- Split View
-
Views
-
Cite
Cite
Monia Ginevrino, Enza Maria Valente, The multiple faces of TOR1A: different inheritance, different phenotype, Brain, Volume 140, Issue 11, November 2017, Pages 2764–2767, https://doi.org/10.1093/brain/awx260
- Share Icon Share
This scientific commentary refers to ‘TOR1A variants cause a severe arthrogryposis with developmental delay, strabismus and tremor’, by Kariminejad et al. (doi:10.1093/brain/awx230).
Flipping through the pages of any textbook of genetics, the reader would soon come across the seminal distinction between ‘simple’ Mendelian (or monogenic) disorders, caused by specific defects in a single gene, and ‘complex’ multifactorial disorders, which arise from the combination of many genetic and environmental factors. Monogenic disorders are further subdivided into dominant or recessive, depending on whether only one copy or both copies of the gene need to be mutated in order for the disease to manifest. Yet, this simple paradigm ‘gene - mutation - phenotype’ has been challenged in recent years, and it is now obvious that monogenic diseases are much more complex than previously thought. For instance, most dominant disorders are characterized by variable expressivity and incomplete penetrance: individuals who carry the same genetic mutation may either develop a full-blown clinical phenotype, present milder manifestations of the disease, or even remain completely asymptomatic. The mechanisms underlying these phenomena are largely unknown, although the presence of genetic modifiers in the same or in distinct genes has been advocated in some cases. The advent of next-generation sequencing techniques has further highlighted the complexity of monogenic disorders, often expanding the phenotypic spectrum associated with variation in a single gene to encompass distinct conditions. But even more challenging to the principles of Mendelian genetics is the observation that mutations in the same gene may behave either in a dominant or a recessive way in distinct pedigrees. In this issue of Brain, Kariminejad and co-workers provide one such example, by revealing that individuals who are homozygotes for mutations in TOR1A display a genetic disorder which is clearly different from that observed in heterozygous carriers (Kariminejad et al., 2017).
TOR1A, encoding the protein TorsinA, was the first gene to be associated with isolated early-onset dystonia (termed DYT1-dystonia). In all affected individuals, the causative mutation was found to be a heterozygous 3 bp GAG deletion in exon 5, resulting in the loss of one of a pair of glutamic acid residues towards the C-terminus of the protein (p.Glu303del). A few other heterozygous variants have been reported in dystonic patients, but their pathogenic significance remains to be established. DYT1-dystonia is an autosomal dominant disorder with variable expressivity and incomplete penetrance: more than 70% of individuals carrying the GAG deletion remain asymptomatic throughout their lives, while 20–30% of mutation carriers manifest the dystonic phenotype, which is usually severe and generalized but can also remain focal or segmental in a minority of patients (Ozelius and Lubarr, 1993). By using a whole exome sequencing approach, Kariminejad and colleagues identified homozygous pathogenic variants of TOR1A (including the same p.Glu303del variant and a novel p.Gly318Ser missense variant) in four children from three unrelated consanguineous families presenting an overall different and more severe disorder, characterized by congenital contractures of multiple joints, dysmorphic features, severe psychomotor delay, increased muscle tone, and limb tremor. At their most recent examination, between 2 and 4 years of age, all children had developed strabismus, moderate to severe intellectual disability and progressive microsomia. Interestingly, both tremor and contractures seemed to improve over time (with tremor disappearing by age 18 months in two siblings), while strabismus was reported to worsen. None of the affected children had any obvious sign of dystonia, but two of them developed stereotypic hand movements, facial grimacing and neck extension movements, which could possibly be dystonic in nature (Kariminejad et al., 2017).
TOR1A is only the latest example of an expanding group of genes in which mutations can be found either on one or both alleles. In true dominant disorders, pathogenic variants result in the same phenotype regardless of their presence in the heterozygous or homozygous state. This is the case for LRRK2-associated Parkinson’s disease, which is clinically indistinguishable in patients who are heterozygous or homozygous for the common p.Gly2019Ser missense variant (Ishihara et al., 2006). In other cases, the dominant or recessive inheritance of distinct mutations may depend on their different pathogenic impact on the protein product, resulting in variable phenotypic manifestations. One such example is GNAL, encoding the α-subunit of the heterotrimeric G protein Golf (Gαolf). Heterozygous loss-of-function variants of this gene underlie a form of adult onset focal or segmental dystonia with dominant inheritance and incomplete penetrance. Recently, a GNAL homozygous missense variant was reported in two consanguineous siblings with early onset generalized dystonia. This variant did not result in a complete loss of function, but was found to selectively impair Gαolf functional coupling to dopamine D1 receptors (Masuho et al., 2016). Yet, the variable effects of gene mutations cannot be solely explained by their functional consequences at the protein level. An enlightening example is myotonia congenita, a hereditary muscle channelopathy caused by mutations in the CLC-1 chloride channel gene, and clinically characterized by delayed relaxation of skeletal muscle following voluntary contraction. Two forms of myotonia congenita exist, a dominant form (Thomsen disease) and a more severe recessive form (Becker disease). It was initially thought that dominant myotonia congenita was caused by missense mutations acting via a dominant negative mechanism to impair the voltage-dependent gating activity of the dimeric chloride channel, while the recessive form was associated with biallelic loss-of-function mutations resulting in a total impairment of channel activity. This hypothesis was subsequently challenged by the identification of recessive mutations yielding functional CLC-1 channels, and even more so by the discovery of ‘semi-dominant’ CLC-1 mutations, associated with either Thomsen or Becker disease according to their heterozygous or homozygous state (Tang and Chen, 2011).
The study by Kariminejad and colleagues indicates that the common TOR1A GAG deletion also behaves as a semi-dominant mutation, which accounts for generalized early onset dystonia with reduced penetrance in heterozygotes and for the more severe phenotype of congenital arthrogryposis with intellectual impairment and strabismus in homozygotes. An intriguing question is how these two apparently distinct phenotypes can be explained by impairment of the same gene. The diagnosis of arthrogryposis multiplex congenita is a descriptive one, insofar as any factor that interferes with normal foetal movements may in principle result in multiple congenital contractures. As a consequence, the pathogenic mechanisms underlying arthrogryposis are plentiful and diverse, and may include myopathic defects, neuropathic defects (at the central or peripheral nervous system level), neuromotor endplate abnormalities, connective tissue disorders, limitation of space in utero, decreased blood flow to the placenta or the foetus, and teratogenic exposure. Thus, it is not surprising that arthrogryposis has been reported in a very large number of genetic syndromes, and a search for this term in the Online Mendelian Inheritance in Man catalogue (http://omim.org/) yields as many as 75 clinical synopses. In all patients reported by Kariminejad and colleagues, contractures and tremor significantly improved in the first years of life. This observation, along with the reported muscular hypertonia, suggests that TOR1A-related arthrogryposis is the result of a neurological defect in controlling muscle activity rather than a primary defect of the muscle or the connective tissue. Interestingly, a recently reported gene ontology analysis of the over 300 genes with mutations that have been associated with various forms of arthrogryposis identified several enriched pathways that could be relevant also for TOR1A/TorsinA function, such as CNS development, developmental growth, endoplasmic reticulum (ER) activity, and synaptic transmission (Hall and Kiefer, 2016).
TorsinA is a ubiquitous member of the AAA+ (ATPases associated with various cellular activities) superfamily, a group of proteins implicated in a wide variety of cellular processes. Within cells, TorsinA resides in the endoplasmic reticulum and the nuclear envelope, where it regulates many important functions, including protein quality control and folding, lipid metabolism and membrane trafficking. Moreover, at the synapse level, TorsinA was found to interact with proteins of the SNARE complex to modulate vesicle docking and fusion (Fig. 1). Cellular models overexpressing mutant TorsinA show peculiar damage to the nuclear envelope, ER stress, abnormal synaptic vesicle cycling, altered lipid metabolism and impaired cytoskeletal organization (Cascalho et al., 2017).
Dominant negative mutation: A mutation on one allele that interferes with the function of the wild-type protein encoded by the other allele.
Expressivity: The range of phenotypic severity associated with a particular genetic cause.
Loss-of-function mutation: A mutation that results in the gene product having reduced or no function (being partially or wholly inactivated).
Penetrance: The proportion of the time that a genetic cause results in an observable clinical phenotype.
Recessive mutation: A mutation that needs to be present on both alleles in order for the mutant phenotype to manifest.
Semi-dominant mutation: A mutation that is associated with a certain phenotype when present in heterozygosity, and with a more severe phenotype when present in homozygosity.
True dominant mutation: A mutation that is associated with the same phenotype regardless of its presence in heterozygosity (on one allele) or in homozygosity (on both alleles).
Of note, the Tor1AΔGAG/ΔGAG knock-in mouse, similarly to the constitutive knock-out model, dies soon after birth, and shows nuclear membrane defects that are restricted to certain neuronal populations, mirroring the purely neurological manifestations of DYT1-dystonia (Goodchild et al., 2005). This was recently explained by the observation that TorsinA and its close paralogue TorsinB, which has redundant activities and is therefore protective against TorsinA loss or malfunction, are spatially and temporally regulated during development in neuronal cells, with TorsinB undetectable prenatally and expressed at progressively higher levels in the first weeks of postnatal life (Bahn et al., 2006; Tanabe et al., 2016). Thus, it is not surprising that in the presence of homozygous TOR1A mutations, the most disruptive effects of lack of functional TorsinA manifest during pre-natal and early neonatal life. These include severe neurological deficits manifesting with increased muscle tone, contractures and involuntary movements, which then progressively improve after birth paralleling the rise of TorsinB. The moderate-to-severe intellectual impairment observed in these patients is possibly related to defects affecting synaptic plasticity and functioning during a crucial time for the development of the CNS, which cannot be rescued by the delayed expression of TorsinB. Further studies are needed to better understand the pathogenic impact of homozygous mutant TorsinA in the developing human brain and the correlation between its dysfunction and the occurrence of congenital arthrogryposis and intellectual impairment.