Oral-facial-digital (OFD) syndromes are rare heterogeneous disorders characterized by the association of abnormalities of the face, the oral cavity and the extremities, some due to mutations in proteins of the transition zone of the primary cilia or the closely associated distal end of centrioles. These two structures are essential for the formation of functional cilia, and for signaling events during development. We report here causal compound heterozygous mutations of KIAA0753/OFIP in a patient with an OFD VI syndrome. We show that the KIAA0753/OFIP protein, whose sequence is conserved in ciliated species, associates with centrosome/centriole and pericentriolar satellites in human cells and forms a complex with FOR20 and OFD1. The decreased expression of any component of this ternary complex in RPE1 cells causes a defective recruitment onto centrosomes and satellites. The OFD KIAA0753/OFIP mutant loses its capacity to interact with FOR20 and OFD1, which may be the molecular basis of the defect. We also show that KIAA0753/OFIP has microtubule-stabilizing activity. OFD1 and FOR20 are known to regulate the integrity of the centriole distal end, confirming that this structural element is a target of importance for pathogenic mutations in ciliopathies.
Cilia, which include the long-known motile flagella of swimming protists and sperm cells, have gained a great interest in the biology of human development and diseases during the last 15 years, because defects in these organelles are linked to an increasing number of genetic syndromes now called ciliopathies (1,2).
In vertebrates, cilia exist as primary cilia (PC), immotile solitary protrusions at the surface of many cell types, and as multiple motile cilia (MMC) at the surface of specialized epithelia such as the oviduct and the respiratory tract, which produce directional fluid flows of biological importance. A transient and solitary motile cilium is also present at the embryonic node to break left–right symmetry (3). PC are required, to various extent, for the cellular response to several signaling molecules including WNT, Hedgehog and platelet-derived growth factor. An alteration of this function underpins the developmental defects observed in patients with syndromic ciliopathies. Ciliopathies have various but overlapping phenotypes, such as polydactyly, intellectual disability, obesity, renal cystic disease, craniofacial abnormalities and retinitis pigmentosa.
The anchoring unit and seeding platform of a newly formed cilium are the modified centriole known as the basal body. PC are extending from basal bodies derived from the mother centriole, and MMC originate mainly from basal bodies amplified from daughter centriole-associated deuterosomes (4,5). Building a cilium requires the initial recruitment of post-Golgi vesicles at the distal end of centrioles by distal appendages that subsequently fuse together to initiate the formation of a ciliary shaft (6,7).
The construction and maintenance of a functional cilium rely on the activity of several molecular complexes. Remarkably, all of these complexes can be targeted in ciliopathies by gene mutation affecting individual components. The intraflagellar transport (IFT) complex associates with bidirectional molecular motors and carries building blocks, such as tubulin, into the cilium for extension and removes constituents to maintain a steady state (8,9). IFT assembly and turnover are regulated by the BBSome complex, which also regulates the ciliary trafficking of several signaling molecules (10,11). The BBS4 BBSome subunit associates with pericentriolar satellites (PSs) to regulate the recruitment of the BBSome into the cilium (12). PSs also associate with many centriole and basal body (CBB) proteins and may be assembly points for cilia proteins (13). Additional interlinked complexes of ciliopathy proteins were recently identified, localizing at the cilia transition zone and regulating its function as a ciliary gate to allow or prevent the selective entry of some membrane proteins (14).
Among ciliopathies, oral-facial-digital (OFD) syndromes are rare, clinically and genetically heterogeneous disorders characterized by the association of abnormalities of the face, the oral cavity and the extremities (15). OFD syndromes follow an autosomal recessive pattern of inheritance, except for the OFD I type that shows dominant X-linked inheritance and lethality in males. The OFD VI type is distinguished from the other types by the specific occurrence of cerebellar malformation, also described as molar tooth sign (MTS), and by metacarpal abnormalities. Because of these phenotypes, OFD VI is included into the Joubert syndrome-related disorders (JSRDs) (16). To date, causal mutations in nine ciliary genes (OFD1, C2CD3, TBC1D32/C6orf170, SCLT1, DDX59, WDPCP, C5orf42, TCTN3 and TMEM216) have been identified in OFD patients (17–21), and particularly OFD1, C5orf42 and TMEM216 in the OFD VI type. Interestingly, the best-characterized OFD proteins, OFD1 and C2CD3, bind together and colocalize at PSs and at the distal end of the centriole where they cooperate to regulate centriole length, providing an additional pathogenic mechanism in ciliopathies (22–24).
Here, we have identified two causal heterozygous mutations in the KIAA0753 gene in a newborn female presenting with an OFD VI syndrome. KIAA0753 is a conserved centrosome and PS protein that binds directly to FOR20, a protein required for the structural integrity of CBB distal ends and membrane anchoring (25,26). We show that a ternary complex is formed between FOR20 and OFD1 and KIAA0753 [hence subsequently named OFD1 and FOR20 interacting protein (OFIP)], and may be a prerequisite for recruitment onto PC and centrosomes, and for ciliogenesis.
KIAA0753 is mutated in OFD
Exome sequencing of blood cells DNA of a newborn female born with OFD syndrome type VI identified two heterozygous mutations in the KIAA0753 gene (RefSeq NM_014804.2) (see whole-exome statistics in Supplementary Material, Table S1). The female was born from non-consanguineous parents at 38 weeks of gestation. Birth measurements appeared normal, but she presented with facial dysmorphism (flat facial profile, straight palpebral fissures, hypertelorism, a wide nasal bridge with upturned nares, posteriorly low-set left ear and normal right ear), lobulated tongue, clefting of the alveolar ridges, left hand postaxial polydactyly, broad right and left hallux duplication and intermittent respiratory difficulties. Brain magnetic resonance imaging (MRI) evidenced vermis hypoplasia with MTS (Fig. 1A and 1B), agenesis of corpus callosum, ventricular dilatation and several other cerebral abnormalities. Abdominal ultrasound revealed bilateral hydronephrosis. OFD syndrome type VI was thus diagnosed due to the association of oral defects, polydactyly and MTS.
Both KIAA0753 mutations, one nonsense variant (c.1891A>T; p.Lys631*) and one substitution in Intron 8 (c.1546-3C>A), were confirmed by Sanger sequencing, as well as the maternal heterozygous status (Fig. 1C). The sporadic occurrence of the c.1546-3C>A variant was confirmed by samples concordance.
We hypothesized that the intronic substitution (c.1546-3C>A) caused a splicing defect and analyzed the effect of this variant on the splicing products. cDNA Sanger sequencing revealed a heterozygous skipping of Exon 8 that caused a frameshift, changed the amino acids sequence and led to the occurrence of a premature stop codon (Fig. 1C and D).
Because both mutations appeared truncating and probably compound heterozygous, they were considered as potentially causal.
Since the allelic OFD VI and JBSs are due to mutations in several genes, we also sequenced the KIAA0753 genes in 32 OFD VI or JBS individuals (Supplementary Material, Table S2). The secondary functional results (see below) led us to also sequence the FOR20 gene in the same cohort (Supplementary Material, Table S2). No additional mutation was identified.
KIAA0753 is a centrosome and PS protein interacting with FOR20
In parallel to clinical studies, our work to identify proteins involved in ciliogenesis had led to the identification of KIAA0753 as a binding partner of FOR20 (FOPNL), a recently characterized PS and basal body protein, using a two-hybrid screening approach (Supplementary Material, Fig. S1). KIAA0753 was subsequently named OFIP (see below). To further characterize OFIP, we produced a panel of monoclonal antibodies resulting from the immunization of rats with the 300 carboxy-terminal residues expressed in bacteria as a GST fusion protein. A mix of two selected antibodies detected OFIP as a ∼115 kDa protein in the soluble fraction of Radio-immunoprecipitation assay (RIPA) lysate of RPE1 cells (Fig. 2A). In contrast, OFIP was detected in the Triton-X100-insoluble fraction of RPE1 and lymphoblastoid KE37 cells, as expected for proteins associated with centrosomes and PSs (25). OFIP was indeed detected in a cellular extract enriched in centrosomes (Fig. 2A). IF stainings were done on methanol-fixed RPE1 cells to localize OFIP at different phases of the cell cycle. OFIP was detected at centrosomes and PSs and perfectly colocalized with FOR20 in interphasic and mitotic cells (Fig. 2B). PSs are typically defined by a staining with anti-PCM1 antibodies. We confirmed colocalization of OFIP and FOR20 on PSs by triple-labeling experiments on Hela Kyoto cells expressing Green fluorescent protein (GFP)-tagged OFIP [TransgeneOmics cells (27), see ‘Materials and Methods’ section] (Fig. 2C). We also co-labeled OFIP and PCM1 in pre-embedding immunoelectron microscopy experiments in RPE1 cells. Analysis of ultrathin sections showed the presence of OFIP and PCM1-specific grains on PSs (Fig. 2D).
We used Hela Kyoto TransgeneOmics cells that express GFP-tagged mouse OFIP or FOR20 at physiological level to confirm the interaction between the two proteins: immunoprecipitations with anti-GFP antibodies demonstrated the association of endogenous FOR20 with OFIP-GFP and of endogenous OFIP with FOR20-GFP (Fig. 3A).
To determine the portions of each protein involved in the interaction, co-immunoprecipitation (CoIP) experiments were done in Cos1 cells following ectopic expression of full-length and truncated or mutated proteins (see Fig. 5A for OFIP constructs). A first set of experiments demonstrated that the 83 C-terminal residues of OFIP (885–967) were necessary and sufficient to bind FOR20 (Fig. 3B and Supplementary Material, Fig. S2A). In a second set of experiments, various Myc-tagged FOR20 constructs (25) (Fig. 3C) were co-expressed with full-length GFP-tagged OFIP. Anti-Myc immunoprecipitates revealed that the poorly conserved C-terminus of FOR20 was dispensable for the interaction with OFIP (Fig. 3D). In contrast, the conserved FOR20 TOF domain and some residues surrounding the PLL motif were required for the interaction, but the integrity of this motif was not, because a mutation of the first leucine to an arginine in PLL did not preclude the interaction. In addition, mutation in the FOR20 LisH domain known to inhibit homodimerization (25) also affected the CoIP of the two proteins (Fig. 3D).
The use of human proteins as baits and preys in yeast two-hybrid experiments allows the detection of direct interaction in most cases, especially when the proteins are not conserved in yeast, which was the case in our study. Ponceau-S staining of immunoprecipitated Myc-FOR20 associated with GFP-tagged OFIP C-terminus after sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) and transfer onto membrane demonstrated the sole presence of the two overexpressed proteins in the immunoprecipitates, together with Ig chains (Supplementary Material, Fig. S2B). Thus, the interaction is very likely to be direct.
OFIP overexpressed in cells decorates microtubules (MTs) (see below for more details). Therefore, the association of OFIP and FOR20 was also ascertained in Cos1 cells by the capacity of overexpressed GFP–OFIP to recruit FOR20 onto MTs. As expected, this recruitment did not require the C-terminus of FOR20, but was dependent on OFIP C-terminal tail (Supplementary Material, Fig. S3).
OFIP, FOR20 and OFD1 form a complex associated with PSs and centrosomes
We next used a proteomic approach to identify additional proteins interacting with OFIP. Non-transfected cells and cells expressing GFP only were used as controls. In addition, OFIP constructs that could associate with FOR20 (OFIP 300–967 and 667–967) were co-expressed with mCherry-tagged FOR20 to identify proteins binding to OFIP indirectly through FOR20 or to the OFIP-FOR20 binary complex. GFP-tagged and copurifying proteins were isolated from lysates using anti-GFP nanobodies (see ‘Materials and Methods’ section), digested with trypsin, and analyzed by mass spectrometry (see Supplementary Material, Table S3 for complete results). Interestingly, we could specifically identify PCM1 associated with OFIP N-terminal region (1–299) and OFD1 with the C-terminal region (300–967 and 667–967) only when overexpressed with FOR20 (Table 1).
aGFP-tagged OFIP fragments expressed in RPE1 cells for pull-down and mass spectrometry analysis.
bGFP-tagged OFIP fragment were co-expressed with FOR20-mCherry.
cPercentage of coverage.
Total number of peptides.
CBB, centriole and basal body; IFT, intraflagellar transport; JBS, Joubert syndrome; JSRD, Joubert syndrome-related disorders; MMC, multiple motile cilia; MRI, magnetic resonance imaging; MT, microtubule; MTS, molar tooth sign; OFD, oral facial digital; PC, primary cilium; PS, pericentriolar satellite.
We used Hela Kyoto TransgeneOmics cells expressing mouse GFP-tagged OFIP or PCM1 to confirm the interaction. We could demonstrate CoIP of endogenous OFD1 and PCM1 with OFIP–GFP and of endogenous OFIP with PCM1–GFP (Fig. 4A).
To better define a putative OFIP-FOR20-OFD1 complex, different combinations of proteins tagged with various recognition sequences were transfected in Cos1 cells, and the associations were probed following immunoprecipitation and immunoblotting. The results are shown in Figure 4B and Supplementary Material, Figure S4. The conclusion of this series of experiments is that a weak interaction between the OFIP C-terminus and OFD1 was strongly increased by the presence of FOR20 in the complex (see Lanes 5 and 6 in Fig. 4B, right panel). Remarkably, OFD1 and FOR20 did not associate in the absence of OFIP (Lane 7 in Fig. 4B, right panel). The conserved N-terminus of OFD1 was necessary and sufficient to strongly interact with OFIP C-terminus in the presence of FOR20.
In support of our biochemical observations, overexpressed full-length OFD1 or its conserved N-terminus colocalized with OFIP on MTs in transfected cells in the absence or presence of FOR20 (Supplementary Material, Fig. S5A). In addition, the three proteins expressed together colocalized perfectly. The lack of interaction of OFD1 and FOR20 in the absence of OFIP was also confirmed by the lack of colocalization of the two proteins when overexpressed in Cos1 cells. The requirement of the most C-terminal 82 residues of OFIP and OFD1 N-terminus for complex formation was also confirmed in colocalization experiments (Supplementary Material, Fig. S5B). Thus, OFD1 can interact with OFIP in a FOR20-independent manner, but FOR20 increases the stability of the interaction.
Finally, we showed that OFIP, OFD1 and FOR20 have similar localization at PSs and centrioles (Fig. 4C).
Taken together these results suggest that OFIP, FOR20 and OFD1 form an independent ternary complex that may associate with PSs through the binding of OFIP N-terminus to PCM1.
The central region of OFIP stabilizes MTs
The subcellular localization of full-length and truncated OFIP constructs (Fig. 5A) was assessed by IF after expression in RPE1 (Supplementary Material, Fig. S6) and NIH3T3 cells (not shown). The N-terminus (residues 1–299) was sufficient for localization to PSs and centrosomes. Strikingly, overexpressed full-length OFIP and its isolated central region (residues 300–666) and truncated proteins retaining this region strongly decorated a population of cytoplasmic MTs (Supplementary Material, Fig. S6). Expression of these latter OFIP constructs induced MT resistance to depolymerization by high concentrations of nocodazole (Fig. 5B). MT-pelleting assays showed a quantitative increase of insoluble MTs from nocodazole-treated cells expressing GFP-OFIP, but not from control cells, in two different experimental settings, confirming that OFIP protects from depolymerization and stabilizes MTs (Fig. 5C). The stabilization effect was also observed in the presence of the central 300–666 OFIP fragment (Fig. 5C, right panel).
Depletion of individual proteins of the complex affects centrosomes, PSs and ciliogenesis
Each protein of the above-identified protein complex was down-regulated by siRNA transfection in RPE1 cells and cells were analyzed by IF. PCM1-labeled PS were poorly affected by depletion of OFIP, OFD1 and FOR20 (Fig. 6A and B). In contrast, PCM1 depletion suppressed PS localization of all three proteins as expected. Remarkably, however, the localization of all three proteins in cytoplasmic PSs and in close proximity of the centrosome showed strong inter-dependence, mainly at PSs (Fig. 6A and B). This indicates that an intact protein complex is required for localization at the PSs, and possibly at the centrosomes. When PC were studied in serum-starved siRNA-treated cells, a 15–20% decrease in cilium length was observed (Fig. 6C).
The OFIP-OFD1-FOR20 complex is defective in OFD patient cells
The heterozygous compound mutations of OFIP in the OFD VI patient predict OFIP products lacking the C-terminus (Fig. 1). In agreement with this prediction, the immunoblot analysis of control and patient extracts from Epstein–Barr virus (EBV)-immortalized cells with anti-OFIP mAb recognizing the C-terminus barely detected OFIP in patient cells (Fig. 7A). This was confirmed by IF staining of patient cells, which showed faint OFIP at the centrosomes, contrasting with OFIP protein abundance at the centrosomes and PSs in control cells (Fig. 7B). This indicates that the de novo splicing site mutation somehow allows minimal normal splicing. Because of the lack of antibodies reacting with the N-terminus of OFIP, it was not possible to control the production of truncated proteins, which could anyway be absent due to nonsense-mediated mRNA decay.
Importantly, in agreement with the experiments in RPE1 cells, the patient cells showed a strong reduction of FOR20 and OFD1 at the centrosome and PSs (Fig. 7B). CEP290, another transition zone and centriolar protein associated with PSs was not affected, excluding a more general effect (Supplementary Material, Fig. S7). This reduction is likely the consequence of lower levels cellular pools of OFD1 and FOR20 (Fig. 7A), suggesting that the formation of a ternary complex is required for the stability and subcellular localization of these proteins.
OFIP is an evolutionary-conserved protein in ciliated species
To find further evidence of OFIP association with the cilium, we examined its phylogenetic distribution, i.e. is OFIP present only in species with cilia? Cilia are ancient eukaryotic organelles that, together with their genes, have been lost multiple times in evolution like in most Fungi and in the flowering plants. Comparative genomics has been proved to be powerful in finding new ciliary proteins, including ones associated with ciliopathies (28). We performed sensitive sequence similarity searches to identify OFIP homologs. This resulted in the identification of OFIP homologs in animals, but also in distantly related ciliated species like Chlamydomonas reinhardtii, Paramecium tetrauralia, Leishmania major and one of the few ciliated fungi, the Chytrid Batrachochytrium dendrobatidis. No homologs could be detected in species without cilia, like most Fungi and flowering plants, and in some species with minimal cilia like Plasmodium falciparum and Cryptosporidium parvum. Species that do have a homolog of OFIP have only a single copy indicating that the proteins are likely orthologs. These observations suggest that OFIP is a cilia specific protein.
OFIP does not appear to have any known protein domains. We applied IUpred (29) and observed that OFIP consists of a large extent of disordered non-globular structures. The C-terminus of the protein is one of the few regions that are predicted to be globular. It is the C-terminus of the protein, which is conserved among orthologs (Fig. 8). It contains a peculiar pattern of alternating conserved negatively charged residues and hydrophobic residues that are predicted to form an α-helix by PhD (30). From such alternating charged and hydrophobic residues one might expect an amphipathic helix, but helical wheel projections did not show enrichment of either charged or non-charged residues on one side of the helix (data not shown).
To further study potential interactions between OFIP and other proteins linked to OFD (OFD1, C2CD3, TMEM107, TMEM216, C5orf42, WDPCP, DDX59, TCTN3, SCLT1 and TBC1D32) as well as FOR20, we examined whether their genes have similar phylogenetic distributions. Although almost all genes in this group occur only in ciliated species (with the exception of DDX59), their phylogenetic distributions are not very similar to OFIP, except for TMEM107 and FOR20, suggesting that either, or both are somehow functionally linked to OFIP (Supplementary Material, Fig. S8). Interestingly, OFIP and FOR20 appear to form an evolutionary module with other known transition zone proteins (TMEM107, TCTN3 and TMEM216), which would be consistent with a previously described role of FOR20 in the assembly of the transition zone (26,31), suggesting that OFIP may be part of the same assembly pathway.
The present study extends both our knowledge of molecular players of ciliogenesis and their contribution to the pathogenesis of human diseases. Among human syndromic diseases of genetic origin, ciliopathies have recently emerged as a prominent group characterized by developmental abnormalities linked to defects in primary and motile cilia (2,32). They can be categorized according to the combination of phenotypes they display, which are overlapping over different syndromes. Accordingly, some genes are targets for mutations in specific syndromes, but the involvement of one gene in two or more syndromes is becoming common from the results of sequencing of more and more exomes (33). OFD syndromes are characterized by mouth and craniofacial abnormalities and polydactyly, which can be associated with other features such as kidney diseases and central nervous system malformations. To date, 13 types have been classified based on this variability (15). OFD syndrome I is the most common and well-characterized OFD syndrome and is caused only by mutation in the OFD1 gene, which is also mutated in JBS (18,34,35). Genes causal to other OFD types can be found mutated in JBTS, Bardet–Biedl and Meckel syndromes, emphasizing the genetic and phenotypic complexity of ciliopathies. Our work has identified a causal compound heterozygous mutation of the OFIP/KIAA0753 gene in one case of OFD type VI. Distinctive features of this subtype are cerebellar abnormalities with MTS and central polydactyly. This is the first described mutation in this gene, which codes for a recently identified centrosomal and PS protein (36,37). This rarity of OFIP mutations may be due to an embryonic lethal effect. The case studied here may result from a hypomorphic mutation, due to leakage of the correctly spliced OFIP product. We previously reported CBB defects in patients and mutant mice with OFD syndromes caused by mutations in the centrosome and PS proteins OFD1 and C2CD3, resulting in undocked basal bodies and short centrioles, respectively (24,38).
Preliminary functional characterization of OFIP indicated a possible role in centriole duplication (37). We have here extensively characterized OFIP, showing that it is a direct interactor of FOR20, a protein involved in primary ciliogenesis in human cells and required for assembly of the cilium transition zone and basal body anchoring at the cell cortex in Paramecium (25,26). We show that OFIP directly interacts with OFD1 in the absence of FOR20, but the interaction between OFIP and OFD1 is much stronger when FOR20 is present in the complex. FOR20 and OFD1 have a similar localization at the distal part of CBB and are necessary for basal body anchoring (22,26,38), in agreement with their inclusion in a unique complex. Cells lacking either FOR20 or OFD1 have abnormalities at the CBB distal ends (22,26). However, these observations have been made in different model systems, i.e. Paramecium and human cells, respectively. Remarkably, OFD1 also interacts with C2CD3, and the two proteins are proposed to play antagonistic role in regulating the length of centriole distal ends (24). C2CD3 may thus be associated with the ternary complex we have described, but was not identified in our proteomic analysis of OFIP- and/or FOR20-associated proteins. That OFD1, C2CD3 and OFIP are targets of causal mutations in OFD syndromes confirms the hypothesis that affecting CBB distal end length and/or structure by disrupting this molecular complex is a pathogenic mechanism of some ciliopathies.
The mutations described here disrupt the integrity of the complex as they truncate the C-terminus of OFIP, which interacts with FOR20 and OFD1, and strongly down-regulate OFD1 in patient cells. Remarkably, the C-terminal end is the conserved region among the OFIP orthologs which, together with the similar conservation of FOR20 and OFD1 (Supplementary Material, Fig. S8), stresses out the biological importance of this complex. Overexpressed OFIP decorates and stabilizes a population of MTs, but does not organize in rod-like fashion at the centrosome, like an overexpressed C2CD3 (24). It will be interesting to know if these structures, which correspond to elongated centrioles, require and/or recruit OFIP to stabilize longer MT triplets. The centrosomal protein FOP is evolutionary related to FOR20 and OFD1 and interacts with the CEP350 protein, which stabilizes Golgi-associated and centriolar MTs (39,40). An emerging scheme of mechanism may be that related molecular complexes at CBB associate MT-stabilizing/-binding activity coupled with an unknown structural or functional role of FOR20, OFD1 and FOP. In support of this, the molecular determinant of the OFIP-FOR20 and CEP350-FOP interaction are very similar, involving the last C-terminal residues on one partner and the conserved TOF and LisH-domain containing N-terminus on the other. The activity of the OFIP-associated complex is likely to take place at the centriole distal end to regulate MT triplet length and assembly of the adjacent transition zone. As recently established, several protein complexes involving ciliopathy proteins localized at the transition zone to regulate its assembly and function as a cellular gate to organize cilium structure and signaling pathways (14,41–43). How the distal end of centriole participates in the assembly of the transition zone is unclear. A number of transition zone proteins are already present on undocked centrioles indicating that transition zone is partly organized before cilium formation (44). Recently, the CEP162 centriole distal end protein has been shown to promote assembly of the transition zone by connecting core transition zone component to ciliary MTs (45). It will be of interest to study transition zone assembly in the absence of OFIP and its partners to assign it a role similar to CEP162 in ciliogenesis.
Our study also points out to a role of complex formation in the stability of OFD1 and to a lower extent of FOR20. An effective quality control mechanism exists in the cytoplasm to recognize and eliminate misfolded proteins by the ubiquitin-proteasome-dependent pathway (46). It is possible that correct folding of OFD1 and FOR20 is dependent on their interaction with OFIP, but the possible involvement of chaperone to assist this process, like in BBSome formation, is not excluded (47).
Recently, >20 proteins were characterized as associated with PSs (48,49). Most of them are also centrosome and basal body proteins, raising the possibility that PSs play a role in ciliogenesis. However, the depletion of PCM1 in differentiating mouse trachea cells does not inhibit centriole multiplication and MMC formation (50). In contrast, several studies point to a role of PSs in the formation of PC and an emerging picture is that (i) PCM1 depletion inhibits primary ciliogenesis (12,51,52), (ii) PSs adopt a regulated dynamics during primary ciliogenesis, a process regulated by TALPID3 and CEP290 (53) (iii) PS proteins, such as CEP72, CEP290 and SSX2IP control the accumulation of RAB8 and the BBSome complex, two key effectors of cilia formation (12,54) and (iv) PSs are an assembly point for proteins affected in ciliopathies such as ODF1, CEP290 and BBS4 (13). PSs and PCM1 are only present in metazoa and possibly arose during evolution to regulate ciliogenesis in the context of complex tissues. Strikingly, however, no mutations in PCM1 were detected in ciliopathies. Like for OFIP, PCM1 mutations may be incompatible with the development of an embryo. Both OFIP (this study) and OFD1 are able to interact with PCM1 (13), the major constituent of PS. Because OFIP and OFD1 are conserved in ciliated protists, the two proteins have probably evolved to acquire the capacity of binding PCM1 and PS in metazoa. The ternary complex we have identified is likely a cargo of PSs to be addressed to CBB. PSs are indirectly associated with the dynactin subunit p150glued and move along MTs in a dynein–dynactin-dependant manner (55–57). It has been reported that kinesin 1 has higher affinity for acetylated and detyrosinated (58,59) and that tubulin glutamylation, another feature of stable MTs, regulates axonemal dynein (60). It is thus possible that local stabilization of MTs by PS-associated OFIP is important for the dynamics of PSs.
The recent plethora of information on the phenotypic and genetic complexity of ciliopathy does not fit with a simple genotype–phenotype correlation and raises the central question of the qualitative and quantitative signaling defect caused by a given mutation or combination of mutations. For a subset of ciliopathies, this is underpinned by the hypothesis that mutations in different genes affect the gating capacity of the transition zone to similar extent, thus giving similar phenotypes. KIAA0753 and also FOR20 are thus candidate genes for most non-motile ciliopathies described to date.
In conclusion, we have uncovered a previously unidentified protein complex involved in centriole and cilium regulation, and affected in ciliopathies, particularly in the OFD syndromes.
Materials and Methods
Exome sequencing was performed on the patient DNA according to standard procedures using the SureSelect Human All Exon V2 kit (Agilent), using three micrograms of genomic to whole-exome capture. The resulting libraries were sequenced on a HiSeq 2000 (Illumina) as paired-end 75 bp reads in accordance with the manufacturer's recommendations. Raw data were processed as previously described (24). BAM files were aligned to a human genome reference sequence (GRCh37/hg19) using BWA (Burrows–Wheeler Aligner; v0.7.6) and potential duplicate paired-end reads were removed by Picard 1.77. Indel realignment and base quality score recalibration were conducted with Genome Analysis Toolkit (GATK; v2.1-10). Variants with quality scores of <30, allele balance of >0.75, sequencing depth of <4, quality/depth ratio of <5.0, length of homopolymer run of >5.0 and strand bias of >−0.10 were flagged and excluded from subsequent analyses. Variants with a quality score of >30 and alignment quality score >20 were annotated with SeattleSeq SNP Annotation (see URLs). Rare variants present at a frequency >1% in dbSNP 138 and the NHLBI GO Exome Sequencing Project or present from 69 local exomes of unaffected individuals were excluded (see URLs). Variant filtering was as follows: (i) variants affecting the coding sequence, (ii) rare variants, absent from public databases (see URLs), (iii) homozygous or compound heterozygous variants and (iv) ciliary/centrosomal genes.
OFIP/KIAA0753 and FOR20 mutation validation
Genomic DNA was amplified by polymerase chain reaction (PCR) using HotStarTaq PCR kit (Qiagen) according to the manufacturer's protocol. OFIP/KIAA0753 and FOR20 primers were designed on the RefSeq NM_014804.2 and NM_ 144600.2) (primers available on request). PCR products were purified by the Agencourt CleanSEQ system (Beckman Coulter) and sequenced with the BigDye Terminator Cycle Sequencing kit, v3.1 (Applied Biosystems) in ABI 3730 sequencer (Applied Biosystems). Sequence data were analyzed using Mutation Surveyor v4.0.9 (Softgenomics).
OFIP/KIAA0753 cDNA analysis
Total RNA was isolated with TRIzol Reagent (Life Technologies) from EBV transformed lymphoblastoid cell line of the patient, according to the manufacturer's instructions. One microgram of RNA was transcribed into cDNA with the QuantiTect Reverse Transcription Kit (Qiagen). Using PCR primers positioned in Exons 6, 10 and 11 in KIAA0753 gene (primers available on request), cDNA was amplified using the HotStartTaq Plus DNA Polymerase Kit (Qiagen) following the instructions provided by the manufacturer.
Replication cohort sequencing
All coding region of OFIP/KIAA0753 and FOR20/FOPNL gene was amplified by PCR long-range with PrimeStar GXL kit (Takara) for 32 patients of replication cohort (primers available on request). PCR products were pooled for each case. Nextera XT DNA Preparation kit (Illumina) was used to create a DNA library tagged and fragmented according to the manufacturer's protocol. Samples obtained were pooled and sequenced in Miseq (Illumina). Resulting data were aligned to the gene reference sequence (GRCh37/hg19) using BWA (Burrows–Wheeler Aligner; v0.7.6). The Genome Analysis Toolkit (GATK; v2.1-10) enabled indel realignment and base quality score recalibration. Variants with a quality score of >30 and alignment quality score >20 were annotated with SeattleSeq SNP Annotation.
The baits were clones into pDBa (Leu) using the Gateway technology (Invitrogen™). The bait plasmids were transformed in MAV03 yeast strain (MATα; leu2-3,112; trp1-901; his3Δ200; ade2-101; gal4 Δ; gal80 Δ; SPAL10UASGAL1::URA3, GAL1::lacZ, GAL1::His3@LYS2, can1R, cyh2R) in accordance with the previously described transformation protocol (61). The self-activation of the resulting yeast strains was tested with the four phenotypes. Mav203 cells were then transformed with the cDNA library of choice (cloned into pEXP502-AD (Trp), Proquest libraries™, Invitrogen™) as described previously (61). Following transformation with the cDNA library, yeast were plated onto synthetic complete (SC) medium minus leucine (−L), minus tryptophane (−W), minus histidine (−H) +25 mm 3-amino-1,2,4-triazole (3-AT) and incubated at 30°C for 4–5 days. Positive clones where then selected and patched onto SC-WHL +3-AT in 96-well plate format and incubated for 3 days at 30°C. Yeast clones are then transferred in liquid SC-WL for 3 days at 30°C with agitation to normalize the yeast cell concentration used for the phenotypic assay. The cells are then diluted 1/20 in water and spotted onto different selective medium (−WHL + 25 mm 3-AT, −WL + 0.2% 5-5-fluoroorotic acid and −WUL). To perform the β-galactosidase assay, the undiluted yeast cells are spotted onto yeast extract peptone dextrose (YPD) medium plate with a nitrocellulose filter.
Plates are incubated at 30°C for 4–5 days, except the YPD plate which is removed after 1 day, and β-galactosidase activity is evaluated as described.
The OFIP/KIAA0753 full-length cDNA in pBluescript II was obtained from the Kazusa DNA Research Institute (Chiba, Japan). The OFD1 ORF cDNA was obtained from Genecopoeia™.
The sequence coding for the full-length and truncated proteins were amplified by PCR using Platinum HiFi polymerase (Invitrogen) and Gateway-compatible primers and transferred into pDONR/ZEO using Gateway technology-mediated recombination. The sequences were subsequently transferred by recombination into modified Gateway-compatible pRK5/Myc, pEGFP-C1 and pmCherry-C1 expression vectors.
Cells, antibodies and reagents
Immortalized human retinal pigment epithelial cells htert-RPE1 (thereafter referred as RPE1) were grown in Dulbecco's modified Eagle's medium F12 (DMEM/F12) (Life Technologies) supplemented with 10% heat-inactivated fetal calf serum (FCS). Hela Kyoto and Cos1 cells were grown in DMEM supplemented with 10% heat inactivated FCS. Lymphocytes (patient and control) cells were grown in RPMI supplemented with 10% heat inactivated FCS.
Anti-human OFIP monoclonal antibodies were obtained following immunization of rats with the OFIP carboxy terminus (residues 667–967) fused to GST protein. Immune spleen cells were fused with the non-secreting myeloma X63-Ag8.653 cells and grown in selective medium. Hybridoma supernatants were first screened by ELISA on GST-OFIP and GST alone. GST-OFIP-only positive populations were screened by immunofluorescence on methanol-fixed RPE1 cells and western blotting of lysates from Cos1 cells overexpressing Myc-tagged OFIP. Selected clones were cloned three times before large-scale antibody production.
A rabbit anti-OFIP/KIAA0753 antibody was from Sigma-Aldrich. Anti-γ-tubulin antibodies were from Sigma-Aldrich (mouse monoclonal GTU-88 and rabbit polyclonal T3559). Rabbit anti-PCM1 was from Bethyl Laboratories. Mouse anti-c-Myc (9E10) was from Santa Cruz Biotechnology. Anti-GFP antibodies were purchased from Abcam (rabbit polyclonal) and Roche (mouse monoclonal).
For plasmids, cells were transfected with Fugene6 or Fugene HD (Roche Applied Science) according to manufacturer's protocol.
siRNA oligonucleotides were transfected into cells using INTERFERin® (Polyplus Transfection) according to the manufacturer's protocol. The sequences are the following: OFIP (5′-GCUCAAAGCUGAAGAAAUG-3′), FOR20 (5′-GAGAGUAUUUAGAAUUCAA-3′) (25), PCM1 (5′-UCAGCUUCGUGAUUCUCAGTT-3′) (55), OFD1 (5′-GCUCAUAGCUAUUAAUUCA-3′) (13), CEP290 (pool of 4 siRNAs: 5′-GAAGUAGAGUCCCUCAGAA-3′, 5′-GAAAGUUAAUGAGCAAUUG-3′, 5′-GGAAUUGACUUACCUGAUG-3′, 5′-GGAUUCGGAUGAAUGAAA-3′). A non-targeting siRNA was used for controls. All siRNAs were purchased from Dharmacon as ON-TARGETplus siRNAs.
Cells were washed in phosphate buffered saline (PBS) and lysed in 20 mm Tris–HCl pH 7.5, 2 mM EDTA, 150 mM NaCl, 1% Igepal CA630 (Sigma) or 50 mM Tris–HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, containing 1% NP-40 and 0.25% sodium deoxycholate (modified RIPA) plus a complete protease inhibitor cocktail (Roche Applied Science) buffer on ice. After centrifugation for 20 min at 13 000 rpm at 4°C, cleared lysates were obtained.
Triton-X100 soluble and insoluble fractions were obtained as follows: cells were washed in PBS, lysed in PHEM buffer (45 mM PIPES, 45 mM HEPES, 10 mM ethylene glycol tetra-acetic acid (EGTA), 5 mM MgCl2, pH 6.9) plus 1% Triton X-100 and protease inhibitors and centrifuged at 400 g for 45 min. The supernatant was collected (soluble fraction) and the pellet was washed in PHEM alone and solubilized in 1% SDS containing buffer (insoluble fraction) and sonicated. Cell extracts separated on polyacrylamide gels were transferred onto Hybond™-C membrane (GE Healthcare Life Science) followed by detection with antibodies.
Immunoprecipitation and western blot analysis
For anti-Myc immunoprecipitation, extracts were incubated with agarose beads conjugated to mouse anti-Myc antibodies (Santa Cruz Biotechnology) for 3 h at 4°C. Beads were pelleted and washed five times with lysis buffer. For anti-GFP immunoprecipitations, extracts were incubated with agarose beads conjugated to rabbit anti-GFP antibodies (Abcam) or with GFP-Trap® alpaca anti-GFP-conjugated agarose beads. Samples were separated using in house SDS–PAGE or NuPAGE 4–12% Novex Bis–Tris gels according to manufacturer's instructions (Invitrogen).
Cells were grown on coverslips, fixed in cold methanol (6–8 min at −20°C) and incubated with the appropriate mixture of primary antibodies for 30 min at room temperature. After washing in PBS 0.1% Tween, they were incubated for another 30 min with the adequate secondary antibodies conjugated to cy2, cy3 or cy5 (Jackson laboratories) to which was added 250 ng/ml DAPI for DNA staining, rinsed and mounted in Prolong Gold anti-fade reagent (Life Technologies). Confocal images were acquired by capturing Z-series with 0.3 µm step size on a Zeiss LSM 510 laser scanning confocal microscope.
Cells grown on Lab-Tek permanox Chamber were permeabilized 6 min in 45 mM PIPES, 10 mM EGTA, 5 mM MgCl2, pH 69 (PEM) 0.25% triton and were prefixed in PEM 0.25% glutaraldehyde 10 min at room temperature. After three washes in PBS containing 0.1% bovine serum albumin (BSA) and 0.1% Tween, the cells were incubated with anti-OFIP (1/20 000) and anti-PCM1 (1/1000) antibodies for 1 h. The cells were washed three times with PBS containing 0.1% BSA and 0.1% Tween-20, and were incubated with goat anti-rat antibody conjugated to 6 nm colloidal gold (1/40; Jackson Laboratories) and goat anti-rabbit antibody conjugated to 18 nm colloidal gold (1/40; Jackson Laboratories) for 1 h. For electron, microscopy cells were fixed with 2.5% glutaraldehyde and processed for dehydration and embedding, as described previously (62).
MT stability assay
Cells were lysed with 2 mµ EGTA, 1 mM MgCl2 (EM) buffer and proteases inhibitor coktail 10 min on ice and homogenized with 10 strokes of potter. Of note, 0.5 M K-Pipes was added for a final concentration of 100 mM. Lysates were centrifuged at 4°C for 1 h at 55 000 rpm. Supernatants were incubated with 1 mM GTP, 1 mM DTT ± 10 µM nocodazole and ±50 µg purified porcine tubulin for 90 min at 37°C and overlaid on PEM buffer (100 mM K-Pipes [pH 6.8], 1 mM EGTA 1 mM MgCl2) containing 10% sucrose. After centrifugation at 25°C for 30 min at 19 000 rpm, supernatants and pellets were analyzed by electrophoresis and western blotting.
The image analysis was performed on maximum intensity projection images that were generated from confocal stacks. Image processing was done using ImageJ.
To quantify the centrosome fluorescence, regions in these images were defined by γ-tubulin staining for each cell using threshold. Centrosome mask was created, and integrated density of individual centrosomes within the mask was measured.
To quantify satellites fluorescence, the centrosome mask was subtracted to mask cell obtained using threshold. Integrated density was measured. All data were analyzed by Prism.
We used BLAST (63) for initial sequence searches for orthologs of OFIP and other OFD-related genes. To identify additional homologs, we employed Jackhammer (64) to perform profile-based homology searches. We compared the OFIP phylogenetic pattern with those of transition zone and centriolar proteins using a predefined set of organisms used previously (Lambacher et al., submitted for publication). UIpred was used to predict disordered regions and globular folds within the protein structure. The protein C-terminal sequences of OFIP orthologs were aligned with clustalx (65) and visualized with JalView (66).
NHLBI Exome Sequencing Project Exome Variant Server, http://evs.gs.washington.edu/EVS/
University of Burgundy Centre de Calcul, https://haydn2005.u-bourgogne.fr/dsi-ccub/Seattle
Seq Annotation tool, snp.gs.washington.edu/SeattleSeqAnnotation137/
Human Gene Mutation Database HGMD, http://www.hgmd.org/
This work was supported by INSERM and Institut Paoli-Calmettes and by grants from the GIS-Institut des Maladies Rares (HTS), the French Ministry of Health (PHRC national 2010-A01014-35 to C.T.-R.), the Italian Telethon Foundation (TGM11CB3 to B.F.), the Regional Council of Burgundy (to C.T.-R.) and the European Community's Seventh Framework Programme FP7/2009 under grant agreement 241955 SYSCILIA. The Marseille proteomic plateform is recognized by IBiSA (Infrastructures en Biologie-Santé et Agronomie) and supported by Institut Paoli-Calmettes, Inserm and Aix-Marseille University.
We thank the patients and their families for their participation. We thank the Integragen society for exome analysis. We also thank the NHLBI GO Exome Sequencing Project (see URLs) and its ongoing studies which produced and provided exome variant calls for comparison: the Lung GO Sequencing Project (HL-102923), the WHI Sequencing Project (HL-102924), the Broad GO Sequencing Project (HL-102925), the Seattle GO Sequencing Project (HL-102926) and the Heart GO Sequencing Project (HL-103010). We are grateful to Aicha Aouane and Jean-Paul Chauvin (Electron Microscopy service, Institut de Biologie du Développement de Marseille, Marseille, France) for assistance on electron microscopy experiments and to Michel Pierres (Centre d'Immunologie de Marseille-Luminy, Marseille, France) for helping with antibody production. We acknowledge the France-BioImaging/PICsL infrastructure (supported by the French National Research Agency, ANR-10-INSB-04-01, ‘Investments for the future’). We thank Andrew Fry (University of Leicester, Leicester, UK) for the gift of OFD1 plasmid and antibody. We are also indebted to Ina Poser and Anthony Hyman (Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany) for the TransgeneOmics cell lines and BAC constructs. J.P.B. is scholar of Institut Universitaire de France.
Conflict of Interest statement. None declared.