Discussion
In this study, we identified three POC1B mutations that cause autosomal-recessive COD or CRD. In two siblings of one family, loss of central vision was observed in childhood, consistent with progressive cone disease; however, in another sibling and in one isolated individual, poor visual acuity and nystagmus were present from early infancy, suggesting a form of ACHM. In the latter two individuals, visual acuity also deteriorated over time, and in one of them, peripheral retinal degeneration was observed in the sixth decade. Although there is overlap in genes associated with either ACHM or COD as a result of mutations in CNGA3 (MIM 600053) and CNGB3 (MIM 605080),5,30,31 there are, to our knowledge, no reports on the natural history of ACHM concerning peripheral degeneration.
POC1B is one of the two POC1 homologs that function together as a highly conserved core centriole and basal body component in vertebrates,15,32–34 invertebrates,35,36 and even Chlamydomonas reinhardtii32 and Tetrahymena thermophila.33 The other POC1 homolog, encoded by POC1A (previously Pix2 [MIM 614783]), shows protein structure and intracellular localization similar to those of POC1B.20,36 Studies in Tetrahymena thermophila suggest that POC1 proteins are essential for both structure and stability of the basal body.15 Depletion studies show that POC1B, unlike POC1A, is necessary for ciliogenesis, and typical ciliopathy-associated developmental defects (e.g., curved body axis, kidney cysts, and laterality defects) were described in poc1b morphant zebrafish. Interestingly, they were also reported to exhibit smaller eyes, but a more detailed ophthalmological analysis was not undertaken.15
In light of the retinal phenotype we observed in affected individuals with POC1B mutations and the reported smaller eyes in poc1b morphant zebrafish, we disrupted poc1b expression by using the same translation-blocking MO used by Pearson et al.15 An accurate evaluation of Poc1b function in the eyes indeed confirmed that the protein is required for normal vision, given that the Poc1b-depleted zebrafish showed a severely decreased OKR in combination with smaller eyes (Figure 4C). Analysis of morphant eyes revealed decreased length of photoreceptor outer segments in the cone-dominated larval retina (Figures 4D and 4E). Poc1b appeared to be present at all basal bodies of vertebrate photoreceptors, suggesting that loss of function affects both rods and cones (Figure 4A; Figure S3). Indeed, knockdown of poc1b reduced immunoreactivity for important proteins in the light-transduction cascade of rod and cones alike (Figure S4C). This corresponds with the decreased visual response of poc1b morphants, measured in the OKR assay. Rescue of the smaller eyes associated with this phenotype was achieved with wild-type human POC1B mRNA.
The affected amino acids identified in this study are moderately or highly conserved in evolution (Figure 1C), and both affect the N-terminal WD40 domain (Figure S1). The third mutation alters the splice site of exon 7 and results in a truncation of the protein within the last WD40 repeat. This WD40 domain, but not the C-terminal region of POC1, has been demonstrated to be sufficient for targeting POC1 localization to centrioles.32 Indeed, whereas wild-type POC1B localized to the basal bodies, as previously reported, both p.Gln67del and p.Arg106Pro variant POC1B revealed a loss of association with the basal body of the cilium (Figure 3). The effect of the variants on Poc1b in zebrafish was addressed by coinjection of a poc1b MO in combination with human POC1B mRNA carrying either one of the mutations. In contrast with coinjection of wild-type mRNA, coinjection of the mutated mRNA could not induce (partial) rescue of the ocular phenotype, confirming the disturbing retinal effect of the variant amino acid residues (Figure 4B).
To provide further insights into the retinal function of POC1B, we aimed to identify retinal proteins interacting with POC1B by using a GAL4-based interaction trap screen in yeast of a retinal cDNA library. Out of the four different bait fragments of POC1B employed, only the coiled-coil region was found to yield one significant interactor: FAM161A. Interestingly, mutations in FAM161A lead to another retinal ciliopathy, autosomal-recessive RP (RP28).27,28 Binding of FAM161A was validated with coimmunoprecipitation and colocalization studies (Figure 5). Although the interaction was initially detected with a fragment containing the coiled-coil region of POC1B, introduction of the two POC1B variants in the WD40 domain of the full-length protein strongly decreased its interaction with FAM161A, reiterating the structural importance of this domain. Because FAM161A was found to be a retinal-ciliopathy-associated protein,21,37 the decreased interaction we observed with this retina-specific protein might induce degeneration of rod photoreceptors as a result of POC1B mutations in individuals with CRD.
Pearson et al. showed that in the absence of Poc1b, zebrafish present with various phenotypes that point toward a syndromic ciliopathy.15 In contrast, the POC1B mutations identified in this study are associated with a much milder, nonsyndromic cone-disease phenotype in two families. Although species-specific differences might contribute to the observed phenotypic heterogeneity, on the basis of the type and combinations of mutations identified and the reduced, but not absent, interaction between the altered POC1B and the retina-specific FAM161A, it is plausible to conclude that individuals with COD have residual POC1B activity. Combinations of more severe and/or loss-of-function POC1B mutations therefore might be associated with syndromic forms of retinal ciliopathies, in line with the wide disease spectrum previously observed for another ciliopathy-associated gene, CEP290 (MIM 160142).38–41
In conclusion, WES led to the identification of POC1B mutations in two unrelated families affected by autosomal-recessive nonsyndromic COD or CRD. These variants were found to disrupt the ciliary basal body localization of POC1B and its interaction with a retina-specific, RP-associated protein, FAM161A. Given that loss of Poc1b in zebrafish furthermore resulted in early-onset retinal dysfunction, this study highlights a basal body protein photoreceptor module that contains POC1B and FAM161A and is required for photoreceptor homeostasis.