SPF45/RBM17-dependent splicing and multidrug resistance to cancer chemotherapy

ABSTRACT

The early splicing complex A occupies at least eighty nucleotides of intron, in which U2AF covers the polypyrimidine tract. SPF45 (RBM17) functionally substitutes for U2AF on a subset of short introns. Since SPF45 expression confers resistance to various anticancer drugs, SPF45-dependent splicing may play a critical role in multidrug resistance.

SPF45 is a U2 snRNP-associated spliceosomal protein (reviewed in ref.¹) and it was characterized as a regulator of alternative splicing.² We recently demonstrated that a subset of human short introns are spliced out by SPF45 (RBM17 as HGNC approved symbol), instead of by the known splicing factor, U2AF heterodimer.³ Thus, it is reasonable to assume that SPF45-mediated changes of alternative and constitutive splicing are involved in cancer-related processes, like proliferation, migration, apoptosis and the cell cycle, because SPF45 has been mechanistically linked to anticancer drug resistance. This scenario enlarges the therapeutic space inhabited by splicing in cancer.

Human pre-mRNA introns vary extensively in length, ranging from under fifty to over a million nucleotides (nt). Human pre-mRNA splicing involves dynamic stepwise reactions in a huge protein-RNA complex, termed the ‘spliceosome’, which includes five kinds of small nuclear ribonucleoprotein particles (U1, U2, U4, U5 and U6 snRNPs) and >170 protein factors (reviewed in ref.¹). The essential splicing signal sequences in pre-mRNA – the 5′ splice site, the branch-site sequence, and the poly-pyrimidine tract (PPT) followed by the 3′ splice site – are bound by U1 snRNP, U2 snRNP and U2AF⁶⁵/U2AF³⁵ (U2AF2/U2AF1 as HGNC approved symbol), respectively, which together constitute the spliceosomal A complex. The globular shape of the A complex fully occupies the length of a 79–125 nt single-stranded RNA, which is about two-fold longer than verified short introns (43–56 nt).^4,5 How can these very short introns accommodate the oversize complex of known essential factors? We postulated that there is a distinct alternate splicing mechanism involved in splicing of human short introns.

In our new study,³ we used HNRNPH1 (heterogeneous nuclear ribonucleoprotein H1) pre-mRNA as it has a 56-nt short intron that is spliced out efficiently in HeLa cells.⁴ We began by screening for essential factors that splice out this 56-nt intron from among 154 human nuclear proteins; we downregulated these proteins’ expression in HeLa cells using small interfering RNAs (siRNAs) and analyzed endogenous splicing activity on this 56-nt short intron. The strongest splicing repression was caused by knockdown of SPF45, which had no effect on a control pre-mRNA with a longer, 366-nt, intron. To further inquire if SPF45 is a general splicing factor for a group of short introns, we performed whole-transcriptome sequencing (RNA-Seq) with total RNA prepared from SPF45-knockdown cells. The most frequent splicing changes in these cells were retained introns (i.e. splicing inhibition) of which we identified 187. Remarkably, these SPF45-dependent introns were strongly biased toward shorter lengths, suggesting that SPF45 is an essential splicing factor for a group of pre-mRNAs with short introns.

Next, we searched for the cis-acting element responding to, the trans-acting factor, SPF45. In classical splicing, the PPT sequence and downstream 3′ splice site are required for binding of the U2AF heterodimer (U2AF⁶⁵/U2AF³⁵). Remarkably, when we truncated the PPT, it could no longer be spliced by U2AF but it could now be spliced by SPF45, implying that a short PPT determines SPF45-dependent splicing (Figure 1). This turned out to be the case as knockdown of the U2AF heterodimer significantly repressed the splicing of conventional introns, whereas SPF45-dependent short introns were spliced out rather efficiently. This indicates that SPF45 out-competes U2AF heterodimer on truncated PPTs, and once installed, SPF45 promotes splicing of the short intron. Finally, biochemical analyses and splicing assays with various SPF45-mutant proteins demonstrated that SPF45 is localized at the truncated PPTs via a specific protein domain’s interaction with the U2 snRNP component, SF3b155 (SF3B1 as HGNC approved symbol), to promote SPF45-dependent splicing (Figure 1).

Figure 1.

U2AF-dependent and SPF45-dependent splicing

U2AF-heterodimer complex cannot form stably on a short intron with a truncated poly-pyrimidine tract (PPT), and thus it is replaced by the SPF45 complex (adopted and modified from Ref.³). The P14–SF3b155–U2AF⁶⁵/U2AF³⁵ complex is remodeled to a P14–SF3b155–SPF45 complex by switching of their UHM-ligand motif (ULM)–U2AF-homology motif (UHM) interactions to promote splicing of pre-mRNAs with short introns. RRM denotes RNA-recognition motif. HGNC-approved symbols of SF3b155 and P14 are SF3B1 and SF3B6, respectively.

Previously, SPF45 was reported to function as a regulator of alternative splicing;^2,6 however, SPF45 is also an essential factor for cell survival and maintenance in vivo.⁷ Our study clarifies this enigma by establishing SPF45 as a novel and distinctive constitutive splicing factor in the early spliceosome, i.e., splicing out of a subset of human short introns with truncated PPTs requires SPF45, but not the previously thought essential U2AF heterodimer (Figure 1).

This finding is not only a groundbreaking achievement in terms of basic research, but it also has potential medical applications, as SPF45 expression is linked to pathological characteristics and poor prognosis in cancer patients. SPF45 expression is low in normal tissues but is highly expressed in many types of cancers, including breast, ovarian and prostate.⁸ Overexpression and phosphorylation of SPF45 enhances migration and invasion of ovarian cancer cells, which is critical for cancer metastasis.⁹ SPF45 is overexpressed in hepatocellular carcinoma and glioma, and its downregulation reduces cell proliferation and induces cell cycle arrest and apoptosis.^10,11

Overexpression of SPF45 in cervical and ovarian cancer cells confers resistance to at least six anticancer drugs with different mechanisms of action, and it is notable that SPF45 overexpression does not affect the levels of intracellular drug accumulation, which is a well-documented mechanism of multidrug resistance.^8,12 By RNA-Seq analysis in SPF45-knockdown cells, we identified not only SPF45-dependent retained introns but also changes in many types of alternative splicing.³ Therefore, there is still much work to be done in linking the upstream factors involved in anticancer drug resistance to SPF45-mediated alternative splicing changes. It was demonstrated that overexpression of MYC plays a critical role in chemotherapy-resistant cancer cells.^13,14 Interestingly, our RNA-Seq analysis in SPF45-knockdown cells identified the down-regulation of cancer-related genes including MYC (unpublished data). Understanding the SPF45-targeted mechanisms may aid in development of effective therapeutic interventions.

Funding Statement

K.F. was partly supported by Grants-in-Aid for Scientific Research (C) [Grant number: JP18K07304] from the Japan Society for the Promotion of Science (JSPS), a Research Grant from the Hori Sciences and Arts Foundation, and a Research Grant from Nitto Foundation. A.M. was partly supported by Grants-in-Aid for Scientific Research (B) [Grant number: JP16H04705] and for Challenging Exploratory Research [Grant number: JP16K14659] from JSPS.

Disclosure statement

No potential conflict of interest was reported by the authors.