Multimorbidity due to novel pathogenic variants in the WFS1/RP1/NOD2 genes: autosomal dominant congenital lamellar cataract, retinitis pigmentosa and Crohn’s disease in a British family

Introduction

Next generation sequencing (NGS) has allowed the identification of digenic or trigenic inheritance. We have encountered a rare instance of coexistent autosomal dominant congenital cataract (CC), retinitis pigmentosa (RP) and Crohn’s disease (CD) in affected members of a pedigree that was initially investigated for CC that presented in a 4 year old. Whole exome sequencing (WES) identified a WFS1 variant causing CC in all the affected individuals, a RP1 variant causing RP in all but one affected family member and a NOD2 variant found in five affected family members with CD, thus far confirmed clinically in one member of the pedigree (figure 1). CC was the first autosomal disease to be genetically mapped in humans and it is detected either at birth or during the first decade of life.1 2 CCs are mostly inherited as autosomal dominant, but recessive and X-linked inheritance also occur. Nearly 50 cataract-causing genes have been identified (https://cat-map.wustl.edu/).1 2 Cataract is, therefore, a clinically and genetically heterogeneous disease, displaying a range of phenotypes.1 2 It can be both isolated and associated with other ocular defects such as anterior segment mesenchymal dysgenesis, glaucoma, microcornea and aniridia, or with syndromic conditions such as congenital adrenal hyperplasia3 and Wolfram syndrome (WS),4 as well as a consequence of metabolic disease such as diabetes.2 Thus far the reported genetic association with CC has involved single genes, but NGS allows more complex inheritance patterns to be identified.

Request permission

Here, we report the multigenic inheritance of three heterozygous variants WFS1 p.V633L, RP1 p.L2115* and NOD2 p.R702W in a British pedigree with lamellar CC, RP and CD. All three genes play crucial roles in ocular tissues. For instance, pathogenic variants in WFS1 cause WS, isolated CC and optic atrophy. RP1 variants cause RP, also a clinically and genetically heterogeneous group of progressive inherited retinal dystrophies in which cataract is also mentioned as a clinical phenotype. Finally, NOD2, a member of a large family of nucleotide oligomerization domain (NOD) like receptors (NLRs) that are multidomain cytosolic receptors that form signalling complexes to mediate major cellular pathways such as autophagy, inflammation and the innate immune response to cellular injury and stress.5 NOD2 is expressed in the eye6 and heterozygous variants can cause cataract.7 The NLR family of receptors are expressed in eye tissues during pathological conditions suggesting a role in dry eye, ocular infection, retinal ischaemia, cataract, glaucoma, age-related macular degeneration, diabetic macular oedema and diabetic retinopathy. Moreover, variants in NOD2 result in a defective response to microbiota in the intestinal tract, which have been proposed to lead to CD, an inflammatory bowel disease8 linking NOD2 mutations to a diverse set of pathologies involving both the colon and the eye. Our data provide genetic evidence for potential overlap in the consequences of mutations in WFS1, RP1 and NOD2 to account for the observed multimorbidities of CC, RP and CD within the family pedigree.

Materials and methods

Results

A five-generation English pedigree with 10 affected, 3 unaffected and 6 spouses presented with bilateral lamellar CC (figure 1, figure 2A–F, table 1 and table 2). This family was first examined clinically nearly 30 years ago and was found then to present with isolated lamellar CC in all the affected individuals. At that time, it was not possible to map by linkage studies, but now three DNA samples (IV-2, V-3 and V-4) became available for WES. Using the Phenopolis genetic variant analysis pipeline, variants were then filtered by allele frequency. Fifty from a total of 3064 rare coding variants in individual IV-2, 79 rare variants out of 2831 in V-3, and 58 rare variants out of 2925 for individual V-4 remained.

Request permission

The top scoring variant by CADD was a rare heterozygous variant NM_006269.2: c.6344T>G; L2115* in exon 4 of RP1 with a score of 35.1 and NM_006005.3: c.1897G>C; p.V633L in exon 8 of WFS1 with a score of 26.3 (figure 2, table 1). Individual (IV-2) harboured WFS1, RP1 and NOD2 gene variants, his son (V-4) also inherited all three variants (figure 2, table 1). His daughter (V-3) inherited the WFS1 and NOD2 variants. Direct Sanger sequencing confirmed both variants (figure 2) and they cosegregated in other affected family members. Subsequently, we found out that many of the affected members of the family had RP (figure 1, table 2). We reviewed our WES data for the genes associated with CD and found a likely pathogenic missense variant NM_022162.3: c.2104C>T; p.R702W with a CADD score of 24.1 in exon 4 of the NOD2 gene in individuals IV-2, IV-5, V-3, V-4 and V-7. Individual V-7 developed very severe CD symptoms in her early 20s (figure 2, table 1,table 2). This variant is highly frequent in the Western population and can be associated with CD, although there might also be population-dependent component, for example, Zaahl et al.9 None of the variants in WFS1, RP1 or NOD2 were found in unaffected family members.

Rare Exome Variant Ensemble Learner tool was used to predict the pathogenicity of missense variants based on a combination of scores from 13 individual tools; Genomic Evolutionary Rate Profiling Neutral Rate corresponds to the neutral rate conservation score of the site and CADD is a score for the deleteriousness of a variant. A CADD score of >15 is considered damaging; Sorts Intolerant Frome Tolerant score (0.0–0.05) to check the deleteriousness of the amino acid substitution on the protein function; * indicates protein truncation as a result of the sequence change.

Discussion

Here, we report heterozygous variants in WFS1, RP1 and NOD2 associated with the multimorbidities of CC, RP and CD. These multimorbidities presented in a five-generation family pedigree that was initially identified some 30 years ago with autosomal dominant CC.

In WS, compound heterozygosity is often observed particularly within the C-terminal tail domain.10 There are three other WFS1 variants Q667X, Q668X and Y669C adjacent to the V633L variant we report here.11 Likewise for RP, the majority of pathogenic variants occur in exon 412 causing both dominant and recessive disease as well as other complex genetic inheritance patterns involving other genes.13 The RP1 variant reported here truncates the protein by just 41 residues and therefore as a result it is only a portion of the very C-terminal disordered domain of the protein that is lost (https://www.uniprot.org/uniprotkb/P56715/entry). How such a small truncation might affect RP1 function remains to be determined. The R702W variant in NOD2 has been variably associated with CD9 10 and in some studies it has been identified as benign (http://mgdd.pasteur.ma/search.php?info=GeneInfo&disease_type=Multifactorial&id_gene=221). Here we hypothesise that these three variants in WSF1, RP1 and NOD2 lead to the current diagnosis of all three diseases (CC, RP and CD, respectively) in a single member of the pedigree, V-7. Four other family members (IV-2, IV-5, V-3 and V-4) have been identified who harbour the same three sequence variants. All have CC and RP, but we are unaware of their full pathogenic spectrum at this point in time (table 2). It may be that CD has yet to develop or perhaps our single contact who provided the family history for the pedigree presented here is uninformed of their CD diagnosis. It is, however, most likely that CD is a multifactorial disease with a greater weight of environmental exposure required for phenotype development in a polygenetic ‘at risk’ individual within the pedigree.

WS was first described by Wolfram and Wagener.14 It is a severe neurodegenerative disease mainly characterised by juvenile-onset diabetes mellitus and optic atrophy along with several other disorders. Wolframin is integral membrane protein within the endoplasmic reticulum (ER),15 suggesting a physiological role in membrane trafficking, secretion, processing and/or regulation of ER calcium homeostasis. WFS1 comprises 8 exons encoding an 890-residue protein called Wolframin. It is a hydrophobic protein with nine transmembrane segments and large hydrophilic regions at both the N-terminal and C-terminal. It can form a tetramer, is glycosylated and locates to the ER.16 Disease-causing variants in WFS1 are known to cause autosomal-recessive WS. Its variable expression in tissues results in a range of clinical complications such as depression, diabetes, hearing and vision loss, bilateral cataract in conjunction with other systemic disorders as well as isolated, dominant bilateral CCs. Wolframin variants are reported to result in ER stress and loss of ER function, disrupting the export of vesicle cargo proteins such as insulin.17 To date four autosomal dominant families (including one in this study) and three sporadic cases have been documented with variants in WFS1 causing isolated cataracts or other WS pathologies18–21 (table 3). The majority of recessive WFS1 inactivating variants cause typical WS, whereas dominant non-inactivating variants are associated with less severe, but more varied, phenotypes. Here, we report a novel missense variant (c.1897G>C; p.V633L) in WFS1 associated with lamellar CC. V633 is highly conserved in WFS1 across species, including human, mouse, monkey, rat and chicken (https://www.ncbi.nlm.nih.gov/nuccore/?term=WFS1). Recently, La Morgia and coauthors suggested that WFS1 variants can affect mitochondrial function and that in turn will affect calcium regulation,22 a known cataract risk factor.23

Furthermore, in the above pedigree, the RP1 variant p.L2115* was found in exon 4, along with the WFS1 and NOD2 variants, causing late onset RP in all family members who harboured this variant. RP1 comprises four exons and encodes a 2156-residue protein. RP is a clinically and genetically heterogeneous group of progressive inherited retinal disorders causing retinal degeneration and affecting 1 in 3000 people.24 The clinical variation reflects differences in the age of onset, progression, retinal appearance and visual outcome.12 Pathogenic variants in RP1 often lead to a truncated protein, but a splice site variant in a long non-coding RNA was linked to Volkmann cataract, where opacities are formed around the lens sutures.25 This indicates the potential importance of RP1 to the lens as well as the retina. Indeed, RP1 is expressed in the postnatal mouse lens epithelium (https://research.bioinformatics.udel.edu/iSyTE/ppi/) and should be considered as a CC causing gene.

Through correspondence with a single family member, we became aware that one individual (V-7) had presented with severe CD in their early 20s. Later, we found four other family members carrying the NOD2 pathogenic variant, but we are currently unaware of any clinical diagnosis of CD as mentioned earlier. CD is a chronic inflammatory disease of the gastrointestinal tract affecting 26–200 per 100 000 in European populations. The exact cause of CD is unknown, though it is likely to involve a disrupted immunological response to gut microbiota in genetically susceptible individuals.26 Genetic studies have begun to shed light on CD. Many of the major genes related to CD, including NOD2, ATG16L1, IL23R and IRGM, are involved in immune system function27 and are induced by the presence of bacteria in the lining of the digestive tract.28 Recently, it has also been shown that store-operated calcium entry regulates immune cell function in inflammatory bowel disease.29 This is a potential link to CD contributing to exacerbation of the symptoms/time of onset as seen in the case of the affected individual (V-7; Figure 1).

With the emergence of NGS technology, multiple disease genes are being found in the same family. Digenic inheritance was first reported in 1994 for human RP30 followed by Bardet-Biedl syndrome.31 Recently, digenic inheritance of disease-causing variants in EPHA2 and SLC26A4 have also been reported in Pendred syndrome.32 Recently, Kinoshita and coauthors33 described trigenic ADH5/ALDH2/ADGRV1 mutations in myelodysplasia with Ushers syndrome.

In summary, herein we report the putative trigenic inheritance of WFS1, RP1 and NOD2 variants causing lamellar CC, autosomal dominant RP and CD as multimorbidity phenotypes in a single British pedigree. No one in the pedigree with CC, who was available for genotyping, presented only with the WSF1 variant. All with CC and the WSF1 variant presented with either RP1 (III-2, IV-7, V-2, V-5 and V-6) or NOD2 (V-3 or both (IV-2, IV-, V-4; V-7) accompanying WSF1 (figure 1). We speculate that there might be a link to CD as a result of disrupted calcium regulation. The literature suggests that ocular phenotypes accompany CD34 identifying multimorbidities as increasingly important for the development of personalised medicines and treatment strategies. However, we cannot at this point in time provide functional studies to support the genetic associations reported in this study.