Broadening INPP5E phenotypic spectrum: detection of rare variants in syndromic and non-syndromic IRD

Abstract

Pathogenic variants in INPP5E cause Joubert syndrome (JBTS), a ciliopathy with retinal involvement. However, despite sporadic cases in large cohort sequencing studies, a clear association with non-syndromic inherited retinal degenerations (IRDs) has not been made. We validate this association by reporting 16 non-syndromic IRD patients from ten families with bi-allelic mutations in INPP5E. Additional two patients showed early onset IRD with limited JBTS features. Detailed phenotypic description for all probands is presented. We report 14 rare INPP5E variants, 12 of which have not been reported in previous studies. We present tertiary protein modeling and analyze all INPP5E variants for deleteriousness and phenotypic correlation. We observe that the combined impact of INPP5E variants in JBTS and non-syndromic IRD patients does not reveal a clear genotype-phenotype correlation, suggesting the involvement of genetic modifiers. Our study cements the wide phenotypic spectrum of INPP5E disease, adding proof that sequence defects in this gene can lead to early-onset non-syndromic IRD.

Fig. 1. Pedigrees of the 12 INPP5E families described in this study.

For each family, the specific IRD phenotype diagnosed is mentioned above each pedigree (LCA, Leber Congenital Amaurosis; RCD, Rod Cone Degeneration). Mildly syndromic families C and K are indicated with a hashtag (#). Affected male and female subjects are represented with black squares or circles, respectively. Probands are indicated by a black arrow. The five unaffected sisters in family L are indicated with the superscript 5–10. Novel variants are indicated in bold. When performed, segregation of the INPP5E variants in other family members is shown. First cousin marriage is indicated by a double line. All presented variants refer to the INPP5E transcript NM_019892.5.

Fig. 2. INPP5E structure and protein variants.

a INPP5E secondary structure, tolerance landscape calculated using MetaDome, and distribution of known causal variants. Multiple sequence alignment-derived INPP5E motifs and catalytic domain were highlighted using different colors, while variants were divided into two groups, depending on whether they were found in syndromic or non-syndromic IRD patients. Variants found in our patients are in bold, while novel variants described in this study are additionally highlighted in red. Variants p.(Ser249Phe) and p.(Arg596Thr), found to be part of the same complex allele, are indicated by boxes. Variant Y543*, present in both syndromic and non-syndromic patients, results from two different nucleotide changes: c.1629C>G (JBTS) and c.1629C>A (IRD). b INPP5E tertiary structure. The tridimensional structure was predicted only for C-terminal 349 amino acids (residues 275–623) available on PDB (ID: 2XSW), as the N-terminal half was classified as a disordered region. Two glycerol molecules, acting as a proxy for the larger ligand of this protein (i.e., phosphatidylinositol polyphosphate), are shown in green. Amino acid residues for which missense variants in our patients were found are highlighted in red, except for Serine 249 located in the un-modeled region. c–e Predicted effect of missense variants p.(Arg621Gln) and p.(Arg621Trp) on ligand binding. In the wild-type protein model, Arginine 621 is located in close proximity (3 Å) to one glycerol molecule, which uniquely interacts by establishing one ion bond, indicated by a blue dashed line (c). Missense variants introducing Glutamine (d) or Tryptophan (e) are predicted to increase distance with the glycerol of 4.7 Å and 5.4 Å, respectively, thus disrupting the ion bond.

Fig. 3. Clinical phenotypes of INPP5E-IRD patients.

Images show fundus photos (left column), fundus autofluorescence (middle column), and OCTs (right column) for a representative subset of individuals. The specific IRD phenotype of each patient is given in brackets (LCA–Leber congenital amaurosis; RCD–Rod-cone degeneration). Novel INPP5E variants are highlighted in bold.

Fig. 4. Meta-analysis of all pathogenic INPP5E variants and their phenotypic correlation.

a Distribution of the INPP5E variants in all reported INPP5E patients. Four different phenotypes of increasing severity were marked by circles (non-syndromic IRD), diamonds (IRD-mild JBTS, reported in this study), squares (JBTS with IRD), and triangles (JTBS without IRD). b Violin plot of the cumulative CADD Phred scores for the INPP5E variants in syndromic (JBTS) and non-syndromic IRD cases.