Sir, In the recent letters by Hedberg et al. (2013) and Pfeffer et al. (2013), the important question of genetic variants determining the pathogenesis of human myopathy with early respiratory failure (HMERF) was discussed. We welcome the attention and scientific discussion about the molecular causes and downstream pathogenic alterations in patients affected by this debilitating disease. We recognize the increasing evidence that mutations in titin domain A150 are the major and often dominant cause of HMERF. However, the conclusions the authors draw are debatable for several reasons.
Firstly, the availability of next-generation sequencing, which was not available 9 years ago, allows us now to confirm from our own analysis that all affected members of the originally reported families in (Lange et al., 2005) with the R279W titin kinase (TK) variant (TK nomenclature of residues as in PDB 1TKI to allow comparison) carry a second titin variant in the A150/Fn3 119 domain, P30091L, as also shown in a subset of these patients by Hedberg et al. (2013). Moreover, by clarifying the segregation in the two families and one sporadic patient, we are now able to show that both variants are located in cis on the same allele and not in trans.
Secondly, in contrast to the letter by Pfeffer et al. (2013) but in agreement with the recent paper by Pfeffer et al. (2014), the P30091L variant is clearly not sufficient by itself to cause HMERF. In Family K in our recent article (Palmio et al., 2014) with P30091L in homozygosity in the proband, we have now been able to carry out additional studies on the heterozygous members in this family. The parents are normal healthy individuals also as judged by the sensitive muscle MRI studies at the ages of 59 and 63 years, respectively, as is the 39-year old sister. This is in full agreement with the fact that P30091L was suggested as ‘a neutral variant’ by Pfeffer et al. (2014), as it was also found in an unaffected sibling of a patient carrying this mutation, and clearly indicates that when this P30091L variant causes disease, it is in combination with an additional titin gene abnormality.
In light of the recent reports on novel HMERF mutations, the following conclusions can be made: (i) the TK variant strictly cosegregates with HMERF in the original Swedish families, making the assumption of a dominant mutation in TK justified at the time, considering that 400 controls were sequenced without finding the TK variant in that population, and the still valid biochemical evidence for pathogenicity (see also Chauveau et al., 2013); (ii) the strict dominance of the P30091L-TK variant combination in the Swedish families is in contrast with the recessive nature in other families with the P30091L variant alone; (iii) dominance of P30091L must therefore be because of a strictly co-inherited modifier; (iv) in case of the originally identified HMERF patients, this modifier is most plausibly the TK variant; (v) it is incorrect to assume that because the R279W variant occurs in single nucleotide polymorphisms, it is not pathogenic—see the recent disruptive TK variant W260R (Chauveau et al., 2013). Rather, single nucleotide polymorphisms also include rare recessive pathogenic variants; and (vi) TK is directly linked to the autophagy pathway through the two autophagy adaptor proteins p62/SQSTM1 and NBR1, and indirectly linked with the ubiquitination pathway through MuRF (Lange et al., 2005; Chauveau et al., 2013); these pathways now also seem to be affected by single dominant A150/FN3 119 variants.
Mechanistically, it is therefore highly plausible that co-inheritance of a mutation challenging muscle protein turnover through autophagy and/or ubiquitination (A150/Fn3 119 variants) and one deregulating it (TK variants) may lead to phenotypic penetrance similar to a single dominant mutation. We propose therefore that recessive variants in A150/Fn3 119 can become penetrant when recessive TK variants cooperate with a ‘second hit’, at least in cis, on protein turnover pathways. It is interesting to speculate that co-inheritance of recessive A150/Fn3 119, or indeed TK variants with mutations in other ubiquitination- or autophagy-related proteins (MuRF, p62/SQSTM1 and NBR1 in particular) may lead to similar scenarios.
Rather than not screening for TK mutations, this scenario would make it essential to do so at least in certain cases to establish whether A150 variants of variable penetrance become phenotypic as a result of either a compound-heterozygous or monoallelic-bimutational genotype. Other recessive TK variants have now been linked to compound-heterozygous scenarios (Chauveau et al., 2013). Also, the 1000 Genomes database contains 34 other TK missense variants, of which 23 have PolyPhen ratings >0.8 (including R279W and W260R) and are structurally predicted to be disruptive (unpublished observations). The potential for co-inheritance of recessive disruptive titin kinase variants with recessive A150/Fn3 119 variants is therefore considerable, and adequate genotyping for diagnostics and genetic counselling would likely fail if this possibility would be discounted.
To elucidate the suggested impact of recessive TK variants, we are working on functional studies, including a TK mutant mouse model, to gain a more comprehensive understanding of the complexity of these different titin functions and alterations.
S.L. is funded by the National Institutes of Health/NHLBI K99/R00 Pathway to independence award (HL107744), B.U. by the Folkhalsan Institute of Genetics Foundation, and M.G. holds the British Heart Foundation Chair of Molecular Cardiology.