- Split View
-
Views
-
Cite
Cite
Mariko Takahara, Yohei Katoh, Kentaro Nakamura, Tomoaki Hirano, Maho Sugawa, Yuta Tsurumi, Kazuhisa Nakayama, Ciliopathy-associated mutations of IFT122 impair ciliary protein trafficking but not ciliogenesis, Human Molecular Genetics, Volume 27, Issue 3, 01 February 2018, Pages 516–528, https://doi.org/10.1093/hmg/ddx421
- Share Icon Share
Abstract
The intraflagellar transport (IFT) machinery containing the IFT-A and IFT-B complexes mediates ciliary protein trafficking. Mutations in the genes encoding the six subunits of the IFT-A complex (IFT43, IFT121, IFT122, IFT139, IFT140, and IFT144) are known to cause skeletal ciliopathies, including cranioectodermal dysplasia (CED). As the IFT122 subunit connects the core and peripheral subcomplexes of the IFT-A complex, it is expected to play a pivotal role in the complex. Indeed, we here showed that knockout (KO) of the IFT122 gene in hTERT-RPE1 cells using the CRISPR/Cas9 system led to a severe ciliogenesis defect, whereas KO of other IFT-A genes had minor effects on ciliogenesis but impaired ciliary protein trafficking. Exogenous expression of not only wild-type IFT122 but also its CED-associated missense mutants, which fail to interact with other IFT-A subunits, rescued the ciliogenesis defect of IFT122-KO cells. However, IFT122-KO cells expressing CED-type IFT122 mutants showed defects in ciliary protein trafficking, such as ciliary entry of Smoothened in response to Hedgehog signaling activation. The trafficking defects partially resembled those observed in IFT144-KO cells, which demonstrate failed assembly of the functional IFT-A complex at the base of cilia. These observations make it likely that, although IFT122 is essential for ciliogenesis, CED-type missense mutations underlie a skeletal ciliopathy phenotype by perturbing ciliary protein trafficking with minor effects on ciliogenesis per se.
Introduction
Cilia are microtubule-based appendages that project from the surfaces of various eukaryotic cells, and play key roles in sensing extracellular stimuli and transducing developmental signals, such as Hedgehog (Hh) and Wnt signaling (1,2). Various proteins are specifically localized within cilia and on the ciliary membrane, including seven-pass transmembrane G protein–coupled receptors (GPCRs) and ion channels. Defects in the assembly and function of cilia result in a broad class of congenital disorders, collectively termed ciliopathies, accompanying a variety of manifestations (3–5). These include Bardet-Biedl syndrome (BBS), Joubert syndrome (JBTS), nephronophthisis, and short-rib thoracic dystrophies (SRTDs). In SRTDs, the skeletal phenotype, affecting the limbs, ribs, and sometimes the craniofacial skeleton, is predominant (6,7). Cranioectodermal dysplasia (CED) has extensive overlap in clinical features with SRTDs.
The intraflagellar transport (IFT) machinery, which is often referred to as ‘IFT trains’ or ‘IFT particles’ based on microscopic analyses of Chlamydomonas flagella (8), is essential for the assembly and maintenance of cilia by the trafficking of ciliary proteins (9,10). IFT is mediated by the bidirectional trafficking of IFT particles, which contain the IFT-A and IFT-B complexes. Anterograde trafficking of ciliary proteins from the ciliary base to the tip along the axoneme, a microtubule-based structure extending from the basal body, is mediated by the IFT-B complex with the aid of kinesin-2 motor proteins, whereas retrograde trafficking is mediated by the IFT-A complex powered by dynein-2 (9–11). In addition to these IFT complexes, the BBSome composed of 8 BBS proteins (12–14) moves in association with the IFT particles (15). The BBSome is thought to mediate not only entry into but also removal of membrane proteins from cilia, by connecting them with the IFT-B complex (16,17).
We and others recently demonstrated the overall architecture of the IFT-B complex composed of 16 subunits, which can be divided into the core subcomplex (composed of 10 subunits) and the peripheral subcomplex (composed of 6 subunits) (18–20). On the other hand, we recently demonstrated the architecture of the IFT-A complex (see Supplementary Material, Fig. S1A) (21). The IFT-A complex can be divided into the core subcomplex, which is composed of IFT122/IFT140/IFT144 and is associated with TULP3, and the peripheral subcomplex composed of IFT43/IFT121/IFT139. The core and peripheral subcomplexes are connected by the interaction between IFT122 and the IFT43–IFT121 dimer (also see Supplementary Material, Fig. S1A). Thus, IFT122 serves as a hub of the IFT-A complex. Mutations in all of the IFT-A subunits are known to cause skeletal ciliopathies; IFT43/CED3, IFT121/SRTD7/CED2, IFT122/CED1, IFT139/SRTD4, IFT140/SRTD9, and IFT144/SRTD5/CED4.
In this study, we first demonstrated that IFT122 forms the trimeric IFT122/IFT140/IFT144 core subcomplex via its C-terminal (CT) region containing the tetratricopeptide repeat (TPR) domain, whereas it interacts with the IFT43–IFT121 dimer via the N-terminal (NT) region containing the WD40 domain and the CT region. We then established IFT122-knockout (KO) and IFT121-KO cell lines of human telomerase reverse transcriptase–immortalized retinal pigment epithelial 1 (hTERT-RPE1) cells and found that whereas IFT121-KO cells have cilia but show defective retrograde trafficking within cilia, IFT122-KO cells have no recognizable cilia; the ciliogenesis defect of IFT122-KO cells was much severer than that of KO cells of IFT144, another IFT-A core subunit that directly interacts with IFT122. Furthermore, the exogenous expression of missense mutants in the IFT122 WD40 domain, which are found in CED1 patients, can restore the ciliogenesis of IFT122-KO cells but still demonstrate an abnormal localization of various ciliary proteins. These results are consistent with the differences in ciliogenesis between Ift122-KO mice (22) and CED1 patients with any of the point mutations of IFT122 (23,24); cells of Ift122-KO mice lack cilia, whereas cells from CED1 patients show mildly reduced cilia number and length.
Results
Interaction modes of IFT122 with other IFT-A subunits
We recently determined the overall architecture of the IFT-A complex using our practical and flexible protein–protein interaction assay (21), which we named the visible immunoprecipitation (VIP) assay (12); the VIP assay can detect not only one-to-one but also one-to-many and many-to-many protein interactions. The IFT-A complex can be divided into the core subcomplex composed of IFT122/IFT140/IFT144, which was originally identified by analysis of the IFT-A complex purified from Chlamydomonas reinhardtii (25), and the peripheral subcomplex composed of IFT43/IFT121/IFT139 (Supplementary Material, Fig. S1A); IFT122 forms a dimer with IFT144 and connects the core subcomplex to the peripheral subcomplex via its interaction with the IFT43–IFT121 dimer. Thus, IFT122 appears to serve as a hub in the IFT-A complex. In this study, we therefore investigated the interaction modes of IFT122 with other IFT-A subunits by the VIP assay and conventional immunoblotting analysis using EGFP- and mCherry (mChe)-fusion constructs.
Consistent with our previous study (21), when coexpressed in HEK293T cells, EGFP-IFT122(WT) was demonstrated to coimmunoprecipitate mChe-IFT144(WT) by the VIP assay and immunoblotting analysis (Fig. 1A, lane 1). When the IFT122 protein was divided into the NT (NT1) and CT halves containing the WD40 and TPR domains, respectively (see Supplementary Material, Fig. S1B, third row), only the CT region containing the TPR demonstrated an interaction with IFT144(WT) (Fig. 1A, lane 3). On the other hand, when the IFT144 protein was divided into the NT and CT (CT1 and CT2) halves (see Supplementary Material, Fig. S1B, sixth row, and Table S1), only IFT144(CT2) demonstrated a robust interaction with IFT122(CT) (Fig. 1A, lanes 4–6), suggesting that an IFT144 region encompassing at least residues 357–653 is involved in its interaction with IFT122.
We then analysed which region of the IFT140 protein is involved in its interaction with the IFT122–IFT144 dimer. IFT140(WT) interacted with IFT122(WT)–IFT144(WT) (Fig. 1B, lane 1), as determined previously (21), and with IFT122(CT)–IFT144(WT) (Fig. 1B, lane 2) in line with the above mentioned data (Fig. 1A). When the IFT140 protein was divided into the NT and CT halves (see Supplementary Material, Fig. S1B, fifth row), only IFT140(CT) demonstrated an interaction with IFT122(CT)–IFT144(WT) (Fig. 1B, lanes 6 and 7). Thus, the IFT122 CT region is responsible for the formation of the IFT122/IFT140/IFT144 trimer. In agreement with our previous study showing that IFT140 exhibits a binary, although weak, interaction with IFT144 (21), IFT140(WT) (Fig. 1B, lanes 4 and 5) or IFT140(CT) (lanes 8 and 9) coprecipitated not only IFT144(CT2) but also IFT144(CT1), although the latter IFT144 construct cannot interact directly with IFT122 (Fig. 1A, lane 5).
Next, we analysed the interaction mode of IFT122 with the IFT43–IFT121 dimer. As described previously (21), IFT122(WT) interacted robustly with the IFT43–IFT121 dimer (Fig. 1C, lane 1). However, when the IFT122 protein was divided into two halves (NT1 and CT), neither half interacted with IFT43–IFT121 (Fig. 1C, lanes 2 and 4). We therefore analysed another NT construct (NT2, see Supplementary Material, Fig. S1B, third row) with a C-terminal extension compared with NT1, but did not observe an interaction with IFT43–IFT121 (Fig. 1C, lane 3). Although we thus failed to narrow down the region of the IFT122 protein responsible for its interaction with IFT43–IFT121, we determined that the WD40-containing NT region of IFT122 participates in its interaction with IFT43–IFT121 (see below).
CED1-associated missense mutations of IFT122 affect its interaction with IFT43–IFT121 and IFT139
The IFT122 gene has been reported to be mutated in CED1 patients (23,24,26,27). The reported point mutations are restricted to the NT region containing the WD40 domain, namely, W7C, S263F, G436R, V443G, and G513V (see Fig. 2A). As the NT region of IFT122 participates in its interaction with the IFT43–IFT121 dimer (see Fig. 1), we analysed whether the CED1-type mutations of IFT122 affect its interaction with IFT43–IFT121.
As shown in Figure 2B, the interaction of IFT122 with IFT43–IFT121 was almost completely abolished by any of the G436R, V443G, and G513V mutations (lanes 4, 5, and 6, respectively). In contrast, IFT122(W7C) and IFT122(S263F) retained the ability to interact with IFT43–IFT121 (lanes 2 and 3, respectively). Thus, at least a part of the WD40-containing region of IFT122 is essential for its interaction with the IFT43–IFT121 dimer.
We then analysed whether mutations in IFT122 affected its interaction with the whole peripheral subcomplex, composed of IFT43, IFT121, and IFT139, although IFT122 does not directly interact with IFT139 (21). When EGFP-IFT122(WT) was coexpressed with mChe-IFT139 in addition to mChe-IFT43 and mChe-IFT121, EGFP-IFT122(WT) coimmunoprecipitated all three mChe-fused peripheral subunits (Fig. 2C, lane 1). As expected from the data shown in Figure 2B, EGFP-IFT122(G436R), EGFP-IFT122(V443G), and EGFP-IFT122(G513V) had greatly reduced abilities to coimmunoprecipitate mChe-IFT43, mChe-IFT121, and mChe-IFT139 (Fig. 2C, lanes 4–6). On the other hand, EGFP-IFT122(W7C) and EGFP-IFT122(S263F) failed to coimmunoprecipitate mChe-IFT139, although they substantially retained the ability to interact with IFT43–IFT121 (Fig. 2C, lanes 2 and 3, respectively). These data were somewhat unexpected in view of the fact that IFT122 does not demonstrate a direct interaction with IFT139 (21) (see Discussion).
Virtually no cilia are formed in IFT122-KO cells
Previous pathophysiological and histological analyses of Ift122-KO mice (22) and mutant mice with a missense mutation in the initiation codon of the Ift122 gene (sopb mice) (28,29) clearly demonstrated that both types of Ift122-defective mice were embryonic lethal and demonstrated severe defects in Hh signaling during embryonic development. At the cellular level, however, these mice showed different phenotypes. Cells derived from Ift122-KO mice apparently lacked cilia (22), whereas cells from sopb mice demonstrated mild ciliogenesis defects and the accumulation of various ciliary proteins at the bulged ciliary tips (28,29); the less severe ciliogenesis defects of sopb cells are reminiscent of those of cells from CED1 patients (23,24).
To pursue the reason for the apparent phenotypic difference between Ift122-KO and sopb cells, we established hTERT-RPE1 cell lines defective in IFT122 using the CRISPR/Cas9-mediated homology-independent knock-in system; experimental details were recently published (30); also see (21,31–33). Among the obtained KO cell lines, we selected two independent cell lines, #122–1-13 and #122–2-1, which were established using distinct target genome sequences (Supplementary Material, Table S3), for the following experiments. PCR analysis of genomic DNA demonstrated that the #122–1-13 and #122–2-1 cell lines have a reverse and forward integration, respectively, of the donor knock-in vector in one IFT122 allele (Supplementary Material, Fig. S2A, lanes 7 and 9, respectively). Direct sequencing of the PCR products (lanes 5, 7, 8, and 9) demonstrated that, in the other IFT122 allele, #122–1-13 and #122–2-1 have a 1-bp deletion and a 1-bp insertion, respectively (Supplementary Material, Fig. S2B and C), probably resulting from error-prone nonhomologous end joining after CRISPR/Cas9-mediated genome DNA cleavage. Depletion of the IFT122 protein in the KO cell lines was confirmed by immunoblotting analysis (Supplementary Material, Fig. S2D) using commercially available antibody used in a recent study of Phua et al. (34).
We then attempted to visualize cilia of control RPE1 cells and the IFT122-KO cell lines by staining for ARL13B and acetylated α-tubulin (Ac-α-tubulin). Control cells demonstrated uniform staining along the entire cilia for both ARL13B (a marker for the ciliary membrane) and Ac-α-tubulin (an axoneme marker) (Fig. 3A–A'′). In most cells of the two independent IFT122-KO cell lines, however, we could not detect typical ciliary staining for ARL13B or Ac-α-tubulin (Fig. 3B–B′′ and C–C′'); relatively faint staining for both ciliary markers were sometimes observed around the basal body depicted by γ-tubulin staining. Thus, the severe ciliogenesis defects in the IFT122-KO cell lines are consistent with those reported for cells derived from Ift122-KO mice (22). However, the ciliogenesis phenotype of IFT122-KO cells are in contrast to the phenotypes of RPE1 cells defective in IFT144 (another IFT-A core subunit) and IFT139 (an IFT-A peripheral subunit), which we previously reported (21); slightly but significantly short cilia in IFT144-KO cells and normal length cilia in IFT139-KO cells, although the cilia of both KO cells have a bulged structure at the distal tips.
Next, we stained control and IFT122-KO cells with a commercially available anti-IFT140 antibody. In control cells, IFT140 staining was mainly found at the ciliary base (Fig. 3D); note that the available anti-IFT140 antibody also stained unidentified punctate structures in the nucleus of RPE1 cells (see the manufacturer’s website; http://www.ptglab.com/Products/IFT140-Antibody-17460–1-AP.htm; date last accessed December 12, 2017). In IFT122-KO cells, however, no apparent IFT140 signals were found around the basal body (Fig. 3E and F). As the loss of IFT-A signals around the basal body is also seen in IFT144-KO cells but not in IFT139-KO cells (21), it is likely that the IFT-A complex does not successfully assemble in the absence of a core subunit, namely, IFT122 or IFT144.
We then analysed the localization of IFT88, a subunit of the IFT-B complex. In control RPE1 cells, IFT88 was detected mainly around the basal body and faintly along cilia (Fig. 3G), as described previously (21). In IFT122-KO cells, the IFT88 signals were still observed around the basal body even in the absence of cilia (Fig. 3H and I). This is in contrast to the phenotypes of IFT139-KO and IFT144-KO cells, in which IFT88 was accumulated at the bulged ciliary tips due to normal anterograde trafficking and impaired retrograde trafficking (21).
We also examined the localization of GPR161 in IFT122-KO cells, because ciliary GPCRs cannot enter the cilia of IFT144-KO cells. In control RPE1 cells, GPR161 was uniformly distributed on the ciliary membrane (Fig. 3J). In contrast, no GPR161 signals were detected around the basal body in IFT122-KO cells (Fig. 3K and L).
Phenotypic differences between IFT121-KO and IFT122-KO cells
In view of the substantial ciliogenesis defects of IFT122-KO cells, we then established IFT121-KO cell lines, and compared their phenotypes with those of IFT122-KO cell, since IFT122 directly interacts with the IFT43–IFT121 dimer of the peripheral subcomplex [(see Supplementary Material, Fig. S1A); see Materials and Methods and Supplementary Material, Fig. S3 for details of establishment of these KO cells lines (#121–1-3 and #121–1-12)]. We could not confirm depletion of the IFT121 protein in the IFT121-KO cell lines due to the lack of available antibodies that work well in immunoblotting and/or immunofluorescence. However, the abnormal phenotype observed in these cell lines is unlikely to result from off-target effects, since two independent IFT121-KO cell lines displayed essentially the same phenotype (see below).
As shown in Supplementary Material, Figure S4B and C, formation of cilia depicted by staining for ARL13B and Ac-α-tubulin appeared to be comparable to control cells (Supplementary Material, Fig. S4A). However, ARL13B staining was often observed at bulged ciliary tips (Supplementary Material, Fig. S4B and C), similarly to ARL13B staining in IFT139-KO cells (21). When localization of the IFT-B subunit IFT88 and the IFT-A subunit IFT140 was analysed (Supplementary Material, Fig. S3D–F and G–I, respectively), both were substantially accumulated at the bulged ciliary tips in IFT121-KO cells, similarly to in IFT139-KO cells (21). These observations indicate that the absence of either of the IFT-A peripheral subunits, IFT121 and IFT139, resulted in impaired retrograde trafficking of the IFT machinery, although anterograde trafficking mediated by the IFT-B complex occurred normally.
To address whether ciliary protein trafficking is indeed impaired in IFT121-KO cells, we then analysed the localization of Smoothened (SMO) and GPR161, both of which are seven-pass transmembrane GPCRs regulating Hh signaling (35). Under basal conditions, SMO is eliminated from cilia whereas GPR161 localizes on the ciliary membrane and represses Hh signaling. When cells are stimulated with Hh or Smoothened Agonist (SAG), however, SMO enters cilia whereas GPR161 exits cilia; consequently, Hh signaling is activated (35). As shown in Supplementary Material, Fig. S5, under basal conditions, SMO was not found within cilia in both control (Supplementary Material, Fig. S5A) and IFT121-KO (Supplementary Material, Fig. S5B and C) cells, whereas GPR161 was found within cilia in both control (Supplementary Material, Fig. S5G) and IFT121-KO (Supplementary Material, Fig. S5H and I) cells. Upon stimulation of control cells with SAG, SMO was imported into cilia (Supplementary Material, Fig. S5D) whereas GPR161 was exported from cilia (Supplementary Material, Fig. S5J). In striking contrast, SMO does not substantially enter cilia (Supplementary Material, Fig. S5E and F) and GPR161 was retained within cilia (Supplementary Material, Fig. S5K and L) when IFT121-KO cells were treated with SAG; the failed ciliary entry of SMO upon SAG treatment observed in the IFT121-KO cells is in agreement with recent studies (36,37), although neither study analysed the localization of GPR161. Two explanations are possible for these observations: one is that both the entry of SMO into cilia and ciliary retrograde trafficking and/or exit of GPR161 were impaired in the absence of IFT121; the other is that as a consequence of impaired ciliary entry of SMO, GPR161 was not cued to exit cilia. We favor the former possibility in view of the marked accumulation of the IFT-A and IFT-B complexes at the ciliary tips in IFT121-KO cells (Supplementary Material, Fig. S4H and I, and E and F, respectively). In any case, the phenotype of IFT121-KO cells is completely different from that of IFT122-KO cells despite the direct interaction between IFT121 and IFT122, but rather resembles the phenotype of KO cells of IFT139 (21), which constitutes the IFT-A peripheral subcomplex together with IFT43 and IFT121.
CED1-type IFT122 mutations have a minor effect on ciliogenesis but substantial effects on the localization of ciliary proteins
We next analysed whether the severe ciliogenesis defect of IFT122-KO cells can be rescued by the exogenous expression of IFT122. As shown in Figure 4, stable expression of EGFP-IFT122(WT) (E) but not EGFP (D) by infection of lentiviral vectors significantly restored the formation of cilia in the #122–1-13 cell line (also see Fig. 4H). Essentially the same results were obtained using the #122–2-1 cell line (Fig. 4H). Thus, these results exclude the possibility that the severe ciliogenesis defect of IFT122-KO cells resulted from an off-target effect of the CRISPR/Cas9 system.
We then investigated whether CED1-type mutants can recover the ciliogenesis of IFT122-KO cells, as cells derived from CED1 patients were reported to form cilia, although slightly less efficiently than in control cells (23,24). When the ciliary axoneme was visualized by staining for Ac-α-tubulin, stable expression of EGFP-IFT122(W7C) (Fig. 4F) or EGFP-IFT122(G513V) (Fig. 4G) significantly recovered ciliogenesis of the #122–1-13 cell line, although slightly less efficiently than by expression of EGFP-IFT122(WT) (Fig. 4H). These CED1-type mutants also restored the ciliogenesis defect of the #122–2-1 cell line (Fig. 4H).
However, we noticed that expression of CED1-type mutants in the IFT122-KO cell lines had somewhat different effects on the localization of ciliary proteins when the ciliary membrane was stained with an antibody against ARL13B. In the #122–1-13 cell line with exogenous expression of EGFP-IFT122(WT), ARL13B was uniformly distributed along cilia (Fig. 5B'), like in control RPE1 cells (Fig. 5A). In contrast, in the #122–1-13 cell line expressing EGFP-IFT122(W7C) (Fig. 5C’) or EGFP-IFT122(G513V) (Fig. 5D'), a significant proportion of ARL13B was found on the bulged structures at the ciliary tips (also see Fig. 5E); these bulged structures are reminiscent of those observed in the cells of CED1 patients (23,24) and sopb mice (28,29). Similar results were obtained by rescue experiments using the #122–2-1 cell line (Fig. 5E).
We previously showed that INPP5E, a phosphoinositide 5-phosphatase, is localized within cilia in a manner dependent on ARL13B (32); both ARL13B and INPP5E are encoded by the causative genes of JBTS (JBTS8 and JBTS1, respectively) (38). We therefore analysed the localization of INPP5E in the IFT122-KO cell lines exogenously expressing WT or mutant IFT122. In the #122–1-13 cell line expressing EGFP-IFT122(WT), INPP5E demonstrated a uniform distribution (Fig. 5G'), similarly to in control cells (Fig. 5F). In striking contrast, the #122–1-13 cell line expressing EGFP-IFT122(W7C) (Fig. 5H') or EGFP-IFT122(G513V) (Fig. 5I') had no discernible INPP5E signals within cilia (also see Fig. 5J). Similarly, INPP5E was not found within cilia in the #122–2-1 cell line expressing EGFP-IFT122(W7C) or EGFP-IFT122(G513V) (Fig. 5J). The most plausible explanation for the localization of ARL13B and INPP5E in IFT122-KO cells expressing the CED1-type mutant is that ARL13B molecules accumulated at the distal tip cannot mediate ciliary entry of INPP5E.
We also noticed that when exogenously expressed in IFT122-KO cells, EGFP-IFT122(WT) was often found around the ciliary base (for example, see Fig. 5B and G), whereas EGFP-IFT122(W7C) or EGFP-IFT122(G513V) did not demonstrate such signals (Fig. 5C, D, H and I).
On the other hand, IFT88 was localized mainly around the base of cilia with faint localization along cilia (Fig. 6B') in the #122–1-13 cells expressing EGFP-IFT122(WT), similarly to in control cells (Fig. 6A). In contrast, in the #122–1-13 cells expressing EGFP-IFT122(W7C) or EGFP-IFT122(G513V), IFT88 was mainly found at the bulged ciliary tips (Fig. 6C', D' and E). Essentially the same results were obtained using #122–2-1 cells (Fig. 6E). These phenotypes of IFT122-KO cells exogenously expressing IFT122(W7C) or IFT122(G513V) resemble those of cells defective in IFT144, another IFT-A core subunit; in the absence of IFT144, the functional IFT-A complex cannot be assembled at the ciliary base, and the IFT-B complex does not undergo retrograde trafficking although it is trafficked anterogradely to the ciliary tips (21).
Next, we analysed the localization of SMO and GPR161 in IFT122-KO cells exogenously expressing IFT122 or its mutant. As described above, in control cells, SMO was not found within cilia under basal conditions (Fig. 7A; also see Supplementary Material, Fig. S5A), whereas it was transported into cilia when cells were treated with SAG (Fig. 7E; also see Supplementary Material, Fig. S5D). In the #122–1-13 cells expressing EGFP-IFT122(WT), as in control cells, SMO is excluded from cilia under basal conditions (Fig. 7B') and enter cilia upon stimulation of cells with SAG (Fig. 7F'). In striking contrast, ciliary entry of SMO in response to SAG was not significantly restored by the exogenous expression of EGFP-IFT122(W7C) or EGFP-IFT122(G513V) (Fig. 7G' and H', respectively). Similar results were obtained when #122–2-1 cells were transfected with IFT122(WT) or its mutant (see Fig. 7I).
In control cells and IFT122-KO cells expressing EGFP-IFT122(WT), GPR161 localized within cilia under basal conditions (Fig. 7, J and K', respectively) was exported from cilia upon the stimulation of cells with SAG (Fig. 7N and O', R). On the other hand, in IFT122-KO cells expressing EGFP-IFT122(W7C) or EGFP-IFT122(G513V), although statistically insignificant, less GPR161 tended to localize within cilia under basal conditions (Fig. 7L' and M', respectively; also see Fig. 7R).
Discussion
In this study, we demonstrated that IFT122 serves as a hub in the IFT-A complex at the molecular and cellular levels, as expected from the IFT-A architecture. We first showed that IFT122 connects the core and peripheral subcomplexes of the IFT-A complex via its CT and NT regions, respectively. The CED1-type mutations of IFT122 located in its NT region almost completely abrogated its interaction with the peripheral subunits. Of the IFT122 missense mutations that were analysed, G436R, V443G, and G513V abolished the direct interaction of IFT122 with the IFT43–IFT121 dimer, whereas IFT122(W7C) and IFT122(S263F) retained the ability to bind IFT43–IFT121 but lost the ability to indirectly interact with IFT139. Although we do not know the exact reason why the W7C and S263F mutations affected the indirect interaction of IFT122 with IFT139, the IFT43–IFT121 dimer in complex with the IFT122 mutant may not adopt a conformation that enables its interaction with IFT139. Anyway, our data indicate that CED1-type mutations of IFT122, in common, impair incorporation of IFT139 into the IFT-A complex.
We then established IFT122-KO cells and found that they demonstrate the most severe ciliogenesis defects compared with KO cells of the other IFT-A subunits. Furthermore, we showed by rescue experiments of the IFT122-KO cells that CED1-associated missense mutations, W7C and G513V, of the IFT122 gene disturb trafficking of ciliary proteins by disrupting the interaction of IFT122 with other IFT-A subunits, although these mutations have minor effects on ciliogenesis per se, as reported for cells from CED1 patients (23,24).
In contrast to KO cells of IFT121 (this study), IFT139 (21), and IFT144 (21), the IFT122-KO cells established in this study and cells derived from Ift122-KO mice (22) demonstrated no recognizable cilia. In agreement with this, an ift122-null zebrafish has been recently reported to display no or shortened cilia in a tissue- and age-dependent manner (39). This complete lack of cilia formation is probably due to the failed transport of tubulins that are cargoes of the IFT machinery and essential for construction of the microtubule-based axoneme (40,41). For example, the disruption of key subunits of the IFT-B complex, which directly binds tubulins and mediates their anterograde trafficking within cilia (40), completely abolishes ciliogenesis (19,30,42–44), whereas the absence of auxiliary IFT-B subunits has no or minor effects on ciliogenesis per se (16,17,31,45). These data suggest that IFT122 is also implicated in tubulin trafficking, although there has been no evidence for direct binding of the IFT-A complex to tubulins.
One possible explanation for this apparent disparity is that IFT122 in some way participates in the loading of tubulins to the IFT machinery. In this context, it is interesting to note the study of Lechtreck and colleagues using live-cell imaging of IFT particles and tubulins in Chlamydomonas flagella. They showed that tubulins are loaded onto the IFT particles in the basal body pool briefly before the particles move across the transition zone into cilia (46,47). Furthermore, their imaging data suggested that, whereas IFT-B proteins from retrograde IFT particles often re-enter the basal body pool, IFT-A proteins and motor proteins are recruited from the cell body source (46). In good agreement with this, we showed that in the absence of IFT122, IFT88 (an IFT-B protein) is still able to localize to the basal body, even though cilia are not formed and IFT140 (an IFT-A protein) is not found around the ciliary base, probably due to failed assembly of the functional IFT-A complex (Fig. 3).
In this study, we also clarified the molecular basis underlying the ciliary dysfunction caused by IFT122 mutations found in CED1 patients. The CED1-type missense mutations located in the IFT122 NT region containing the WD40 domain abrogated the ability of IFT122 to interact with peripheral IFT-A subunits (Fig. 2). When these CED1-type mutants or IFT122(WT) were expressed in IFT122-KO cells, ciliogenesis was restored (Fig. 4). However, unlike IFT122(WT), neither of these mutants was found around the ciliary base (Figs 5 and 6), suggesting their inability to be incorporated into the basal body pool of IFT particles. On the other hand, IFT88 was found at the bulged ciliary tip in IFT122-KO cells expressing the CED1-type mutant (Fig. 6). These observations suggest that, even in the absence of the functional IFT-A complex, the IFT-B complex can be trafficked anterogradely to the distal tips, where it accumulates due to the lack of IFT-A-mediated retrograde trafficking.
In IFT122-KO cells expressing the CED1-type mutant, ciliary entry of SMO upon SAG treatment was significantly impaired (Fig. 7). In view of the fact that the loading of cargo molecules occurs after the assembly of IFT particles in the basal body pool and immediately before their entry into cilia (46), it is possible that ciliary entry of SMO is mediated by IFT particles containing a functional IFT-A complex. Mitchell and colleagues recently reported that SAG-stimulated entry of SMO into cilia across the transition zone is severely impaired in cells derived from Inpp5e-KO mice (48). As ciliary localization of INPP5E was also impaired in IFT122-KO cells expressing the CED1-type mutant (Fig. 5), it is possible that the failed ciliary entry of SMO was an indirect consequence of the lack of INPP5E.
In summary, we have outlined the molecular basis of the ciliary dysfunction caused by CED1-associated mutations of IFT122, by a combination of flexible interaction assays involving the VIP method and analyses of KO cell lines followed by rescue experiments with CED1-type IFT122 mutants that were validated in advance by the VIP-based assays. With the rapid spread of next generation sequencing, there has been not only an increase in the number of disease-causative genes identified, but also an increase in the identification of mutations within known causative genes of various hereditary disorders, including ciliopathies (3,5). The combinatorial use of VIP-based, flexible protein–protein interaction assays and rescue experiments of KO cells with prevalidated disease-associated mutants will provide further information towards achieving a clear understanding of the molecular basis of these hereditary disorders.
Materials and Methods
Plasmids, antibodies, and reagents
Expression vectors for IFT122 and other IFT-A subunits and their deletion constructs used in this study are listed in Supplementary Material, Table S1; several of them were constructed in our previous study (21). The antibodies used in this study are listed in Supplementary Material, Table S2. GST-tagged anti-GFP nanobody (Nb) prebound to glutathione–Sepharose 4B beads were prepared as described previously (12). SAG was purchased from Enzo Life Sciences.
VIP assay and immunoblotting analysis
The VIP assay and subsequent immunoblotting analysis were performed by a previously described method (12,19) with slight modifications (33). Briefly, HEK293T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with high glucose (Nacalai Tesque) supplemented with 5% fetal bovine serum (FBS). Approximately 1.6 × 106 cells in six-well plates were transfected with EGFP and mChe fusion constructs using Polyethylenimine Max (20 µg, Polysciences), and cultured for 24 h. The cells were then lysed in 250 µl of HMDEKN cell lysis buffer (10 mM HEPES [pH 7.4], 5 mM MgSO4, 1 mM DTT, 0.5 mM EDTA, 25 mM KCl, and 0.05% NP-40) containing EDTA-free protease inhibitor cocktail (Nacalai Tesque). After 20 min on ice, cell lysates were centrifuged at 16, 100 × g for 15 min at 4 °C in a microcentrifuge. The supernatants (200 µl) were incubated with 5 µl of GST-fused anti-GFP Nb prebound to glutathione-Sepharose 4B beads for 1 h at 4 °C. The beads were washed three times with 180 µl of lysis buffer, and the precipitated beads were observed using an all-in-one-type fluorescence microscope (BZ-8000, Keyence) using a 20×/0.75 objective lens under constant conditions (sensitivity ISO 400, exposure 1/20 s for green fluorescence; and sensitivity ISO 800, exposure 1/10 s for red fluorescence).
Immunofluorescence analysis
Human retinal pigment epithelial hTERT-RPE1 cells (CRL-4000, American Type Culture Collection) were cultured in DMEM/F-12 (Nacalai Tesque) supplemented with 10% FBS and 0.348% sodium bicarbonate. To induce ciliogenesis, cells were grown to 100% confluence on coverslips, and starved for 24 h in Opti-MEM (Invitrogen) containing 0.2% bovine serum albumin. Subsequent immunofluorescence analysis was performed as described previously (21,49). The cells were fixed and permeabilized with 3% paraformaldehyde at 37 °C for 5 min and subsequently in methanol at −20 °C for 5 min, and washed three times with phosphate-buffered saline. For detection of endogenous IFT140, cells were fixed and permeabilized with methanol at −20 °C for 5 min, and washed three times with phosphate-buffered saline. The fixed/permeabilized cells were blocked with 10% FBS and stained with antibodies diluted with 5% FBS. The stained cells were observed using an Axiovert 200M microscope (Carl Zeiss). Statistical analyses were performed using JMP Pro 12 software (SAS Institute).
Establishment of KO cell lines using the CRISPR/Cas9 system
The strategy for targeted gene disruption of hTERT-RPE1 cells by the CRISPR/Cas9 system using homology-independent DNA repair was recently reported in detail (30); also see (21,31–33). Single guide RNA (sgRNA) sequences targeting the human IFT122 or IFT121 gene (see Supplementary Material, Table S3) were designed using CRISPR design (50). For disruption of the IFT122 and IFT121 genes, we applied the version 1 and version 2 systems, respectively, of a combination of the donor knock-in vector and an all-in-one sgRNA expression vector [pDonor-tBFP-NLS-Neo [Addgene #80766] and pSpCas9(BB)-2A-Puro (Addgene #48139), and pDonor-tBFP-NLS-Neo(Universal)] (Addgene #80767) and peSpCas9(1.1)–2 × sgRNA [Addgene #80768], respectively] (30). Cells grown to approximately 3.0 × 105 cells in a 12-well plate were transfected with 1 µg of the sgRNA and Cas9 containing vector and 0.25 µg of the donor vector using X-tremeGENE9 Reagent (Roche Applied Science), and cultured in the presence of G418 (600 µg/ml). The cells with nuclear tBFP signals were isolated under a microscope. Genomic DNA was extracted from the isolated cells and subjected to PCR using three sets of primers (Supplementary Material, Table S3) to distinguish the following three states of integration of the donor knock-in vector: forward integration, reverse integration, and no integration with a small indel [for the principle, see (30)]. To confirm disruption of the IFT122 and IFT121 genes, genomic PCR products were subjected to direct DNA sequencing.
Preparation of cells stably expressing EGFP-IFT122
The preparation of lentiviral vectors was described previously (49). Briefly, pRRLsinPPT-EGFP-IFT122(WT), pRRLsinPPT-EGFP-IFT122(W7C), or pRRLsinPPT-EGFP-IFT122(G513V) was transfected into HEK293T cells together with the packaging plasmids [pRSV-REV, pMD2.g, and pMDL/pRRE; provided by Peter McPherson, McGill University (51)]. The culture medium was replaced 8 h after transfection, and collected between 24–48 h after transfection. The medium containing viral particles was passed through a 0.45-µm filter and centrifuged at 32, 000 × g at 4 °C for 4 h. Precipitated lentiviral particles were resuspended in Opti-MEM (Invitrogen) and stored at −80 °C until use.
Supplementary Material
Supplementary Material is available at HMG online.
Acknowledgements
We thank Peter McPherson for providing the plasmids for recombinant lentivirus production and Helena Akiko Popiel for critical reading of the manuscript.
Conflict of Interest statement. None declared.
Funding
Grants-in-Aid for Scientific Research on Innovative Areas “Cilia and Centrosome” from the Ministry of Education, Culture, Sports, Science and Technology, Japan (grant number 15H01211 to Ka.N.); grants from the Japan Society for the Promotion of Science (grant numbers 15H04370 to Ka.N., and 15K07929 to Y.K.); and the Astellas Foundation for Research on Metabolic Disorders to Ka.N., and from the Takeda Science Foundation and the Uehara Memorial Foundation to Y.K.
References
Author notes
Mariko Takahara, Yohei Katoh and Kentaro Nakamura contributed equally to this study.