Abstract
Spatial organization in the nucleus is an active layer of RNA regulation in skeletal muscle, yet localization is often interpreted more strongly than the evidence permits. This critical review compares physiological nuclear RNA compartments with pathological repeat-RNA assemblies using a graded evidence framework. NEAT1-associated paraspeckles provide the clearest muscle-linked example of an RNA-marked nuclear body, although isoform-specific architectural causality remains untested in muscle. MALAT1 regulates speckle-associated splicing-factor and chromatin states but is dispensable for gross speckle formation, and muscle studies report opposing differentiation phenotypes. Nucleolar FRG2 transcripts in facioscapulohumeral muscular dystrophy connect spatial RNA organization to rDNA transcription and translational capacity. Expanded DMPK CUG and CNBP CCUG RNAs in myotonic dystrophy sequester processing factors and drive spliceopathy, although many endogenous DM1 puncta are individual retained transcripts instead of bulk condensates. We separate spatial association, molecular participation, functional necessity, in vivo relevance, and therapeutic target engagement. The resulting framework maps intervention strategies to toxic-RNA abundance, RNA-protein binding, repeat-locus transcription, downstream splicing, or compartment state. Progress will require isoform-resolved perturbations, endogenous molecule counting, matched measurements of assembly and RNA processing, and rescue in muscle-relevant models.