Identification of novel and hotspot mutations in the channel domain of ITPR1 in two patients with Gillespie syndrome

Abstract

ITPR1 encodes an intracellular receptor for inositol 1,4,5-trisphosphate (InsP3) which is highly expressed in the cerebellum and is involved in the regulation of Ca2 + homeostasis. Missense mutations in the InsP3-binding domain (IRBIT) of ITPR1 are frequently associated with early onset cerebellar atrophy. Gillespie syndrome is characterized by congenital ataxia, mild to moderate intellectual disability and iris hypoplasia. Dominant or recessive ITPR1 mutations have been recently associated with this form of syndromic ataxia. We performed next generation sequencing in two simplex families with Gillespie syndrome and identified de novo pathological mutations localized in the C-terminal channel domain of ITPR1 in both patients: a recurrent deletion (p.Lys2596del) and a novel missense mutation (p.Asn2576Ile) close to a point of constriction in the Ca^2 + pore. Our study expands the mutational spectrum of ITPR1 and confirms that ITPR1 screening should be implemented in patients with congenital cerebellar ataxia with or without iris hypoplasia.

Highlights

• •De novo ITPR1 mutations were identified in two patients with Gillespie syndrome.
• •The Asn2576Ile mutation is located close to a point of constriction in the Ca^2 + pore.
• •ITPR1 is frequently involved in congenital ataxias with or without iris hypoplasia.

1. Introduction

In 1965 F.D Gillespie described a syndrome consisting of cerebellar hypoplasia, mild to moderate intellectual disability and aniridia, characterized by the absence of the pars pupillaris of the iris and the pupillary border (Gillespie, 1965). Recently de novo missense mutations in the inositol 1,4,5-triphosphosphate receptor type 1 gene (ITPR1) were reported in 12 families with Gillespie syndrome (MIM #206700) whereas compound heterozygosity for inactivating mutations in the same gene were identified in a smaller proportion of cases (McEntagart et al., 2016, Gerber et al., 2016) We used trio-based whole-genome sequencing (WES) or targeted next generation sequencing, in two patients with congenital ataxia, psychomotor delay and iris hypoplasia in which aniridia causing PAX6 and FOXC1 genes (Ansari et al., 2016) were excluded by previous screening.

2. Case report

We recruited patients from two unrelated families with congenital ataxia and iris hypoplasia as part of a cohort of ataxic patients followed-up for more than 15 years at the Bambino Gesù Children's Hospital in Rome (Department of Neurosciences, Unit of Neuromuscular and Neurodegenerative Disorders). Informed consent was obtained from all participating subjects according to the Declaration of Helsinki. For evaluation of the neuropsychological profile, a protocol characterized by the evaluation of cognitive, visuomotor, language and academic abilities was administered. Following the identification of the mutations, all neuroimaging datasets were retrospectively analyzed by expert pediatric neurologists. The clinical features of the probands are summarized in Table 1: early-onset (within the first 6 months of life), nonprogressive cerebellar ataxia, hypotonia, developmental delay, moderate intellectual disability are constant findings. In both patients, early onset horizontal nystagmus with preserved eye movements was present. Slitlamp examination revealed bilateral iris hypoplasia. Visual acuity was bilaterally reduced in Patient 1 (29 years of age at the time of the last examination); not measurable in Patient 2 (2 years of age at the time of examination). In Patient 1, Flash visual evoked potential (VEP) stimulation showed reduced amplitude and delayed time-to-peak responses in both eyes. Flicker VEP showed reduced amplitude response (Fig. S1A). Patient 2 showed normal VEP values (Fig. S1B). Electroretinography (ERG) was normal in both patients. Additional findings in patient 1 included: precocious puberty, autoimmune thyroiditis, scoliosis and pulmonary stenosis, which has been already reported in three patients (McEntagart et al., 2016, Kieslich et al., 2001, Wittig et al., 1988).

Table 1.

Clinical features of the patients.

	Patient 1	Patient 2
Age (years)	29 years	2 years
Gender	F	F
Mutation	p.Asn2576Ile	p.Lys2596del
	Ocular features
Bilateral iris hypoplasia	(+)	(+)
Normal fundus	(+)	(+)
Visual acuity	Reduced (LogMAR 0,15)	n.d
Ptosis	(+)	(−)
Flicker and Flash VEP	Reduced amplitude	Normal
	Neurological signs (+) or (−)
Nystagmus	(+)	(+)
Ataxia	(+) 2 years	(+) 1 years
Postural tremor	(−)	(−)
Slurred speech	(+)	(n.d)
Dysphagia	(−)	(−)
General hypotonia	(+)	(+)
Oculomotor apraxia	(−)	(−)
Pyramidal sign	(+) Reduced DTR	(−)
Intellectual disability	(+) Moderate	(+) Mild
	Motor development
Sit without support	Delayed	Not achieved
Walk unassisted	10 years	Not achieved
	MRI findings
Cerebellar and/or vermis atrophy	(+)	(+)
Brainstem	(−)	(−)
Cerebrum	(−)	(−)
Other findings	Pulmonic stenosis, scoliosis, early puberty	Growth retardation

For the neuropsychological evaluation of patient 1, the WAIS-IV test (Wechsler Adult Intelligence Scale–Fourth Edition) was administered. Global intellectual functioning resulted in a Full Scale IQ score of 37, indicating moderate to severe intellectual disability. Better results were obtained in “Verbal Comprehension” (IQ = 57), “Working Memory” (IQ = 52) i.e. the ability to memorize new information and “Information Processing Speed” (IQ = 50), compared to “Perceptual Reasoning” (IQ = 46) also defined as nonverbal, fluid reasoning. Neuroimaging studies showed moderate cerebellar atrophy, predominant in the vermis in both patients (Fig. 1A I–IV). Mild periventricular hyperintensities adjacent to frontal and occipital horns are present in T2 weighted axial image of patient 2 (Fig. 1A V).

Fig. 1.

A- Neuroimaging studies: T1 weighted mid-sagittal section (I, III) and T2 weighted coronal section (II, IV) of the patients at the age of 5 (patient 1) and 2 years (patient 2), showing moderate cerebellar atrophy, predominant in the vermis. T2 weighted axial image of patient 2 showing mild periventricular hyperintensities adjacent to frontal and occipital horns (white arrowheads) (V).

B-Sanger sequencing chromatograms in the two families.

C-Protein sequence alignment of human ITPR1 with its paralogs.

D-Linear representation of the ITPR1 protein and Gillespie related mutations identified to date. Amino acid numbering is based on GenBank: NP_001161744.1 (Q14643-2; ENSP00000306253.8). Previously reported recessive inactivating mutations are highlighted in black; dominant negative (de novo) mutations are highlighted in red. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3. Methods

In patient 1 ITPR1 (mRNA sequence: NM_001168272.1; protein sequence: NP_001161744.1) was sequenced as part of a small panel of five genes causative of Non Progressive Congenital Ataxia (NPCA) (ITPR1, KCNC3, ATP2B3, CACNA1A, GRM1) using a Truseq Custom Amplicon (TSCA) (Illumina Inc., San Diego, CA) targeted capture and paired end library kit. Targeted resequencing was performed using TSCA enrichment sequenced on an Illumina MiSEQ desktop sequencer (Illumina Inc.). Expected coverage was 99% of the targeted genomic regions of interest, which achieved an average alignment performance of 98% across all samples. The obtained sequences were aligned to the reference genome (GRCh37/hg19) detecting discrepancies such as deletions, insertions and single nucleotide polymorphism (SNPs). Exome capture was carried out in patient 2 and her parents using Nextera rapid capture v1.2 (Illumina), and sequencing was performed on a NextSeq 500 platform (Illumina) (see Supplemental files). Sanger sequencing was used to confirm ITPR1 variants and segregation analysis. Homology modelling analysis of human ITPR1 was based on the Protein Data Bank (PDB) structure 3JAV (representing the ITPR1 channel from rat in a closed state). Since the site of the p.Asn2576Ile mutation and other critical residues described in the text were conserved in the human/rat ITPR1 channels (the two proteins share 98% amino acid identity), the amino acid positions of the rat ITPR1 protein were simply renumbered as in the human homologue. A modelled ITPR1 channel carrying the p.Asn2576Ile amino acid change was made by mutating the chains A and B of the PDB entry 3JAV and leaving unaltered chains C and D in the same structure. The calcium pore was determined with Caver (v3.0) (Chovancova et al., 2012) on the original rat ITPR1 channel structure. Multiple sequence alignment was performed using clustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2/).

4. Results

In patient 1, a novel missense mutation (c.7727A > T; p.(Asn2576Ile)) was identified. The variant was not present in ExAC database of 60,706 control individuals, or in ClinVar. In patient 2, we found a small deletion (c.7786–7788 delAAG; p.(Lys2596del)) previously reported in four families (McEntagart et al., 2016) (Fig. 1B). Both changes were de novo and predicted to impact protein function by dbNSFP and/or CADD, and affected residues were conserved among orthologs and paralogs (Fig. 1C). Asn2576 is part of the Ca^2 + pore tetrameric structure. Hydrogen bonding with the nitrogen atom of Leu2577 causes the side chain of Asn2576 to bend towards the pore wall to avoid the hindering of the path employed by Ca^2 + ions. Conversely, the hydrophobic residue introduced by the Ile2576 mutation cannot form an analogous hydrogen bond, thus its side chain will lean across the pore restricting its diameter and impairing the passage of Ca^2 + ions (Fig. 2).

Fig. 2.

Structural consequences of the Gillespie syndrome-causing p.Asn2576Leu substitution. A: model of the human ITPR1 channel in a closed state showing the location of the site affected by the p.Asn2576Ile mutation and the Ca^2 + pore (the four monomers are in different colors). B: detailed view around Asn2576 (please note that all the four Asn2576 residues contributed by each ITPR1 monomer are proximal to a constriction point in the Ca^2 + pore). C: view through the pore of a modelled p.Asn2576Ile ITPR1 channel mutant (the two upper helices contributing to the pore are wild type and the two lower helices mutated). In the wild type ITPR1 monomers the Asn2576 side chains are bent away from the Ca^2 + path by hydrogen bonding (black dotted lines) with the backbone of Leu2577. Conversely, in the p.Asn2576Ile mutant monomers, an analogous hydrogen bond cannot be formed by the hydrophobic side chain of the replacing Ile2576, causing steric hindrance (highlighted by red circle with arrows) in the Ca^2 + path. The channel closed state is contributed by the side chain of Phe2579 that directly lays in the pore and obstructs it. In the p.Asn2576Ile mutant, conformational changes of Phe2579 to reverse this obstruction (as needed for channel switching to the open state) are impeded owing to hindrance by the Ile2576 side chain. Amino acid numbering is according to the NCBI entry NP_001161744.1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

5. Discussion

ITPR1 is the major neuronal isoform and is part of a postsynaptic molecular complex formed by metabotropic glutamate receptor (mGluR1) and plasma membrane Ca2 + ATPase (PMCA), both mutated in patients with congenital ataxia (Guergueltcheva et al., 2012, Zanni et al., 2012). ITPR1 is particularly expressed in cerebellar Purkinje cells (Matsumoto et al., 1996). Its primary structure consists of three domains, including an InsP3-binding domain (IRBIT) domain, a regulatory/carbonic anhydrase-related protein VIII (CARP)-binding domain, and a transmembrane-ion transporting domain near the C-terminus (Sugawara et al., 2013). Heterozygous missense mutations in the IRBIT domain are most frequently been reported in spinocerebellar ataxia SCA29 (MIM #117360), characterized by early onset, very slow progressive ataxia, and delayed psychomotor development (Sasaki et al., 2015). The two mutations identified in the present study affect residues located in the C-terminal channel domain, as the majority of Gillespie syndrome-causing dominant negative mutations (Fig. 1D). As ITPR1 forms a homotetramer, if a single variant subunit can block channel function, only a minor percentage of tetramers will be functional. Compound heterozygosity for inactivating mutations in ITPR1 are detected in a smaller proportion of cases, pointing out that GS-causing mutations present distinct modes of inheritance and action.

Asn2576 is close to a point of constriction in the Ca^2 + pore, contributed by Phe2579, corresponding to the pore-constricting Phe2586 residue as determined in the rat protein (Fan et al., 2015). In the wild type protein, hydrogen bonding with the nitrogen backbone atom of Leu2577 causes the side chain of Asn2576 to bend away from Phe2579, thus allowing enough conformational freedom to this phenylalanine to unblock the Ca^2 + ion path when the channel is required to switch to the open state. Conversely, in the p.Asn2576Ile mutant the replacing isoleucine bears a hydrophobic side chain, which cannot be hydrogen-bonded to the Leu2577 backbone. Thus, the hydrophobic and larger side chain of Ile2576 will rather lean across the pore causing steric hindrance to Phe2579, preventing it to change its pore-closing conformation. An impaired closed-open state switching could underly the pathogenic mechanism of the p.Asn2576Ile mutation.

Mutations in the C-terminal channel domain (p.Gly2547Ala and p. Ile2550Asn) have been also reported in patients with congenital ataxia or severe pontine and cerebellar hypoplasia without iris abnormalities (Gonzaga-Jauregui et al., 2015, Van Dijk et al., 2017) whereas mutations in the CARP-binding/regulatory domain were identified in both SCA29 and Gillespie syndrome (Gerber et al., 2016, Ohba et al., 2013) thus making genotype-phenotype correlations difficult. Aniridia is a rare congenital disorder of either partial or complete hypoplasia of the iris and can be associated with other eye defects, such as corneal opacification, cataract, glaucoma, lens dislocation, ciliary body hypoplasia, foveal hypoplasia, strabismus, and nystagmus. The ocular features in patients with Gillespie syndrome are distinguished from complete aniridia by presenting a normal fovea and no opacification of the cornea and lens, nor the development of glaucoma. De novo PAX6 mutations were identified in patients with atypical Gillespie syndrome: one patient had partial aniridia, balance disorder, hand tremor, and learning disability. However, brain MRI revealed no cortical or cerebellar abnormalities. Additional ocular features which are not found in Gillespie syndrome included anterior lens capsule opacities, foveal hypoplasia and retinal hypopigmentation (Ticho et al., 2006, Graziano et al., 2007). Another patient showed aniridia, and moderate intellectual disability, her gait was ataxic, tremor of both upper limbs, axial hypotonia, lower limb hypertonia, ptosis, and strabismus. Ophthalmologic examination revealed bilateral symmetric aniridia but lacked the iridolenticular strands and festooned pupillary edge, pathognomonic for Gillespie syndrome. ITPR1 expression in human non-pigmented ciliary epithelium derived from the anterior segment of the eye has been found to be regulated by both FOXC1 and PITX2, two transcription factors generally associated with aniridia and congenital glaucoma (Bheeman et al., 2008). Further studies will be necessary to evaluate the expression of these two genes in ITPR1-mutated cells. Although a systematic evaluation of the cognitive and behavioral profile of ITPR1 mutated patients has not been performed, most patients with SCA29 show a cognitive profile within the lower limit of the normal range whereas individuals with ITPR1-related Gillespie syndrome present a moderate to severe intellectual disability. A GS patient with cerebellar cognitive affective syndrome without global intellectual disability has also been reported (Mariën et al., 2008). Our study expands the mutational and clinical spectrum of ITPR1-related congenital ataxia and confirms that screening of ITPR1 and other genes involved in Ca2 + homeostasis should be implemented in patients with early-onset ataxia with or without iris hypoplasia.

Financial disclosure statement

The authors have no financial relationships relevant to this article to disclose.

Appendix A. Supplementary data

Supplementary material