Past studies suggest potential therapies for newly discovered retinitis pigmentosa mutations in U6 and U4 snRNAs

Sequencing of whole genomes from hundreds of thousands of people has made it possible to track down the genetic causes of many inherited diseases by correlating the presence of rare sequence variants with the disease in affected families (Posey et al. 2019). Only a small fraction of the human genome codes for proteins but much of the genome is transcribed, so disease variants are increasingly being found in noncoding RNAs. A very recent report describes mutations in the genes for two RNA components of the spliceosome, U4 and U6 small nuclear (sn)RNAs, that are associated with the progressive vision disorder retinitis pigmentosa (RP) (Quinodoz et al. 2026). The correspondence of the U6 RP mutations with mutations studied in detail in brewer's yeast suggests possible disease mechanisms and therapies.

RP is the most common form of inherited retinal disease in humans and results from mutation of any one of over 100 different genes (Rivolta et al. 2025). The mutations cause slow, progressive loss of photoreceptor cells (initially rods) typically starting in adolescence, resulting in loss of night vision, progressive tunnel vision, and eventual partial or complete blindness (Verbakel et al. 2018). Previously, all the known RP genes coded for proteins, five of which (Brr2, Prp3, Prp4, Prp8, and Prp31) are present in an RNA-protein subcomplex of the spliceosome called the U4/U6.U5 tri-snRNP. Thus, it made sense to find RP mutations in two RNA components of the tri-snRNP, U4 and U6. About one-third of all RP families exhibit autosomal dominant inheritance (adRP), including all of those with mutations in protein or RNA components of the tri-snRNP. Distinct rare variants in U4 and U5 snRNAs were recently associated with neurodevelopmental disorders (Chen et al. 2024; Greene et al. 2024; Nava et al. 2025), but the RP mutations are the first disease mutations reported in U6 snRNA. There are at least four active copies (paralogs) of the U6 gene in the human genome (Domitrovich and Kunkel 2003), so it was not expected that it could harbor disease mutations, given the redundancy.

U6 snRNA is a key part of the active site of the spliceosome (Fica et al. 2013), which catalyzes intron removal to convert premessenger RNA into mRNA ready for translation into protein. A spliceosome forms on each intron by the ordered assembly of snRNPs and free proteins (Kastner et al. 2019; Wan et al. 2020; Wilkinson et al. 2020). U6 joins the assembling spliceosome extensively base paired to U4 in the U4/U6.U5 tri-snRNP (Fig. 1A, lower right). Catalytic activation of the assembled spliceosome requires displacement of U4 from U6 by the helicase Brr2 (SNRNP200 in humans) under control of the master protein of the spliceosome, Prp8. U6 then base pairs with U2 snRNA to form the catalytic center. After splicing is completed, U2 and U6 snRNAs are separated and U6 snRNP rejoins the U4 snRNP.

FIGURE 1.

(A) Schematic of changes in U6 snRNA secondary structure and RNA interactions during the splicing cycle. The asterisks mark the position of adRP mutations. (ISL) Internal stem–loop. (B) Comparison of wild-type (left) and mutant (right) yeast and human U6 ISLs. Red residues differ between yeast and humans. The green nucleotide is an RP mutation in humans and the purple A is a selected suppressor of the A62G mutation in yeast.

There are two striking features of the adRP mutations in U6 (Quinodoz et al. 2026). First, although found in 52 families so far, there are only two, nearly identical mutations: insertion of a G residue on either side of nucleotide A56. This finding suggests that the mutations alter a specific function of U6, consistent with their dominant phenotype. Second, while each adRP family has a mutation in only one of the four active U6 gene loci, mutations in each of the four loci were found among the 52 families. This finding suggests that the mutant U6 is, on average, one-eighth of the total U6 snRNA in a patient cell, since there are maternal and paternal copies of each locus and only one allele of one locus is mutated. Thus, the adRP mutant U6 phenotype is super dominant.

While only two adRP mutations were also found in U4, one or the other was present in much fewer families (15) and the mutations are of distinct types and positions: either a U > C substitution at position 56 (which pairs with U6-A56 in U4/U6 Stem I) or an insertion of an A residue between G18 and U19. Furthermore, the U4 mutations are found in only one of the two active gene loci, RNU4-2, and this locus is more active in the retina (Quinodoz et al. 2026), so the mutant U4 likely contributes at least 30% of the total U4. All four adRP mutations in U4 and U6 colocalize to the junction between three helices in the U4/U6 di-snRNP and U4/U6.U5 tri-snRNP (Fig. 1A), and all five tri-snRNP proteins mutated in adRP influence the structure of this junction. Prp3, Prp4, Prp8, and Prp31 associate with U4/U6 Stem II and the U4 5′ Stem–loop and Brr2/SNRNP200 catalyzes U4/U6 dissociation during spliceosome activation, starting with U4/U6 Stem I and translocating 3′-to-5′ along U4 snRNA (Kastner et al. 2019; Wan et al. 2020; Wilkinson et al. 2020).

So how do the adRP mutations in U4 and U6 result in progressive photoreceptor cell death? Quinodoz et al. propose they impair U4/U6 di-snRNP assembly, but do not propose a specific mechanism or how this could confer a super dominant phenotype. Mutational studies on yeast U4 and U6 have shown that one must consider the effects of a mutation on all forms of the RNA that occur in the splicing cycle (Fig. 1A). Remarkably, the adRP mutations in human U6 mirror the A62G substitution made decades ago in yeast U6 to test the effect of stabilization of the internal stem–loop (ISL) (Fig. 1B; Fortner et al. 1994). The highly conserved ISL is present in both free U6 snRNP and the catalytic U2/U6 snRNP and competes with formation of U4/U6 base pairs. When A62 in yeast U6 is changed to a G in the sole U6 gene of a haploid yeast cell, it pairs with a C across the ISL and confers cold-sensitive growth, stabilizes the ISL, and inhibits U4/U6 pairing to form the di-snRNP (Fortner et al. 1994). The human U6 adRP mutants likely also form this stable G-C pair by bulging out the A residue before or after the inserted G (Fig. 1B).

More than 100 spontaneous suppressors of the cold-sensitivity of yeast U6-A62G were selected, which yielded 18 unique mutations in the U6 gene, one of which is an insertion of an A residue just before G62 (Fortner et al. 1994). This 61_62insA mutational suppressor of U6-A62G is predicted to alter the U6 ISL in precisely the same way as the human 56_57insG adRP mutation (Fig. 1B). The bulged A residue likely destabilizes the mutant G-C pair, consistent with reversion of the U4/U6 snRNP assembly defect seen in the U6-A62G yeast strain (Fortner et al. 1994). So why would the adRP mutations in human U6 affect U4/U6 snRNP assembly, especially in the relative warmth of the human retina? A likely answer is the greater stability of the extended ISL present in human U6 (Fig. 1A and see below). Also among the yeast U6-A62G-suppressors are 33 unique substitutions in the U6 snRNP protein Prp24 (SART3 in humans) and one in U4 snRNA (Vidaver et al. 1999; Montemayor et al. 2014). All the suppressor substitutions are predicted to destabilize the U6 snRNP and/or stabilize U4/U6 pairing, suggesting U4/U6-A62G snRNP assembly may become rate-limiting for pre-mRNA splicing. Indeed, overexpression of wild-type U4 partially suppresses the cold-sensitivity of yeast U6-A62G (Fortner et al. 1994).

An important difference between the yeast experiments and human cells with a U6 adRP mutation is that in haploid yeast 100% of the U6 snRNA is mutant instead of ∼12% in adRP patients. While it cannot be ruled out that a small reduction in functional U6 snRNA inhibits splicing of a critical photoreceptor pre-mRNA, it seems more likely that the aberrant U6 exerts a dominant negative effect. Photoreceptor cell death could conceivably be due to an apoptotic signal elicited by the accumulation of free U6 and/or U4 snRNPs or a U2/U6-containing complex, if U4/U6 pairing is required to drive U2/U6 disassembly. Indeed, haploid yeast cells that carry the U6-A62G substitution and a destabilizing U4 substitution in Stem II are dead but can be rescued by another substitution in U6 that disrupts binding to Prp24. These cells exhibit little or no U4/U6 di-snRNP and accumulate an unusual U2/U6-containing complex (Burke et al. 2015). Toxic accumulation of a U2/U6 snRNA complex would be consistent with adRP caused by partial loss of function substitutions in the helicase domain of Brr2/SNRNP200 (Zhao et al. 2009), as Brr2 is implicated in unwinding the U2/U6 complex during spliceosome disassembly (Small et al. 2006).

The yeast genetic findings and biochemical studies with the human snRNAs suggest possible therapeutic strategies for patients with the U6 adRP mutations. Despite forming intermolecular base pairs with nearly identical predicted stabilities, the intact human U4/U6 snRNA complex is less stable than yeast U4/U6 (Brow and Guthrie 1988). Deproteinized yeast U4/U6 has a melting temperature over 50°C, while human U4/U6 snRNAs form two complexes, one that dissociates at 30°C–35°C and one that is as stable as yeast U4/U6. The latter was shown to be missing U6 3′-end nucleotides that form a competing intramolecular structure that extends the U6 ISL (Brow and Vidaver 1995). Addition of oligonucleotides that compete with formation of the extended ISL in full-length human U6 drive U4/U6 complex formation in deproteinized human nuclear RNA (Brow and Vidaver 1995). If such oligonucleotides have a similar effect in the context of living cells, they could potentially function as a therapeutic to improve U4/U6 snRNP assembly in patients with adRP due to mutations in the RNA or, perhaps, protein components of the U4/U6.U5 tri-snRNP.

RNA-based therapies for inherited retinal disorders, including adRP, are currently in clinical trials (Girach et al. 2022). The most common method of administration of oligonucleotides is by injection into the vitreous humor of the eyeball. A chemically stabilized RNA oligonucleotide complementary to either strand of the extended ISL of U6 snRNA could potentially enhance U4/U6 snRNP assembly in the presence of an adRP mutation in U6. However, treatment would require repeated injections, possibly every six months (Girach et al. 2022). One-time gene augmentation therapy using an AAV vector is another alternative (Nguyen et al. 2023). Given suppression of the cold-sensitivity of yeast U6-A62G by increased U4 gene dosage (Fortner et al. 1994) or the U4-A16G substitution (Vidaver et al. 1999), providing additional copies of RNU4-2, either wild-type or with U4/U6 Stem II-stabilizing substitutions, could be effective.

Typically, disease associated mutations are first identified in patients and then tested in model organisms to elucidate potential mechanisms and therapies. In the case of the adRP mutations in U6 snRNA, their discovery was anticipated by model organism studies designed to better understand the basic biochemistry of U6 snRNA function. It is unlikely this will be the only such example for mutations in the spliceosome. Indeed, the yeast U4-cs1 cold-sensitive mutation and its suppressors (Li and Brow 1996; Brow 2019) may shed light on the mechanism of the colocalized n.64_65insT mutation in RNU4-2 associated with the autosomal dominant neurodevelopmental disorder ReNU syndrome (De Jonghe et al. 2026).