRV1a and BBS6, BBS10 and BBS12 (Figures 5B–D). We expressed SF-ADGRV1a or SF-CIB2 together with RFP-tagged BBS6, BBS10, BBS12, or centrin-1, respectively, in HEK293T cells and incubated cell lysates with RFP-Trap® beads for RFP-Trap®-precipitations. Western blot analyses of the recovered proteins revealed the co-precipitation of ADGRV1a and CIB2 with all three BBS chaperone-like proteins (Figures 6B–D). In contrast, neither ADGRV1 nor CIB2 was co-precipitated in the control RFP-Trap®-precipitations with the RFP-tagged centrin 1 (Figures 6B–D). An exemplary GST pull-down with bacterially expressed GST-BBS6 and His-tagged ADGRV1_ICD confirmed the direct interaction between BBS molecules and ADGRV1 (Supplementary Figure S5).

Taken together, the present interaction assays revealed that ADGRV1 and CIB2 interact with the three chaperonin-like BBS proteins.