Splicing Factor Mutations and Disease Phenotype: Searching for a Needle in a Haystack.

T he democratization of genomic and transcriptional profiling has revealed widespread mRNA splicing alterations in cancer,[1-4] responsible for dysfunctional gene splicing that can affect disease initiation, propagation, and treatment response.[5] For instance, in myeloid leukemia including clonal hematopoiesis, splicing factor mutations are initiating events acquired in the most immature hematopoietic compartment, whereas in chronic lymphoid leukemia or breast cancer, mutations in splicing factors are considered secondary hits that contribute to drug treatment resistance.

In clonal hematopoiesis, myeloproliferative neoplasms, acute myeloid leukemia and myelodysplastic syndrome (MDS), core spliceosomal factors are recurrently mutated, and at least 1 out of 2 MDS patients harbor mutations in a splicing factor.[6-8] Most introns are spliced by the major spliceosome and only less than 1% of human introns are spliced by the minor spliceosome. Mutations in the major (U2-type introns) and minor (U12-type introns) spliceosomes include mutations in 4 core spliceosomal factors: SF3B1, SRSF2, the small subunit of the U2AF heterodimer U2AF1, and the component of minor spliceosome, ZRSR2.

One of the major challenges with understanding the role of splicing factors in causing disease is to discern, among the hundreds of misspliced transcripts observed, which are responsible for the disease phenotype.

While SF3B1, SRSF2, and U2AF1 are essential to the major spliceosome, ZRSR2 is the only 1 out of the 4 to be mainly involved in the minor spliceosome functions. Inoue et al[9] exploited this feature of ZRSR2 to shed light on the role of minor intron retentions in driving clonal expansion and disease propagation.

The authors generated mice models with conditional knock out of Zrsr2 restricted to the hematopoietic compartment and surprisingly, contrary to previous models that evaluated the effect of hotspot mutations in splicing factors of the major spliceosome, the loss of Zrsr2 promoted hematopoietic stem cell (HSC) self-renewal. Zrsr2-null HSCs showed enhanced clonogenic capacities in vitro and out-competed Zrsr2 WT cells in vivo. Using eCLIP-sequencing to map RNA binding targets and RNA-sequencing from The Cancer Genome Atlas (TCGA) they showed that only a third of U12-type introns are sensitive to ZRSR2 loss and that these introns are characterized by a 3′ splice site-proximal adenosine branch point that closely resembles the U12-snRNA consensus with a weak or absent polypyrimidine tract. Subsequently a functional genomics screen revealed that LZTR1, a cullin-3 adaptor regulating ubiquitin-mediated suppression of RAS-related GTPases, is the target of the minor spliceosome. Indeed, loss of ZRSR2 impairs LZTR1 minor intron excision and promotes clonal expansion. Interestingly, loss of function of LZTR1 was previously reported in glioblastoma, schwannomatosis and in Noonan Syndrome, a RASopathy. The authors confirmed that LZTR1’s loss of expression happens through activation of the nonsense mRNA decay, in MDS-ZRSR2 mutant bone marrow cells. They also showed in one pedigree with autosomal recessive Noonan syndrome that LZTR1’s mutation occurs in the branch point region of the minor intron and induces a loss of protein expression. Finally, using TCGA they investigated LTZR1’s minor intron splicing across cancers and of note observed alternative splicing of LTZR1 in a substantial number of tumors, even though components of the minor spliceosome were not mutated.

These findings promote a general acceptance that aberrant splicing is a pan-cancer hallmark driving disease progression. It also highlights the benefit of using HSCs and MDS models to characterize splicing factors’ impact on disease propagation not only in hematological malignancies but also across cancers. It is interesting to observe that despite conferring a clear clonal advantage, ZRSR2 is the least frequently mutated of the 4 core spliceosomal factors (SF3B1, SRSF2, U2AF1, and ZRSR2) in MDS. Furthermore, it was intriguing to see that loss of Zrsr2 could rescue impaired clonogenic capacities of Sf3b1 mutant mice, which is in strong contrast with the lethal phenotype of combined Sf3b1 and Srsf2 mutant mice. Indeed, it is rare to identify more than one RNA splicing factor mutation in individual patients.[10]

As our journey in unveiling key splicing events progresses, many questions arise. Minor introns are highly conserved in evolution; however, it is less the case for major introns and mice models struggle to recapitulate key splicing events identified in primary human samples, slowing down our understanding of the disease’s cause. This observation is likely to be due to ribonucleic sequence discrepancies between human and mouse. Besides, splicing events are dependent on the expression of variants, which might vary between lineages and tissues, showing different phenotypes during differentiation. Likewise, it is probable that cumulative missplicing events are necessary to reveal an overt disease phenotype.

Progress in genomics and transcriptomics has revealed widespread mRNA splicing alterations in cancers. We are just at the dawn of understanding how mutations in splicing factors, RNA binding protein or intronic regions can induce alternative splicing with dramatic impacts on cell biology.

Disclosures

The authors have no conflicts of interest to disclose.