GNAL mutation in isolated laryngeal dystonia

Abstract

Background

Up to 12% of patients with laryngeal dystonia report familial history of dystonia, pointing to involvement of genetic factors; however, its genetic causes remain unknown.

Methods

Using Sanger sequencing, we screened 57 patients with isolated laryngeal dystonia for mutations in known dystonia genes TOR1A (DYT1), THAP1 (DYT6), TUBB4A (DYT4) and GNAL (DYT25). Using functional MRI, we explored the influence of the identified mutation on brain activation during symptomatic task production.

Results

We identified one patient with laryngeal dystonia who was a GNAL mutation career. Compared to 26 patients without known mutations, the GNAL carrier had increased activity in the fronto-parietal cortex and decreased activity in the cerebellum.

Conclusions

Our data show that GNAL mutation may represent one of the rare causative genetic factors of isolated laryngeal dystonia. Exploratory evidence of distinct neural abnormalities in the GNAL carrier may suggest the presence of divergent pathophysiological cascades underlying this disorder.

Introduction

Isolated laryngeal dystonia (LD), or spasmodic dysphonia, is a focal adult-onset dystonia primarily affecting speech production. LD is characterized by involuntary spasm-inducing voice breaks with strained and strangled voice quality in the adductor form (ADLD) and excessive breathiness in the abductor form (ABLD). Despite well-described clinical symptoms, the underlying causes of this disorder remain unknown, as challenges associated with traditional genetic studies have hindered the identification of LD-specific causative genes. On the other hand, familial history of dystonia in about 12% of LD patients ^{1, 2} points to the contribution of genetic risk factors. Laryngeal involvement has also been reported in the cohorts of patients with generalized or segmental dystonias who are carriers of DYT1, DYT4, DYT6 and, most recently, DYT25 mutations ^3-5. However, it is rare that any of these known gene mutations result in isolated LD.

In this study, we sought to investigate the contribution of DYT1, DYT4, DYT6 and DYT25 mutations as possible genetic causes of isolated sporadic and hereditary LD using Sanger sequencing of the corresponding coding regions. Because genes have direct influence on brain organization ⁶, we conducted an exploratory study to examine brain activation in an identified mutation carrier compared to healthy controls as well as to sporadic and familial LD cases using functional MRI during symptomatic speech and syllable production.

Materials and Methods

A total of 57 patients with isolated LD were recruited for genetic screening of known dystonia genes TOR1A (DYT1), THAP1 (DYT6), TUBB4A (DYT4) and GNAL (DYT25), which were previously found in dystonias with laryngeal involvement. Forty-one patients (23 ADLD/18 ABLD) were sporadic and 16 patients (12 ADLD/4 ABLD) had family history of LD and/or other primary dystonias (see demographic details in Table 1). All patients except 1 ABLD (African-American/Jewish) and 1 ADLD (African-American) were of a Caucasian descent. None had any present or past history of other neurological, psychiatric, or laryngeal problems. All patients had isolated LD, which was confirmed by fiberoptic nasolaryngoscopy, without involvement of any other body regions. The average age of onset was 40.9±12.3 (mean±standard deviation) in all sporadic LD patients and 40.9±13.1 in all familial LD patients (see further demographic details per clinical phenotype in Table 1).

Table 1.

Demographics of participants

	Type of LD	Gender	Age at exam (mean ± s.d.)	Age at onset (mean ± s.d.)	Ethnicity
	All (n = 41)	33 F/8 M	57.6±10.6	40.9±12.3	39C/1AA/1AA-J
Sporadic LD	ADLD (n = 23)	16 F/7 M	57.9±10.4/59.7±9.2	41.2±12.6/41.9±13.1	22C/1AA
	ABLD (n = 18)	17 F/1 M	58.8±9.4/70	41.5±12.4/58	17C/1AA-J
	All (n = 16)	13 F/3 M	61.2±11.1	40.9±13.1	16C
Familial LD	ADLD (n = 12)	10 F/2 M	56.9±10.7/58.6±10.4	40.6/13.3±16.8	12C
	ABLD (n = 4)	3F/1M	58.7±10.9/63	40.9±15.6/31	4C
Controls	All (n = 12)	18F/9M	53.9±9.4	N/A	12C
Handedness	All: Right on Edinburgh Inventory
Language	All: Monolingual English speakers
Cognitive status	All: Mini-Mental State Examination ≥ 27 points

The average age (± standard deviation) Healthy controls – no history of neurological, psychiatric or laryngeal problems. LD – laryngeal dystonia; ADLD – adductor laryngeal dystonia; ABLD – abductor laryngeal dystonia; C – Caucasian; AA – African-American: J – Jewish; F – female; M – male; N/A – not applicable; s.d. – standard deviation.

As an exploratory study to investigate possible neural correlates of gene mutations in LD, 27 ADLD patients from the same cohort, including 13 sporadic ADLD, 13 familial ADLD and 1 mutation carrier ADLD (18 females/9 males; 58.9±9.6 years old), as well as 27 age- and gender-matched healthy controls (18 females/9 males; 53.9±9.4 years old) underwent brain functional MRI during symptomatic task production. The patients who received regular botulinum toxin injections for symptom management participated in the MRI study only if they were symptomatic, at least 3-4 months after their last injection. All subjects gave written, informed consent, which was approved by the Institutional Review Board of the Icahn School of Medicine at Mount Sinai.

Genomic DNA was isolated from blood following the Purgene procedure (Gentra Systems). PCR amplification across the GAG deletion region of the TOR1A gene was performed as previously described ⁷. All exons and flanking regions of the THAP1, TUBB4A and GNAL genes were sequenced as previously described ^{5, 8, 9}. PCR products were enzymatically cleaned and sequenced by Sanger sequencing ⁸.

MRI data were acquired on a 3 Tesla Philips scanner with an 8-channel head coil. Whole-brain functional brain images were obtained using a gradient-weighted echo planar imaging (EPI) pulse sequence (TR = 2 s per volume and 10.6 s between volumes, TE = 30ms, FA = 90, FOV = 240 mm, voxel size = 3.75 × 3.75 mm, 36 slices with 4-mm slice thickness) and an event-related sparse-sampling experimental design during the production of LD-symptomatic sentences (e.g., “Are the olives large?”), syllables /i?i/ and resting baseline, as described earlier ¹⁰. A high-resolution T₁-weighted image was obtained for anatomical reference using magnetization prepared rapid gradient echo (MPRAGE) sequence (TR = 7.5 ms, TE = 3.4 ms, TI = 819 ms, FA = 8 degrees, FOV = 210 mm, 172 slices with 1-mm slice thickness). Following standard image pre-processing and smoothing with 4-mm full-width at half-maximum Gaussian kernel, multiple linear regression was used to analyze task-related responses with a single regressor for the task convolved with a canonical hemodynamic response function and six motion parameters, including three translations along the XYZ axes and three rotations (pitch, roll, and yaw) as covariates of no interest. Statistical comparisons were performed using two-sample t-tests between all 27 ADLD patients and 27 healthy controls to confirm the presence of brain abnormalities as reported earlier ^10-12, and between one ADLD mutation carrier and 26 ADLD patients without known mutations to explore the pattern of abnormalities in the mutation carrier. The statistical significance was set at a family-wise error (FWE)-corrected p ≤ 0.05.

Results

Among 57 LD patients, none were carriers of TOR1A (DYT1), THAP1 (DYT6) or TUBB4A (DYT4) mutations. However, one sporadic ADLD patient was a carrier of a novel coding variant in the GNAL (DYT25) gene. Clinically, this patient (Caucasian male, 37 y.o. at the time of initial evaluation with the onset of LD at 36 years) presented with isolated ADLD without past or present familial history of dystonia or any other movement disorders. At the four-year follow-up after the onset of disorder, the patient continued to exhibit isolated LD only, although a possibility of future spread of dystonia to other body regions cannot be ruled out. The mutation caused a G→A change in the coding region of the GNAL gene (at genomic position chr.18:11868562; hg19/GRCh37) resulting in an amino acid substitution, p.V234I (isoform NM_001142339) (Fig. 1-I). This variant is predicted by PolyPhen-2 ¹³ to be probably damaging and by SIFT ¹⁴ to be deleterious. In addition, cross-species alignment using MutationTaster ¹⁵ (accessed on 6/11/2015) showed that this variant gives rise to an amino acid substitution in a region of GNAL that is highly conserved throughout evolution. This mutation is not present in any of the variant databases including dbSNP 142 (http://www.ncbi.nlm.nih.gov/SNP/), the Exome Variant Server, NHLBI GO Exome Sequencing Project (ESP), Seattle, WA (URL: http://evs.gs.washington.edu/EVS/, accessed on 6/11/2015), or the Exome Aggregation Consortium (ExAC), Cambridge, MA (URL: http://exac.broadinstitute.org, accessed on 6/11/2015). This same variant was also present in his unaffected mother, confirming the reduced penetrance of GNAL mutations as shown previously ^{5, 16}.

Figure 1. (I) Mutation identified in the GNAL gene in ADLD patient.

Schematic of the exon-intron structure of the short isoform of GNAL with a mutation shown in red. Protein sequence alignment of G_αolf across species is obtained from RefSeq database and aligned using MutationTester. Altered residue is colored in red. (II) Differences in brain activation during symptomatic voice production in the GNAL mutation carrier compared to healthy controls and patients with sporadic and familial LD. (A) Statistically significant differences in brain activation during symptomatic sentence and syllable production between all 27 LD patients, including the GNAL mutation carrier, sporadic and familial cases without known mutations, and 27 healthy controls are presented on a series of axial brains slices in the standard Talairach space (B) with the BOLD percent signal change in each individual shown on the bar charts at an FWE-correction of p ≤ 0.05. (C) Significant differences between one GNAL mutation carrier and 26 sporadic and familial LD patients are shown on a series of axial slices in the standard Talairach space with (D) the corresponding bar charts, depicting the individual levels of BOLD percent signal change during the production of symptomatic sentences and syllables at an FWE-correction p ≤ 0.05.

At the neural level, comparisons between healthy controls and all ADLD patients, including both sporadic and familial patients, as well as the GNAL carrier, showed typically increased brain activation during symptom production in the primary sensorimotor cortex and superior parietal lobule as reported previously ^{10, 11} (Fig. 1-IIA,B). However, when comparing the GNAL mutation carrier with a group of sporadic and familial LD patients (both groups without known dystonia mutations), significant activation increases were identified in the supplementary motor area (SMA), middle frontal gyrus (MFG) and supramarginal gyrus (SMG) that were distinctive of the GNAL mutation carrier, whereas the cerebellum was over-activate in sporadic/familial ADLD patients (Fig. 1-IIC). Mean BOLD percentage signal change in the significant clusters showed that the GNAL carrier resided significantly outside the range of values of all sporadic and familial ADLD patients (Fig. 1-IID).

Discussion

Mutations in GNAL (DYT25), the first gene identified in adult-onset dystonia ⁵, have been reported in approximately 0.4-1.7% of both sporadic and familial patients with predominantly cervical or cranio-cervical segmental dystonias ^{5, 16-23}. Our finding of one GNAL carrier in 57 LD patients (1.8%) demonstrates that rare mutations in this gene do cause isolated LD, thus broadening the range of clinical phenotypes of dystonia associated with GNAL mutations. Our findings further indicate that gene mutations may underlie even sporadic presentations of dystonia due to reduced penetrance, thus stratification of patients into sporadic and familial cases remains somewhat arbitrary, pending the discovery of novel genetic factors contributing to this disorder. As none of patients from our cohort had coding mutations in TOR1A, THAP1, or TUBB4A genes, these mutations are perhaps either absent or as rare as 0.4% ²⁴ in patients with isolated LD.

GNAL encodes the stimulatory α subunit of the G protein, G_olf, and has been linked to the mediation of odorant signaling in the olfactory epithelium ²⁵, striatal dopaminergic signaling via coupling with D₁ receptors of the direct pathway, and adenosine A2A receptors of the indirect pathway ^26-31, as well as being co-localized with corticotropin-releasing hormone receptors in the cerebellar Purkinje cells ¹⁶. Relevant to dystonia pathophysiology, abnormal dopaminergic function and altered structural and functional organization of the cerebellum have been previously reported in isolated LD ^{10-12, 32-34}. As dopamine is one of the main modulators of brain function during cognitive and executive processes, the effects of striatal dopaminergic abnormalities may be reflected in aberrant fronto-parietal cortical activity, leading to additional alterations at the preparatory and sensorimotor integrative stages of motor sequence execution in the GNAL mutation carrier compared to other LD patients. On the other hand, a similar level of abnormalities in the primary sensorimotor cortex appeared to be a shared feature of brain changes across all LD patients compared to healthy controls, whereas greater cerebellar alterations in sporadic and familial cases without known genetic causes compared to the GNAL mutation carrier are suggestive of the distinct contribution of this structure to the pathophysiology of different forms of dystonia. Although this fMRI study compared a single GNAL mutation carrier to a larger group of isolated LD cases with no known mutations, our results should be interpreted with caution as they offer only initial clues about the potential links between a particular pattern of brain activity and genetic status in LD and suggest that future studies of GNAL mutation carriers in LD and other forms of isolated dystonia are warranted.