Targeting Poly(U) Binding Splicing Factor 60 (PUF60): A Small-Molecule Inhibitor Shows Anti-Leukemic Activity and Impacts Cell Cycle in Leukemia Models

The emerging role of altered RNA splicing in the development of myeloid neoplasms including MDS and AML has received wide attention. Facilitating genes, including the frequently mutated splicing factors, SF3B1, U2AF1, and SRSF2, have been reported to play a key role in leukemogenesis. Pharmacological intervention targeting cells harboring these mutations has been pursued, but limited agents have been reported. Rare events of more than one hit in splicing factors were found in MDS and AML patients' samples. This suggests that inhibiting another target in the splicing process may overcome the threshold of survival in these vulnerable cells carrying splicing factor mutations and induce toxicity. The concept has been demonstrated in the case of leukemia cells carrying U2AF1 mutations being vulnerable to inhibitors modulating wild-type SF3B1 function. The primary function of U2AF1 is to form a complex with U2AF2 for 3‘ splice site (SS) definition in RNA splicing and lower frequency U2AF2 mutations are also detected in MDS patients, underlying the significance of dysfunctional 3‘ SS definition as a potential facilitator of leukemogenesis.

Here, we hypothesize that compounds inhibiting proteins participating in 3' SS selection may disrupt isoform patterns in leukemia cells. These agents may act alone or serve as a second hit in the RNA processing to invoke additional vulnerabilities to leukemia cells carrying the most frequent splicing factor mutations.

Definition of 3' SS by U2AF2/U2AF1 is mediated by the recognition of the polypyridine tract (PPT) by U2AF2 and the binding of U2AF1 to the conserved AG dinucleotides at the 3' SS. Like U2AF2, PUF60 contains two zinc-finger domains, recognizing the polyuridine tract preceding 3' SS, and a U2 homology motif (UHM) domain that engages in protein-protein interaction with partner proteins including U2AF2, SF1, and SF3B1. Different from the U2AF1/U2AF2 complex, PUF60 does not have a protein domain to bind to the exon intron boundary junction at the 3' SS. PUF60 has been reported to enhance the splicing activity of U2AF1/U2AF2, but also directly regulates alternative splicing of a subset of genes. A recent finding indicated that knockout of PUF60 leads to alternative splicing switching of CDC25 from isoform A to isoform C to induce cell cycle arrest at G2/M in solid tumors. Leukemia patients carrying PUF60 overexpression have also been found to have less progression free survival (https://ualcan.path.uab.edu/cgi-bin/TCGA-survival1.pl?genenam=PUF60&ctype=LAML) compared to patients with low expression of PUF60.

To support our hypothesis, we undertook an inhibitor development campaign to discover small-molecule compounds selectively targeting PUF60. Based on a relatively less selective and weak inhibitor to the UHM domain, SF-153, we conducted chemical modifications and obtained a PUF60 inhibitor, SF2-69, with >15-fold selectivity against U2AF1, RBM39, SPF45, and U2AF2. We found SF2-69 had an IC50 value of ~3 mM in NKM-1 and K562 leukemia cell lines. In the cell cycle analysis, we found SF2-69 increased S and subG1 cell population in NKM-1 at 24h, but induced G1 arrest in K562 at 48h in a dose dependent manner. In comparison, E7820, an RBM39 degrader, increased G2/M phase in both NKM-1 and K562. RNA-seq analysis of NKM-1 treated with SF2-69 after 24h revealed that SF2-69 upregulated a set of genes participating in cholesterol biosynthesis, NME1-NME2 (a tumor suppressor kinase) and downregulated PLK1, a kinase that phosphorylates and activates CDC25C in G2/M transition. Because K562 is p53-null and NKM-1 has wild-type p53, SF2-69 likely induced p53-dependent and p53-independent responses in NKM-1 and K562 cells, respectively.

In summary, SF2-69 is a new selective PUF60 inhibitor which disrupts cell cycle progression in leukemia cell lines by modulating p53-mediated responses. Upregulation of cholesterol biosynthesis may be a feedback loop to counteract the inhibition of PUF60 by SF2-69. Further optimization of SF2-69 will allow us to develop more effective and selective chemical probes for studying the role of PUF60 overexpression in leukemia and the potential use of PUF60 inhibitors in MDS and AML cells carrying splicing factor mutations.

No relevant conflicts of interest to declare.