We describe the clinical and molecular features of 26 subjects with TRPV4-related bone dysplasias (MD, SMDK, and brachyolmia). These data highlight that these chondrodysplasias constitute an allelic "family" in which there is a spectrum of overlapping severity ranging from lethal MD to mild AD brachyolmia, which we contend is the mildest phenotypic expression of SMDK, as all cases have radiographic evidence of metaphyseal involvement.
We found 9 mutations in 22 patients: 4 are previously described mutations, 5 are novel.
This study has also identified the first TRPV4 insertion: a de novo c.1566_68dup (p.L523dup) found in patient 17, who has a typical SMDK phenotype. The CCT insertion after the second base of codon 523 maintains the leucine at codon 523 and inserts an additional leucine codon (CCG) at this position. The sequence thereafter remains in-frame.
Two MD patients had a very severe/lethal phenotype: patient 8, with mutation P799S and patient 14, with mutation Q239H. Some authors [2
] proposed different inheritance modes for MD, with a lethal, possibly recessive, type of MD, and a dominant "classic" MD. Others, however,[3
] have shown TRPV4
heterozygous mutations in very severe/lethal MD.
Our data confirm dominant mutations underlie very severe/lethal MD. This information is crucial for genetic counselling and recurrence risk assessment as this disorder was previously considered a recessive trait with 25% recurrence risk.
The family with the Y591C mutation (pts 19,20,21) was particularly illustrative in the fact that the same bone dysplasia expert made different diagnoses in half brothers (AD brachyolmia/SMDK respectively), not knowing they came from the same family. These cases, and the fact that neither of the patients (24 and 25) with purely isolated spinal changes showed TRPV4
mutations, prompted us to critically review the AD brachyolmia phenotype. The first cases of brachyolmia with TRPV4
mutations reported in the literature showed, in association with the typical vertebral abnormalities, mild metaphyseal changes in the femoral heads and delayed bone age [9
]. Dai et al. (2010) wrote about "two SMDK patients who did not show overt metaphyseal changes and were considered to be of intermediate severity between SMDK and Brachyolmia" [14
]. Kozlowski questioned the existence of brachyolmia as a separate entity, saying that all patients with brachyolmia had at least some mild metaphyseal changes, with shortening of femoral necks representing the minimal diagnostic changes, and for that reason had to be considered as spondylometaphyseal dysplasia patients [16
]. Given these observations and the data from our study, we propose that "true" brachyolmia, i.e. with changes limited only to the spine, is the recessive Hobaek type, while "autosomal dominant brachyolmia" should be considered as the mildest end of the spectrum of SMDK. SMDK cases with very mild metaphyseal changes have been described [17
] as have metaphyseal irregularities (especially in the femurs) in AD brachyolmia [18
As the molecular data on TRPV4 mutations are becoming more available, it is feasible to try to find valid genotype-phenotype correlations as it is increasingly evident that these TRPV4 skeletal dysplasias, have considerable phenotypic overlap with blurred boundaries.
Our data confirm that P799 and R594 are the most commonly mutated codons in the TRPV4 skeletal dysplasias; however, as others have reported [14
], it is not always easy to make a definite association between a specific codon and a specific phenotype. Codon 799 was mutated in 6 subjects: this was already considered a mutational hotspot, and different amino acid changes had been described. All our patients with a mutation in codon 799 had MD, consistent with the reported experience [3
]. Nine subjects had a mutation at codon 594: R594H is a recurrent mutation and the finding of a different amino acid change (R594S) confirms this codon as another mutational hotspot. In contrast to the published data, we found this mutation in MD subjects as well as in SMDK. We therefore suggest that the recurrent mutation R594H should be considered when an apparent bridging phenotype between SMDK and MD is observed.
For some of the mutations there seems to be a good genotype-phenotype correlation, based on the cases described in the literature, like P799L or F617L, which are associated respectively with classical MD and "mild" MD/"severe" SMDK. Our data show, however, that this is not true for all the mutations described and it is not even possible, at present, to find a true association between a specific protein domain and bone dysplasia disease severity. Mutations are spread throughout the protein and mutations causing MD, for example, are found in the ankyrin repeat domain as well as in the C-terminal domain.
Some of the mutations described in the literature have been tested functionally and found to be gain of function mutations. We know TRPV4 is a very complex protein and many authors have demonstrated with in vitro
experiments that different changes at the same codon can affect the channel response to agonists differently [19
]. In the last year the picture became even more complex when mutations in the TRPV4 protein were shown to cause not only the above-mentioned bone dysplasias but also neurologic conditions such as congenital distal Spinal Muscular Atrophy (SMA), scapuloperoneal SMA and hereditary motor and sensory neuropathy 2C (also known as Charcot-Marie-Tooth type 2C) [20
]. Recently, other bone dysplasias have been associated with TRPV4
mutations too, like spondylo-epiphyseal dysplasia, Maroteaux type (pseudo-Morquio syndrome type 2) and parastremmatic dysplasia [23
For a full understanding of the pathogenic mechanisms that lead from mutations in this complex gene to such a variety of different conditions, more functional data on the mutant channels is required as well as more information about how interactions between the TRPV4 channel and other proteins are disrupted in these conditions and how downstream physiological processes are perturbed.
In conclusion, the data from this study strongly suggest that: a) the condition previously considered autosomal recessive, lethal MD represents the severe end of the MD phenotypic spectrum, as demonstrated by our findings and by previous data from the literature [3
]; b) MD and SMDK, should be considered part of a TRPV4
bone dysplasia spectrum with significant overlapping clinical and radiographic features, even within families; c) AD brachyolmia is not a separate phenotypic entity and should be considered as the mildest expression of the SMDK spectrum, while AR brachyolmia has different radiographic features and, probably, a different genetic cause (Figure ).
Figure 2 Different perspectives in classification of the TRPV4 bone dysplasias. On the top the classification into separate conditions; on the bottom the idea that lethal MD is the severe end of the MD spectrum; there is a region where the clinical features of (more ...)