Expanding SPTAN1 monoallelic variant associated disorders: From epileptic encephalopathy to pure spastic paraplegia and ataxia

Purpose Nonerythrocytic αII-spectrin (SPTAN1) variants have been previously associated with intellectual disability and epilepsy. We conducted this study to delineate the phenotypic spectrum of SPTAN1 variants. Methods We carried out SPTAN1 gene enrichment analysis in the rare disease component of the 100,000 Genomes Project and screened 100,000 Genomes Project, DECIPHER database, and GeneMatcher to identify individuals with SPTAN1 variants. Functional studies were performed on fibroblasts from 2 patients. Results Statistically significant enrichment of rare (minor allele frequency < 1 × 10–5) probably damaging SPTAN1 variants was identified in families with hereditary ataxia (HA) or hereditary spastic paraplegia (HSP) (12/1142 cases vs 52/23,847 controls, p = 2.8 × 10–5). We identified 31 individuals carrying SPTAN1 heterozygous variants or deletions. A total of 10 patients presented with pure or complex HSP/HA. The remaining 21 patients had developmental delay and seizures. Irregular αII-spectrin aggregation was noted in fibroblasts derived from 2 patients with p.(Arg19Trp) and p.(Glu2207del) variants. Conclusion We found that SPTAN1 is a genetic cause of neurodevelopmental disorder, which we classified into 3 distinct subgroups. The first comprises developmental epileptic encephalopathy. The second group exhibits milder phenotypes of developmental delay with or without seizures. The final group accounts for patients with pure or complex HSP/HA.


Introduction
The αII-spectrin gene, SPTAN1 (OMIM 182810), encodes a membrane scaffolding protein that plays an important role in the maintenance of integrity of myelinated axons, axonal development, and synaptogenesis. 1Heterozygous SPTAN1 pathogenic variants have been previously reported with variable phenotypes, most frequently causing mild to severe developmental epileptic encephalopathy (DEE) and developmental delay (DD) 2 and rarely with hereditary motor neuropathy and autosomal recessive hereditary spastic paraplegia (HSP). 3,4A mouse model harboring αII-spectrin missense variant (p.Arg1098Gln) was reported to develop progressive ataxia with global neurodegeneration and seizures. 5On the basis of these findings, we carried out a SPTAN1 gene enrichment analysis in the 100,000 Genomes Project (100K GP) 6 and identified a statistically significant enrichment for rare probably damaging variants in hereditary ataxia (HA) and HSP groups.In this study, we present an extended phenotypic spectrum of neurologic syndromes caused by pathogenic variations of SPTAN1 gene.

Patients
Our initial cohort comprised 100K GP neurology patients. 6All 100K GP genomes were previously screened for single nucleotide variants, small insertions/deletions, structural variants (SVs) or copy number variants (CNVs), and short tandem repeats in relevant genes from the PanelApp virtual gene panels (Genomics England). 7We then screened DECIPHER database cohort for patients carrying single nucleotide variants and/or SVs/CNVs in SPTAN1 gene. 8Additional families were subsequently recruited through GeneMatcher. 9All coding variants reported in this article are with reference to SPTAN1 RefSeq: NM_001130438.3transcript.All procedures adhered to the principles set out in the Declaration of Helsinki and all patients/their guardians included in the study consented to participation according to ethical approval of the recruiting center.

Gene enrichment analysis
Case-control gene enrichment analysis was performed within the rare disease component of the 100K GP.Cases were defined as all 100K GP probands recruited under HA/ HSP, whereas controls were all remaining probands recruited into the 100K GP except for those with neurologic and neurodevelopmental disorders or metabolic disorders.Enrichment of SPTAN1 rare, probably damaging variants in cases compared with controls was assessed via a two-sided Fisher exact test.The contributing variants were defined as rare (minor allele frequency < 1 × 10 -5 ) and either proteintruncating variants or missense variants predicted to be pathogenic by 2 in silico tools (Combined Annotation Dependent Depletion [CADD] 10 and Polymorphism Phenotyping [PolyPhen] 11 ).

Functional studies
Fibroblasts derived from patient 1 (p.Arg19Trp) and patient 29 (p.Glu2207del) were used to test the functional effects of SPTAN1 variants on protein expression compared with that of healthy unrelated controls.Western blot analysis, immunocytochemistry, and confocal microscopy were performed as previously described. 3ructural modeling of SPTAN1 missense variants Three-dimensional protein modeling was used to evaluate the effect of reported SPTAN1 missense variants.Although the crystal structure for full-length αII-spectrin is unknown, crystal structures of the N-terminal tetramerization site and 2 spectrin repeat unit of chicken brain αII-spectrin have been solved (Protein Data Bank: 3F31 and 3Fb2). 12,13We used the Protein Homology/analogY Recognition Engine V 2.0 (Phyre2) predicted models for C-terminal and spectrin repeats 13 to 20 of αII-spectrin protein. 14DynaMut software was used to predict variant effect. 15For simulating amino acid substitutions and visualization, UCSF Chimera built-in tools were used. 16In addition, in silico pathogenicity prediction analysis of all missense variants identified in the study and those previously reported in literature was conducted.

Results
SPTAN1 heterozygous damaging variants are enriched in families with HSP or HA SPTAN1 was investigated as a candidate gene for HA or HSP using gene enrichment analysis in the rare disease component of the 100K GP, which has a total of 35,422 rare disease families, including 1142 HA/HSP probands as cases and 23,847 non-neurologic/non-metabolic unrelated individuals as controls.A case-control analysis revealed a statistically significant enrichment of rare probably damaging heterozygous variants of SPTAN1 in probands with HA or HSP (12/1142 cases vs 52/23,847 controls, p = .00002846,odds ratio = 4.8594, 95% CI = 2.5867-9.1290)(Supplemental Table 1).None of the SPTAN1 variants found in controls were protein-truncating variants and none overlapped with any of the missense variants described in the study.
Subsequently, we screened 100K GP neurology cohort (16,014 individuals with neurodevelopmental disorders) for probably damaging SPTAN1 variants in families with spasticity and ataxia in addition to the previously described phenotypes of seizures and/or intellectual disability (ID).We identified 11 patients from 9 families (Table 1, Figures 1 and 2).Patients 1 to 4 had pure HSP phenotype and shared the same SPTAN1 variant, p.(Arg19Trp).Later-onset and a more complex phenotype was noted in patient 8 who harbored p.(Ser2448Phe) variant.Although patient 7 was recruited under early-onset dystonia phenotype, she presented with abnormal eye movements, ataxia, myoclonus, and dyspraxia and had SPTAN1 variant, p.(Arg2124Cys).Patient 10, who had pure ataxia, harbored a heterozygous splice alteration in

Abnormal eye movement
Convergent strabismus    SPTAN1 (NC_000009.12[SPTAN1_v001]:c.3519+2T>G).SPTAN1 gene was also screened for CNVs/SVs in the 100K GP and 1 deletion was identified.Patient 9 carried a large heterozygous in-frame deletion (DEL1), encompassing exons 25 to 27 (Supplemental Figure 1) and presented with pure HA.Both patients 9 and 10 carried sporadic SPTAN1 variants because the de novo nature could not be confirmed owing to the unavailability of family members.Additional 3 probands with SPTAN1 variants presenting with seizures, ID, and ataxia/spasticity were identified.By screening DECIPHER database, 8 4 additional variants were identified; 1 missense variant in patient 22 and 3 de novo microdeletions in the SPTAN1 gene.DEL2 in patient 11 is a 4.46-kilobase deletion that removes exons 36 to 40.This patient presented with ataxia and severe DD.Patients 23 and 26, who carried DEL4 (exons 14-20) and DEL5 (exons 20-27), respectively, presented with DD and seizures.
A total of 13 additional SPTAN1 families were identified through GeneMatcher. 9Patients 5 and 6 shared the same de novo missense variant, p.(Arg19Trp), and HSP phenotype as patients 1 to 4. Nevertheless, patient 5 presented with complex HSP, learning disability, and seizures.Three frameshift variants were identified in 5 patients (patients 12-14, 20, and 25) with DD/Seizures.Dominant inheritance was noted in 3 patients (patients 12-14).Five patients (patients 15-18 and 21) with ID/seizures carried nonsense SPTAN1 variants.The last variant was an in-frame deletion (DEL3), which was identified in patient 19, a 2-year-old presented with DD.This deletion is almost 9 kilobase, encompasses exons 22 to 27, and overlaps with DEL5.All variants reported in this study were classified according to the guidelines of the American College of Medical Genetics and Genomics and Association for Molecular Pathology. 17Supplemental Table 2).

Clinical phenotypes
Detailed clinical information was collated for 31 individuals from 26 unrelated families carrying heterozygous variants in SPTAN1 (Table 1).Common phenotypes including ID/ learning disability and motor delay were reported in 73.5% and 58.8% of our cohort, respectively.Remarkably, around half of the patients (15/31) manifested ataxia, and seizures were reported in almost one-third of the cases (12/31).
Recently reported phenotypes of HSP/HA were identified in a subgroup of our patients.A total of 8 individuals from 6 families presented with pure HSP/HA and further 2 families with complex HSP (Supplemental Video 1).All patients with HSP showed typical features.Most of them had lower limb hyperreflexia and/or ataxia (6/8) whereas none had sensory abnormalities.In contrast, 2 patients with HA showed a pure phenotype.Abnormal eye movement, a common condition in patients with ataxia, was noted in 70% of this HSP/HA group.Severe phenotype of DEE was   reported in 5 patients.All had seizures in early months of life and had severe DD.The remaining 16 patients from 13 families presented with varying degrees of DD.

Functional consequences of SPTAN1 variants in patient-derived fibroblasts
We analyzed fibroblasts derived from 2 patients with the missense variants p.(Arg19Trp) and p.(Glu2207del) (Figure 3).Although western blot did not show a quantitative difference in protein expression between patient 1 and controls, a quantitative reduction of protein expression in patient 29 was noted.Nevertheless, a high immunofluorescence brightness and intense immunoreactivity of αIIspectrin was observed in both the studied patients.In patient 1, αII-spectrin was ubiquitously expressed throughout the fibroblast cells, whereas there was a more localized aggregation of αII-spectrin in the plasma membrane with a Figure 2 Schematic structure of SPTAN1 gene and its coding protein highlighting variants identified in this study.Coding exon numbers (NM_001130438.3) are reported on the gray bar.Truncating variants are indicated on the top.Missenses, in-frame deletion/ insertion, and splice variants are on the bottom.Deletion 1, 3, and 5 (green) remain in frame, whereas predictions for deletions 2 and 4 (orange) are not available.p.Gln1448= (c.4344G>A) is predicted to affect exon 33 donor splice site, based on maxENTScan (predicting splice sites using 'Maximum Entropy Principle') (maxENT score wild-type 6.99 → 3.84 mutant).The splice altering variant (NC_000009.12(SPTAN1_v001):c.3519+2T>G)predicted to alter exon 25 canonical donor splice site (maxENT score wild-type 10.28 → 2.63 mutant).Variants identified in patients presenting with HSP/HA are highlighted in red.All other variants are represented in black.
relatively higher immunofluorescence brightness observed in the fibroblasts of patient 29.

Structural modeling of SPTAN1 missense variants
The effect of missense variants on αII-spectrin protein structure (Q13813-1) was investigated using homology modeling of experimentally validated models [12][13][14] (Supplemental Figure 2).Variant p.(Arg19Trp) had the most deleterious effect on protein structure.This variant is located within the Nterminal tetramerization domain and results in steric clashes with 2 leucine residues in the beta chain.Less severe effect was noted in modeling of p.(Arg1464Trp) and p.(Arg2204Gln) variants.Structural modeling showed no structural effect for p.(Glu2271Lys), p.(Arg2124Cys), and p.(Ser2448Phe); however these would likely affect αII-spectrin heterodimerization with its partner, β-spectrin.Overall, most (4/6) of the reported missense variants led to protein destabilization (Supplemental Table 3).In addition, according to in silico pathogenicity prediction, almost all missense variants identified in this study are potentially pathogenic (Supplemental Table 4).

Discussion
Identification of an enrichment of SPTAN1 heterozygous variants in patients presenting with HA and HSP confirms SPTAN1 involvement in a wide phenotypic spectrum.We suggest that SPTAN1 is a genetic cause of neurodevelopmental disorders, with 3 phenotypic subgroups (Table 1).The first group comprises patients with DEE presenting with severe phenotype (OMIM 613477).DEE was identified in 5 families, consistent with previous reports. 2A total of 16 patients manifested milder phenotype of DD with or without childhood-onset seizures forming the second phenotypic group.The final group consists of patients (n = 10) with pure or complex HSP/HA.Involvement of SPTAN1 variants in peripheral nervous system abnormalities was previously reported for heterozygous variants causing hereditary motor sensory neuropathy 3 and biallelic variants associated with autosomal recessive HSP. 4 The phenotype of HA and HSP is further supported by a previously reported SPTAN1 mouse model presenting with unsteady gait and spasticity 5 (Supplemental Video 2) and the recently published study reporting de novo and dominant variants of SPTAN1 in patients with ataxia and patients with spastic paraplegia. 18henotypic heterogeneity may be explained by the involvement of SPTAN1 pathogenic variants in different mechanisms of pathogenicity as it was described with other structural proteins. 19Syrbe et al 2 concluded that variants in the last 2 αII-spectrin repeats are associated with severe phenotype due to protein aggregation with dominant negative effect.This mechanism is supported by our observation in αII-spectrin immunocytochemistry experiment performed on fibroblast of patient 29.However, further experiments on multiple cell lines would be imperative to support this hypothesis.
We noted that the excess of truncating variants in the milder category of our cohort (DD +/seizures) is in agreement with the proposed mechanism of quantitative defect of αII-spectrin protein leading to a milder phenotype. 20We suggest that truncating variants are responsible for a mild DD with or without epilepsy.
In our HSP/HA group, we have detected accumulation of αII-spectrin in fibroblasts of patient 1 with the recurrent p.(Arg19Trp) variant, indicating abnormal protein function.Such interesting finding adds further evidence to Van de Vondel et al 18 report of the recurrent p.(Arg19Trp) variant detected in 7 families with spastic paraplegia.All other identified variants in this group of our patients were missense variants except for 1 splice altering variant.All are predicted to have a moderate protein effect except for p.(Arg19Trp), which is localized at an essential position of the N-terminal tetramerization domain. 1 An arginine to tryptophan change, p.(Arg35Trp), at the N-terminus has been reported previously in erythrocytic α-spectrin gene (SPTA1), where it prevented N-terminal domain of α-spectrin to form heterotetramers with its beta partner.We suggest a similar mechanism for SPTAN1 variant, p.(Arg19Trp). 21In a previous report, we showed that ataxia and/or HSP cases may be accounted by hypomorphic pathogenic variants in genes known to manifest with severe phenotypes when mutated. 22The p.(Arg2124Cys) and p.(Ser2448Phe) variants identified in families 5 and 6 are missense variants that are predicted to introduce mild structural alterations in the αII-spectrin protein potentially explaining the milder neurologic impairment.In family 8, with mild and late-onset ataxia, a splice altering variant (c.3519+2T>G) was detected.The resulting predicted inframe deletion rather than a loss of function might explain the mild neurologic phenotype in this family.
It is interesting to notice that αII-spectrin forms heterotetramers with each of the 4 nonerythrocytic β-spectrins 1 to 4 at precise membrane domains. 23The first 3 β-spectrin genes (SPTBN1, SPTBN2, and SPTBN4) are responsible of multiple neurologic disorders, depending on the gene and inheritance pattern. 19Particularly, βIII spectrin (SPTBN2) is specific to Purkinje cells and is involved in a relatively pure late-onset ataxia phenotype. 24αII-Spectrin broad distribution in the neurons is therefore a plausible explanation of the pleiotropic consequences of SPTAN1 variants.Molecular diagnosis of patients with neurodevelopmental disorders is often challenging owing to both phenotypic and genetic heterogenicity.We were able to expand the phenotypic and genetic spectrum of SPTAN1 variants, shedding light on the critical role that αII-spectrin has in maintaining brain health.

Figure 1
Figure 1 Pedigrees of reported families with SPTAN1 variants showing disease segregation.

Figure 3
Figure 3 Representative images of αII-spectrin protein expression and staining pattern in fibroblast cells derived from 2 patients and unrelated controls.A. Western blot. 1.Western blotting of protein extracted from fibroblast cell lines of patients 1 and 29 and 3 wildtype age-matched controls.2. Densitometric analysis of western blot using BioRad Image Lab software after relative normalization to actin as a housekeeping protein.The analysis showed no change in protein expression in patient 1 but showed a quantitative reduction of protein expression in patient 29.B. Immunocytochemical staining of αII-spectrin expression in primary fibroblasts of patients 1 and 29 and unrelated control individual with Alexa Fluor 488 conjugated secondary antibody (green) and Hoescht 33342 nuclear staining (blue).Scale bar represents 50 μm.Immunocytochemical staining showed high immunofluorescence brightness and intense immunoreactivity and aggregation of αII-spectrin in both studied patients compared with the healthy unrelated control.Ctr, control.

Table 1
Clinical and genetic findings of 31 patients reported in this study

Table 1
Continued General Information