Background Optineurin (OPTN), a causative gene of hereditary primary open-angle glaucoma, has been recently associated with amyotrophic lateral sclerosis (ALS) with mainly autosomal recessive, but also dominant, traits. To further define the contribution of OPTN gene in ALS, we performed a mutational screening in a large cohort of Italian patients.
Methods A group of 274 ALS patients, including 161 familial (FALS) and 113 sporadic (SALS) cases, were screened for OPTN mutations by direct sequencing of its coding sequence. All patients fulfilled the El Escorial criteria for probable or definite ALS and were negative for mutations in SOD1, ANG, TARDBP and FUS/TLS genes.
Results The genetic analysis revealed six novel variants in both FALS and SALS patients, all occurring in an heterozygous state. We identified three missense (c.844A→C p.T282P, c.941A→T p.Q314L, c.1670A→C p.K557T), one nonsense (c.67G→T p.G23X) and two intronic mutations (c.552+1delG, c.1401+4A→G). The intronic c.552+1delG variant determined a splicing defect as demonstrated by mRNA analysis. All mutations were absent in 280 Italian controls and over 6800 worldwide glaucoma patients and controls screened so far. The clinical phenotype of OPTN-mutated patients was heterogeneous for both age of onset and disease duration but characterised by lower-limb onset and prevalence of upper motor neuron signs.
Conclusion In this cohort, OPTN mutations were present both in FALS (2/161), accounting for 1.2% cases, and in SALS patients (4/113), thereby extending the spectrum of OPTN mutations associated with ALS. The study further supports the possible pathological role of optineurin protein in motor neuron disease.
- motor neuron disease
- clinical neurology
Statistics from Altmetric.com
Amyotrophic lateral sclerosis (ALS) is a devastating and fatal adult-onset disorder characterised by motor neuron degeneration in the cerebral cortex, brainstem and spinal cord leading to progressive muscle weakness, wasting and paralysis. Approximately 5% of ALS cases are familial (FALS), mostly inherited as autosomal dominant traits.1 Mutations in the Cu/Zn superoxide dismutase 1 (SOD1) gene are responsible for about 20% of FALS, while variants in the TARDBP and FUS/TLS genes account for an additional 10% of cases.2 Moreover, pathogenic mutations causing rare and atypical forms of ALS have been described in several additional genes such as ALS2, SETX, VAPB, ANG, DNCT1, SPG11, FIG4, DAO and VCP.3–6 Although more than 90% of cases occur without family history, sporadic ALS (SALS) is thought to have a significant genetic component. Indeed, mutations in genes known to cause FALS have been described in apparently sporadic forms, suggesting that FALS cases are often underestimated because of misleading factors and incomplete collection of anamnestic data.7 8
Optineurin (OPTN), a causative gene of primary open-angle glaucoma (POAG; MIM: 137760),9 has been recently involved also in the pathogenesis of ALS (ALS12; MIM: 613435).10 By screening a large Japanese ALS population, three novel OPTN mutations were found in four different families and in one sporadic case, showing both recessive and dominant patterns of inheritance.10 More recently, another Japanese study confirmed the presence of OPTN mutations in FALS (3.8%) and SALS (0.29%) patients, while two independent groups reported a 1.5–2% mutation rate in FALS cases of European descent.11–13
To analyse and further clarify the contribution of OPTN gene to the etiopathogenesis of ALS, we performed a mutational screening in a cohort of 274 unrelated Italian ALS patients with a prevalence of FALS cases (161 FALS and 113 SALS).
Materials and methods
Patients and controls
The 274 ALS patients included in this study were collected by six ALS Centres participating to the Italian SLAGEN Consortium. All patients were of Italian descent, with the exception of a single FALS case of Egyptian origin. The diagnosis of ALS was made according to the El Escorial revised criteria.14 Familial history was considered positive if the proband had one (‘probable’ FALS) or more (‘definite’ FALS) first- or second-degree relatives with ALS, according to the recent criteria proposed for FALS classification.8 We also included as ‘possible’ FALS 32 individuals with third-degree or more distantly related affected relatives.8 Our ALS cohort was characterised by a male:female sex ratio of 2.1:1 and by an age of onset of 52.9±13.7 in FALS and 58.1±13.2 in SALS cases. Mutations in SOD1, ANG, TARDBP and FUS/TLS genes were excluded in all patients enrolled for this study as previously reported.15–23 The control group consisted of 280 Italian healthy volunteers with no report of neurological disorders.
Standard protocol approval and patient consent
We received approval for this study from the local ethical committees on human experimentation of the participating Institutions. Written informed consent was designed according to the Declaration of Helsinki and obtained from all patients and healthy subjects participating in the study (consent for research).
Genomic DNA was isolated from peripheral blood according to standard protocols. The entire coding region of OPTN (from exon 4 to exon 16) and the intron/exon boundaries (including at least of 40 bp of adjacent intronic sequences) were amplified by PCR as recently reported10 and directly sequenced using BigDyeTerminator v3.1 cycle sequencing kit on 3100 ABI Prism Genetic Analyser (Applied Biosystems, Foster City, California). All identified nucleotide changes were confirmed by sequencing an independent PCR product. Control DNA samples were tested by direct sequencing (exons 4, 6, 9, 13, 16) or by HRM (high-resolution melting) analysis (exon 10) using LightCycler 480 (Roche, Basel, Switzerland) as described.24 Information on HRM conditions is available upon request.
Nucleotide numbering of OPTN gene mutations reflects cDNA numbering with +1 corresponding to the A of the ATG translation initiation codon of the GenBank reference sequence NM_001008211.1. The initiation codon is Met 1.
The effect of the identified OPTN missense mutations on protein structure or function was analysed in silico with three prediction programs: PolyPhen (http://genetics.bwh.harvard.edu/pph/), PMUT (http://mmb2.pcb.ub.es:8080/PMut) and SNPs3D (http://snps3d.org). The effect of missense, synonymous and intronic variants on splicing process was analysed with Spliceview (http://zeus2.itb.cnr.it/∼webgene/wwwspliceview.html) and NNSplice (http://www.fruitfly.org/seq_tools/splice.html) programs. The presence of the Exonic Splice Enhancer (ESE) was checked using ESE Finder (http://rulai.cshl.edu/tools/ESE) and RESCUE ESE (http://genes.mit.edu/burgelab/rescue-ese) softwares.
To evaluate the potential effect of mutations on mRNA folding, the secondary structure of OPTN transcripts was predicted by MFOLD (http://bioweb.pasteur.fr/seqanal/interfaces/mfold-simple.html) using full-length wild-type mRNA sequence as a control.
Total RNA was isolated from muscle tissue with Eurozol reagent (EuroClone, Milan, Italy) according to the manufacturer's protocol. The cDNA was generated through reverse transcription using Ready-To-Go RT-PCR kit (GE Healthcare Life Sciences, Little Chalfont, UK) and analysed by PCR amplification with two primer sets spanning exons 5–9 and exons 5–11, respectively. The primer sequences are: Ex5_for: 5′-TGAGTCATGAGAATGAGAAAT-3′; Ex9_rev: 5′-CCTCTGTCTGGGTTTCAATC-3′; Ex11_rev 5′-CCACTTGTAGCTCTAACT-3′. PCR products were subcloned into TOPO-TA vector (InVitrogen, Carlsbad, California). Several independent clones were analysed by direct sequencing using commercial T3 and T7 primers flanking the vector cloning site.
A cohort of 274 ALS-affected individuals was considered for mutational screening of OPTN gene. Our cohort consisted of 161 FALS and 113 SALS cases, including six ALS patients with fronto-temporal dementia (FTD), all previously screened and tested negative for SOD1, ANG, TARDBP and FUS/TLS gene mutations.15–23
The genetic analysis of OPTN coding region (exons 4–16) revealed six novel variants predicted to affect protein sequence in six ALS patients (two FALS and four apparently SALS). In particular, we identified three missense mutations (c.844A→C p.T282P, c.941A→T p.Q314L, c.1670A→C p.K557T) and one nonsense (c.67G→T p.G23X) mutation, and two intronic changes (c.552+1delG, c.1401+4A→G), all occurring in a heterozygous state (table 1). The p.G23X and p.K557T mutations were identified in two FALS individuals, while the other four variants were identified in four SALS patients (supplemental figure 1A). No OPTN variants were found in our ALS/FTD patients.
We could not prove segregation of the identified mutations with the disease in our FALS pedigrees because no other affected family members were available for DNA sampling. The occurrence of such mutations in healthy controls was excluded by screening 280 Italian individuals with no reported history of neurological disorders and additional 75 controls from North Africa for exon 16.
To test the presence of exon 5 deletion, previously identified by Maruyama et al10 in two affected siblings from consanguineous parents, we performed a long-range PCR amplification according to their conditions. No genomic rearrangement in this region was detected in our population.
Among the six novel variants identified in our cohort, the missense p.Q314L and p.K557T mutations were predicted to be pathological and probably affecting protein function by three distinct prediction softwares (PMUT, PolyPhen and SNPs3D), while the p.T282P was scored as neutral by the same programs (supplemental table 1). Both Q314 and K557 represent highly conserved amino-acidic residues along evolution from Homo sapiens to Gallus gallus and even to Danio rerio (K557), while T282 shows a lower level of conservation (supplemental figure 1b).
The nonsense p.G23X mutation introduces a very premature stop codon in optineurin protein synthesis, while the two intronic changes c.552+1delG and c.1401+4A→G affect very conserved nucleotides in splice donor consensus sequences. Bioinformatic analyses by NNSplice and SpliceView programs predicted that both changes potentially affect the splicing process by abolishing the original donor site sequences (data not shown). To confirm the in silico predictions, we performed RT-PCR on an mRNA sample obtained from muscle tissue available from patient no. C1538 carrying the c.552+1delG change in intron 6. PCR amplification using primer pairs spanning exons 5–9 showed a 429 bp long amplicon besides the expected 540 bp fragment (figure 1A). The presence of an alternatively spliced form was further confirmed by a different set of PCR primers spanning exons 5–11 (figure 1A). Cloning and sequencing of these amplicons revealed that the c.552+1delG variant determined the activation of an exonic cryptic donor site (c.442) leading to an in-frame 111 nucleotide-long skipping of exon 6 (c.442_c.552del) corresponding to a 37-amino-acid deletion (p.148_184del) (figures 1B,C). Unfortunately, no tissue/cells were available from patient no. G5305 and the in silico prediction for the c.1401+4A→G change on the correct splicing of intron 13 could not be proven.
The main clinical features of ALS patients carrying the identified OPTN mutations are reported in table 1, while a more exhaustive clinical description is provided as supplemental appendix 1. All OPTN-positive patients showed a lower-limb onset. The age of onset was extremely variable, ranging from 24 to 71 years of age, and the progression also showed variability, including very aggressive forms (<1 year) and very slow disease courses (over 10 years) with no differences between FALS and SALS cases (table 1). It is noteworthy that four out of six patients carrying OPTN mutations were characterised by a prevalence of upper motor neuron (UMN) signs (supplemental appendix 1).
We also identified three synonymous mutations in one FALS (c.1107A→G p.L369L), two FALS (c.1704A→G p.L568L) and one healthy control (c.885T→C p.V295V) (table 2). These variants have never been described before in POAG patients and in their associated controls as well as in the other ALS patients screened so far.10–13 25 Interestingly, a bioinformatic analysis using the MFOLD program predicted a conformational change in mRNA secondary structure for the synonymous c.1704A→G p.L568L variant (data not shown).
Our study also detected the four exonic single nucleotide polymorphisms rs2234968 (c.102G→A p.T34T), rs11591687 (c.123G→A p.L41L), rs11258194 (c.293T→A p.M98K) and rs113811959 (c.489A→G p.E163E), already reported and analysed in glaucoma association studies in different populations.25 We found no significant differences between ALS cases and controls for p.T34T, p.L41L and p.E163E, while allele frequencies for p.M98K in ALS patients were in agreement with data in the NCBI dbSNP database (supplemental table 2). The c.102G→A change associated with p.T34T is predicted to modify an ESE site by in silico analysis (ESE Finder program, data not shown), probably affecting exon 4 splicing. Interestingly, the two ALS patients carrying the nonsense p.G23X and the intronic c.1401+4A→G mutations, respectively, also harboured p.T34T in a heterozygous form. Using allele-specific PCR, we found that the c.67G→T (p.G23X) and the c.102G→A (p.T34T) changes are located in cis in patient no. G3906. By contrast, the genomic distance between c.102G→A and c.1401+4A→G changes did not allow to settle the cis/trans condition in patient no. G5305.
Here we report a mutational screening of OPTN gene in a cohort of 274 Italian ALS individuals affected by both familial and apparently sporadic forms. We identified six novel mutations, all occurring in the heterozygous state in both FALS and SALS cases. Mutations in OPTN gene were previously described to account for about 16% of hereditary POAG.9 Recently, mutations in OPTN were shown to be associated also with ALS in the Japanese population with mainly autosomal recessive, but also dominant traits.10 11 In the first study including 16 patients from consanguineous marriages, 76 FALS and 597 SALS, Maruyama et al10 described three novel mutations in a total of eight ALS individuals (seven FALS from four distinct families and one SALS). They identified the nonsense p.Q398X mutation and the exon 5-deletion in homozygosity, while the missense p.E478G mutation was found in heterozygosity in two different families, each with two affected siblings with slow-progression forms. The p.E478G mutation was additionally identified by Iida et al11 in heterozygous form in one SALS patient and in homozygosity in one FALS subject. By screening 687 SALS and 26 FALS they also found two novel missense variants (p.A93P and p.R271C), which were absent in 940 controls.
More recently, two other studies confirmed the presence of OPTN mutations in heterozygous state in ALS populations of European descent, finding three novel variants (p.R96L, p.A481V, c.1242+1G→A_insA) and the c.382_383insAG mutation, already described in POAG patients, with an overall mutational frequency of 1.5–2%.12 13 26
The present study shows that OPTN gene mutations occur in heterozygous form in our cohort of patients accounting for 1.2% (2/161) of FALS and 3.5% (4/113) of apparently SALS cases, all negative for other ALS-causing genes. As already observed for other ALS-causing genes, the identification of OPTN mutations in sporadic patients raises the issue of whether they represent truly SALS forms.7 In fact, we cannot exclude the possibility that such SALS patients carrying OPTN mutations indeed represent unrecognised FALS cases because of several misleading factors, including unavailability of accurate family history, premature death of mutation carriers, misdiagnosis of other affected relatives or low penetrance of gene mutations.8 In this perspective, genetic counselling may be particularly challenging.27
Although in FALS cases we could not prove the segregation of mutations with the disease, the six variants reported in this study had not been described before either in the large cohort of POAG patients and controls screened so far (more than 6800 subjects of different ethnic origins, including at least 1600 Caucasian individuals)10 25 28 or in our 280 Italian controls.
Two out of the three novel missense mutations affect very evolutionary conserved amino-acid residues in OPTN protein, with the Q314 residue being located in a coiled-coil region and K557 in the highly conserved C-terminal zinc-finger domain (figure 2). The C-terminal region of the protein was shown experimentally to interact with different factors, such as huntingtin,29 30 myosin VI,31 ubiquitin32 and RIP1 (receptor-interacting protein 1).33 In particular, the mutant K557 residue identified in our study resides within the 172-amino-acid-long C-terminal region responsible for the interaction with the adenoviral E3-14.7K protein, an inhibitor of tumour necrosis factor α-induced apoptosis.34 Interestingly, the E478 and A481 residues, found to be mutated in ALS patients,10 12 are located in the same region, probably disrupting the ubiquitin-binding domain) motif.
The c.552+1delG variant, shown to affect exon 6 splicing, would generate a 37-amino-acid in-frame deletion in OPTN protein (p.148_184del) that would nearly completely abolish the N-terminal leucine-zipper domain (figure 2). This deletion might compromise the interaction with the GTPase Rab8 and, consequently, prevent OPTN translocation from the Golgi to the nucleus upon apoptotic stimulus.35 Although we could not prove a splicing defect in the ALS patient carrying the intronic c.1401+4A→G change, the in silico analysis predicted the generation of an aberrant transcript.
In contrast to studies in the Japanese population,10 11 all OPTN mutations were found in the heterozygous state in our Italian patients, as also described in the Canadian and French ALS cohorts.12 13 Maruyama et al10 hypothesised two different disease mechanisms for OPTN mutations: a loss of function for the nonsense homozygous mutations, possibly leading to a reduced protein level by a nonsense-mediated decay mechanism, and a gain of function for the missense p.E478G heterozygous variant. We cannot exclude, in particular for the ALS case carrying the p.G23X mutation, the contemporary presence on the other allele of another variant affecting OPTN expression, such as the presence of large genomic deletions other than the tested exon 5 one or nucleotide changes in non-coding sequences. Nevertheless, we may also hypothesise that, in a different genetic background, OPTN nonsense mutations cause the disease in the heterozygous state too.
On the other hand, functional studies are needed to verify whether the three missense variants identified in this study, together with the other four novel variants recently described,11–13 may result in a gain of function of the OPTN protein as already demonstrated for the heterozygous p.E478G mutation in the Japanese ALS population.10
A genotype/phenotype correlation is hard to define in our cohort, since the identified OPTN mutations are private. We found that both age of onset and disease duration were highly variable. However, the clinical phenotype of OPTN-mutated patients was constantly associated with a lower-limb onset and characterised by a prevalence of UMN signs.
Our ALS patients neither suffered from glaucoma/visual problems nor reported any history of glaucoma in their families. It is very unlikely that our OPTN-mutated cases represent preclinical POAG patients and they are only affected coincidentally by ALS. In fact, given that the POAG prevalence in the adult Italian population is about 2%,28 that hereditary POAG accounts for 50% of cases9 and that mutations in the OPTN gene were found to be responsible for about 16% of familial POAG,26 we would expect to find about 0.16% of ALS patients with a POAG phenotype associated with OPTN. By contrast, in our cohort, we have identified 2% of ALS cases (six patients) with mutations in the OPTN gene. Of interest, the mutations we found in ALS patients were not identified in more than 850 Caucasian controls (including our 280 Italian controls) or in more than 3900 patients with POAG analysed so far.9 25 36 The common glaucoma-causing mutations described in different ethnic groups affect both the N-terminal region (p.H26D, 2 bp-insertion), including the bZIP domain (p.E50K), and the C-terminal coiled-coil region of OPTN protein (p.H486R), where the ALS-causing mutations are also located (figure 2). It is possible therefore that different mutations in OPTN gene in association with distinct susceptibility factors may cause such divergent disorders, such as glaucoma and motor neuron disease. Interestingly, no mutations have been identified recently in a large cohort of patients with FTD,37 and we have not found any variants in our ALS/FTD patients.
In conclusion, these data confirm that OPTN gene mutations are associated with ALS, extending the spectrum of OPTN mutations and supporting the possible pathological role of optineurin in motor neuron disease, but not in FTD. As optineurin was found to colocalise with SOD1- and TDP-43-positive inclusions in ALS-affected brain tissues10 and with FUS in basophilic inclusions,38 this protein certainly represents a novel and important neuropathological hallmark for ALS. However, further mutational and functional studies are needed to assess the role of the OPTN gene in ALS pathogenesis, given all optineurin interactors identified so far and its involvement in membrane trafficking and NF-kB signalling.39
We thank all patients and their families for participating in this study. EuroBioBank project QLTR-2001-02769 is gratefully acknowledged.
RDB and CT contributed equally to the work.
Funding VP, BC, GS, CC, SD, CG, GPC and VS are financially supported by Italian Ministry of Health; CT, AR, NT, CG and VS by AriSLA (Grants EXOMEFALS and RBPALS); LC received a fellowship from ‘Amico Canobio’ Association.
Competing interests None.
Ethics approval Ethics approval was provided by the Local ethics committees of each participating Institutions.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.