On the other hand, the database survey reveals that deletions leaving “fractional STRs” are more commonly associated with BMD than those yielding predicted possible “hybrid STRs”, contrary to our expectations (Table ​(Table2).2). Two deletion patterns strongly associated with mild clinical phenotype (ex45-47 and ex45-49) make up more than half the cases in which fractional Inhibitors,research,lifescience,medical STRs are predicted. It would seem, therefore, that we must look for other features of these deletions to account for the severity of disease with which they are associated. Besides the primary binding site of

F-actin in the N-terminus of Dystrophin, actin also binds the rod domain between STRs 11 to 17 (encoded by ex31-45) (32, 33). It is not known whether different deletion patterns differentially affect actin binding, however, it seems that the presence/absence of the actin-binding site does not correlate simply with the different phenotypes in each pattern in the database, Inhibitors,research,lifescience,medical since deletions starting from exon 45 result in milder phenotypes than those starting from later exons (Table Inhibitors,research,lifescience,medical ​(Table22). A more general rule is that deletions starting from exon 45 and ending before hinge 3 tend to be associated with mild phenotypes irrespective

of the predicted STR structure (Table ​(Table2).2). It is also noteworthy that, exon Inhibitors,research,lifescience,medical 45-55 deletions, which remove hinge 3 are strongly associated with mild clinical phenotype (3) but deletions whose breakpoints fall adjacent to hinge 3, such as D50/51, or D49-53 usually lead to DMD rather than BMD. This conundrum is not readily explicable, but it might suggest that there are limits on the size of the rod domain if the hinge domain is lacking, perhaps because hinge domains are needed to give flexibility to rod structures beyond a certain length (34). Clearly, we need a better understanding of

these differences if we are to Inhibitors,research,lifescience,medical make rational choices of exon-skipping targets that will generate the most functionally effective “quasi-dystrophins”. In this isothipendyl quest for a better understanding of the relevant factors, perhaps the most relevant evidence is to be found in a detailed study of the genotype/phenotype correlations in the human DMD and BMD populations. This requires the assembly of a reliable database founded on the unison of strictly defined clinical criteria with comprehensive and detailed information on the nature of the mutation ideally including the genomic, and the transcript data, and, where applicable, the exons Selleck AG-14699 represented in any partial dystrophin protein produced. It is encouraging to note that a number of databases approaching this level of detail are currently being organized. Acknowledgements Authors thank Drs Akinori Nakamura, Shin-ichi Takeda, and Eric Hoffman for useful discussions.