Skip to main content
NCBI home page
American Journal of Human Genetics logoLink to American Journal of Human Genetics
. 2026 Feb 19;113(3):616–626. doi: 10.1016/j.ajhg.2026.01.015

Bi-allelic variants in FSD1L cause retinitis pigmentosa with or without neurological involvement

Siying Lin 1,2,3,4,22, Francesca Cancellieri 5,6,22, Yexuan Cao 7,22, Andrew J Lotery 8,9, Abigail R Moye 5,6, Veronika Vaclavik 10, Fabienne Perren 11,12, Andrzej B Poplawski 13, Elena R Schiff 3,4, Mukhtar Ullah 5,6, Ana Belen Iglesias-Romero 5,6, Karolina Kaminska 5,6, Aleksandr Jestin 4, Marc Folcher 5,6, Sandrine Wallerich 5,6, Mariana M Ribeiro 5, Vincent Hahaut 5, Simone Picelli 5, Debarshi Mustafi 14,15, Aleksander Tworak 7, Roman Smidak 7, Yumei Li 7, Jiaxiong Lu 7, Meng Wang 7, Omar A Mahroo 3,4,16, Shyamanga Borooah 17, Mathieu Quinodoz 5,6,18, Krzysztof Palczewski 7,19,20,21, Andrew R Webster 3,4, Carlo Rivolta 5,6,18,23, Rui Chen 7,23, Gavin Arno 3,4,13,23,∗
PMCID: PMC13087403  NIHMSID: NIHMS2165957  PMID: 41720099

Summary

Retinitis pigmentosa (RP) is an inherited retinal disease (IRD) characterized usually by progressive photoreceptor degeneration, leading to night blindness, peripheral visual field loss, and can progress to central vision impairment in some individuals. Despite advances in genomic diagnostics, many individuals with RP remain without a molecular diagnosis. We identified bi-allelic ultra-rare variants in fibronectin type II and Spry domain-containing protein 1-like (FSD1L) in six individuals with RP with or without neurological features from four unrelated families. FSD1L encodes a cytoplasmic protein, variants of which have not previously been associated with Mendelian disease. The gene is expressed in both human and mouse retinas that are enriched in cone and rod photoreceptors. Immunofluorescence and ultrastructure expansion microscopy show that FSD1L localizes along the photoreceptor microtubule axoneme, including the connecting cilium and outer segment, supporting a possible role in intracellular trafficking. A retina-enriched isoform of FSD1L includes an alternatively spliced exon (exon 10b), which we characterize as absent in minigene assays and affected individual-derived lymphocytes due to a deep intronic 26 nt deletion. Together, these findings support the association between bi-allelic disruption of FSD1L and IRD.

Keywords: inherited retinal dystrophy, retinitis pigmentosa, FSD1L, cilia, genome sequencing, retinal transcriptomics, single-cell transcriptome, splice variant, retina-enriched isoform, nanopore sequencing

Bi-allelic ultra-rare variants in FSD1L are identified as a cause of retinitis pigmentosa. One variant identified disrupts the splicing of a retina-enriched isoform. FSD1L localizes to the photoreceptor axoneme, implicating disruption of intracellular trafficking as a possible mechanism for photoreceptor degeneration.

Main text

Inherited retinal disease (IRD) describes a clinically and genetically heterogeneous group of disorders characterized typically by progressive retinal degeneration, leading to visual impairment and blindness. IRDs are a leading cause of vision loss among children and working-age adults and affect an estimated 5.5 million individuals worldwide,1,2 representing a significant global healthcare burden.

Retinitis pigmentosa (RP [MIM: 268000]) is the most common form of IRD, affecting approximately 1 in 3,500 individuals in the USA and Europe.3,4 RP typically presents with nyctalopia (night blindness), followed by progressive peripheral visual field loss, and, in many cases, progresses to central visual impairment. RP, and IRDs more generally, may occur as an isolated disease, limited to ocular involvement, or as part of syndromic conditions involving additional tissues, organs, or systems, including hearing loss, neurodevelopmental disease, or metabolic dysfunction.5

Despite advances in high-throughput sequencing technologies and diagnostic pipelines, up to 40%–50% of IRD-affected individuals remain molecularly undiagnosed,6 limiting access to accurate prognostic information, genetic counseling, and emerging gene-directed therapies. This paucity of data suggests that novel mechanisms of degenerative retinal disease remain to be identified.

Here, we report the clinical and genetic findings of six affected individuals, two males and four females aged 14–51 years, from four unrelated families. All six individuals manifested RP, while some had additional clinical findings, as summarized in Table 1 (further details are available in the supplemental notes: clinical findings). All six individuals had bi-allelic ultra-rare variants (AF [allele frequency] < 0.00015; gnomAD v.4.1.0) in FSD1L (MIM: 609829), which encodes for the fibronectin type II and Spry domain-containing protein (FSD)1-like (FSD1L) protein. Variants were identified by exome sequencing or genome sequencing (GS) analysis performed as part of large-scale projects aimed at elucidating the etiology of genetic disease, including the UK 100,000 Genomes Project (individuals A.III-1 and A.III-3), the UK National Health Service (NHS) Genomic Medicine Service (GMS) (individual B.II-1),6,7 and ongoing research studies at the University of California, Irvine (UCI), and the University of Washington (individuals C.II-1 and C.II-2) and the Institute of Molecular and Clinical Ophthalmology, Basel (individual D.II-2).

Table 1.

Clinical features for individuals with bi-allelic FSD1L variants (GenBank: NM_001145313.3)

Family A (GC4822) B (GC17709) C D (CHlaus0427)
Individual A.III-1 A.III-3 B.II-1 C.II-1 C.II-2 D.II-2
Allele 1 c.1049G>A (p.Arg350Gln) c.1049G>A (p.Arg350Gln) c.488G>A (p.Arg163His) c.488G>A (p.Arg163His) c.488G>A (p.Arg163His) c.1037_1038delinsT (p.Pro346Leufs∗8)
Allele 2 c.1428del (p.Phe476Leufs∗22) c.1428del (p.Phe476Leufs∗22) c.745C>T (p.Arg249∗) c.226_227del (p.Ser77Argfs∗4) c.226_227del (p.Ser77Argfs∗4) c.1025+624_1025+649del
Country of recruitment UK UK UK USA USA Switzerland
Ethnicity White British White British White British White American White American White Swiss
Sex male female female female male female
Age at last examination 36 years 32 years 32 years 14 years 12 years 51 years
Diagnosis RP with mild neurological involvement RP with possible mild neurological involvement RP, no known neurological involvement RP, no known neurological involvement RP, no known neurological involvement RP, no neurological involvement
Ocular features
Initial symptoms (age of onset) nyctalopia (9 years) nyctalopia (13 years) nyctalopia (5 years) nyctalopia (14 years) asymptomatic (diagnosed after sibling diagnosis) nyctalopia, high myopia (10 years)
BCVA (Snellen) OD: 20/40
OS: 20/200		 OD: 20/120
OS: 20/200		 OD: 20/50
OS: 20/80		 OD: 20/50
OS: 20/50		 OD: 20/32
OS: 20/32		 OD: 20/1,200
OS: 20/400
Refraction OD: +0.50/−3.50 × 11°
OS: +0.25/−3.50 × 176°		 OD: +0.75/−2.75 × 16°
OS: 1.00/−5.75 × 147°		 emmetropia emmetropia emmetropia OD: −0.5/−2.25 × 63°
OS: −0.75/−1 × 43° (pseudophakic)
Lens status phakic with clear lenses bilaterally bilateral posterior subcapsular lens opacities bilateral pseudophakia (cataract surgery OD 30 years; OS 27 years) bilateral pseudophakia (cataract surgery OD 13 years; OS 13 years) phakic with clear lenses bilaterally bilateral pseudophakia (cataract surgery OU 36 years)
Fundus hypopigmented fundus with prominent choroidal vasculature, macular atrophy, pale optic discs, attenuated vessels, no pigmentary retinopathy hypopigmented fundus with prominent choroidal vasculature, pale optic discs, and attenuated vessels, pigmentary retinopathy pale optic discs, attenuated retinal vessels, and mid peripheral pigmentary retinopathy bone spicule-type pigmentation superonasally, normal optic discs, loss of foveal reflex with macular edema bilaterally normal fundus except for loss of foveal reflex with macular edema bilaterally hypopigmented posterior pole macular atrophy, pale optic discs, attenuated vessels, pigmentary retinopathy
Autofluorescence small hyper-AF ring at the macula mid-peripheral hypo-AF with a central ring of hyper-AF widespread hypo-AF with small central macular hyper-AF hypo-AF peripherally without central hyper-AF increased AF at the posterior pole with a hyper-AF ring at the macula mid-peripheral and posterior pole hypo-AF, area of normal AF within arcades
OCT mild ERM; perifoveal EZ loss with subfoveal sparing; no CME perifoveal EZ loss with subfoveal sparing; no CME perifoveal EZ loss with subfoveal sparing; no CME perifoveal EZ loss with subfoveal sparing; CME perifoveal EZ loss with subfoveal sparing; CME thinning of retinal layers, EZ not visible, no CME
ERG (age of examination) consistent with rod-cone dystrophy with marked macular involvement (19 years) consistent with rod-cone dystrophy with marked macular involvement (15 years) severely reduced rod and cone responses (16 years) NP NP undetectable scotopic and photopic ERGs (42 years)
Visual fields severely constricted to confrontation severely constricted to confrontation mid-peripheral scotoma (Goldman) mid-peripheral constriction (no scotoma in the central 10°) mid-peripheral constriction (no scotoma in the central 10°) severely constricted (<10°)
Other ocular features – previous CME, decompensated left exophoria previous CME – – –
Non-ocular features
Neurological findings mild learning disability, spastic diplegia mild learning disability none reported none reported none reported normal neurological examination (50 years)
Neuroimaging (age of examination) MRI: slight hypoplasia of the splenium of the corpus callosum and the cerebellar vermis, and mild parietal increase of subarachnoid spaces (29 years) MRI: normal (19 years) NP NP NP MRI: normal (50 years)
Other systemic features high BMI, premature adrenarche (age 10 years) high BMI no no no normal weight
Open in a new tab

AF, autofluorescence; BCVA, best corrected visual acuity; BMI, body mass index; CME, cystoid macula edema; ERG, electroretinogram; ERM, epiretinal membrane; EZ, ellipsoid zone; MRI, magnetic resonance imaging; NP, not performed; OCT, optical coherence tomography; OD, right eye; OS, left eye; OU, both eyes; RP, retinitis pigmentosa; -, not present.

This study adhered to the Declaration of Helsinki and was conducted in accordance with the ethical standards of the institutional and national research committees on human experimentation. Ethical approval was obtained from the following institutional review boards: Moorfields Eye Hospital and the Northwest London Research Ethics Committee (12/LO/0141); the University of California, San Diego (UCSD); the University of Washington; the UCI; the Ethikkommission Nordwest-und Zentralschweiz; and the Commission cantonale d'éthique de la recherche sur l'être humain (CER-VD). Written informed consent for participation and publication was obtained from participants (and parents where appropriate). Self-identified racial and ethnic categories were collected from all individuals as part of their standard clinical care.

Individuals A.III-1 and A.III-3 (Figure 1A) are a brother and sister from a White British family, born to unrelated parents, with one additional unaffected sister. Both were diagnosed with RP and had features suggestive of a mild learning disability (Figure 1B). Additionally, individual A.III-1 was diagnosed with spastic diplegia in childhood, initially attributed to perinatal complications. GS was performed for individuals A.III-1 and A.III-3 and their unaffected parents (A.II-1 and A.II-2) as part of the UK 100,000 Genomes Project.6 Initial clinical-grade variant interrogation failed to identify a pathogenic or likely pathogenic genotype in genes listed on the PanelApp “posterior segment abnormalities” panel (encompassing 174 genes with established associations to IRDs).8 Subsequent research analysis focused on rare (AF < 0.001) bi-allelic protein-altering genotypes shared by both siblings and identified only a single candidate compound heterozygous genotype in FSD1L c.1049G>A (GenBank: NM_001145313.3) (p.Arg350Gln) and c.1428del (GenBank: NM_001145313.3) (p.Phe476Leufs∗22), confirmed to be inherited in trans (based on parental GS analysis).

Figure 1.

Figure 1

Open in a new tab

Pedigrees, representative retinal imaging, and schematic overview of FSD1L retinal isoforms, including the variants identified in this study

(A) Pedigrees of families A–D, showing genotypes of affected individuals and segregation of FSD1L variants where familial data were available. The diamond symbol shows 6 additional siblings where the sex is not relevant to this study.

(B) Representative multimodal retinal imaging from individuals A.III-1 (age 36 years) and A.III-3 (age 32 years), illustrating features typical of RP observed across all affected individuals. Ultra-wide-field (UWF) pseudocolor images and fundus autofluorescence (FAF) (Optos, Dunfermline, UK) show attenuated retinal vessels and mid-peripheral pigmentary changes (A.III-3), with a central ring of hyperautofluorescence demarcating surviving outer retinal structures. Macular optical coherence tomography (Spectralis OCT, Heidelberg Engineering, Heidelberg, Germany) demonstrates loss of outer retinal structures with preserved central macular structure, correlating with the FAF findings.

(C) Schematic “lollipop” plot of FSD1L showing the location of variants identified in this study, overlaid on relevant retinal isoforms. Exons are represented as boxes and introns as horizontal lines. Protein domains and variant types are color coded. Isoforms GenBank: NM_001287191.2 and NM_001330739.2 differ only by 3 bp at the beginning of exon 11, which are also present in the canonical transcript GenBank: NM_001145313.3. The lollipop plot was drawn with ProteinPaint.

Through an expanded interrogation of the Genomics England National Genomic Research Library7 and international collaborative efforts, four additional unrelated individuals (individuals B.II-1, C.II-1, C.II-2, and D:II-2) with non-syndromic RP and candidate bi-allelic FSD1L variants were identified (Figure 1A). Individual B.II-1 (Figure 1A) is a White British female and the only child of unaffected parents. Individuals C.II-1 and C.II-2 (Figure 1A) are affected siblings from a non-consanguineous White American family. Individual D.II-2 (Figure 1A) is a White Swiss female with no siblings. Across all families, there is no reported consanguinity and no known antecedents affected with a genetic eye disease. For each individual, the FSD1L variants identified represented the only plausible disease-associated genotype. Full details of sequencing methodology and variant prioritization for all six individuals are provided in the supplemental methods.

Five of the six affected individuals reported childhood-onset night vision difficulties (one was asymptomatic and diagnosed following identification of an affected sibling), and all exhibited an ocular phenotype consistent with rod-cone dystrophy or RP. In view of the mild neurological features described in individuals A.III-1 and A.III-3, as well as a contemporaneous report describing a severe neurological syndrome associated with bi-allelic FSD1L variants (in this issue of AJHG9), individual D.II-2 underwent a full neurological evaluation, including brain neuroimaging, which revealed no evidence of central nervous system involvement. Similarly, no neurological features were observed or reported in the remaining individuals. Additional clinical findings are detailed in Table 1 and Figures S1A–S1C. A total of seven FSD1L variants were identified, including two missense variants (c.1049G>A [p.Arg350Gln] and c.488G>A [p.Arg163His]), three frameshift variants (c.1428del [p.Phe476Leufs∗22], c.226_227del [p.Ser77Argfs∗4], and c.1037_1038delinsT [p.Pro346Leufs∗8]), one nonsense variant (c.745C>T [p.Arg249∗]), and one deep intronic deletion (c.1025+624_1025+649del [p.?]). All variants were either absent from gnomAD v.4.1.0 or observed at extremely low AFs, with no homozygous individuals identified (variant details are summarized in Figures 1C and S2A–S2C and Table 2). Segregation analysis, where available (families A, C, and D), confirmed autosomal-recessive inheritance with the two FSD1L alleles present in trans in affected individuals (Figure 1A). No bi-allelic null genotypes were observed (Figure 1A). Notably, the p.Arg163His variant was observed in two unrelated individuals, each in trans with a different predicted loss-of-function (pLoF) variant.

Table 2.

FSD1L variants identified in this study

Variant ID (family)a Variant
 Variant type gnomAD AF In silico predictionsb
 ClinVar (ID) ACMG/ACGS classification (evidence)c
Genomic coordinates (GRCh38) Nucleotide Protein REVEL AlphaMissense MutScore SpliceAI Pangolin
m1 (A) chr9:105534516G>A c.1049G>A p.Arg350Gln missense 0.00004455 0.15 0.677 0.25 0.27 0.24 absent VUS (PM2_supp, PM3)
m2 (A) chr9:105539309CT>C c.1428del p.Phe476Leufs∗22 frameshift 0.000001326 N/A N/A N/A 0.01 0.05 absent LP (PVS1_mod, PM2_supp, PM3)
m3 (B, C) chr9:105484404G>A c.488G>A p.Arg163His missense 0.0001259 0.29 0.587 0.398 0.06 0.05 VUS (2302875) VUS (PM2_supp, PM3_mod)
m4 (B) chr9:105506557C>T c.745C>T p.Arg249∗ stop-gain 0.00000258 N/A N/A N/A 0.01 0.18 absent VUS (PVS1_mod, PM2_supp, PM3_supp)
m5 (C) chr9:105468210ACT>A c.226_227del p.Ser77Argfs∗4 frameshift absent N/A N/A N/A 0 0.05 absent LP (PVS1_mod, PM3_mod, PP2_supp)
m6 (D) chr9:105534504CA>T c.1037_1038delinsT p.Pro346Leufs∗8 frameshift absent N/A N/A N/A 0.15 N/A absent LP (PVS1_mod, PM3_mod, PP2_supp)
m7 (D) chr9:105513556TT
GTAAAACAGTTT
CTTAACAGTTGCC>T		 c.1025+624_1025+649del p.? intronic 0.000001337 N/A N/A N/A 0.07 0.27 absent LP (PS3_mod, PM3_mod, PM2_supp)
Open in a new tab

FSD1L reference transcript GenBank: NM_001145313.3. AF, allele frequency; gnomAD, Genome Aggregation Database v.4.1.0; LP, likely pathogenic; N/A, not available; VUS, variant of uncertain significance.

a

Variant IDs correspond to labels used in Figure 1A.

b

REVEL is an ensemble score based on 13 individual scores for predicting the pathogenicity of missense variants.10 AlphaMissense scores can be interpreted as the approximate probability of a variant being clinically pathogenic.11 MutScore integrates qualitative features of DNA substitutions with new additional information derived from positional clustering.12 SpliceAI and Pangolin delta scores can be interpreted as the probability that the variant affects splicing at any position within a ±500 bp window around it.13,14 Scores for REVEL, AlphaMissense, MutScore, SpliceAI, and Pangolin range from 0 to 1, with higher scores indicating a higher probability of the variant being damaging or having a splice-altering effect.

c

Variant classification and evidence codes follow American College of Medical Genetics and Genomics and Association for Molecular Pathology (ACMG/AMP) guidelines15 with strength modifiers (e.g., _mod and _supp) applied according to the Association for Clinical Genomic Science (ACGS) Best Practice Guidelines.16

Consistent with the observed photoreceptor degenerative phenotype, FSD1L is expressed in the retina, as shown by single-cell RNA sequencing (RNA-seq) of human and mouse retina (Figure S3).17,18 In the human retina, FSD1L is most highly expressed in cone photoreceptors, with lower expression observed in rod photoreceptors, retinal ganglion cells (RGCs), and horizontal cells (HCs). Minimal to no expression is detected in amacrine cells (ACs), bipolar cells (BCs), and Müller glia (MGs) (Figures 2A and 2B). A similar expression pattern is observed in the mouse retina, with strong Fsd1l expression in cone cells, followed by rods, HCs, RGCs, and MGs (Figures 2C and 2D).

Figure 2.

Figure 2

Open in a new tab

Single-cell transcriptomic profiling of FSD1L expression in human and mouse retinas

(A and B) Violin plot (A) and dot plot (B) showing FSD1L expression across retinal cell types in human retina.

(C and D) Violin plot (C) and dot plot (D) showing Fsd1l expression across retinal cell types in mouse retina.

AC, amacrine cells; BC, bipolar cells; HC, horizontal cells; MG, Muller cells; RGC, retinal ganglion cells; RPE, retinal pigmented epithelium.

Within the human retina, FSD1L is expressed not only as the canonical (or Matched Annotation from NCBI and EBI, MANE) transcript (GenBank: NM_001145313.3) but also as at least two alternative isoforms, GenBank: NM_001330739.2 and NM_001287192.2 (Figure 1C). The non-canonical transcripts do not include exon 2, and GenBank: NM_001330739.2 includes an alternatively spliced exon, referred to here as exon 10b. This exon is conserved in mammals and shows coding constraint, supportive of a role as a functional protein-coding element.19 Disruption of exon 10b, therefore, may have important functional consequences, particularly in the context of the retina-enriched isoforms.

RNA-seq data20 obtained from human retina show that inclusion of exon 10b is above 60% in the peripheral retina (Figures S4A and S4B), with only skeletal muscle tissue having more inclusion (Figures S4C and S4D). At a cellular level, long-read single-cell sequencing of mouse retina shows exon 10b inclusion in 100% of reads from rods (128 reads) and cones (506 reads) and in 50% of reads from Müller cells (10 reads), and there is no detectable inclusion in BCs (Figure S5A). Additionally, in human macula long-read single-nuclei sequencing data, exon 10b is included in 40.05% of reads in rods (1,623 reads) and 55.19% in cones (270 reads) (Figure S5B).

One variant identified in this study, c.1025+624_1025+649del (GenBank: NM_001145313.3), is located close to the splice acceptor site of exon 10b, c.930-39_930-14del (GenBank: NM_001330739.2). In silico predictions suggest that both the acceptor and donor splice sites of this exon are weak (SpliceAI scores of 0.07 and 0.02, respectively), and the deletion abolishes the splice acceptor strength at the exon 10b junction (SpliceAI score of 0).13 The deleted region also overlaps a predicted branchpoint,21 representing a potential disruption of key splicing regulatory elements required for exon 10b inclusion.

To investigate the impact of the FSD1L c.1025+624_1025+649del variant on splicing, wild-type (WT) and mutant FSD1L minigenes encompassing exons 10 and 10b and nearby intronic sequences were generated (Figure 3A). These minigenes were transfected into ARPE-19 cells, and splicing patterns were subsequently analyzed.

Figure 3.

Figure 3

Open in a new tab

Minigene assay shows exon skipping caused by the c.1025+624_1025+649del variant

(A) Schematic representation of the design of FSD1L pCI-NEO-RHO exon3,5/DEST minigene construct for wild type (WT) and mutant (Mut), including the variant c.1025+624_1025+649del (m7).

(B) Resolution of splicing PCR products derived from ARPE-19 cells transfection with WT or Mut FSD1L minigenes; untransfected cells (Unt) were used as a control. Sanger sequencing confirmed the presence of the illustrated products. The filled boxes indicate exons, connected by introns. Blue color indicates the genomic fragment area cloned into the plasmid, with the vector backbone depicted in violet. Arrows indicate primer binding sites used for amplification of the cDNA. The black dotted lines refer to the amplified area on the cDNA, with their corresponding sizes. bp, base pairs; Mrk, size marker.

Transfection with the WT FSD1L minigene produced two transcripts of 314 and 281 bp. Sanger sequencing confirmed that these corresponded to transcripts with and without exon 10b, respectively (Figure 3B). As predicted, the mutant FSD1L minigene yielded a single 281 bp product, consistent with skipping of exon 10b; no product including exon 10b was detected. Thus, the c.1025+624_1025+649del variant leads to exon 10b skipping in ARPE-19 cells.

In tandem, we examined endogenous FSD1L transcripts in lymphocytes from individuals D.II-2 and D.I-1 and three controls. RT-PCR was performed on peripheral blood-derived RNA using primers spanning exons 10–14. Long-read nanopore sequencing was performed on resultant amplicons to investigate the splicing pattern, according to previous methods developed in our laboratory for low-level transcript analysis.22,23,24 The maximal read depth generated from individual samples was between 51,000 and 99,000. These data showed that exon 10b is included in approximately 3%–14% of FSD1L transcripts in the three unrelated control samples, measured as the read depth of exon 10b vs. the canonical exon 10. In contrast, it was absent in the affected individual, indicating that the intronic deletion disrupts normal splicing of this exon in lymphocytes (Figure S6). Phasing of the transcript reads from individual D.II-2 covering exon 11 showed a low proportion of reads derived from the trans allele carrying the c.1037_1038delinsT pLoF variant (883/5,936 reads, 15%). This suggests that transcripts derived from the pLoF allele are undergoing nonsense-mediated decay (NMD), leading to the skewed representation of the alleles. This may therefore explain why, in individual D.II-2, even though the exon 10b mis-splicing variant is heterozygous, the normally spliced exon 10b on the trans allele is absent due to NMD, leading to the complete absence of exon 10b in the sequencing reads.

The missense variant identified in individuals A.III-1 and A.III-3 (c.1049G>A [p.Arg350Gln]) affects a residue on an unstructured loop in the predicted 3D model (Figure S2C) but also had a SpliceAI high recall score (Δ0.27, Table 2). Examination of endogenous transcripts in lymphocytes from A.III-3, compared to her unaffected father (A.II-1) carrying c.1428del (p.Phe476Leufs∗22), showed a low level of skipping of exon 11 (8.9% of canonical transcripts, Figure S7), which would lead to an out-of-frame truncation terminating in exon 12. These data suggest that the effect of this allele may be a combination of LoF and the amino acid substitution.

Together, these findings suggest a weakening of splicing of exon 10b consequent upon the intronic deletion and specifically affecting transcript GenBank: NM_001330739.2, which may represent a retina-enriched isoform. Of note, given the identification of a severe neurodevelopmental phenotype association with bi-allelic pLoF and severe variants in this gene (in this issue of AJHG9), disrupted splicing of this retina-enriched exon may represent a mechanism for the sparing of non-retinal neurons. For other families, subtle neurological features may not yet be evident, given the young age of some affected individuals, and longitudinal follow-up will be important to determine whether additional manifestations emerge with time.

FSD1L, located on chromosome 9q31.2, was originally identified based on its sequence homology to FSD1 (MIM: 609828), a gene on chromosome 19q13.33 that shows preferential expression in the brain. FSD1L encodes a cytoplasmic protein that closely resembles its paralog FSD1, sharing approximately 50% amino acid identity and 65%–70% similarity across the full-length protein.25 Both FSD1L and FSD1 contain an N-terminal coiled-coil domain, a central fibronectin type III (FN3) motif, and a C-terminal SPRY domain—highly conserved elements implicated in protein-protein interactions.25

Although FSD1L has not been functionally characterized in detail, its homolog FSD1 (also known as MIR1) has been reported to bind and stabilize microtubules, potentially via its C-terminal SPRY domain; it is also described as exhibiting dynamic localization throughout the cell cycle—associating with centrosomes in interphase and redistributing during mitosis—consistent with a role for FSD1 in centrosomal positioning and cytoskeletal regulation.26 Given the high degree of sequence similarity and conserved domain architecture, it is plausible that FSD1L may exhibit similar properties. Indeed, both FSD1 and FSD1L exhibit similarity to ubiquitin ligases of the TRIM (Ring-B-box-coiled-coil, RBCC) family.27 In addition, FSD1L shows preferential expression in ciliated cells and localizes to ciliary structures in human bronchus and fallopian tubes.28

To gain insight into the localization of the FDS1L protein in the human retina, we performed immunofluorescence staining and ultrastructure expansion microscopy (U-ExM). These techniques, using antibodies that detect most retinal isoforms of FSD1L (including that encoded by GenBank: NM_001330739.2 containing exon 10b), revealed localization of FSD1L along the photoreceptor microtubule axoneme (Figures 4 and S8A–S8C). Using rhodopsin as the outer segment marker, FSD1L was observed to colocalize with tubulin from the basal body throughout the length of the connecting cilium and into the outer segment axoneme. Localization was also observed in what appears to be the ciliary pocket, an inner segment region surrounding the connecting cilium, using one of the two antibodies specific for FSD1L.

Figure 4.

Figure 4

Open in a new tab

FSD1L localizes to the photoreceptor cilium in human retina

3D shadow projections of confocal z stacks from human rod photoreceptors labeled with tubulin (magenta), rhodopsin (cyan), and FSD1L (yellow) after expansion.29 BB, basal body; CC, connecting cilium; CP, ciliary pocket; DC, daughter centriole; OS, outer segment. Scale bars: 1.25 μm corrected for 4× expansion.

To our knowledge, variants in FSD1L have not previously been implicated in heritable disease. In this study, we identify bi-allelic FSD1L variants in four unrelated families with RP. We show that FSD1L is expressed in the retina, particularly in photoreceptors, and that the protein localizes to the photoreceptor microtubule axoneme, including the connecting cilium and outer segment. Furthermore, we show that a deep intronic variant identified in an individual with non-syndromic RP disrupts splicing of a conserved, retina-enriched exon (exon 10b), supporting a tissue-specific isoform effect as a potential mechanism for isolated retinal disease in this individual. In summary, these findings provide evidence supporting bi-allelic disruption of FSD1L as a mechanism of IRD. Notably, individuals A.III-1 and A.III-3 are affected with both retinal dystrophy and a mild neurodevelopmental disorder, supporting a role for FSD1L in both retinal and neurological function.

FSD1L-associated disease may therefore span a broad phenotypic spectrum, ranging from severe neurodevelopmental syndromes to, at its mildest, non-syndromic retinal dystrophy, where the specific combination and functional severity of the two alleles likely determines the clinical outcome. Here, the phenotype observed in affected individuals in families A–C is likely to be driven by non-LoF alleles and may be expected to have a milder effect on the protein. Affected individuals B.II-1, C.II-1, and C.II-2, who harbor a missense variant with a pLoF allele, likely retain partial FSD1L function, resulting in a retinal phenotype. Individuals A.III-1 and A.III-3 harboring a pLoF allele in trans with a missense variant shown to cause low-level exon skipping and out-of-frame truncation exhibit the retinal phenotype with an associated mild neurological presentation. A subtle or late-onset neurological component in individuals B.II-1, C.II-1, and C.II-2 cannot be excluded, particularly as formal neuroimaging could not be obtained, and some affected individuals are still in early adolescence. For D.II-2, the intronic deletion specifically disrupting the retina-enriched exon 10b provides a plausible mechanism for the absence of neurological findings.

Taken together, these observations highlight the importance of considering isoform- and allele-specific effects in the interpretation of FSD1L variants. We recommend that disruption of both the canonical (GenBank: NM_001145313.3) and retina-enriched (GenBank: NM_001330739.2) isoforms be considered in diagnostic variant assessment, given the full spectrum of FSD1L-associated disease—from isolated retinal dystrophy to syndromic neurodevelopmental presentations.

While our findings identify a role for FSD1L disruption in IRDs, the specific molecular mechanisms leading to retinal degeneration remain to be elucidated. Given the reliance of photoreceptors on microtubule-based trafficking for the delivery of proteins to the outer segment and maintenance of cell polarity,30 disruption of this process through FSD1L dysfunction provides a plausible mechanism for the retinal degeneration observed in affected individuals with bi-allelic FSD1L variants. Functional studies in additional model systems may further clarify the role of FSD1L in retinal health and inform how disruption of specific isoforms or protein domains contributes to retinal disease pathogenesis.

Data and code availability

  • •

    Affected individual data supporting the findings of this study are confidential and subject to ethical restrictions. Access to data can be requested from the corresponding author, subject to appropriate approvals and compliance with data protection regulations.

  • •

    Research on the de-identified individual data used in this publication can be carried out in the Genomics England Research Environment, subject to a collaborative agreement that adheres to individual-led governance. All interested readers will be able to access the data in the same manner as the authors did. For more information about accessing the data, interested readers may contact research-network@genomicsengland.co.uk or access the relevant information on the Genomics England website: https://www.genomicsengland.co.uk/research.

Acknowledgments

The authors thank the individuals and their families for their participation in this study. This study was supported by the Medical Research Council (Clinician Scientist Fellowship, grant reference UKRI440 [S.L.]), Fight for Sight UK (Early Career Investigator Award, grant no. 5045/46 [G.A.]), National Institutes of Health (NIH; grant nos. NIH-P20GM139769 [G.A.], EY022356 [R.C.], and EY018571 [R.C.]), the Retinal Research Foundation (R.C.), the Wellcome Trust (grant no. 206619/Z/17/Z [O.A.M.]), the National Institute for Health and Care Research (NIHR) Manchester Biomedical Research Centre (BRC) (NIHR203308), and the NIHR-BRC at Moorfields Eye Hospital and the UCL Institute of Ophthalmology. C.R. was supported by the Swiss National Science Foundation (SNSF) grant no. 310030_204285. M.Q. was supported by the RetinAward 2021. The authors acknowledge support to the Gavin Herbert Eye Institute at the University of California, Irvine, from an unrestricted grant from Research to Prevent Blindness and from NIH core grant P30 EY034070. We thank the Imaging Core Facility (IMCF, Biozentrum, University of Basel) for use of the Stellaris 8 Falcon microscope. We also thank Dr. Anna Pichiecchio for her assistance in preparing the neuroimaging figures during the early stages of this work. This research was made possible through access to data and findings in the National Genomic Research Library; please see the supplemental information for the full acknowledgments. The funding organizations had no role in the design or conduct of this research. The views expressed are those of the authors and do not necessarily represent those of the funding organizations, NHS, NIHR, or the Department of Health.

Author contributions

Conceptualization (ideas, formulation, or evolution of overarching research goals and aims), C.R., R.C., and G.A.; formal analysis (application of statistical, mathematical, computational, or other formal techniques to analyze or synthesize study data), Y.C., A.R.M., A.B.I.R., M.Q., M.W., and S.L.; investigation (conducting a research and investigation process, specifically performing the experiments, or data/evidence collection), S.L., F.C., Y.C., A.J.L., A.R.M., V.V., F.P., A.B.P., E.R.S., M.U., A.B.I.R., K.K., A.J., M.F., S.W., M.M.R., V.H., S.P., D.M., A.T., R.S., Y.L., J.L., M.W., O.A.M., S.B., M.Q., K.P., A.R.W., C.R., R.C., and G.A.; resources (provision of study materials, reagents, materials, patients, laboratory samples, animals, instrumentation, computing resources, or other analysis tools), M.F. and S.W.; writing – original draft (preparation, creation, and/or presentation of the published work, specifically writing the initial draft [including substantive translation]), S.L., F.C., Y.C., M.Q., A.R.M., C.R., R.C., and G.A.; writing – review & editing (preparation, creation, and/or presentation of the published work by those from the original research group, specifically critical review, commentary, or revision, including pre- or post-publication stages), S.L., F.C., Y.C., A.J.L., A.R.M., V.V., F.P., A.B.P., E.R.S., M.U., A.B.I.R., K.K., A.J., M.F., S.W., M.M.R., V.H., S.P., D.M., A.T., R.S., Y.L., J.L., M.W., O.A.M., S.B., M.Q., K.P., A.R.W., C.R., R.C., and G.A.; visualization (preparation, creation, and/or presentation of the published work, specifically visualization/data presentation), S.L., Y.C., A.R.M., M.Q., F.C., A.B.I.R., C.R., R.C., and G.A.; supervision (oversight and leadership responsibility for the research activity planning and execution, including mentorship external to the core team), C.R., R.C., and G.A.; funding acquisition (acquisition of the financial support for the project leading to this publication), C.R., R.C., and G.A.

Declaration of interests

The authors declare no competing interests.

Published: February 19, 2026

Footnotes

Supplemental information can be found online at https://doi.org/10.1016/j.ajhg.2026.01.015.

Web resources

Supplemental information

Document S1. Figures S1–S8, supplemental notes, and supplemental methods
mmc1.pdf (11.8MB, pdf)
Document S2. Article plus supplemental information
mmc2.pdf (17.9MB, pdf)

References

  • 1.Hanany M., Rivolta C., Sharon D. Worldwide carrier frequency and genetic prevalence of autosomal recessive inherited retinal diseases. Proc. Natl. Acad. Sci. USA. 2020;117:2710–2716. doi: 10.1073/pnas.1913179117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Liew G., Michaelides M., Bunce C. A comparison of the causes of blindness certifications in England and Wales in working age adults (16–64 years), 1999–2000 with 2009–2010. BMJ Open. 2014;4 doi: 10.1136/bmjopen-2013-004015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Chiang J.P.w., Trzupek K. The current status of molecular diagnosis of inherited retinal dystrophies. Curr. Opin. Ophthalmol. 2015;26:346–351. doi: 10.1097/ICU.0000000000000185. [DOI] [PubMed] [Google Scholar]
  • 4.Cross N., van Steen C., Zegaoui Y., Satherley A., Angelillo L. Retinitis Pigmentosa: Burden of Disease and Current Unmet Needs. Clin. Ophthalmol. 2022;16:1993–2010. doi: 10.2147/OPTH.S365486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Karuntu J.S., Almushattat H., Nguyen X.T.A., Plomp A.S., Wanders R.J.A., Hoyng C.B., van Schooneveld M.J., Schalij-Delfos N.E., Brands M.M., Leroy B.P., et al. Syndromic retinitis pigmentosa. Prog. Retin. Eye Res. 2025;107 doi: 10.1016/j.preteyeres.2024.101324. [DOI] [PubMed] [Google Scholar]
  • 6.100000 Genomes Project Pilot Investigators. Smedley D., Smith K.R., Martin A., Thomas E.A., McDonagh E.M., Cipriani V., Ellingford J.M., Arno G., Tucci A., et al. 100,000 Genomes Pilot on Rare-Disease Diagnosis in Health Care - Preliminary Report. N. Engl. J. Med. 2021;385:1868–1880. doi: 10.1056/NEJMoa2035790. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Genomics England (2024). The National Genomic Research Library. 10.6084/m9.figshare.4530893. [DOI]
  • 8.Martin A.R., Williams E., Foulger R.E., Leigh S., Daugherty L.C., Niblock O., Leong I.U.S., Smith K.R., Gerasimenko O., Haraldsdottir E., et al. PanelApp crowdsources expert knowledge to establish consensus diagnostic gene panels. Nat. Genet. 2019;51:1560–1565. doi: 10.1038/s41588-019-0528-2. [DOI] [PubMed] [Google Scholar]
  • 9.Serpieri V., Vezain-Mouchard M., Orsi A., Lecointre M., Mazzotta C., Marguet F., Garbelli A., Marcorelles P., Celli L., Goldenberg A., et al. Bi-allelic variants in FSD1L cause a neurodevelopmental disorder overlapping with L1 syndrome. American Journal of Human Genetics. 2026:113. doi: 10.1016/j.ajhg.2026.01.014. [DOI] [PubMed] [Google Scholar]
  • 10.Ioannidis N.M., Rothstein J.H., Pejaver V., Middha S., McDonnell S.K., Baheti S., Musolf A., Li Q., Holzinger E., Karyadi D., et al. REVEL: An Ensemble Method for Predicting the Pathogenicity of Rare Missense Variants. Am. J. Hum. Genet. 2016;99:877–885. doi: 10.1016/j.ajhg.2016.08.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Cheng J., Novati G., Pan J., Bycroft C., Žemgulytė A., Applebaum T., Pritzel A., Wong L.H., Zielinski M., Sargeant T., et al. Accurate proteome-wide missense variant effect prediction with AlphaMissense. Science. 2023;381 doi: 10.1126/science.adg7492. [DOI] [PubMed] [Google Scholar]
  • 12.Quinodoz M., Peter V.G., Cisarova K., Royer-Bertrand B., Stenson P.D., Cooper D.N., Unger S., Superti-Furga A., Rivolta C. Analysis of missense variants in the human genome reveals widespread gene-specific clustering and improves prediction of pathogenicity. Am. J. Hum. Genet. 2022;109:457–470. doi: 10.1016/j.ajhg.2022.01.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Jaganathan K., Kyriazopoulou Panagiotopoulou S., McRae J.F., Darbandi S.F., Knowles D., Li Y.I., Kosmicki J.A., Arbelaez J., Cui W., Schwartz G.B., et al. Predicting Splicing from Primary Sequence with Deep Learning. Cell. 2019;176:535–548.e24. doi: 10.1016/j.cell.2018.12.015. [DOI] [PubMed] [Google Scholar]
  • 14.Zeng T., Li Y.I. Predicting RNA splicing from DNA sequence using Pangolin. Genome Biol. 2022;23:103. doi: 10.1186/s13059-022-02664-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Richards S., Aziz N., Bale S., Bick D., Das S., Gastier-Foster J., Grody W.W., Hegde M., Lyon E., Spector E., et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet. Med. 2015;17:405–424. doi: 10.1038/gim.2015.30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Durkie, M., Cassidy, E.-J., Berry, I., Owens, M., Turnbull, C., Scott, R.H., Taylor, R.W., Deans, Z.C., Ellard, S., Baple, E.L. and McMullen D.J. (2024). ACGS Best Practice Guidelines for Variant Classification in Rare Disease 2024. https://www.genomicseducation.hee.nhs.uk/wp-content/uploads/2024/08/ACGS-2024_UK-practice-guidelines-for-variant-classification.pdf
  • 17.Li J., Choi J., Cheng X., Ma J., Pema S., Sanes J.R., Mardon G., Frankfort B.J., Tran N.M., Li Y., Chen R. Comprehensive single-cell atlas of the mouse retina. iScience. 2024;27 doi: 10.1016/j.isci.2024.109916. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Li J., Wang J., Ibarra I.L., Cheng X., Luecken M.D., Lu J., Monavarfeshani A., Yan W., Zheng Y., Zuo Z., et al. Integrated multi-omics single cell atlas of the human retina. Nat Genet. 2026 doi: 10.1038/s41588-025-02454-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Blanchette M., Kent W.J., Riemer C., Elnitski L., Smit A.F.A., Roskin K.M., Baertsch R., Rosenbloom K., Clawson H., Green E.D., et al. Aligning multiple genomic sequences with the threaded blockset aligner. Genome Res. 2004;14:708–715. doi: 10.1101/gr.1933104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Schumacker S.T., Coppage K.R., Enke R.A. RNA sequencing analysis of the human retina and associated ocular tissues. Sci. Data. 2020;7:199. doi: 10.1038/s41597-020-0541-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Desmet F.O., Hamroun D., Lalande M., Collod-Béroud G., Claustres M., Béroud C. Human Splicing Finder: an online bioinformatics tool to predict splicing signals. Nucleic Acids Res. 2009;37:e67. doi: 10.1093/nar/gkp215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Jurkute N., Cancellieri F., Pohl L., Li C.H.Z., Heaton R.A., Reurink J., Bellingham J., Quinodoz M., Yioti G., Stefaniotou M., et al. Biallelic variants in coenzyme Q10 biosynthesis pathway genes cause a retinitis pigmentosa phenotype. NPJ Genom. Med. 2022;7:60. doi: 10.1038/s41525-022-00330-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Daich V.M., Schiff E., Malka S., Wright G., Mahroo O.A., Webster A.R., Michaelides M., Arno G., PHYH c. 678+5G>T Leads to In-Frame Exon Skipping and Is Associated With Attenuated Refsum Disease. Investig. Ophthalmol. Vis. Sci. 2024;65:38. doi: 10.1167/iovs.65.2.38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Chandrasekhar S., Lin S., Jurkute N., Oprych K., Estramiana Elorrieta L., Schiff E., Malka S., Wright G., Michaelides M., Mahroo O.A., et al. Investigating Splice Defects in USH2A Using Targeted Long-Read Sequencing. Cells. 2024;13 doi: 10.3390/cells13151261. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Carim-Todd L., Escarceller M., Estivill X., Sumoy L. Characterization of human FSD1, a novel brain specific gene on chromosome 19 with paralogy to 9q31. Biochim. Biophys. Acta. 2001;1518:200–203. doi: 10.1016/s0167-4781(01)00178-6. [DOI] [PubMed] [Google Scholar]
  • 26.Stein P.A., Toret C.P., Salic A.N., Rolls M.M., Rapoport T.A. A novel centrosome-associated protein with affinity for microtubules. J. Cell Sci. 2002;115:3389–3402. doi: 10.1242/jcs.115.17.3389. [DOI] [PubMed] [Google Scholar]
  • 27.Marín I. Origin and diversification of TRIM ubiquitin ligases. PLoS One. 2012;7 doi: 10.1371/journal.pone.0050030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Ivliev A.E., 't Hoen P.A.C., van Roon-Mom W.M.C., Peters D.J.M., Sergeeva M.G. Exploring the transcriptome of ciliated cells using in silico dissection of human tissues. PLoS One. 2012;7 doi: 10.1371/journal.pone.0035618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Mercey O., Kostic C., Bertiaux E., Giroud A., Sadian Y., Gaboriau D.C.A., Morrison C.G., Chang N., Arsenijevic Y., Guichard P., Hamel V. The connecting cilium inner scaffold provides a structural foundation that protects against retinal degeneration. PLoS Biol. 2022;20 doi: 10.1371/journal.pbio.3001649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Gerdes J.M., Katsanis N. Microtubule transport defects in neurological and ciliary disease. Cell. Mol. Life Sci. 2005;62:1556–1570. doi: 10.1007/s00018-005-5007-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Figures S1–S8, supplemental notes, and supplemental methods
mmc1.pdf (11.8MB, pdf)
Document S2. Article plus supplemental information
mmc2.pdf (17.9MB, pdf)

Data Availability Statement

  • •

    Affected individual data supporting the findings of this study are confidential and subject to ethical restrictions. Access to data can be requested from the corresponding author, subject to appropriate approvals and compliance with data protection regulations.

  • •

    Research on the de-identified individual data used in this publication can be carried out in the Genomics England Research Environment, subject to a collaborative agreement that adheres to individual-led governance. All interested readers will be able to access the data in the same manner as the authors did. For more information about accessing the data, interested readers may contact research-network@genomicsengland.co.uk or access the relevant information on the Genomics England website: https://www.genomicsengland.co.uk/research.

Articles from American Journal of Human Genetics are provided here courtesy of American Society of Human Genetics

RESOURCES