Abstract

Main Text

Heterotrimeric G proteins trigger a signal transduction cascade composed of α, β, and γ subunits. They are associated with G protein-coupled receptors (GPCRs) in modulating an array of cellular functions, including release of a multitude of hormones and growth factors, regulation of cell contraction and migration, and cell growth and differentiation during development.^{1, 2, 3, 4} G protein-coupled signaling plays a crucial role in neuronal communication, including regulation of the antagonistic effects of the parasympathetic and sympathetic branches of the autonomic nervous system throughout the body. We report a genetic disorder caused by mutations affecting GNB5 (MIM: 604447), encoding guanine nucleotide-binding protein subunit beta-5, and with disease manifestation in multiple systems.

We identified nine affected individuals (six females and three males) from six unrelated families presenting with a clinical overlap of neurological and cardiac conduction defects; all subjects were found to have variation in the same gene, GNB5, and share a similar rare phenotype. This work results from exome and phenotype data aggregation among independent groups engaged in studying the molecular basis of yet unsolved human genetic rare disease traits. Shared phenotypic features representing the cardinal characteristics of the syndrome include global developmental delay, seizures, generalized hypotonia, retinal disease, and the uncommon feature of early-onset sinus-node dysfunction (Table 1). Additional clinical investigations and diagnostic studies did not show any evidence of structural CNS, ocular, or cardiac anomalies. Affected individuals from four of the six families (families A–D) demonstrated the severe end of the disease spectrum, including substantial cognitive deficits, delayed motor development, severe hypotonia, retinal disease, pathological gastro-esophageal reflux, and sinus-node dysfunction. Affected individuals in families E and F presented with a milder phenotype, including mild intellectual impairment, language delay, and bradycardia (Figure 1, Table 1, Supplemental Note).

Open in a separate window

Figure 1

Pedigrees from the Six Families Investigated in this Study

Affected members of families A to D (upper red-lined panel) and E to F (lower blue-lined panel) show severe and mild manifestation of the core symptoms of the syndrome defined in this study. Filled symbols represent individuals with severe sinus sick syndrome (SSS; top left quarter), intellectual disability (ID; top right quarter), hypotonia (bottom left quarter), and seizures (bottom right quarter), whereas a light-gray top left quarter indicates the presence of mild ID. Genotypes are specified according to GenBank: NM_006578.3.

Given that no potentially pathogenic genomic structural abnormalities were identified by array comparative genomic hybridization and karyotyping of the affected subjects, we applied whole-exome sequencing to all the affected individuals and their healthy parents. Families were recruited in Italy (family A), Brazil (B and F), the United States (C and D), and the Netherlands (E). The institutional review boards of the IRCCS Casa Sollievo Della Sofferenza Hospital, the Hospital das Clínicas da Universidade de São Paulo, the Baylor College of Medicine, the Amsterdam Academic Medical Center, and the University of Lausanne approved this study. Participants were enrolled after written informed consent was obtained from parents or legal guardians. The clinical evaluation included medical history interviews, a physical examination, and review of medical records. To uncover genetic variants associated with the complex phenotype shown by the nine affected subjects, we sequenced their exomes and those of their parents. DNA libraries were prepared from blood-derived genomic DNAs according to standard procedures. Exomes were captured and sequenced with different platforms to reach 50- to 120-fold coverage on average. Variants were called as previously described.^{5, 6, 7} Variants were filtered on the basis of inheritance patterns, including autosomal recessive, X-linked, and de novo and/or autosomal dominant. Variants with MAF < 0.05% in control cohorts (dbSNP, the 1000 Genome Project, NHLBI GO Exome Sequencing Project, the Exome Aggregation Consortium database, and our in-house databases) and predicted to be deleterious by SIFT,⁸ PolyPhen-2,⁹ and/or UMD-Predictor¹⁰ were prioritized.

Given a potential history of consanguinity reported in some families (families B, C, and F [Figure 1, Table 1]), we filtered variants by using Mendelian expectations for the assumption of a rare autosomal-recessive trait. We found GNB5 to be compliant with Mendelian expectations and bearing bi-allelic putative deleterious variants in all affected individuals (Figure 1, Table S1). Sanger sequencing in each family confirmed the anticipated segregation of the GNB5 variants. Strikingly, the variants found in the severely affected individuals (families A–D) were predicted to be loss-of-function (LoF) alleles, whereas the more mildly affected individuals from families E and F were homozygous for the same missense variant, c.242C>T (p.Ser81Leu [GenBank: NM_006578.3]) (Figures 1 and S1A). In families B, C, and D the affected individuals were homozygous for splice variants (c.249+1G>T [p.Asp84Leufs31] and c.249G+3A>G [p.Asp84Valfs31]) and a nonsense variant (c.906C>G [p.Tyr302]), respectively (Figures 1 and S1A, Table S1). In family A, the affected siblings were compound heterozygous for a maternally inherited nonsense variant (c.994C>T [p.Arg332]) and a paternally inherited splice-site change (c.249G>A [p.(=)]), which gives rise to an aberrantly spliced isoform containing an additional 25 nucleotides of the intervening intron 2 (Figure S2A). We experimentally show that the transcripts from both alleles are targeted by nonsense-mediated mRNA-decay (Figure S2B).

The five GNB5 LoF variants identified in families A–D are either not present or present with MAF ≤ 8.25 × 10⁻⁶ in ExAC (Exome Aggregation Consortium, v.0.3.1) (Table S1). Correspondingly, LoF variants in GNB5 are underrepresented in comparison to expectation in this database; specifically, ExAC reports 8 LoF variants whereas 19 were expected. The c.242C>T (p.Ser81Leu) missense variant identified in family E, of Moroccan ancestry, and family F, of Brazilian ancestry, has a MAF < 5 × 10⁻⁵ (6/121,000) in the human population and 4.3 × 10⁻⁴ in Latinos (5/11,574). A sample of individuals from Morocco identified a prevalence of 1 out of 1,260 (7.94 × 10⁻⁴) for this allele. We estimated the prevalence of the c.242C>T (p.Ser81Leu) variant in the Moroccan population by genotyping a total of 630 Moroccan individuals, including 394 Moroccans and 235 Dutch citizens of Moroccan descent by real-time PCR. Pathogenicity of this variant is further supported by three-dimensional representation of the encoded protein complexed with RGS9, a member of the R7 subfamily of regulators of G-protein signaling (RGS) proteins and common binding partner of GNB5. GNB5 is folded into essentially identical seven-bladed β-propellers (WD40 repeated domains) with equivalent N-terminal helical extensions.¹¹ Replacement of the evolutionarily conserved serine 81 (Figure S1B) by leucine will induce localized structural changes in the immediate vicinity of this residue, which could impair both the central pore of the β-propeller and the binding kinetics of RGS proteins (Figures S3– S5).

In line with the clinical presentation of affected individuals, Gnb5 ablation in mice resulted in marked neurobehavioral abnormalities, including learning deficiencies, hyperactivity, impaired gross motor coordination, abnormal gait,¹² defective visual adaptation,¹³ and perturbed development and functioning of retinal bipolar cells.¹⁴ Correspondingly, mice lacking Rgs6, the GNB5-dependent RGS protein enriched in heart tissue, exhibit bradycardia and hypersensitivity to parasympathomimetics.^{15, 16} To independently investigate the functional effects of variation of GNB5 in the full phenotypic spectrum of subjects reported herein, we engineered a zebrafish model knocked out for gnb5.

CRISPR/Cas9 genome editing was used to generate zebrafish with LoF mutations in gnb5a and gnb5b. This teleost has two GNB5 paralogs as a result of an ancient genome duplication event¹⁷ (Figure S6). We identified stable lines with a 7 bp insertion in gnb5a and a 8 bp deletion and 15 bp insertion in gnb5b, causing a frameshift and premature truncation of the encoded proteins, respectively (Figure S7). It was anticipated that gnb5a and gnb5b might have redundant functions, which was confirmed by the absence of overt phenotypes in embryos homozygous for either LoF mutations. As a consequence, a double knockout was generated to ensure complete loss of functional Gnb5. In-crosses of gnb5a and gnb5b double heterozygotes resulted in clutches of embryos containing the expected 6.25% of gnb5a^−/−/gnb5b^−/− double mutants (henceforth referred to as gnb5 mutants). Consistent with syndrome manifestations of affected individuals, zebrafish mutant embryos had no striking dysmorphologic features (Figure S7D). However, the larvae showed impaired swimming activity, remained small, and generally died 7–14 days post fertilization (dpf), most likely as a result of their inability to feed.

To assess the putative involvement of GNB5 in autonomic nervous system functions, we investigated the GNB5-RGS-GIRK channel pathway. As GNB5 recruits RGS proteins to G protein-coupled inward rectifier potassium (GIRK) channels involved in the hyperpolarization of cell membranes,^{16, 18} we investigated whether LoF of GNB5 could delay GIRK channel deactivation kinetics, increase hyperpolarization time of cell membranes, and impair cell responsiveness to new stimuli. Carbachol (PubChem CID: 5831) is a parasympathomimetic compound that activates acetylcholine receptors of the heart and the GNB5-RGS-GIRK channel pathway. Treatment of gnb5 mutant larvae with carbachol resulted in a strong decrease of the heart rate, whereas it had little effect on wild-type and sibling larvae (Figure 2), consistent with loss of negative regulation of the cardiac GIRK channel by GNB5-RGS. In contrast, treatment with the sympathetic agonist isoproterenol resulted in an increased heart rate that was similar in wild-type, sibling, and gnb5 mutant larvae (Figure 2). These results indicate that GNB5 is crucial for parasympathetic control of heart rate, but not for sympathetic control, suggesting that lack of GNB5 is associated with extreme bradycardia at rest. Correspondingly, affected individuals present with severe bradycardia at rest (minimal observed heart rates of <25 bpm [beats per minute]) combined with a normal chronotropic response (maximum heart rates >150 bpm).

Open in a separate window

Figure 2

Cardiac Function in gnb5 Mutant Zebrafish

(A–D) Box-whisker plots demonstrate the heart rate response and the relative heart rate change of 5 dpf wild-type (WT), sibling (SIB), and gnb5 mutant (MT) larvae. Embryos at 5 dpf were embedded in 0.3% agarose prepared in E3 medium containing 16 mg/ml Tricaine. Basal heart rates were recorded first. Then, (A and C) 400 μM of the parasympathetic agonist carbachol (CCh; Sigma-Aldrich C4382) (WT n = 10, SIB n = 39, MT n = 14) or (B and D) 100 μM of the sympathetic agonist isoproterenol hydrochloride (ISO; Sigma-Aldrich 1351005) (WT n = 12, SIB n = 22, MT n = 9) was added and incubated for 30 min and heart rates were measured. Recordings were performed at 150 frames per second and were analyzed with ImageJ. The relative heart rate change is the percentage change between the basal heart rate measured and the heart rate after addition of CCh or ISO.

n denotes the number of fish used per dataset. Differences between two groups were analyzed via the Student’s t test. Differences between more than two groups were analyzed via one-way ANOVA with Tukey’s post-hoc test. Data are shown as mean ± SEM, and p < 0.05 was considered significant. p < 0.05, p ≤ 0.01, p ≤ 0.001, p ≤ 0.0001. p > 0.05 was considered not significant (n.s.). bpm, beats per minute.

The severe muscle hypotonia reported in affected individuals could result from GIRK-mediated hyperpolarization of neurons controlling skeletal muscle tone. gnb5 mutant embryos hatched normally from their chorion, a process that requires muscle contraction, but their swimming behavior appeared abnormal at 3 dpf. To investigate whether this abnormal behavior was linked to neurologic dysfunction and hypotonia, we examined the touch-evoked escape response. We anticipated that neurons would only become fully hyperpolarized after an initial stimulus and thus presented the embryos with three consecutive tactile stimuli. Whereas wild-type larvae rapidly swam away in response to repeated tactile stimuli, gnb5 mutants showed a significant decrease in swimming distance and swimming speed at stimuli two (p ≤ 0.0001) and three (p ≤ 0.01), but not after the first stimulus (Figures 3A–3C). Accordingly, gnb5 mutant larvae were predominantly unresponsive to repeated tactile stimuli (Movies S1 and S2). To test whether this abnormal escape response is the consequence of neurologic dysfunction rather than reduced muscle function, we performed a tail movement assay. 5 dpf larvae were given a strong tactile stimulus while we recorded the movement of the tail (Figures 3D and 3E). No significant difference in the maximum tail angle was detected between wild-type and gnb5 mutant larvae (Figure 3E). These results indicate that the tail muscles of gnb5 mutants are fully functional and that the abnormal escape response is associated with neurological dysfunction and possibly muscle hypotonia.

Open in a separate window

Figure 3

Neurologic Function in gnb5 Mutant Zebrafish

(A–C) Touch-evoked escape response assay in which three consecutive tactile stimuli were applied. Embryos at 3 dpf were placed in the middle of a standard 88 mm petri dish containing E3 medium. Three consecutive tactile stimuli were applied by touching the tail of the embryo with an insect pin. Stimuli were only applied when the embryo was still. Behavior was recorded with a standard camera (30 fps) and analyzed with ImageJ (NIH) and the plugin MTrackJ.¹⁹ (A) shows representative responses of 3 dpf wild-type and gnb5 mutant embryos to a touch stimulus. Scale bar, 0.5 cm. Box-whisker plots show quantification of the (B) swimming distance and (C) swimming speed in TL wild-types (n = 19), siblings (n = 46), and gnb5 mutants (n = 27).

(D and E) Analysis of maximum tail movement at 5 dpf. Larvae at 5 dpf were sedated in E3 containing 16 mg/ml Tricaine and embedded in 0.5% UltraPureTM agarose (Invitrogen 16500-500) in a 35 mm glass bottom dish. After setting, the agarose was cut away caudal to the swimming bladder, leaving the tail free to move. The dish was filled with E3 medium and embryos were left to recover from the sedation for 10 min at 28°C. Next, a maximal escape response was elicited by repeatedly touching the head of the embryo with an insect pin. Recordings were performed at 280 fps, for 30 s, with a high-speed CCD camera (Hamamatsu Photonics K.K., C9300-221) and analyzed with ImageJ (angle tool). (D) shows representative minimum projection images of tail movement in wild-type and gnb5 mutant embryos, including tail angle analysis. The tail angle represents the angle between the head-tail midline axis in resting state and a line that was drawn from just caudal of the swimbladder to the tip of the tail at maximal tail movement. (E) Tail angle quantification is displayed in box-whisker plots (wild-type n = 10, gnb5 mutants n = 10). fps, frames per second.

n denotes the number of fish used per dataset. Differences between two groups were analyzed via the Student’s t test. Differences between more than two groups were analyzed via one-way ANOVA with Tukey’s post-hoc test. Data are shown as mean ± SEM, and p < 0.05 was considered significant. p < 0.05, p ≤ 0.01, p ≤ 0.001, p ≤ 0.0001. p > 0.05 was considered not significant (n.s.).

Given that affected individuals have visual problems, including nystagmus, we investigated the visual system by measuring the optokinetic response (OKR) of gnb5 mutant larvae. When wild-type larvae were placed in a drum with a rotating light stimulus (Figure S8A), the OKR consisted of smooth pursuit eye movements followed by rapid rest saccades in the opposite direction (Figure S8B, Movie S3). In contrast, OKR was completely absent in gnb5 mutant larvae although their eyes showed no morphological abnormalities and could make eye movements (Figure S8C, Movie S4). This indicates that the eye muscles are functional in gnb5 mutants but that proper eye-movement control depends on GNB5. Overall these data show that gnb5 mutants faithfully recapitulate the phenotypic spectrum of affected individuals, including cardiac, neurologic, and ophthalmologic abnormalities.

These results provide evidence for a direct role of GNB5 in the control of heart rate, motor capacity, and vision. Whereas GNB1 (MIM: 139380), GNB2 (MIM: 139390), GNB3 (MIM: 139130), and GNB4 (MIM: 610863) are widely expressed and encode highly homologous proteins,²⁰ GNB5 is preferentially expressed in the brain and nervous system and encodes a peptide with less homology with its four paralogs.^{21, 22}

Germline de novo GNB1 variants cause severe neurodevelopmental disability,²³ hypotonia, and seizures. GNB3 bi-allelic LoF has been linked to congenital stationary night blindness (MIM: 610445, 163500, 610444, 613830, 616389, 310500, 257270, 613216, 614565, 615058, 300071, and 610427) and recessive retinopathy in humans,^{24, 25} retinal degeneration in chickens,²⁶ and reduced cone sensitivity and mild bradycardia in mice.^{27, 28} A SNP in GNB3 was associated with postural tachycardia syndrome²⁹ and incidence of cardiovascular disease and stroke.³⁰ Similarly, GNB2 and GNB4 map to loci governing heart rate on chromosomes 7 and 3, respectively.^{31, 32} We hereby demonstrate that bi-allelic LoF and missense variants in GNB5 cause a multisystem syndrome with features that include global developmental delay, sinus-node dysfunction, seizures, eye abnormalities, gastric problems, and generalized hypotonia. We highlight the importance of GNB5 for neuronal signaling, including the regulation of the antagonistic effects of the parasympathetic and sympathetic nervous system.

Conflicts of Interest

J.R.L. has stock ownership in 23andMe, is a paid consultant for Regeneron Pharmaceuticals, has stock options in Lasergen, is a member of the Scientific Advisory Board of Baylor Miraca Genetics Laboratories, and is a co-inventor on multiple United States and European patents related to molecular diagnostics for inherited neuropathies, eye diseases, and bacterial genomic fingerprinting. Baylor College of Medicine (BCM) and Miraca Holdings have formed a joint venture with shared ownership and governance of the Baylor Miraca Genetics Laboratories (BMGL), which performs clinical exome sequencing. The Department of Molecular and Human Genetics at BCM derives revenue from the chromosomal microarray analysis and clinical exome sequencing offered in the BMGL (http://www.bmgl.com/BMGL/Default.aspx/website). G.M. is a paid consultant for Takeda Pharmaceutical Company.

Notes

Published: August 11, 2016; corrected online: September 1, 2016

Footnotes

References