The finding that the nature of the expanded repeat at the DM2 locus is similar (though not identical) to that producing classic myotonic dystrophy provides strong support for the view that it is the mutations themselves, rather than the actual genes involved, that are central to pathogenesis. Evidence supporting this was already growing before identification of the DM2 mutation (Davis et al, 3), with evidence from transgenic mice with a long CTG repeat in the skeletal muscle actin gene showing key features of myotonic dystrophy pathology, including electrophysiologic myotonia and histologic muscle changes similar to human myotonic dystrophy (Mankodi et al, 13).
The ZNF9 gene, already identified by earlier studies (Venter et al, 29), shows no homologies to myotonic dystrophy protein kinase (DMPK) (Liquori et al, 12) or to any of the other genes in the myotonic dystrophy region of chromosome 19; this fact supports the role of the mutation itself. In this respect, it is interesting that a comparable (though not identical) situation has proved to be the case for the polyglutamine disorders (e.g., Huntington disease) due to CAG repeat expansions in different genes (Harper, Perutz, 8). Unlike these conditions, the expanded repeats responsible for the two types of myotonic dystrophy are not translated into the proteins; indeed, the ZNF9 gene is not known to produce any protein. However, both DNA repeats are transcribed into mRNA as CUG or CCUG repeats. However, as the function of ZNF9 is not known, there may yet be links between these two loci in terms of the pathways in which they are involved.
The importance of an expanded CUG repeat in terms of affecting CUG-binding proteins and thus interfering with the normal splicing mechanisms for other types of RNA has now been fully recognized for myotonic dystrophy (Mankodi et al, 14, Savkur et al, 23) and strongly supported by studies of both affected human muscle and transgenic mice carrying an expanded CUG repeat. The presence of nuclear RNA inclusions is a feature of both human myotonic dystrophy and the animal model (Mankodi et al, 13), but there is at least one mouse model that has the inclusions and no phenotype (Seznec et al, 25).
Study of muscle from patients with the DM2 mutation has now shown similar changes (Fardaei et al, 5; Liquori et al, 12), strongly supporting the existence of a common pathogenetic process for the two forms of myotonic dystrophy, despite the widely different nature of the genes involved and the different chromosome location.
If, indeed, the main features of myotonic dystrophy can result from pathology at the RNA level, with the mutation located in such different genes, this raises a question: Which elements of pathogenesis, if any, can be attributed to the function of the genes themselves or to that of contiguous genes? Absence of significant muscle pathology in DMPK knockout mice (Jansen et al, 1997) has already suggested that DMPK does not have a primary role in determining this aspect; however, the effects on cardiac conduction seen in such mice (Berul et al, 2) are strongly suggestive of a direct role in myotonic dystrophy. Since such cardiac conduction problems are seen at a clinical level in both PROMM/DM2 patients and in classic myotonic dystrophy, it is possible that DMPK is involved at the RNA level in producing cardiac pathology.
The similarity of the DM1 and DM2 locus phenotypes (including cataract) also calls into question the need for invoking contiguous genes in the pathogenesis of the disorder, since there is no evidence for genes adjoining the DM2 locus with a comparable structure or function to those immediately adjacent to the DM1 locus (e.g., SIX5 and DMWD). However, as the function of ZNF9 is not known, there may yet be links between these two loci in terms of the pathways in which they are involved. In a Drosophila mutant, it has clearly been shown that the equivalent gene to SIX5 (D-Six4) is essential for normal development of both muscle and gonad. In fact, adult flies with mutations in D-Six4 demonstrate gonadal atrophy and other abnormalities (Kirby et al, 11).
A further point that needs to be taken into account is the distribution of the mRNA containing the repeat expansions and the pathologies associated with DM1 and DM2. DMPK is only expressed in skeletal, cardiac, and smooth muscle. Invoking the RNA-mediated pathologies in nonmuscle tissues is therefore difficult, and the degree of pleiotropy evident in the phenotypes associated with DM1 is extreme. It is therefore certain that there are other mechanisms contributing to the DM1 phenotype outside muscle tissues; they almost certainly involve the genes flanking the repeat at this locus. In DM2, the ZNF9 gene is from EST databases apparently ubiquitously expressed (Johnson, unpublished) and yet the phenotype so clearly recapitulated many of the specific features of DM1. Until the function of ZNF9 is elucidated from animal models, we will have to wait to find out the extent to which the RNA-mediated effects explain these overlapping features.
Resolving these and other questions will require a detailed reassessment of the clinical similarities and differences between the two forms of myotonic dystrophy. Features given little attention until now point to which forms of RNA and protein are involved in the pathogenetic process, as well as more accurately distinguishing the two forms. Likewise, at a temporal level, genes involved in muscle and brain development could, through their RNAs, be involved in the causation and specific features of congenital myotonic dystrophy.
It can thus be seen that (1) the recognition of a second form of myotonic dystrophy (PROMM, type 2 myotonic dystrophy), (2) the identification of its specific locus on chromosome 3 (DM2), and (3) the fact that the responsible mutation has proved to be an expanded CCTG (CCUG) repeat are developments with profound implications for our understanding of myotonic dystrophy as a whole, quite apart from the importance of distinguishing accurately those patients with the DM2 mutation. By the strong support given to the primary importance of expanded RNA repeats (CUG and CCUG) in causing disease pathology, a novel mechanism of pathology has been recognized; the general importance and extent of this mechanism can be considerable. Likewise, looking ahead to therapeutic approaches, this work will prove to be of critical importance if the specific proteins involved at the RNA level are to be accurately identified and targeted in future studies designed to minimize their dysfunction in myotonic dystrophy.