Highlights
- •NPHP5-LCA patient-derived fibroblasts and RPE display abnormally elongated cilia
- •Outer segment protein localization is impaired in patient-derived photoreceptors
- •CEP290 protein reduction is observed across all NPHP5-LCA patient-derived cells
- •NPHP5 augmentation improves disease phenotypes in patient retinal organoids
Summary
Mutations in the IQ calmodulin-binding motif containing B1 (IQCB1)/NPHP5 gene encoding the ciliary protein nephrocystin 5 cause early-onset blinding disease Leber congenital amaurosis (LCA), together with kidney dysfunction in Senior-Løken syndrome. For in vitro disease modeling, we obtained dermal fibroblasts from patients with NPHP5-LCA that were reprogrammed into induced pluripotent stem cells (iPSCs) and differentiated into retinal pigment epithelium (RPE) and retinal organoids. Patient fibroblasts and RPE demonstrated aberrantly elongated ciliary axonemes. Organoids revealed impaired development of outer segment structures, which are modified primary cilia, and mislocalization of visual pigments to photoreceptor cell soma. All patient-derived cells showed reduced levels of CEP290 protein, a critical cilia transition zone component interacting with NPHP5, providing a plausible mechanism for aberrant ciliary gating and cargo transport. Disease phenotype in NPHP5-LCA retinal organoids could be rescued by adeno-associated virus (AAV)-mediated IQCB1/NPHP5 gene augmentation therapy. Our studies thus establish a human disease model and a path for treatment of NPHP5-LCA.
Introduction
Once considered vestigial, the non-motile primary cilium has emerged as a key organelle in highly specialized sensory signal transduction in most eukaryotic cells (Breslow and Holland, 2019; Reiter and Leroux, 2017; Sang et al., 2011). Mutations in genes associated with cilia biogenesis and/or function lead to diseases termed ciliopathies, which encompass a plethora of varying phenotypes including developmental malformations, kidney cysts, intellectual disability, and sensory dysfunctions (Reiter and Leroux, 2017). In vertebrate photoreceptors, primary cilia acquire a unique architecture consisting of an axoneme surrounded by ciliary membrane with exquisitely organized stacks of discs that are critical for high-efficiency photon capture and signal transmission. Thus, cilia formation and functional maintenance are critical for vision, and photoreceptor defects and/or degeneration are commonly associated with ciliopathies (Chen et al., 2019).
Mutations in the gene encoding IQ calmodulin-binding motif containing B1 (IQCB1/NPHP5) are the most common cause of renal-retinal Senior-Løken syndrome (SLSN) (Otto et al., 2005) and are also detected in patients with non-syndromic Leber congenital amaurosis (LCA) (Estrada-Cuzcano et al., 2011; Stone et al., 2011). Both SLSN and LCA are genetically and clinically heterogeneous. Patients with LCA exhibit an early onset of retinal photoreceptor dysfunction that is accompanied by nystagmus, photophobia, and other clinical findings, whereas SLSN additionally includes nephronophthisis (NPHP), a kidney cystic disease leading to chronic renal failure (Hildebrandt et al., 2009). Notably, patients with SLSN exhibit variable onset of kidney dysfunction, whereas visual function defects are observed in early childhood and have a significant impact on patients’ quality of life, making photoreceptors an important target for therapy development.
NPHP5 protein of 598 amino acids is required for cilia biogenesis and interacts with multiple cilia-associated proteins including RPGR (Otto et al., 2005) and CEP290 (NPHP6) (Barbelanne et al., 2013; Schafer et al., 2008). NPHP5 is localized in the ciliary transition zone with CEP290, where it modulates the integrity of the BBSome protein complex involved in ciliary transport (Barbelanne et al., 2015), and the assembly of basal feet during cilia formation (Hossain et al., 2020). Of particular interest is CEP290, since patients with LCA carrying NPHP5 or CEP290 mutations exhibit overlapping clinical phenotypes (Cideciyan et al., 2011; Otto et al., 2005), suggesting involvement in similar functions within cilia. CEP290, however, interacts with many proteins, exhibits domain-specific functions, and is associated with a broader range of ciliopathies (Coppieters et al., 2010; Drivas and Bennett, 2014; Rachel et al., 2012, Rachel et al., 2015). Patient mutations in IQCB1 cluster around the coiled-coil domains in the C-terminal region, with the SLSN-associated mutations leading to an earlier truncation of the protein than those responsible for LCA (Estrada-Cuzcano et al., 2011). Retinal imaging in patients with NPHP5-LCA reveals rapid and widespread rod photoreceptor degeneration but relatively preserved central region of non-functional cone photoreceptors likely lacking outer segments (Cideciyan et al., 2011; Downs et al., 2016). These observations suggest that the remaining cones could be a viable target for gene replacement therapy (Cideciyan et al., 2011).
Loss of function of Nphp5 in mouse and dog models results in impaired outer segment formation, absence of rod or cone responses on electroretinogram, and early-onset rod photoreceptor loss (Downs et al., 2016; Ronquillo et al., 2016). These models have been used for in vivo assessment of prospective therapies (Aguirre et al., 2021; Hanke-Gogokhia et al., 2018). However, animal models do not completely recapitulate human genetic diversity and features of retinal development (Hoshino et al., 2017; Yan et al., 2020). Furthermore, efficacy of transduction with gene therapy vectors differs between model organisms and human retina (Gonzalez-Cordero et al., 2018; Wiley et al., 2018). Retinal organoids derived from patient induced pluripotent stem cells (iPSCs) complement the in vivo studies using animal models by providing a human genetic context (Kruczek and Swaroop, 2020).
Retinal organoids are in vitro self-formed aggregates from PSCs displaying many key characteristics of the native tissue (Nakano et al., 2012; Zhong et al., 2014). To model genetic heterogeneity of LCA, elucidate disease mechanisms, and evaluate possible treatment paradigms, we recently established retinal organoid culture systems from iPSCs of patients with LCA with mutations in CEP290 and CRX (Kruczek et al., 2021; Shimada et al., 2017). Here, we report the phenotypic modeling of NPHP5-LCA using patient dermal fibroblasts, iPSC-derived retinal pigment epithelium (RPE) cells, and retinal organoids, which exhibit abnormal cilia morphology and reduced levels of CEP290. The photoreceptors in patient organoids demonstrate impaired protein localization and aberrant extension of outer segments. We also show that this phenotype can be rescued by adeno-associated virus (AAV)-mediated delivery of the correct IQCB1/NPHP5 sequence. Together with the animal models (Aguirre et al., 2021; Hanke-Gogokhia et al., 2018), our studies validate gene augmentation as a prospective treatment approach for NPHP5-LCA.
Results
Patients and experimental design
To examine ciliopathy phenotypes associated with mutations in the IQCB1 gene causing NPHP5-LCA, we recruited 4 affected patients as well as 3 healthy familial controls from 3 families (Tables S1 and S2; Figure S1). These individuals carried 6 different mutant alleles across the IQCB1/NPHP5 coding sequence (Figures 1A and 1B ; Table S1). Unaffected carriers (controls) carry a single mutant allele in a heterozygous state, whereas patients are compound heterozygotes for two mutant alleles (Figure 1B; Table S1). Three of the identified mutations are present in known protein domains: (1) c.421_422delTT (p.F141fsX6) localizes to the N-terminal region containing the BBS interaction site (1–157 amino acids), (2) c.1036G>T (p.E346X) occurs within a coiled-coil domain (340–373 amino acids), and (3) c.1516_1517delCA (p.H506fsX13) is present just before the CEP290 binding region (509–529 amino acids) (Figure 1A). We obtained skin biopsies from these patients to model disease pathology and evaluate treatment strategy.
