MDS is a heterogeneous group of clonal haematopoietic stem cell (HSC) disorders characterized primarily by deficiencies in the myeloid, erythroid, and/or megakaryocytic lineages, mono- or multilineage dysplasia, clonal dominance of abnormal immature cells, and an increased risk of secondary AML. The French, American and British divisions of AML subtypes are categorized as M0 through M7 based on the type of cell the leukaemia develops from and how mature the cells are. The World Health Organization system classification is AML with certain genetic abnormalities. The phenomenon of AML emerging from MDS has prompted numerous researchers to suggest that the two disorders are related, with similar pathogenic mechanisms and treatments (20–24).

Heterozygous somatic mutations in splicing factor genes (SF3B1, SRSF2, U2AF1 and ZRSR2) occur in >50% of all patients with MDS (25–29), with U2AF1 mutations ranging from 7–11% and having the highest incidence in AML and the worst prognosis among the four mutations (30,31). Structural changes caused by mutations in U2AF1 occur in two hotspots-S34 and Q157-in the first and second zinc finger domains, as illustrated in Fig. 3. These mutations are capable of inducing gain-of-function changes, which modify the 3′ SS preference in a sequence-specific manner, determined by the nucleotides surrounding the AG dinucleotide responsible for the 3′ SS core consensus motif (10,15,32–34). This suggests that U2AF1 mutations directly affect RNA binding and consequently influence exon inclusion. It has also been observed that the effects of different mutations may vary on different exons; for example, the U2AF1 S34 mutation prefers C or A more frequently in the-3 position when analysed in AML patients, whereas the Q157 mutation, involving the first zinc finger, shows preferential recognition of G over A in the +1 position. By combining comprehensive data from RNA sequencing and U2AF1 mutations with 3′ SS recognition ability, a clearer understanding of the mechanism by which U2AF1 mutations affect splicing in patients with MDS can be established (10,12,35,36).

Dysregulation of cell growth and immunity is a hallmark of MDS (37). In accordance with several research studies, aberrantly spliced genes identified in MDS with U2AF1 mutations converge on disrupted cellular differentiation processes, including RNA splicing, RNA localization/transport, RNA binding, protein translation processes, ribosomal pathways, and mitochondrial dysfunction (31,38–40). Subsequently, an analysis of upstream transcriptional regulators using the Ingenuity Pathway Analysis (IPA) tool suggested the significant enrichment of genes regulated through transcriptional regulators (RICTOR, E2F1, HNF4A, MYC, MYCN and RB1, the major controllers of cell growth/cell cycle) in the aberrantly spliced genes (38,41,42) (Table I). Additionally, these studies revealed that MDS with U2AF1 mutations has ten significant transcriptional pathways, with the sirtuin signalling pathway leading the rankings (38,43). Sirtuins, including sirtuin1, sirtuin2 and other isoforms, are a class of proteins that regulate diverse physiological processes and are considered to be partially involved in MDS and other haematological malignancies (44–46). Furthermore, mitochondrial dysfunction has generally been associated with MDS pathophysiology. In one study, mice with mitochondrial dysfunction exhibited characteristics of MDS, including macrocytic anaemia (47). Mitochondrial dysfunction can be induced by mitochondrial DNA mutations, electron respiratory chain leakages, increased oxidative stress, dysfunctional TCA cycle enzymes and abnormal tumour suppressive signalling pathways (48). Park et al (49) reported that cells expressing the U2AF1 S34F gene had defective autophagy along with mitochondrial dysfunction and increased production of reactive oxygen species, causing genomic instability and an elevated frequency of spontaneous mutations. These sirtuins in the mitochondria, which additionally function as histone deacetylases, are involved in metabolism, inflammation, genome stability and cell proliferation control. In addition, their role in ageing, cancer and survival has been demonstrated in some studies (50,51).

The NF-κB signalling pathway has a significant role in determining the quiescent or active state of HSCs, and aberrantly active NF-κB signalling may promote malignant HSPCs (52). Balka et al (53) identified that U2AF1 mutations resulted in aberrant splicing of IRAK4, a serine/threonine kinase involved in toll-like receptor downstream signalling, giving rise to a longer isoform named IRAK4-L, which preferentially assembles signal transducers (for example, MyD88 and IRAK kinases) into a myddosome, causing a higher degree of NF-κB activation. Moreover, IRAK4-L participated in innate immunity and was found to be associated with a poorer prognosis. Subsequent studies also revealed that mutations in SF3B1 and SRSF2 had a similar effect, leading to hyperactive NF-κB signals via aberrant splicing of MAP3K7 and CASP8, respectively (19,54–56) (Fig. 4). Overall, the aforementioned evidence suggested that mutations in various splicing factors alter different targets at the pre-mRNA splicing level but share a common downstream signalling node leading to heightened innate immune signalling.

The dysregulation of splicing factor genes is associated with the formation of R loops, a pathogenic aberrant RNA-DNA hybrid structure derived from nascent invasion. Crossley et al (57) showed that the expression of U2AF1 S34F mutations in human cancer cells led to R-loop accumulation. Furthermore, treatment of U2AF1 S34F-expressing cells with ATR inhibitors has been identified to induce DNA damage and cell death, which can be enhanced by spliceosome inhibitors (58,59). Chen et al (60) and Pellagatti et al (38) reported ERCC8 and FANCM mutations in MDS patients with U2AF1 mutations that may be involved in the suppression/regulation of R-loop formation and the DNA damage response. Chen et al (61) found that impaired transcription inhibits release at transcription start sites (TSSs), likely due to the inability to extract P-TEFb, a kinase that phosphorylates the RNA polymerase II (Pol II) C-terminal domain, from the inhibitory 7SK complex to transcriptionally engaged Pol II at TSSs in SRSF2-mutant MDS, resulting in enhanced R loops. Interestingly, Pellagatti et al (38) and Boultwood et al (62) identified aberrant splicing of several genes, including SETX and ATR, involved in the suppression/regulation of R-loop formation in bone marrow CD34⁺ cells of MDS patients with splicing factor mutations. Collectively, these data raise the possibility that R-loop-induced DNA damage may lead to deleterious mutations that promote clonal advantage in MDS pathogenesis.

PMF is classified as a chronic myeloproliferative neoplasm characterized by bone marrow tissue fibrosis, enlargement of the spleen, and the occurrence of anaemia associated with the presence of nucleated and teardrop-shaped red blood cells. In PMF, the osseous components of bone marrow undergo a process known as osteosclerosis. Furthermore, fibroblasts secrete collagen and reticulin proteins, collectively referred to as fibrosis. Both of these pathological processes disrupt the usual functioning of bone marrow, leading to reduced production of various blood cells, including erythrocytes (red cells), granulocytes and megakaryocytes; the latter cells are responsible for platelet production. Thus, the combined effects of osteosclerosis and fibrosis in PMF significantly impair the normal haematopoietic activities of bone marrow (63–65).

Most cases of PMF are associated with mutations in the JAK2, CALR or MPL genes (Table I) (66). According to sequencing results, >15–20% of patients with PMF suffer from U2AF1 mutations, of which the two major mutation sites are Q157 and S34. These findings contrast with MDS, in which the incidence of S34 and Q157 mutations is low. Interestingly, the age distribution of patients with U2AF1 mutations is also different in MDS and PMF, with MDS patients tending to be younger and PMF patients tending to be older (67). In terms of clinical features, U2AF1 mutations are associated with anaemia, thrombocytopenia and poor disease prognosis. In MDS, U2AF1S34 mutations are more specifically associated with thrombocytopenia, while anaemia is more specifically associated with Q157 mutations (Fig. 3). Both types of mutations in PMF may cause anaemia, and Q157 mutations are more strongly associated with thrombocytopenia (68–70).

CMML is a haematopoietic disorder that involves clonal proliferation of cells, presenting with diverse myeloproliferative and myelodysplastic characteristics. This complexity contributes to the heterogeneous nature of CMML. The tyrosine kinase inhibitor imatinib remains the primary targeted drug for the treatment of CMML, thus rendering it inevitable that resistance to this inhibitor will decrease its efficacy (71,72).

An analysis of a large cohort of patients with CMML revealed that the occurrence rate of U2AF1 mutations is ~9%. Patnaik et al (73) reported that the primary sites of U2AF1 mutations are S34F, Q157P/R and R156H. Among patients with U2AF1 mutations, >50% carried the S34F mutation, while a subset of patients exhibited the Q157P and Q157R mutations, and only a minority had the R158H mutation (Fig. 3). Additionally, it has been demonstrated that other gene mutations, such as ASXL1, TET2, CBL, RAS, EZH2, TP53, IDH1/2, DNMT3A, RUNX1 and UTX, interfere with spliceosome genes and induce distinct amino acid missense mutations in U2AF1 (73–75). These insights suggest that targeting U2AF1 through drug design may hold promise for improving survival rates among patients with U2AF1 mutations.

HCL is a rare B-cell malignancy characterized by pancytopenia and splenomegaly, with an incidence of ~2% in adults. It is generally recognized as a subtype of chronic lymphocytic leukaemia and can be divided into two categories: Variant HCL (vHCL) and classical HCL (cHCL). vHCL is more prevalent, accounting for nearly 10% of HCL patients, whereas cHCL is rare among mature B-cell malignancies (76).

The presence of various markers, such as CD29, CD20, CD22, CD25, tartrate-resistant acid phosphatase (TRAP), annexin 1A (Anx-A1) and BRAF V600E mutations has been documented in the immune response of leukaemia cells and summarized in Table I. In the case of vHCL, several similarities with cHCL are notable, yet vHCL lacks expression of CD25, CD123, ANXA1, TRAP and BRAF V600E. Additionally, vHCL has been found to be more aggressive and less responsive to purine analogues.

vHCL exhibits several similarities to HCL. However, vHCL lacks the expression of CD25, CD123, ANXA1, TRAP and BRAF V600E. Furthermore, vHCL has been observed to be more aggressive and less purine analogue-sensitive (77). By detecting vHCL patients, Durham et al (78) found that the spliceosome U2AF1 was mutated, leading to its loss of function, and the major mutation sites were S34F and R156H, as demonstrated in Fig. 3. This is in contrast to cHCL, which only included BRAF V600E mutations. Thus, there is an important genetic difference between cHCL and vHCL. The aforementioned research provides evidence for potential novel therapeutic targets for vHCL, which may have implications for the future personalized treatment of the disease.

As revealed in separate studies, there are recurrent U2AF1 mutations in LC, especially in lung adenocarcinoma (LUAD). U2AF1 mutations have been observed in ~3% of patients with LUAD, and the incidence of U2AF1 mutations in other LC subtypes appears to be lower, including squamous cell carcinoma and small cell carcinoma. In contrast to the aforementioned neoplastic disorders, the S34F mutation is the most pervasive hotspot in LC targeting the first zinc finger domain. Additionally, this mutation has been linked to altered splicing of genes such as GNAS, H2AFY, STRAP and PICALM in both LUAD and non-small cell LC patients. Based on an analysis of A549 cells, U2AF1 S34F mutations can lead to aberrantly spliced DNA damage repair factors (32,79).

In terms of pathogenesis, Esfahani et al (32) reported that U2AF1 S34F preferentially binds to and regulates splicing of introns containing CAG trinucleotides at their 3′ splice junctions compared with its wild-type counterpart. Moreover, overexpression of U2AF1 S34F upregulates the expression of genes associated with epithelial-mesenchymal transition, thereby inducing cell invasion, which is mediated by preferential splicing of SLC34A2-ROS1 long isoforms (Table I) (32,80). These studies have demonstrated that in lung tumours harbouring U2AF1 S34F, a wild-type allele of the gene is invariably retained and is critically dependent on cell survival. However, a previous in vitro study observed that in cells harbouring multiple U2AF1 alleles, a mutation in one allele stimulates the emergence of the U2AF1 S34F mutation, leading to the development of U2AF1 ‘wild-type’ cells. Consequently, these cells exhaust their wild-type alleles and give rise to isogenic U2AF1 mutant cells (80). This means that mutations in U2AF1 appear to be an important contributing factor in the formation of LC. Targeted therapies directed at U2AF1 mutations could represent a promising clinical option in the treatment of LC.

Prostate cancer is the second most commonly occurring cancer in men. A recent study identified U2AF1 mutations in prostate tumours, suggesting that this may be a putative biosignature for this disease (81). According to a recent study, downregulation of U2AF1 has been linked to poor prognosis and a correlation with androgen receptor variant 7 (ARV7). This downregulation can stimulate the proliferation of prostate cancer cells and lead to bicalutamide resistance by regulating intron splicing of ARV7. MAPK pathways have also been identified as having an administrative impact on U2AF1 and its regulation of ARV7 splicing (Table I). In summary, downregulation of U2AF1 can cause androgen resistance and increased proliferation of prostate cancer cells through the regulation of ARV7 splicing and an increase in MAPK1 expression, predicting a poor prognosis in patients with this cancer (82). A further study based on prostate tumour biopsies suggested, however, the overexpression of U2AF1 genes in association with EZH2 localized in mRNA and MAPK pathways (83).

Borderline ovarian tumours account for ~15–20% of all ovarian malignancies and arise from the tissue that encapsulates the ovary. However, in clinical practice, it is difficult to detect borderline mucinous tumours until advanced stages. In a case study of one patient with a borderline mucinous tumour, a novel mutation of U2AF1 was observed, with an alteration in amino acid serine in place of tyrosine, although such mutations are uncommon in ovarian tumours (occurring in <0.2% of cases) (84). On the other hand, TP53 mutations remain highly common in ovarian cancers. Moreover, U2AF1 and TP53 mutations can predict poor prognosis and a lack of remission. Consequently, inhibitors targeting U2AF1 are essential for these patients.

The small-molecule inhibitor NSC 194308 was synthesized by combining metallic Na and dry MeOH under an exothermic reaction with the evolution of hydrogen. Subsequently, (−)-cis-myrtanylamine was added and incubated for 16 h at room temperature before being reacted with isopropyl alcohol (iPrOH) and diethyl ether to form 2-chloro-N-(−)-cis-myrtanyl acetamidinium hydrochloride. The solution was stirred in ethyl (Et) alcohol and combined with Na2S2O3 in water R, and the compound's formation was monitored using thin layer chromatography (85). As a result, NSC 194308 was successfully synthesized and characterized as demonstrated in Fig. 5.

The experiments of NSC 194308 in HeLa cells revealed that its addition directly reduced the splicing of the adenovirus major late promoter transcript substrate. An in-depth analysis of the intermediate stages of spliceosome assembly revealed that NSC 194308 repressed the pre-mRNA splicing process by inhibiting the gathering of spliceosomes at a U2AF-dependent checkpoint before tri-snRNP enlistment and catalytic activity. Subsequent chemical synthetic treatment with NSC194308 in K562 leukaemia cells, which had S34F mutations of U2AF1, was followed by high cell apoptosis (85). Collectively, these data suggested that NSC194308 is a promising agent for targeted therapy of leukaemia with U2AF1 S34F and other related mutations.

The Julia-Kocienski olefination is of paramount importance in regard to the synthesis of sudemycin D6. This form is derived from natural product spliceosome modulators (FR901464) through intricate chemical reactions, culminating in the formation of the (S,Z)-5-(((1R,4R)-4-((2E,4E)-5-((3R,5S)-7,7-dimethyl-1,6-dioxaspiro[2.5]octan-5-yl)-3-methylpenta-2,4-dien-1-yl)cyclohexyl)amino)-5-oxopent-3-en-2-yl-methylcarbamate structure of sudemycin D6. Notably, the biocatalytic synthesis of sudemycin D6 was first achieved using the Julia-Kocienski olefination methodology. In this way, the Julia-Kocienski olefination is a pivotal step in the successful synthesis of sudemycin D6 (Fig. 5) (86).

Haematopoietic progenitor cells expressing CD34 were shown to be sensitive to sudemycin D6 treatment, resulting in altered pre-mRNA splicing and reduced survival of K562 cells with U2AF1 (S34F) mutants. The effects of this alteration on the cell cycle profile were examined in vivo using transplanted mice treated with sudemycin D6. The results demonstrated that expansion of U2AF1 mutant progenitors was attenuated, and a beneficial effect on the bone marrow of U2AF1-mutant mice was observed. Notable improvements were observed in biological pathways related to the immune response, leukocyte differentiation and cell proliferation regulation (87). The aforementioned study provided evidence that sudemycin D6 is a promising therapeutic strategy for targeting U2AF1 mutants and may have broader implications for the treatment of leukaemia.

ATR is an essential protein-coding gene activated by DNA damage and DNA replication stress. The two known ATR inhibitors are VX-970 and AZD6738. VX-970, also known as berzosertib, has been used in clinical trials of patients with ovarian serous tumours, especially in U2AF1-mutant cancers. Its chemical name is 3-[3-[4-(methylaminomethyl) phenyl]-5-isoxazolyl]-5-(4-propan-2-ylsulfonylphenyl)-2-pyrazinamine. Furthermore, compound AZD 6738 (named ceralasertib) is a potent and effective sulfoximine morpholinopyrimidine ATR inhibitor. The detailed chemical structure of AZD 6738 is 4-(4-(1-((S(R))-S-methylsulfonimidoyl)cyclopropyl)-6-((3R)-3-methyl-4-morpholinyl)-2-pyrimidinyl)-1H-pyrrolo(2,3-b)pyridine (Fig. 5). Moreover, AZD6738 can inhibit ATR through the loss of downstream phosphorylation of CHK1 at Ser345 (88).

Adding ATRi to U2AF1S34F-expressing HeLa and OCI-AML3 cells demonstrated an increased sensitivity to inhibit cell growth by inducing DNA damage. Furthermore, HeLa cells incubated with splicing modulators have been shown to generate a dependency on ATR to counteract DNA damage, and the modulation of RNA splicing can make tumour cells more susceptible to ATR inhibitors. Human primary CD34⁺ haematopoietic cells exposed to ATRi exhibit inhibited viability in a dose-dependent manner. The findings of Nguyen et al (59) revealed that the expression of U2AF1S34F in human primary CD34⁺ haematopoietic cells induces a dramatic decrease in R-loop formation, making them vulnerable to ATR inhibition. These findings should be further investigated and validated in larger clinical trials.

A study on U2AF1 mutation and IRAK4 isoforms indicated that transfecting U2AF1 (S34F) into HEK293 IRAK4exon4 cells generated an increase in IRAK4 exon 4 products from the wild-type AAG splicing reporter compared with wild-type U2AF1. However, it did not affect IRAK4 exon 4 from the mutant TAG splicing reporter. This mutant U2AF1 was found to induce the expression of IRAK4, which plays a role in the activation of MAPK and NF-κB signalling to mediate the innate immune pathway (54). It was therefore suggested that inhibiting IRK4 activity may be a useful therapeutic strategy for diseases resulting from mutations in U2AF1 S34F.

IRAK1/4-Inh and CA-4948 are two distinct inhibitors targeting IRAK4. Structurally, the chemical compound of IRAK1/4-Inh is N-[1-(2-morpholin-4-ylethyl)-benzimidaz-zol-2-yl]-3-nitrobenzamide, while the clinical name of CA-4948 is emavusertib, formally named 6′-amino-N-[2-(4-morpholinyl) oxazolo [4, 5-b] pyridin-6-yl]-[2, 3′-bipyridine]-6-carboxamide (Fig. 5). The efficacy of both inhibitors was studied in K562 cells expressing U2AF1 S34F mutations. Cells incubated with IRAK1/4-Inh exhibited inhibited growth, which was dependent upon the deactivation of NF-κB signalling, while primary cells derived from AML or MDS patients with U2AF1 mutations incubated with CA-4948 underwent myeloid and erythroid differentiation. Taken together, these results strongly suggest that U2AF1 (S34F) malignant haematopoietic cells are sensitive to concentrations of IRAK4 inhibitors (89). Thus, inhibiting IRAK4 may be effective in diseases linked to U2AF1 S34F mutations.

At present, only a few relevant U2AF1 clinical studies have been included in clinical trials, namely, NCT04447651, NCT04278768 and NCT02060409. Of these clinical trials, NCT04278768 was carried out in Phase 1 and Phase 2 for different trial diseases, while NCT02060409 was completed in clinical trials (Table II). These studies provide a promising outlook for the clinical application of U2AF1 in the future. Through rigorous study design and appropriate controls, this treatment may bring new opportunities for both diagnosis and therapy.

NCT04447651 aimed to explore the use of spliceosome mutational markers (PRISMMs) in the treatment of patients with metastatic solid tumours and haematological malignancies or lymphoma who may benefit from immunotherapy agents. NCT04278768 was a dose alteration trial of the monotherapy combination of CA-4948 and venetoclax or azacitidine in the treatment of AML or MDS patients with mutations in U2AF1, SRSF2, ZRSR2 and SF3B1. As a response to the increasing use of hypomethylating agents in MDS treatment, NCT02060409 investigated the prognostic impacts of U2AF1, SRSF, and ZRSR2 mutations after treatment with decitabine. Unfortunately, consistent prognostic relevance could not be established in this trial (90).