Early-onset diabetes is often monogenic, but a substantial fraction of cases remain genetically unexplained even after genome sequencing. This diagnostic gap has focused attention on the non-coding genome, where variants can disrupt gene regulation and RNA processing without altering protein sequence.
The minor spliceosome is one such system. It processes a small subset of introns, known as U12-type introns, found in around 700 human genes. Although these represent less than 0.5% of all introns, many of the affected genes have essential biological functions. Disruption of this pathway can lead to intron retention, altered transcripts, and downstream loss of function.
In this study, Johnson et al. identify bi-allelic variants in two non-coding small nuclear RNA genes, RNU6ATAC and RNU4ATAC, as a cause of early-onset autoimmune diabetes with immune dysregulation1. By combining genetic screening with transcriptomic and immune profiling, the authors link defects in minor intron splicing to impaired immune development and β-cell autoimmunity.
The findings add to the spectrum of early-onset diabetes and provide evidence that non-coding RNA genes can contribute to autoimmune disease through disruption of core cellular processes.
The authors investigated individuals with infancy-onset or early childhood diabetes in whom known monogenic causes had been excluded. Genome sequencing in three affected individuals each identified a different ultra-rare homozygous variant in the non-coding gene RNU6ATAC, which encodes a catalytic component of the minor spliceosome. Expanding this analysis, the study identified:
Together, these defined a cohort of 19 individuals with defects in minor spliceosome components. All presented with early-onset diabetes, typically within the first months of life (median onset 17–20 weeks), and required full insulin replacement.
Across both groups, additional immune dysregulation was observed in 12 out of 19 individuals, including hypogammaglobulinaemia, B cell lymphopenia, recurrent infections, and autoimmune conditions such as hypothyroidism and alopecia. Islet autoantibody testing was available for 10 individuals, of whom half were positive, supporting an autoimmune mechanism.
There were also differences between the two gene groups. Individuals with RNU4ATAC variants frequently showed developmental abnormalities, including microcephaly and growth restriction, while these features were less apparent in those with RNU6ATAC variants.
To assess the functional impact, the authors performed RNA sequencing on affected individuals. This revealed widespread intron retention affecting 274 genes, of which 258 (94%) were known U12-containing genes. The remaining 16 likely represent previously undescribed U12 genes, based on similar retention patterns seen across both cohorts. Most retained introns occurred in known U12-containing genes, consistent with disruption of minor spliceosome function.
Gene expression analysis identified enrichment of pathways related to immune function, particularly B cell signalling and development. Follow-up immune profiling supported these findings:
Taken together, the data support a shared mechanism in which bi-allelic variants in RNU6ATAC or RNU4ATAC disrupt minor intron splicing, leading to transcriptome-wide intron retention, impaired B cell development, and immune dysregulation that results in autoimmune diabetes. This establishes a monogenic link between RNA splicing defects and β-cell autoimmunity.
This study extends the genetic architecture of early-onset diabetes beyond coding regions, showing that non-coding RNA genes can act as causes of monogenic autoimmune disease. Variants in RNU6ATAC and RNU4ATAC do not alter protein sequence, but disrupt a core RNA processing system, leading to downstream effects across 274 genes with significant intron retention.
The findings also place the minor spliceosome within the biology of immune regulation. Although it processes a small fraction of introns, the genes it affects appear to be functionally enriched in pathways linked to immune development. The consistent pattern of U12 intron retention, coupled with convergent immune pathway signals, suggests that even partial disruption of this system can have system-wide consequences.
A key insight is the link between splicing defects and B cell biology. The study shows reduced naïve B cells and impaired maturation, pointing to a failure in normal immune development. This provides a potential route to autoimmunity, though the mechanism is not fully resolved. Only half of the tested individuals were autoantibody positive, and the precise role of B cells in type 1 diabetes pathogenesis remains debated. The authors note that at least one individual with a complete absence of B cells has previously been reported to develop autoimmune diabetes, indicating that B cells may not be strictly required for disease development. The data therefore support an autoimmune process, but do not establish whether B cell dysfunction is causal or secondary.
More broadly, the study reinforces a shift in rare disease genetics. Moving beyond protein-coding regions is increasing diagnostic yield and revealing new biological mechanisms. In this case, it links RNA splicing, immune development, and β-cell autoimmunity within a single genetic framework.
Despite advances in genome sequencing, many rare disease cases remain unsolved, suggesting that clinically relevant variation in non-coding regions is still being missed. Systematic analysis of non-coding genes, particularly those with important cellular functions, may help close part of this gap.
The findings also raise questions about how broadly minor spliceosome dysfunction contributes to human disease. Only a small fraction of introns are processed through this pathway, yet disruption affects genes of which many have essential functions. Similar mechanisms could underlie other unexplained syndromes, particularly those involving multiple systems such as immunity and development.
From a disease perspective, the study adds a new layer to the biology of autoimmune diabetes. It links defects in RNA processing to immune dysregulation and β-cell autoimmunity, but the pathway from splicing disruption to immune targeting remains incomplete. Larger cohorts and functional studies will be needed to define which of the affected genes drive this process and whether specific pathways can be targeted.
There are also practical implications for diagnosis. Current pipelines are still heavily focused on coding variation. The findings suggest expanding clinical analysis to include non-coding genes such as RNU6ATAC and RNU4ATAC could improve diagnostic yield in early-onset diabetes, particularly where immune features are present.
More broadly, this work reflects a shift in human genetics. Understanding disease increasingly depends on integrating coding and non-coding variation with functional data. Studies that combine genomics with transcriptomic and immune profiling, as shown here, are likely to play a central role in resolving currently unexplained cases and refining disease mechanisms.