RNU genes have garnered considerable attention over the past two years as major monogenic causes of intellectual disability. These genes (RNU4-2, RNU2-2, etc) are small nuclear RNAs (snRNAs), a subclass of noncoding RNAs which are key components of the spliceosome. Their naming is a bit complicated because the genes that encode major spliceosome snRNAs (U1, U2, U4, U5, and U6) have multiple functional copies in the human genome. The most famous of these, RNU4-2, is the “second” of two identical genes which encode U4. About 18 months ago, two studies published in Nature family journals determined that recurrent de novo mutations in RNU4-2 were the cause of a relatively frequent neurodevelopmental syndrome.
Convergent Discoveries in Genomics England
It turns out that both groups were working with a highly overlapping set of undiagnosed patients in the Genomics England database. Greene et al (Nature Medicine, July 2024) initially identified 47 cases with recurrent de novo variants in this gene, while Chen et al (Nature, August 2024) initially identified 46. Both groups searched other cohorts, ultimately identifying more than 100 patients. A third group (Nava, Cogne et al) soon replicated this in an independent cohort of French rare disease patients, and collaborated with other European countries to add 145 new RNU4-2 patients to the literature. It soon became apparent that the condition, now called ReNU syndrome, is remarkably prevalent. By some estimates, these recurrent variants in RNU4-2 explain ~0.5%, or 1 in 200 patients with intellectual disability.
How RNU4-2 Took So Long To Discover
Perhaps the most surprising thing about RNU4-2 and ReNU syndrome is that it eluded detection for so long. There are several factors contributing to this, first and foremost being the fact that RNU genes encode noncoding RNAs rather than proteins. Both of the genes that encode U4 (RNU4-1 and RNU4-2) are in “noncoding space” near the SIRT4 gene at at chr12q24.3:
RNU4-1 lies in the first intron of SIRT4, and RNU4-2 lies about 1.5 kbp upstream, just to the left of the 5′ UTR. Such regions are generally not targeted by exome capture kits which were the basis of Mendelian disease diagnostics and research for about a decade. Even when genome sequencing was performed on these patients, RNU genes were often assigned lower priority by most protein-centric annotation pipelines. For example, RNU4-2 variants often are annotated as SIRT4 UTR variants because annotation tools consider a protein effect more likely to be deleterious than a noncoding RNA effect. In defense of such methodologies, there are numerous types of noncoding RNAs (miRNAs, snRNAs, snoRNAs, lncRNAs, etc) and very few of them have been linked to highly penetrant, Mendelian disorders.
Other RNU Genes in Mendelian Disease
However, since the landmark discovery of RNU4-2, other gene-disease associations for RNU genes have emerged (see Stylianos Antonarakis, Nature Genetics 2025 for a fantastic review). In April and May of last year (2025), two groups screening RD cohorts for snRNA genes published RNU2-2 as the cause of a dominant neurodevelopmental syndrome with prominent epilepsy. Though not as common as ReNU syndrome — Greene et al found 25 cases, while Jackson et al reported 9 cases with RNU2-2 and 6 with RNU5B-1 — it was a similar pattern: many recurrent (often de novo) variants in critical subregions of the RNU2-2 gene.
In November of last year, an international collaboration (Quinodoz et al 2025) found that heterozygous variants in RNU4-2 (a different subregion) or the four genes encoding U6 (RNU6-1, RNU6-2, RNU6-3, and RNU6-4) cause retinitis pigmentosa, an inherited form of progressive vision loss. Several large adRP families that I studied for years with my colleagues at the University of Texas were explained by U4 or U6. In retrospect, this was not surprising: several other well-known dominant RP genes (e.g. PRPF3 and PRPF31) encode protein components of the spliceosome complex. Again, exome sequencing was the assay of choice for many of us, and did not yield coverage of these genes.
Recessive RNU2-2 Variants Cause A Distinct Syndrome
This month, three papers published in Nature Genetics add a new dimension to the RNU story: a clinically and genetically distinct recessive neurodevelopmental disorder caused by biallelic variants in the RNU2-2 gene. Two of those came from teams working with the Genomics England dataset. The study I found most illuminating, however, was led by Caroline Nava’s group in France (Leitao et al, Nat. Genet 2026). They began with a systematic survey of spliceosomal snRNA genes in the human genome. Starting with the set of 2,094 snRNA genes in the Ensembl database for GRCh38, they prioritized genes that were hypermutable, had approved HGNC symbols, and/or overlapped proximal cis-regulatory elements. This winnowed the list to 200 putative functional snRNA genes.
The authors next searched the PFMG cohort — WGS data from ~34,000 patients with rare diseases — for variants in these 200 snRNA genes to see if any were significantly enriched for variation in their RD probands. In the enrichment analysis of de novo variants, there were two obvious outliers:
Both RNU4-2 and RNU2-2 were significantly enriched for de novo variants, with around 85 and 45 patients, respectively. These are essentially built-in true positives for the analysis, so their results are encouraging. Less encouraging is the fact that beyond these two genes there’s not a third locus with anywhere near the same significance. In this large cohort, in other words, the “easy” RNU genes in which de novo mutations cause disease have already been found. It gets more interesting when the authors did the enrichment for biallelic (that is, homozygous or compound-heterozygous) variants:
RNU2-2 again shows a significant enrichment, this time for a different mode of inheritance. Strikingly, there are even more patients with biallelic variants in RNU2-2 than de novo variants in RNU2-2. The authors went on to show that:
- Recessive RNU2-2 disease is at least twice as frequent as dominant RNU2-2 disease.
- Many patients harbor an inherited variant in trans with a de novo mutation, highlighting the high mutability of this gene
- Transcriptome and methylation signatures of RNU2-2 variant carriers show subtle variant effects on splicing and methylation.
- Considering the spectrum of dominant and recessive inheritance modes, RNU2-2 syndrome is nearly as prevalent as RNU4-2 (ReNU) syndrome
For a long time, I’ve written that our protein-centric view of genetic variation is one of the reasons that comprehensive genetic testing fails to provide a diagnosis for more than half of rare disease patients. The challenges are knowing where to look (i.e. what loci are disease-associated) and having the correct data type (genome sequencing) to do so. A growing number of pathogenic variants in RNU4-2, RNU2-2, and other RNU genes are being reported to databases such as ClinVar. As rare disease researchers and diagnostic laboratories query their undiagnosed patient genomes for these genes, I imagine that thousands of new RNU-opathy patients will be diagnosed over the coming years.
Meanwhile, I’m sure the GEL and French teams are searching for the next big ncRNA gene. Maybe they’ll work together this time.
References
Leitão, E., Santini, A., Cogne, B. et al. Systematic analysis of snRNA genes reveals frequent RNU2-2 variants in dominant and recessive developmental and epileptic encephalopathies. Nat Genet 58, 782–797 (2026). https://doi.org/10.1038/s41588-026-02547-5
Jackson, A., Blakes, A.J.M., Alhaddad, B. et al. Biallelic variants in RNU2-2 cause a remarkably frequent developmental and epileptic encephalopathy. Nat Genet 58, 798–809 (2026). https://doi.org/10.1038/s41588-026-02551-9
Greene, D., Mendez, R., Lees, J. et al. Biallelic variants in RNU2-2 cause the most prevalent known recessive neurodevelopmental disorder. Nat Genet 58, 774–781 (2026). https://doi.org/10.1038/s41588-026-02539-5
Leave a Reply