|Home | About | Journals | Submit | Contact Us | Français|
Missense mutations in TUBB3, the gene that encodes the neuronal-specific protein β-tubulin isotype 3, can cause isolated or syndromic congenital fibrosis of the extraocular muscles, a form of complex congenital strabismus characterized by cranial nerve misguidance. One of the eight TUBB3 mutations reported to cause congenital fibrosis of the extraocular muscles, c.1228G>A results in a TUBB3 E410K amino acid substitution that directly alters a kinesin motor protein binding site. We report the detailed phenotypes of eight unrelated individuals who harbour this de novo mutation, and thus define the ‘TUBB3 E410K syndrome’. Individuals harbouring this mutation were previously reported to have congenital fibrosis of the extraocular muscles, facial weakness, developmental delay and possible peripheral neuropathy. We now confirm by electrophysiology that a progressive sensorimotor polyneuropathy does indeed segregate with the mutation, and expand the TUBB3 E410K phenotype to include Kallmann syndrome (hypogonadotropic hypogonadism and anosmia), stereotyped midface hypoplasia, intellectual disabilities and, in some cases, vocal cord paralysis, tracheomalacia and cyclic vomiting. Neuroimaging reveals a thin corpus callosum and anterior commissure, and hypoplastic to absent olfactory sulci, olfactory bulbs and oculomotor and facial nerves, which support underlying abnormalities in axon guidance and maintenance. Thus, the E410K substitution defines a new genetic aetiology for Moebius syndrome, Kallmann syndrome and cyclic vomiting. Moreover, the c.1228G>A mutation was absent in DNA from ~600 individuals who had either Kallmann syndrome or isolated or syndromic ocular and/or facial dysmotility disorders, but who did not have the combined features of the TUBB3 E410K syndrome, highlighting the specificity of this phenotype–genotype correlation. The definition of the TUBB3 E410K syndrome will allow clinicians to identify affected individuals and predict the mutation based on clinical features alone.
Microtubules serve essential cellular functions critical for the development and maintenance of the mammalian nervous system. They provide structure and generate forces needed for neurons to migrate and to develop axonal and dendritic processes, and they provide organized scaffolds on which kinesin and dynein molecular motor proteins transport cargo. In recent years, it has been found that genetic mutations in their building blocks, the α- and β-tubulin isotypes, alter microtubule function in humans and cause neurological disorders (Keays et al., 2007; Abdollahi et al., 2009; Jaglin et al., 2009; Kumar et al., 2010; Poirier et al., 2010; Tischfield et al., 2010).
We previously reported that eight hot-spot missense mutations in TUBB3, which alter six amino acid residues in the neuronal β-tubulin isotype 3 (TUBB3), cause congenital fibrosis of the extraocular muscles (CFEOM), a complex congenital ocular dysmotility disorder characterized by ptosis and strabismus, with or without additional neurological findings (Tischfield et al., 2010). We provided a brief overview of the clinical manifestations among 29 mutation-positive families and noted apparent correlations between specific missense mutations and the resulting phenotypes. We also argued that the phenotypes were at least in part caused by a dominant effect of mutant TUBB3 protein on axon guidance and maintenance and showed that the TUBB3 R262C substitution, which causes isolated CFEOM in humans, causes errors in axon guidance in a mouse model of the disease. Furthermore, we showed that all eight mutations altered microtubule dynamics in vitro, while a subset also perturbed binding of kinesin to the microtubule.
Within our initial description of the CFEOM-associated TUBB3 syndromes, we briefly summarized the phenotype of six individuals harbouring the de novo c.1228G>A TUBB3 mutation, which directly alters a kinesin-binding site residue located within the H12 helix of TUBB3 and results in a TUBB3 E410K amino acid substitution. We reported that these individuals had severe CFEOM, facial weakness and developmental delay with variable thinning of the corpus callosum and anterior commissure, while the oldest may have been developing a sensorimotor polyneuropathy. We have now performed detailed phenotyping of these six patients and two additional previously unreported individuals, each harbouring a de novo TUBB3 c.1228G>A missense mutation. We report that, in addition to the findings aforementioned, this mutation causes Kallmann syndrome, cyclic vomiting, vocal cord paralysis and dysmorphic facial features, including midface hypoplasia. This constellation of findings defines a distinctive ‘TUBB3 E410K syndrome’ that can be recognized at the bedside, leading to the rapid genetic diagnosis of affected individuals.
Approximately 400 mutation-negative probands with isolated or syndromic congenital complex ocular dysmotility and/or congenital facial weakness and ~200 probands with isolated or syndromic Kallmann syndrome, were screened for the TUBB3 c.1228G>A missense mutation (p.E410K). Eight probands were identified and are included in this report. After informed consent, participants enrolled in an ongoing research study at Boston Children’s Hospital, and a subset also enrolled in ongoing research studies at Massachusetts General Hospital and Mount Sinai School of Medicine, with approval by the corresponding institutional review boards. Mutation analysis of six subjects was previously reported (Tischfield et al., 2010). Two previously unreported subjects provided a blood sample from which genomic DNA was extracted using Gentra® Puregene® Blood Kits (Qiagen), and the TUBB3 gene was sequenced as previously described (Tischfield et al., 2010). Maternity and paternity were confirmed by the appropriate inheritance of at least 10 informative polymorphic markers.
Detailed medical and family histories and clinical and neuroimaging data were obtained from participants, medical records and dedicated research visits. Olfaction was evaluated by administration of either the Smell Identification Test (Philadelphia, PA) or the Alcohol Sniff Test (San Diego, CA) when feasible, or documented by history. Facial 3D images for morphometric analysis were attained using a 3DMD system (http://3dmd.com), and measurements were made using a freehand point selection option and repeated to ensure reliability. Coding exons and intron–exon boundaries of previously known Kallmann syndrome genes CHD7, KAL1, PROK2, PROKR2, NELF, FGF8 and FGFR1 were amplified from genomic DNA from five of the eight subjects using published primer sequences and PCR conditions (Jongmans et al., 2009; Sykiotis et al., 2010), and PCR products were sequenced and analysed.
Four male and four female individuals of diverse ethnic backgrounds ranging from 7 to 25 years of age were found to harbour the TUBB3 c.1228G>A mutation (p.E410K). In each, the heterozygous mutation arose de novo. Two pedigrees (ZY and AER) are previously unpublished (Fig. 1), whereas six were presented previously (Tischfield et al., 2010). At the time of enrolment, seven had been diagnosed with Moebius syndrome and one with bilateral oculomotor and facial nerve palsies. Subjects are indicated by roman numerals in the order of chronological age from youngest to oldest, and each subject’s clinical findings are highlighted later in the text and summarized in Table 1.
All eight subjects had ptosis, exotropia and facial weakness at birth, and diagnosed within the first 5 months of life. Three male subjects were diagnosed with cryptorchidism and/or microphallus during infancy. Four subjects had stridor as infants, of whom three underwent direct lanyngoscopy for tracheomalacia or laryngomalacia, two were found to have bilateral vocal cord paralysis and one required tracheostomy for the first 7 years of life.
Seven subjects had short stature (3rd–25th percentile) until adolescence, after which the older male subjects failed to undergo the pubertal growth spurt and dropped below the third percentile. Three subjects had microcephaly. Growth charts are displayed in Fig. 2. All subjects have required special education for delayed intellectual and/or social development, and those tested have full-scale IQ scores between 50 and 90. Subject II developed complex partial seizures at age 3 years, and EEG revealed left centrotemporal and right frontal sleep-activated spikes. Five subjects had fine and gross motor delays. Six had hypotonia and one had hypertonia early in life. Five developed lower-extremity clonus.
Five subjects underwent 3D photography and were found to have midface flattening, a non-prominent and short nose, a short smooth philtrum and low set and/or posteriorly rotated ears (Table 2, representative images shown in Fig. 3). Subjects IV and VII had lumbar and thoracic scoliosis. None had cleft palate, renal agenesis, vertebrae or limb defects or mirror movements. Baseline cardiac, respiratory and abdominal examination results were unremarkable.
Subjects demonstrated clinical abnormalities of the olfactory, oculomotor, facial and vagal nerves as summarized in Table 1.
The seven subjects tested had anosmia, and neuroimaging of all eight subjects revealed olfactory sulcus and olfactory bulb dysgenesis (Fig. 4).
All subjects had a severe form of CFEOM, with congenital bilateral ptosis, globe infraduction and exotropia ranging from 20 to 100 prism dioptres and minimal to no eye movements (Fig. 3). Residual eye movements in four subjects were aberrant (Table 3). Oculomotor nerves were hypoplastic or not as visualized by neuroimaging (Fig. 5), similar to individuals harbouring TUBB3 R262C or D417N substitutions who have oculomotor and optic nerve hypoplasia, and variable extraocular muscle atrophy (Demer et al., 2010). The severe exotropia and hypotropia of the subjects’ eyes limited the ability to assess trochlear and abducens function, and imaging data did not permit definitive evaluation of the presence or absence of these nerves.
All eight subjects had bilateral facial paralysis including the forehead, and facial nerves were not visible by neuroimaging (Fig. 5). Salivation was intact in all subjects, and gustatory sensation was normal in the two subjects tested.
Two subjects were diagnosed with vocal cord paralysis as infants.
No subject was noted to have optic nerve hypoplasia or retinal anomalies; none, however, had visual acuity of 20/20 or better, and all but one tested subject had visual acuity of 20/40 or worse in the better-seeing eye (Table 3). Slit-lamp examination results were normal, with the exception of corneal exposure in five subjects. All had normal facial sensation, hearing, chewing and swallowing, as well as normal midline palate elevation, strong sternocleidomastoid muscles and symmetric tongue movement with no signs of tongue atrophy or fasciculation.
The two oldest subjects reported sensory and/or motor changes in their distal lower extremities. Nerve conduction studies and/or electromyography studies were performed on five subjects between the ages of 9 and 21 years. Three had mild sensory and motor nerve conduction study abnormalities, a fourth had primarily sensory abnormalities and the youngest had essentially normal nerve conduction studies (Table 4). The abnormal conduction velocities were generally ~80% of the lower limit of normal, which is the threshold distinguishing demyelinating from axonal polyneuropathies.
All four male subjects and the two oldest female subjects were diagnosed with Kallmann syndrome, a distinctive phenotypic form of idiopathic hypogonadotropic hypogonadism that is defined by the combined occurrence of gonadotropin-releasing hormone deficiency and defective olfactory function (Bianco and Kaiser, 2009; Crowley, 2011). Two female subjects were too young to be evaluated for hypogonadotropic hypogonadism, but both had anosmia. The subjects diagnosed with Kallmann syndrome had absent or delayed puberty, low gonadotropin and sex steroid levels and otherwise normal pituitary function (Table 5). Three male subjects also had micropenis and/or cryptorchidism, consistent with neonatal gonadotropin-releasing hormone deficiency. Hypogonadotropic hypogonadism is genetically heterogeneous and can be oligogenic (Balasubramanian et al., 2010; Tornberg et al., 2011). To exclude oligogenic inheritance, we sequenced the known Kallmann syndrome genes in the five subjects for whom we had sufficient DNA. Polymorphisms, but no other putative disease-causing variants, were identified (Table 6).
The two oldest male subjects (VII and VIII), were diagnosed with cyclic vomiting syndrome based on recurrent attacks of severe nausea and vomiting beginning at ages 4 and 20 years, respectively, that had no clear causal aetiology and were separated by periods of baseline health (Li et al., 2008). Each had multi-day episodes of severe vomiting triggered by foods, infections or strong emotions. Episodes would be heralded by a prodrome of several hours, after which vomiting would occur every few minutes, accompanied by nausea, photophobia, phonophobia and incontinence, but no headache. Subject VII also had signs suggestive of dysautonomia, including sinus arrhythmia during vomiting episodes, difficulty with body temperature regulation and excessive sweating. EEG monitoring, gastrointestinal anatomical and gastric emptying study results were normal during and between events.
For both subjects, prophylactic valproic acid at a therapeutic serum level of 50–100 mcg/ml reduced the frequency and severity of episodes. Episodes could often be aborted by aprepitant during the prodromal phase. A combination of intranasal sumatriptan, promethazine suppository and rectal diazepam controlled symptoms if administered immediately after onset of vomiting; otherwise, intravenous chlorpromazine, fluids and sedation were required.
The two other male subjects had features of cyclic vomiting syndrome. Subject VI had vomiting episodes several times a year triggered by illness and lasting for several days. Subject II had sporadic vomiting episodes triggered by anxiety and characterized by early-morning awakening with abdominal discomfort, conjunctival injection and nausea, followed by a brief period of unresponsiveness and limpness with subsequent vomiting.
We have determined that the TUBB3 c.1228G>A missense mutation, which typically arises de novo, causes a constellation of congenital and degenerative phenotypes that we have named the ‘TUBB3 E410K syndrome’. The TUBB3 E410K syndrome consists of intellectual disability, Kallmann syndrome, CFEOM, facial weakness, tracheomalacia, vocal cord paralysis and later-onset cyclic vomiting and progressive peripheral neuropathy. Neuroimaging reveals thin corpus callosa and anterior commissures, as well as hypoplastic to absent olfactory sulci, olfactory bulbs and oculomotor and facial nerves, which are consistent with abnormalities in axon guidance and maintenance. These findings expand significantly our initial report of the clinical manifestations resulting from this mutation and permit its recognition as a human malformation syndrome distinct from the phenotypes resulting from other reported TUBB3 amino acid substitutions, including D417H, D417N and R262C that, like E410K, also alter kinesin–microtubule interactions (Tischfield et al., 2010). These findings also highlight the specificity of the TUBB3 E410K genotype–phenotype correlation; none of the ~600 TUBB3 mutation-negative individuals who had either Kallmann syndrome or ocular and/or facial dysmotility disorders had the combined features of the TUBB3 E410K syndrome.
Moebius syndrome (Moebius, 1888) is defined by congenital limitations of eye abduction and facial movement. Affected individuals classically have eyes that are straight or esotropic in the primary position, full vertical eye movements and limitations in abduction. Seven of the subjects in our study cohort carried the diagnosis of Moebius syndrome at the time of enrollment, but had atypical features because their eyes were exotropic and infraducted bilaterally with minimal to absent vertical eye movements and severe limitations in adduction that exceeded limitations in abduction.
A series of case reports have described genetically undefined individuals diagnosed with Moebius syndrome who, like our cohort of eight subjects, have ptosis, horizontal and vertical gaze limitations and facial weakness accompanied by hypogonadotropic hypogonadism and peripheral neuropathy (Olson et al., 1970; Rubinstein et al., 1975; Abid et al., 1978; Kawai et al., 1990; Brackett et al., 1991; Jennings et al., 2003), intellectual disability (Olson et al., 1970; Abid et al., 1978; Brackett et al., 1991), thin corpus callosum (Brackett et al., 1991) and/or facial dysmorphism with small nose (Rubinstein et al., 1975; Brackett et al., 1991). Thus, these individuals likely had the TUBB3 E410K syndrome. We anticipate that our definition of this syndrome will lead to its increased recognition and subsequent testing for the TUBB3 c.1228G>A mutation among Moebius patient populations worldwide.
Clinically, the TUBB3 E410K substitution affects specific cranial nerves and, in some instances, selective nerve branches. Hypoplastic oculomotor and facial nerves are consistent with congenital cranial nerve maldevelopment, and clinical signs of aberrant extraocular muscle innervation support axon misguidance as the pathological basis for at least some of the TUBB3 E410K phenotypes. We note that some individuals also had vocal cord paralysis, consistent with dysfunction of the recurrent laryngeal branch of the vagal nerve. Preserved control of pupillary constriction, salivation and gustatory sensation suggests that the somatic/branchial motor efferent cranial nerve fibres are more vulnerable than visceral efferent or special sensory afferent fibres. Although facial sensation, taste and hearing appear to be intact, the TUBB3 E410K substitution does not spare all sensory nerves, as dysfunction of the olfactory special sensory afferents results in anosmia.
We have identified that the TUBB3 E410K substitution is a novel genetic aetiology of Kallmann syndrome (Kallmann et al., 1944), as the subjects had idiopathic hypogonadotropic hypogonadism and clinical and/or radiological evidence of abnormal olfaction. Hypogonadotropic hypogonadism results from a failure of pulsatile secretion of gonadotropin-releasing hormone required for humans to initiate puberty (Balasubramanian et al., 2010). Male subjects with the E410K substitution had microphallus/cryptorchidism as neonates, while subjects of both sexes had a gonadotropin-releasing hormone deficiency phenotype at the time of puberty. Subject VII displayed a form of partial pubertal development known as the ‘fertile eunuch’ variant of idiopathic hypogonadotropic hypogonadism, characterized by spontaneous testicular maturation in the face of overt hypogonadism and implying a degree of testicular development (Pitteloud et al., 2001, 2002). Thus, the TUBB3 E410K subjects display some phenotypic heterogeneity at the time of puberty, as is typical of most genetic causes of idiopathic hypogonadotropic hypogonadism (Balasubramanian et al., 2010). Although synergistic interactions between multiple mutated disease genes are thought to contribute to this phenotypic heterogeneity, we did not find mutations in other genes, and the TUBB3 E410K substitution alone appears sufficient.
The TUBB3 E410K substitution likely causes Kallmann syndrome by disrupting olfactory axon guidance and gonadotropin-releasing hormone cell migration from the nasal placode to the olfactory bulb. This would be consistent with the proposed aetiology for other genetic forms of Kallmann syndrome (Balasubramanian and Crowley, 2011) and with the deleterious effect of TUBB3 E410K mutant protein on the development of other cranial nerves in humans, and on the guidance of oculomotor axons that we reported in TUBB3 R262C homozygous mice (Tischfield et al., 2010).
The TUBB3 E410K substitution implicates TUBB3 as the first cytoskeletal protein mutated in Kallmann syndrome and suggests a common final pathway for Kallmann syndrome pathogenesis that converges on the microtubule. Notably, one of six TUBB3 mutations reported to cause polymicrogyria in the absence of CFEOM was harboured by a foetus, and pathological findings included olfactory bulb agenesis (Poirier et al., 2010). Moreover, some TUBA1A mutations have been reported to cause olfactory sulcus abnormalities in the setting of lissencephaly (Fallet-Bianco et al., 2008). Thus, mutations in other tubulin isotypes may also cause Kallmann syndrome.
Cyclic vomiting syndrome is a rare disorder, with a childhood prevalence of 0.4–1.9% (Li and Balint, 2000; Ertekin et al., 2006). Its aetiology and pathophysiology are poorly understood. Mitochondrial DNA mutations altering cellular energy production have been proposed to cause both cyclic vomiting syndrome and migraine headaches (Boles et al., 2003; Wang et al., 2004), and one family with cyclic vomiting was reported to harbour a maternally inherited mutation in the mitochondrial transfer RNA gene MT-TL1 (Salpietro et al., 2003). Other proposed mechanisms include autonomic dysfunction (Chelimsky and Chelimsky, 2007) and dysregulation of the hypothalamic corticotropin-releasing hormone response to stress (Tache, 1999). Although cyclic vomiting can occur in isolation or with additional neurological findings such as cognitive disorders, cranial nerve dysfunction and seizures (Boles et al., 2006), to our knowledge, it has not been previously identified as a feature of a defined clinical syndrome.
The TUBB3 E410K substitution is the first genetic and non-mitochondrial aetiology for cyclic vomiting identified in multiple unrelated individuals. The TUBB3 E410K substitution disrupts microtubule function in both central and peripheral neurons, and thus it is unclear whether it causes cyclic vomiting by perturbing autonomic nerve function, affecting CNS circuitry or a combination of both. However, cyclic vomiting appears to be a highly specific manifestation of neuronal dysfunction associated with the TUBB3 E410K substitution. Subjects with other TUBB3 mutations who have similar or more severe abnormalities do not report cyclic vomiting, and we are not aware of cyclic vomiting syndrome associated with mutations in any other neurodevelopmental disease genes.
We previously suggested that the TUBB3 E410K substitution might cause peripheral neuropathy based on the oldest subject’s clinical symptoms (Tischfield et al., 2010). We have now established that the TUBB3 E410K syndrome includes progressive polyneuropathy with onset of nerve conduction abnormalities as early as the first decade of life, and clinical signs and symptoms in the third decade. The subjects’ nerve conduction and electromyography studies demonstrated that the TUBB3 E410K polyneuropathy affects sensory fibres. Although motor fibres are likely also involved based on clinical presentation, more definitive evidence from needle electromyography is needed.
The E410K substitution now becomes the fourth TUBB3 substitution, along with R262H, D417H and D417N, recognized to cause a polyneuropathy, with age of onset and severity varying by amino acid substitution (Tischfield et al., 2010). While TUBB3 is highly expressed in neurons in both the developing central and peripheral nervous systems (Katsetos et al., 2003), high levels of expression are maintained only in peripheral neurons (Jiang and Oblinger, 1992). This may account, in part, for this degenerative phenotype. Moreover, these four TUBB3 substitutions share specific characteristics: each permits mutant TUBB3 protein incorporation into microtubules at wild-type or near wild-type levels in HeLa cells, directly or indirectly disrupts tubulin–motor protein binding sites in the tubulin H12 helix and perturbs microtubule–kinesin interactions in yeast (Tischfield et al., 2010). Thus, the neuropathy likely arises from a dominant mechanism that disrupts specific aspects of motor protein–based axonal transport.
We can only speculate on the aetiology of TUBB3 E410K facial dysmorphisms. Although neural crest gives rise to facial cartilage and bone, TUBB3 is reportedly expressed only transiently in neural crest cells destined to become neurons (Haendel et al., 1996), so this seems an unlikely aetiology. Although congenital muscle paralysis may alter skeletogenesis (Sharir et al., 2011), the facial dysmorphology of subjects with the TUBB3 E410K substitution is dissimilar to that of subjects with facial weakness and absent facial nerve secondary to homozygous HOXB1 mutations (Webb et al., 2012). Given that affected individuals also have Kallmann syndrome, it is intriguing to consider that TUBB3 E410K could disturb nasal patterning by altering development of neural elements within the nasal placode, which normally secretes cues that influence the development of mesenchyme that gives rise to the nasal conchae and nasal bone (Szabo-Rogers et al., 2009).
Our previous report demonstrated that CFEOM-causative TUBB3 mutations alter axon guidance in the central and peripheral nervous system. Many of the findings we report here (selective cranial nerve abnormalities, Kallmann syndrome, peripheral neuropathy and intellectual disability associated with corpus callosum and anterior commissure defects in the absence of known cortical migration deficits) are consistent with abnormalities in both central and peripheral axon guidance and peripheral axon maintenance. It is more difficult to explain how the additional findings of stereotyped facial dysmorphisms and cyclic vomiting are the consequence of disrupted axon guidance or maintenance. The co-segregation of these two features with the other findings suggests that either undefined defects in axon guidance or maintenance contribute to their pathogenesis, or that they result from additional dysfunction of E410K mutant TUBB3 that is yet to be characterized.
The individuals that harbour the TUBB3 E410K substitution demonstrate a set of highly selective phenotypes that result from a single amino acid substitution in a protein expressed in all neurons, and thus highlight the vulnerability of specific populations of cells to alterations in TUBB3 function. The characteristic findings are also so distinctive that we believe clinicians will be able to identify affected individuals and predict the TUBB3 E410K substitution based on clinical features alone. This, in turn, will assist in distinguishing this population from those with typical Moebius syndrome and allow clinicians to provide a more accurate diagnostic and prognostic picture to patients and families. Further characterization of the TUBB3-related syndromes and studies of their associated mutations should continue to improve clinical diagnosis and provide additional insight into the role that TUBB3 plays in the development and maintenance of neurons and the mechanism by which changes in its function may cause disease.
National Eye Institute of the National Institutes of Health [R01EY12498 to E.C.E.]; the Boston Children’s Hospital Intellectual and Developmental Disabilities Research Centre [HD018655 to S.L.P., E.C.E.]; the National Institute of Child Health & Human Development of the National Institutes of Health [U54 HD28138 and T32 HD07396 to W.F.C.]; Children’s Hospital Ophthalmology Foundation [to E.C.E., D.G.H.]; the Moebius Syndrome Foundation [to E.C.E.]. S.C is funded by the HHMI Medical Fellowship and Harvard Medical School. E.C.E. is a Howard Hughes Medical Institute Investigator.
The authors would like to thank the study participants and their families for their generosity and time, and Drs. Max Tischfield, Joseph Mazjoub, Anna Minster, and Athos Bousvaros for helpful discussions.