Findings

Background

Joubert syndrome (JS, MIM213300) is a congenital disorder diagnosed by the presence of a peculiar midbrain-hindbrain malformation (the “molar tooth sign”, MTS), that consists of cerebellar vermian hypodysplasia, thickened mal-oriented superior cerebellar peduncles, and a deepened interpeduncular fossa. The typical neurological features of pure JS include hypotonia, ataxia, psychomotor delay, abnormal ocular movements, intellectual impairment of variable degree, and often breathing abnormalities. This phenotype may be complicated by defects of the kidneys (nephronophthisis), eyes (retinal dystrophy or colobomas), liver (congenital fibrosis), skeleton (mainly polydactyly), and orofacial defects (cleft lip and/or palate, tongue hamartomas), resulting in wide clinical variability [1].

JS is recessively inherited and genetically heterogeneous, with 24 known genes that overall account for about half cases. All genes encode for proteins of the primary cilium, and indeed there is clinical and genetic overlap with other ciliopathies. In particular, JS shares 13 genes with Meckel syndrome (MS, MIM249000), a lethal condition characterized by cystic kidneys, bile duct proliferation of the liver, encephalocele and polydactyly. Other malformations frequently include cleft lip and palate, bowing of long bones and other skeletal defects, and situs inversus [2].

Identification of MKS1 and B9D1 mutations in JS patients

As part of a large screening of ciliopathy genes in 260 JS patients, we identified novel pathogenic mutations in two genes not previously implicated in this condition.

Two patients carried mutations in the MKS1 gene [GenBank:NG_013032.1], a 44-year-old man with JS and retinal dystrophy (COR340), and a 2-year-old child with a pure JS phenotype (COR413). Mutations in the B9D1 gene [GenBank:NG_031885.1] were identified in two other patients, a 9-year-old boy (COR363) and a 7-year-old girl (COR346), both presenting with pure JS. All identified mutations were inherited from heterozygous healthy parents, were not reported in public databases, and affected highly conserved residues (Figure 1). Missense mutations were predicted as pathogenic by prediction web tools. Clinical features of the four patients, compared with the phenotypes of the six JS subgroups [1], are summarized in Table 1. Individual case reports and details on genetic analysis are described in the Additional file 1.

Figure 1
figure 1

Brain magnetic resonance imaging and characterization of mutations of the four patients. A) Parasagittal (left) and axial (right) magnetic resonance imaging sections showing thickened and elongated superior cerebellar peduncles (arrows), and the molar tooth sign (circles). Parasagittal images are not available for patient COR340. B) Electropherograms showing the identified mutations; C) conservation of affected amino acid residues among orthologues (for missense mutations or single amino acid deletions), or prediction of aberrant splicing (for splice-site mutations).

Full size image
Table 1 Phenotypic comparison of the four patients presented here with JS clinical subgroups
Full size table

Discussion

Pathogenic mutations in MKS1 and B9D1 have been reported only in MS fetuses. MKS1 is mutated in about 7-14% of MS patients, with increased frequency in northern European countries due to a founder mutation [3–7]. Several studies have highlighted that mutations in MKS1 are associated with a particularly severe MS phenotype, with high occurrence of polydactyly, bone dysplasia, encephalocele and other central nervous system anomalies [4–6]. To date, B9D1 was found mutated only in one MS fetus with cystic dysplastic kidneys, encephalocele, shortened limbs and ambiguous genitalia [8]. Conversely, the four JS patients described here all had a relatively mild presentation characterized by a pure neurological phenotype, with the exception of retinal dystrophy in patient COR340. The degree of intellectual impairment was variable, and patient COR413 even presented with normal intellectual abilities, a rare occurrence in JS [9]. None of the patients showed involvement of the organs that are typically affected in MS, namely the kidneys, liver and skeleton, although a future renal disease can be safely excluded only in the adult patient (COR340).

This wide phenotypic variability associated with mutations in the same genes remains an intriguing open question. Genotype-phenotype correlates have been proposed for some genes (such as RPGRIP1L, TMEM67, CCD2D2A and TCTN3), with biallelic null mutations causative of lethal phenotypes, and at least one hypomorphic missense mutation found in JS [10–13]. This could also hold true for MKS1 and B9D1. In fact, most MS fetuses are known to carry two null mutations in these genes [4, 8]; conversely, three of our JS patients have at least one mutation not resulting in protein truncation, and the fourth is homozygous for a splice-site mutation involving the penultimate exon of MKS1, whose pathogenetic impact on the protein remains to be determined (See Additional file 1) (Figure 1). Interestingly, a previous study identified two hypomorphic mutations in the MKS1 gene (a missense change and a single aminoacid deletion) in a 2-year-old Turkish patient with Bardet-Biedl syndrome, another non-lethal ciliopathy partly overlapping with JS, supporting this hypothesis [14]. Yet, these genotype-phenotype correlates are unlikely to fully explain the extreme phenotypic variability of these allelic ciliopathies, and other mechanisms, such as the presence of modifier variants in other genes, need to be explored.

MKS1, B9D1 and B9D2 proteins are known to interact physically [15], and are main components of the “B9” or “Tectonic” complex residing at the ciliary transition zone, that includes many other proteins mutated in JS and/or MS [16]. In our large JS cohort, MKS1 and B9D1 mutations each account for less than 1% cases. We failed to identify mutations in B9D2, but we cannot exclude that this gene may also be rarely mutated in JS.

In conclusion, we expand the genetic basis of JS to include MKS1 and B9D1, delineate genotype-phenotype correlates, and further outline JS and MS as the two ends of a common spectrum. These findings have implications for genetic testing and counselling of JS patients and their families.