Abstract
Recurrent somatic mutations in the key spliceosome component, SF3B1, have been identified at various frequencies across several cancer types. The most common hotspot mutation is the K700E missense mutation, and while its effects on splicing have been well characterised at the molecular level, the mis-spliced genes that contribute to cancer progression and/or dictate responses to therapy are still unclear. Here, we used we use cell line modelling to assess the impact of the SF3B1K700E mutation on the cellular response to various apoptosis-inducing agents. Our data suggest that the SF3B1K700E mutation leads to reduced cFLIP levels, along with defects in the splicing and translation of BCL2, causing a shift in the balance of pro- and anti-apoptotic genes and proteins, which confers greater sensitivity to the bivalent SMAC mimetic, BV-6. As such, BV-6 may represent a therapeutic opportunity for patients with SF3B1 mutant cancers.
Introduction
The spliceosome is a multi-megadalton ribonucleoprotein complex that catalyses nuclear RNA processing and maturation [1]. This involves the removal of intronic sequences from precursor messenger RNA (mRNA), followed by the ligation of remaining exons to form mature protein-coding mRNA transcripts. Given the crucial role of the splicing machinery in the processing of mRNA transcripts, most of the genes encoding components of the spliceosome are intolerant to loss-of-function mutations [2]. Instead, many human cancers exhibit abnormal splicing patterns due to the altered functionality of mutated splicing factors.
SF3B1 (splicing factor 3b subunit 1), which aids in the recognition of consensus sequences in introns to promote efficient mRNA splicing, is the most frequently mutated splicing factor in cancer, with particular enrichment in various haematological malignancies—including acute myeloid leukaemia [3], myelodysplastic syndromes and chronic lymphocytic leukaemia [4], as well as uveal melanomas [5, 6] and other solid tumours such as skin [7] and breast cancers [8]. SF3B1 mutations are almost exclusively missense mutations that alter the ability of SF3B1 to recognise 3’ splice sites, often leading to cryptic splice site usage that can result in disruption of target gene open reading frames. Approximately half of the aberrant transcripts produced are subject to nonsense-mediated decay (NMD), resulting in reduced transcript levels and protein expression [9].
While the connection between SF3B1 mutations and their effects on splicing has been well-characterised at the molecular level [10], the precise ways in which the dysregulated gene expression influences carcinogenesis or affects response to treatment remain unclear. Nonetheless, the ability to therapeutically target SF3B1 mutant cells would be clinically useful for the treatment of multiple cancer types. To address this, we utilised our previously generated isogenic SF3B1K700E cell line model [11], and screened for apoptosis-inducing compounds with selective activity in SF3B1K700E cells, identifying the bivalent SMAC mimetic, BV-6. Our data suggest that the SF3B1K700E mutation causes a subtle shift in the balance of pro- and anti-apoptotic transcripts and proteins—particularly cFLIP and BCL-2, resulting in apoptotic priming and resulting in increased sensitivity to BV-6. As such, BV-6 may represent a therapeutic opportunity for patients with SF3B1 mutant cancers.
Results
The cancer-associated SF3B1K700E mutation confers enhanced sensitivity to the SMAC mimetic BV-6
SF3B1 mutations change how the spliceosome component recognises 3’ splice sites, resulting in cryptic splice site usage. This often leads to exon skipping or intron retention that can interfere with the open-reading-frame of target genes, culminating in disrupted gene expression [9, 11]. With this in mind, we performed GSEA on normalised gene counts derived from the sequencing of total RNA extracted from SF3B1WT and SF3B1K700E K-562 cells. GSEA identified ‘Apoptosis’ as an enriched pathway/hallmark associated with the SF3B1K700E mutation (Fig. 1A). Consequently, sensitivity to 89 apoptosis-inducing agents (Supplementary Table 1) was assessed in our isogenic K-562 cell model to investigate whether the SF3B1K700E mutation conferred any therapeutic vulnerabilities. Cell viability was assessed using CellTox Green after 72 h with Z-score analysis identifying two compounds of interest, BV-6 and MI-2, which preferentially targeted SF3B1K700E cells (Fig. 1B). Validation of the two “hits” was carried out with CellTiter-Glo (Fig. 1C, D).
A Gene set enrichment analysis (GSEA) plot depicting enrichment of the Apoptosis Hallmark SF3B1K700E cells (n = 3). NES normalised enrichment score, FDR false discovery rate. B Z-scores plotted for each of the 89 apoptosis-inducing agents screened against our SF3B1WT (red) and SF3B1K700E (black) cells. Dose-responses following a 72 h incubation with (C) BV-6 in SF3B1 wild-type (SF3B1WT) and mutant (SF3B1K700E) K562 cells. (Mean ± SEM; n = 3; 2-way ANOVA with Šidák’s multiple comparisons) and (D) MI-2 (Mean ± SEM; n = 2; 2-way ANOVA with Šidák’s multiple comparisons). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
BV-6 is a synthetic small-molecule peptide that mimics second mitochondria-derived activator of caspases (SMAC) or direct inhibitor of apoptosis (IAP)-binding protein with low pI (DIABLO), as it is otherwise known. SMAC is a mitochondrial protein that is released into the cytosol where it binds to IAPs and thereby shifts the cells' homeostatic balance in favour of caspase activation and programmed cell death [12]. In line with our screening data (Fig. 1B), SF3B1K700E cells proved to be significantly more sensitive to the bivalent SMAC mimetic (SM), BV-6, when compared to their WT counterparts (Figs. 1C and S1A, C). To determine whether the observed SF3B1K700E-mediated sensitivity was specific to BV-6, we tested another bivalent SM–Birinapant in our isogenic model, which also exerted SF3B1K700E-specific cytotoxic effects (Fig. S1B, D). This suggests that the SF3B1K700E mutation primes cells for SM-induced cell death.
Similarly, a lung mesothelioma-derived cell line (H-2595), a pancreatic cancer cell line (Panc05.04) and an isogenic ALL model (Nalm-6), all harbouring the SF3B1K700E mutation also exhibited enhanced sensitivity to BV-6 when compared to matched SF3B1WT lines (Fig. S1E–G), demonstrating that SF3B1K700E-mediated sensitivity to BV-6 is not cell-line-specific.
MI-2 induces specific chromatin changes to disrupt the interaction between menin and MLL1, thereby blocking MLL fusion protein-mediated leukaemic transformation. In contrast to BV-6 (Fig. 1C), MI-2 exerted cytostatic effects on the SF3B1K700E cells while enhancing proliferation in the SF3B1WT cells (Fig. 1D). Consequently, only BV-6 was further investigated.
To confirm that BV-6 inhibits both cellular IAP (cIAP) and X-linked IAP (XIAP), expression levels of these proteins were assessed following treatment with BV-6 (Fig. S1H). As expected, cIAP1 and XIAP were degraded in a dose-dependent manner, and more so in the K700E cells (Fig. S1I, J). This is unsurprising given that cIAPs and XIAP are constitutively organised and stabilised in heteromeric complexes [13]. cIAP2 was undetectable in our K-562 cells with two independent antibodies (data not shown).
BV-6-induced cytotoxicity is dependent on autocrine TNFα production
To help ascertain the mechanism of BV-6-induced cell death, the transcriptomic profiles of baseline and BV-6-treated SF3B1WT and SF3B1K700E cells were assessed. GSEA identified TNF-α signalling via NF-κB to be of interest, with positive enrichment of the pathway in BV-6-treated SF3B1K700E cells (Fig. S2A). In line with this, BV-6 treatment upregulated TNF, NFKB1, RELA, NFKB2, and RELB, with greater impact in the SF3B1K700E cells compared to SF3B1WT cells (Fig. 2A). Fractionated Western blot analysis of NF-κB1 and NF-κB2 proteins provided orthogonal validation of upregulated NF-κB signalling and shows greater contribution of the non-canonical NF-κB signalling pathway in SF3B1K700E cells (Fig. S2B), with clear translocation of RelB into the nucleus of our SF3B1 mutant model. To determine whether altered TNFR1 surface expression could account for enrichment of the TNF-α signalling pathway, TNFR1 protein levels were assessed via flow cytometry but were not significantly altered in SF3B1K700E cells (Figs. 2B and S3A, B).
A Heatmap of the mean normalised counts of selected genes from within the TNF-α signalling and NF-κB pathways for SF3B1WT and SF3B1K700E cells at baseline and following a 48 h incubation with 1 μM BV-6 (n = 3). B Characterisation of TNFR1 surface expression for SF3B1WT and SF3B1K700E cell lines, represented as the percentage of AF488-positive cells, with non-specific signal ascertained using an IgG1 isotype control. C, D Autocrine (soluble) TNFα levels within the cell culture media of SF3B1WT and SF3B1K700E cells, following (C) 24- and (D) 48-h incubations with 0.1% DMSO vehicle or 1 μM BV-6 (Mean ± SEM; n = 3; 2-way ANOVA with Tukey’s multiple comparisons). E Viability assessment of SF3B1WT and SF3B1K700E cells pre-treated with vehicle alone or 1 μg/mL TNFα neutralising antibody, followed by a 48 h incubation with 1 μM BV-6 (Mean ± SEM; n = 3; 2-way ANOVA with Šidák’s multiple comparisons). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
Typically, cells can be grouped into three general classes based on their response to SMs. Class-I cells upregulate autocrine TNF-α production and cell death following exposure to SMs [14,15,16]. Class-II cells exhibit increased autocrine TNF-α production in response to SM treatment, but the cell population is otherwise largely unaffected, and in class-III cells, SMs have no effect on autocrine TNF-α production or on cell death [17]. Furthermore, it has been proposed that the extent of autocrine TNF-α induction and the relative levels of pro- and anti-apoptotic proteins may distinguish class-I from class-II cells [17]. Based on this, we assessed secreted TNF-α levels in the media of cells treated with BV-6 and observed significant dose-dependent upregulation of autocrine TNF-α in the K562 and pancreatic SF3B1K700E models, following BV-6 treatment (Figs. 2C, D and S3C). Interestingly, while the NALM-6 SF3B1K700E model experienced significantly more TNFα secretion when untreated, both these cells and the H2595 SF3B1K700E line, did not experience any dose-dependent increase in TNFα secretion upon BV-6 treatment (Fig. S3C). This suggests that these cells may be more reliant on another death receptor ligand (TRAIL or FasL) to induce programmed cell death.
To provide conclusive evidence for whether TNF-α is required by our K562 and pancreatic SF3B1K700E models for BV-6-induced cytotoxicity, cells were pre-treated with a TNF-α-neutralising antibody prior to BV-6 treatment. This resulted in significant rescue of BV-6-induced cell death, which is particularly evident in the SF3B1K700E cells (Fig. 2E). Taken together, these data indicate that BV-6-induced cytotoxicity involves the interplay of the TNF superfamily death receptor ligands, with greater sensitivity always observed in the SF3B1K700E models.
BV-6 induces programmed cell death via a RIP(K1)-dependent extrinsic apoptotic pathway
Pharmacological inhibitors of programmed cell death were utilised to establish the mechanism(s) underlying BV-6-induced cytotoxicity. Given the association of SF3B1 mutations with abnormal splicing of genes involved in iron metabolism in haematopoietic cells [18, 19], we pre-treated the cells with ferrostatin-1 but found no significant contribution of ferroptosis to BV-6-induced cell death (Fig. S4A). However, pre-treatments with necrostatin-1, which blocks the kinase activity of RIP(K1), and the pan-caspase inhibitor, z-VAD-FMK, both conferred significant protection against BV-6-induced cell death, particularly in SF3B1K700E cells (Fig. 3A). In contrast, pre-treatment with another necroptosis inhibitor, necrosulfonamide, which blocks the Mixed Lineage Kinase ___domain-Like (MLKL) protein, was unable to rescue cell viability (Fig. S4A). To examine this further, we blotted for components of the necro(pto)some complex but could not detect the presence of RIPK3 (data not shown), which is essential for necroptosis [20]. Combined, this suggests an inability of K-562 cells to undergo necroptosis, and instead points towards the induction of RIP(K1) kinase activity-dependent apoptosis by BV-6.
A Viability assessment of SF3B1WT and SF3B1K700E cells pre-treated for 2 h with DMSO vehicle control, 10 μM necrostatin-1, or 10 μM z-VAD-FMK, followed by a 48 h incubation with 0.1% DMSO vehicle or 1 μM BV-6 (Mean ± SEM; n = 3; 2-way ANOVA with Dunnett’s multiple comparisons). B Caspase-8 activity following 24 h treatment with 0.1% DMSO vehicle or 1 μM BV-6 in SF3B1WT and SF3B1K700E cells (Mean ± SEM; n = 3; 2-way ANOVA with Tukey’s multiple comparisons). C Representative Western blot analysis of cFLIP and RIP(K1) protein expression in SF3B1WT and SF3B1K700E cells following treatment with the indicated concentrations of BV-6 for 24 h. D Co-immunoprecipitation assay of RIP(K1), pro-caspase-8, cFLIP and FADD following immunoprecipitation of caspase-8 from SF3B1WT and SF3B1K700E cells treated with the indicated concentrations of BV-6 for 24 h in the presence of 20 μM z-VAD-FMK. E Activity levels of caspases-3/7 in SF3B1WT and SF3B1K700E cells following BV-6 treatment (Mean ± SEM; n = 3; 2-way ANOVA with Tukey’s multiple comparisons). F Proportion of Annexin V+ cells following a 48 h incubation with the indicated concentrations of BV-6 in SF3B1WT and SF3B1K700E cells (Mean ± SEM; n = 4; 2-way ANOVA with Tukey’s multiple comparisons to compare the total proportions of apoptotic cells, i.e., combined Annexin V+/7-AAD+ and Annexin V+/7-AAD− cells). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
As caspase-8 is one of the initiator caspases associated with the extrinsic apoptotic pathway, we next assessed caspase-8 activity within both SF3B1WT and SF3B1K700E cells following BV-6 treatment, which showed significantly more caspase-8 activity in the SF3B1K700E cells compared to the WT control (Figs. 3B and S4B). Additionally, we assessed the cleavage of two target substrates of caspase-8: (1) the endogenous caspase-8 inhibitor and suppressor of extrinsic apoptosis, cFLIP21; and (2) RIP(K1)22. Indeed, proteolytic cleavage of the long isoform of cFLIP (cFLIPL) and RIP(K1) appeared to increase in a BV-6 dose-dependent manner (Fig. 3C). Importantly, we observed reduced expression of cFLIP in SF3B1K700E cells, as well as increased BV-6-induced cleavage of cFLIPL in these cells compared to their SF3B1WT counterparts (Fig. 3C). Furthermore, we assessed the formation of the TNF complex II in SF3B1WT and SF3B1K700E cells following BV-6 treatment using caspase 8 immunoprecipitation assays. Whilst not a quantitative method, this revealed substantial formation of the cell death-inducing signalling complex in SF3B1K700E cells at 0.5 µM BV-6 treatment, in comparison to SF3B1WT cells (Figs. 3D and S4C). In contrast, there is slightly less TNF complex II precipitated from SF3B1K700E cells treated with 1 µM BV-6 compared to with SF3B1WT counterparts. This is likely due to a combination of increased SF3B1K700E cell death (evidenced by increased CASP8 cleavage) and complex II formation saturation with this dose of BV6. In line with this, activity of the executioner caspases-3/7 was significantly enhanced in SF3B1K700E cells compared to SF3B1WT cells following BV-6 treatment (Figs. 3E and S4B). Moreover, the proportions of apoptotic cells were significantly elevated in SF3B1K700E cells compared to their WT counterparts (Figs. 3F and S5A, B).
SF3B1K700E cells contain reduced levels of cFLIP, predisposing them to BV-6-induced cytotoxicity
As previously mentioned, reduced cFLIP protein expression was observed in the SF3B1K700E cells when compared with their WT counterparts (Fig. 3C). Given the role of SF3B1 in pre-mRNA splicing, we examined the transcription and splicing of the cFLIP encoding gene CFLAR. Indeed, normalised gene counts derived from total RNA demonstrate downregulation of CFLAR expression in SF3B1K700E cells compared to their WT counterparts (Fig. 4A). Additionally, we have previously shown that SF3B1 mutation-driven splicing defects can result in impaired nuclear-to-cytoplasmic export of mis-spliced transcripts [11]. In line with this, CFLAR transcripts were significantly downregulated in the cytoplasmic fraction of SF3B1K700E cells compared to their SF3B1 wild-type counterparts (Fig. 4A). Moreover, analysis of RNA sequencing (RNA-seq) data following treatment with an SMG1 inhibitor to inhibit NMD highlighted the inclusion of cryptic splice sites at the 3’ end of exon 5 and the 5’ end of exon 6 within the CFLAR transcripts. These transcripts are usually degraded by NMD (Fig. S6A), which likely accounts for the reduced CFLAR transcript in the cytoplasm and subsequent depletion of cFLIP protein expression in SF3B1K700E cells at baseline. To validate this finding, we designed custom primers to detect the inclusion of intronic sequences, which create the cryptic splice sites at exons 5 and 6. This primer set is only able to detect an amplicon in the SF3B1K700E mutant line following treatment with SMG1 inhibitor (Fig. S6B). Usage of this cryptic splice site was detected in all SF3B1 mutant cell lines tested (Fig. S6B).
A Normalised total, nuclear and cytoplasmic CFLAR gene counts from SF3B1WT and SF3B1K700E cells (Mean ± SEM; n = 3; Unpaired t-test). B Viability assessment of parental and cFLIPL- or cFLIPS-overexpressing SF3B1WT and SF3B1K700E cells treated with BV-6 for 48 h (Mean ± SEM; n = 5; 2-way ANOVA with Dunnett’s multiple comparisons). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
In line with this, the H-2595 lung mesothelioma cells, carrying the SF3B1K700E mutation, expressed significantly lower levels of the CFLAR transcript and cFLIP protein compared to matched SF3B1WT H-2591 cells (Fig. S6C, D). Combined, these data suggest that SF3B1K700E causes altered splicing of the CFLAR gene, leading to NMD, resulting in reduced cFLIPL and cFLIPS protein expression.
To investigate whether this reduced cFLIP expression contributed to the sensitivity of SF3B1K700E cells to BV-6, we stably expressed cFLIPL or cFLIPS transgenes in our isogenic K-562 cell lines (Fig. S6E). Surprisingly, whilst ectopic expression of cFLIPL rescued the modest induction of cell death observed in the SF3B1WT cells, it also completely rescued SF3B1K700E cell sensitivity to BV-6 (Fig. 4B). In contrast, ectopic expression of cFLIPS failed to protect either cell line from BV-6-mediated cell death (Fig. 4B).
SF3B1 K700E cells exhibit reduced BCL-2, promoting the intrinsic apoptotic pathway and further predisposing the cells to BV-6-induced cytotoxicity
In line with the GSEA data (Fig. 1A), the striking downregulation of BCL2 along with partial upregulation of pro-apoptotic BAX, BAD, and BID in SF3B1K700E cells (Fig. 5A) suggests mutation-induced priming of the K-562 cells towards apoptosis. Additionally, given the concurrent downregulation of DIABLO (protein: SMAC) (Fig. 5A), the cytotoxic effects exerted by the bivalent SMs, BV-6 (Figs. 1C and S1A, B) and Birinapant (Fig. S1D, E), on SF3B1K700E cells are unsurprising.
A Heatmap highlighting the mean log2 fold change of various pro- and anti-apoptotic genes in SF3B1WT and SF3B1K700E cells (n = 3). B qRT-PCR validation of the transcriptomic data for BCL2, BCL2L1, and MCL1 genes in SF3B1WT and SF3B1K700E cells at baseline (Mean ± SEM; n = 3; Multiple unpaired t-tests with Welch’s correction). C Evaluation of BCL2 gene transcripts in SF3B1WT and SF3B1K700E cells following a 6 h incubation with the indicated concentrations of NMDI-14 (Mean ± SEM; n = 3; 2-way ANOVA with Dunnett’s multiple comparisons; Paired t-test). D Representative Western blot showing caspase-9 cleavage in SF3B1WT and SF3B1K700E cells following treatment with the indicated concentrations of BV-6 for 24 h. E Viability assessment of parental SF3B1WT and SF3B1K700E cells and cells stably overexpressing BCL2 following a 48 h treatment with BV-6 (Mean ± SEM; n = 4; 2-way ANOVA with Dunnett’s multiple comparisons). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
BCL-2 downregulation in SF3B1K700E cells was validated at both the mRNA and protein levels (Figs. 5B and S7A). To investigate the potential involvement of differential splicing in the marked reductions of BCL2 expression, we used EventPointer to assess our RNA-seq data but were unable to detect any events associated with the BCL2 gene (Supplementary Table 2). Despite this, we treated SF3B1WT and SF3B1K700E cells with the NMD inhibitor, NMDI-14, and found that the 100 nM dose significantly upregulated BCL2 expression within SF3B1K700E cells relative to their vehicle control, and partially rescued BCL2 expression back towards the baseline transcript levels observed in SF3B1WT cells (Fig. 5C). This suggests that rapid degradation of aberrant BCL2 mRNA via NMD is at least partly responsible for the reduced transcript and protein levels. Moreover, BV-6 dose-dependent cleavage of caspase-9 is enhanced in SF3B1K700E cells (Figs. 5D and S7B), indicating involvement of the intrinsic apoptotic pathway to amplify the overall cytoplasmic apoptotic signalling.
To complement the RNA-seq and qRT-PCR data derived from our isogenic SF3B1 cell model, similar analyses were conducted on CD34+-enriched bone marrow samples taken from MDS patients [21] (Fig. S7C) and matched lung mesothelioma cell lines (Fig. S7D), with and without the SF3B1K700E mutation. Notably, and in keeping with the in vitro results (Fig. 5A, B), BCL2 transcripts were significantly decreased in the patient samples and lung mesothelioma cells with the K700E mutation, compared to those without (Fig. S7C, D). Furthermore, Western blot analysis demonstrated reduced BCL-2 protein levels within H-2595 (SF3B1K700E) cells compared to matched WT control cells (Fig. S7E, F), presumably due to SF3B1K700E-mediated mis-splicing of BCL2.
To determine the impact of SF3B1K700E-mediated BCL-2 downregulation on cellular sensitivity to BV-6, we stably expressed a BCL-2 trans-gene in our isogenic SF3B1 model (Fig. S7G). Interestingly, overexpression of the BCL2 trans-gene partially rescued SF3B1K700E cells from BV-6-induced cytotoxicity (Fig. 5E).
Discussion
Despite the essential role of the splicing machinery in the processing and maturation of RNA transcripts, recurrent somatic mutations have been identified in various splicing factors across a range of cancers [22]. This implies functional and pathophysiological consequences of spliceosome mutations in carcinogenesis [23, 24]. The SF3B1K700E mutation has been shown to alter the factor’s recognition of canonical 3’-splice sites, often leading to cryptic splicing of target genes, along with loss or qualitative differences in function at the protein level [10]. Despite this, the precise ways in which SF3B1K700E contributes to cancer progression and/or dictates responses to therapy, remain unclear.
Using our previously generated SF3B1K700E isogenic model [11], we assessed the impact of the SF3B1K700E mutation on response to apoptosis-inducing agents. From this, we identified the bivalent SM, BV-6, as a compound of interest. SMs are synthetic compounds that were modelled on the N-terminal IAP-binding motif of the mitochondrial protein SMAC/DIABLO [25] to antagonise IAPs, thereby enabling them to activate and/or modulate programmed cell death pathways. SMs, such as BV-6, trigger the auto-ubiquitination and subsequent proteasomal degradation of the E3 ligase, cIAP1. cIAP1 regulates extrinsic apoptosis and cellular responses to TNFα by ubiquitinating RIP(K1) [26]. The loss of cIAP1 disrupts the addition of ubiquitin chains to RIP(K1) and promotes the formation of the cytosolic TNF-complex-II, leading to programmed cell death [27]. Further processing of RIP(K1) through caspase-8-mediated proteolytic cleavage [28] is required to proceed with the apoptotic programme [29], which is enhanced in the absence of the endogenous caspase-8 regulator, cFLIPL [30]. Interestingly, we have shown reduced protein levels of cFLIP in the SF3B1-mutant setting, along with altered splicing and reduced CFLAR transcript levels in the cytoplasm of SF3B1K700E cells. This is consistent with our previous reports demonstrating reduced splicing and nuclear export of a large subset of genes in SF3B1K700E cells [11]. Therefore, we propose that the cFLIP-deficient conditions caused by SF3B1K700E enhance caspase-8 activation, contributing to the significant cytotoxicity observed in SF3B1K700E cells following BV-6 treatment. To validate the role of cFLIP in the enhanced sensitivity of SF3B1-mutant cells to BV-6, we stably overexpressed both the long (cFLIPL) and short (cFLIPs) isoforms of cFLIP. Interestingly, only ectopic expression of cFLIPL substantially rescued the cell death observed in our mutant SF3B1 model following BV-6 exposure, with ectopic expression of the cFLIPs isoform having no impact on BV-6-induced cytotoxicity in either cell line. This was somewhat surprising as cFLIPs were previously shown to play a role in inhibiting caspase-8 activation and preventing apoptosis [31]. However, further exploration of the literature highlighted a role for cFLIPs in necroptosis through promotion of the ripoptosome [32]. Therefore, we surmise that the inability of cFLIPS to influence cell viability following BV-6 treatment in our isogenic model was due to its role being redundant, as our K-562 SF3B1 mutant cells do not express RIPK3, which is an important component in ripoptosome formation and therefore necroptosis [20].
Hashimoto and colleagues identified that the combined therapeutic inhibition of XIAP and BCL-2 promotes cell death in aggressive AML [33]. XIAP, unlike other IAP family members, is the only caspase inhibitor with the ability to directly inhibit caspases-3/-7 and -9, thereby enabling effective antagonism of both the intrinsic and extrinsic pathways [34]. In agreement with this, we identified downregulation of anti-apoptotic BCL-2 as an additional mechanism contributing to the enhanced sensitivity of the SF3B1-mutant cells to BV-6, which inhibits both cIAP1/2 and XIAP. We have shown that the downregulation of BCL-2 transcript and protein levels is a result of NMD in SF3B1K700E cells. Unsurprisingly, NMD inhibition did not protect against BV-6-induced cytotoxicity (data not shown), presumably due to the translation of aberrant dysfunctional BCL-2 protein. Considering the influence of the SF3B1K700E mutation on both CFLAR (FLIP) and BCL2 splicing and expression, along with its broader effects on apoptotic gene regulation, the SF3B1K700E variant appears to exert a polygenic impact on apoptosis, ultimately increasing susceptibility to bivalent SMAC mimetics.
Downregulation of both cFLIPL and BCL-2 observed in the mutant SF3B1 models significantly augments activation of executioner caspases in SF3B1K700E cells in response to BV-6. Moreover, it is widely accepted that the ability of SMs to induce apoptosis in vitro is dependent upon autocrine death receptor ligand secretion [33]. The upregulated TNF-α signalling in our SF3B1K700E cells may serve as a potential means to support cell survival and compensate for the imbalance between pro- and anti-apoptotic proteins at baseline. Taken together, the shifted balance of pro- and anti-apoptotic proteins likely primes SF3B1K700E cells for BV-6-induced cell death by facilitating crosstalk between the extrinsic and intrinsic apoptotic pathways (Fig. 6). Importantly, this enhanced sensitivity to BV-6 treatment in cells harbouring the SF3B1K700E mutation provides compelling evidence that patients with cancers carrying this mutation would be good candidates for SM therapy with BV-6 or Birinapant.
(1) The bivalent SMAC mimetic, BV-6, serves as a dual cIAP1 and XIAP inhibitor. (2) BV-6 causes cIAP1 to auto-ubiquitinate thereby targeting cIAP1 for proteasomal degradation. (3) Loss of cIAP1 promotes non-canonical NF-κB signalling and upregulation of NF-κB target genes including TNF. (4) This may promote death receptor signalling through autocrine production of death receptor ligands such as TNF-α (5) Disruption of RIP(K1) ubiquitination promotes the formation of cytosolic TNF Complex II that (6) activates downstream executioner caspases-3/7. (5) Reduced levels of the endogenous caspase-8 inhibitor, cFLIP, in SF3B1K700E cells further promote caspase-8 activation within TNF Complex II, enabling proteolytic cleavage of BID. (7) The resulting truncated BID (tBID) can then be bound and sequestered by pro-survival proteins (e.g., BCL-2, BCL-XL, and MCL-1), however, when these pro-survival proteins are saturated or absent, (8) as is the case for BCL-2 levels in SF3B1K700E cells, tBID and other pro-apoptotic factors can promote mitochondrial outer membrane permeabilisation (MOMP) and the release of apoptogenic factors. In conjunction with (9) BV-6-mediated inhibition of XIAP, this leads to enhanced activation of the mitochondrial (intrinsic) pathway that, in turn, (10) activates downstream executioner caspases-3/7 and culminates in RIP(K1)-dependent apoptosis. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
Methods
Cell lines and culture
The SF3B1 isogenic model was generated and maintained as previously described 11. NCI-H2591 (RRID:CVCL_A543) (SF3B1WT) and NCI-H2595 (RRID:CVCL_A545) (SF3B1K700E) cells were obtained from Cosmic-CLP and maintained as per the supplier’s guidance. Polyclonal, stable K-562 cells overexpressing the long and short splice variants of c-FLIP*, and BCL-2** were generated by retroviral and lentiviral transductions, respectively, followed by selection with 10 μg/mL puromycin (ant-pr-1; InvivoGen). All cell line stocks have been authenticated by short tandem repeat profiling and were regularly verified as Mycoplasma-free.
Plasmids
*pBABE-puro plasmids (RRID:Addgene_1764) [35] containing the long and short isoforms of c-FLIP were gifts from Daniel Longley [36, 37].
**pCDH-puro-Bcl2 was a gift from Jialiang Wang (RRID:Addgene_46971) [38].
Drug screening
SF3B1WT and SF3B1K700E K-562 cells were screened, and Z-score values were calculated as previously described [39, 40].
Compounds
BV-6 (S7597), Xevinapant (S2754), Ferrostatin-1 (S7243), Necrostatin-1 (S8037), and z-VAD-FMK (S7023) were all purchased from Selleckchem. Birinapant (TL32711) (A4219-APE) was purchased from APExBio. Necrosulfonamide (S8251; Selleckchem) was a gift from Richard Turkington. MG-132 (J63250.MA) was purchased from Thermo Fisher Scientific.
Cell viability assay
Cell viability was assessed using the CellTiter-Glo 2.0 luminescent assay (G9242; Promega) following the manufacturer's guidelines, and luminescence was measured using a BioTek Synergy2 plate reader (BioTek).
Caspase-Glo assay
Protease activity levels for caspases-3/7, -8, and -9 were assayed using the Promega Caspase-Glo 3/7 (G8090), Caspase-Glo 8 (G8200), and Caspase-Glo 9 (G8210) assay systems according to the manufacturer’s instructions.
Flow cytometry
Samples were analysed on a BD Accuri C6 Plus flow cytometer and BD CSampler Plus software (version 1.0.23.1).
TNFR1 cell surface staining
SF3B1WT and SF3B1K700E cells were collected, counted, and diluted to 107 cells/mL with eBioscience Flow Cytometry Staining Buffer (#FAB225G; Bio-Techne). Samples were Fc-blocked with 5 μL of Human TruStain FcX (422302; BioLegend) for 15-min at room temperature (RT). A 10 μL volume of AF488-conjugated TNFR1 antibody (Ex: 488 nm, Em: 533/30 nm) was added and incubated for 30-min at RT in the dark. Cells were resuspended in 0.5 mL of eBioscience Flow Cytometry Staining Buffer following washing. Five microliters of 7-AAD viability dye (#420404; Thermo Fisher Scientific) were added to each sample, followed by a 5-minute incubation at RT in the dark prior to data acquisition. All experiments were carried out with an IgG1 isotype control antibody (#IC002G; Bio-Techne).
Annexin V/7-AAD staining
Cell surface phosphatidylserine expression was assessed using the FITC Annexin V Apoptosis Detection Kit with 7-AAD (Cat #640922), purchased from BioLegend, according to their guidelines.
Enzyme-linked immunosorbent assay (ELISA)
TNFα levels in cell culture supernatants were quantified using the ultrasensitive Human TNF alpha ELISA (#KHC3013; Thermo Fisher Scientific). Samples were run neat and assayed as per the manufacturer’s instructions.
Immunoprecipitation
Cells were lysed in CHAPS buffer (30 mM Tris pH 7.5, 150 mM NaCl, 1% CHAPS). Two micrograms of caspase-8 antibody (#ALX-804-242; Enzo) were conjugated to 25 μL Pierce Protein A/G Magnetic beads (88802; Thermo Fisher Scientific). One milligram of protein lysate was immunoprecipitated overnight at 4 °C. Mouse IgG1 isotype control (X093101; Dako) was purchased from Agilent. Co-immunoprecipitation experiments were analysed by western blotting.
Western blotting
Whole-cell extracts (WCE) were prepared by lysing cells in two volumes of erythrocyte lysis buffer (ELB) (250 mmol/L NaCl, 5 mmol/L EDTA, 50 mmol/L HEPES, 0.1% v/v NP-40; pH 7.5) supplemented with 1X Halt protease and phosphatase inhibitor cocktail (78440; Thermo Fisher Scientific). Thirty to 60 μg of WCE was resolved on 4–12% Bolt Gels (Thermo Fisher Scientific). Protein was transferred to nitrocellulose membranes (Thermo Fisher Scientific) and blotted using the following antibodies: NF-κB2 p52/p100 (sc-7386), BCL-2 (sc-509), and β-actin (sc-47778) from Santa Cruz Biotechnology; Caspase-8 (ALX-804-242-C100, Enzo); cIAP1 (10022-1-AP, Proteintech); BCL-XL (2762), FADD (2782), MCL-1 (5453), XIAP (2045), FLIP (56343), and RIP (3493) from cell signalling technology (CST). Anti-mouse (7076) and anti-rabbit (7074) HRP-conjugated secondary antibodies were also purchased from CST.
RNA-seq, gene set enrichment analysis (GSEA) and differential splicing
RNA-seq libraries were prepared using RNeasy Plus Mini Kits (741434; QIAGEN). Sequencing was performed in the Genomics Core Technology Unit at Queen’s University Belfast. Reads were aligned using STAR [41], and transcripts were quantified using DESeq2 [42]. EventPointer [43] was used to assess aberrant splicing events between SF3B1K700E and WT controls.
GSEA was performed using the GSEA v4.1.0 software (www.broadinstitute.org/gsea/). Cytoplasmic RNA preparations, sequencing, and downstream data analysis, including differential splicing analysis, using dSpliceType, were carried out as described previously [11].
Quantitative PCR
cDNA was synthesised using the High-Capacity cDNA Reverse Transcription Kit (#4368814; Thermo Fisher Scientific) according to the manufacturer’s instructions. qPCR was carried out using FastStart Universal SYBR Green Master (Rox) mix (#4913914001; Merck Life Science UK Ltd), QIAGEN QuantiTect Primer Assays***, and the LC480 thermal light cycler (Roche) according to the manufacturer’s instructions.
***Hs_BCL2_1_SG (QT00025011); Hs_BCL2L1_1_SG (QT00236712); Hs_CFLAR_1_SG (QT00064554); Hs_MCL1_1_SG (QT00094122)
Statistical analysis
GraphPad Prism (version 9.1.2) was used to determine all statistical comparisons. One-way and two-way ANOVA tests were performed with a multiple comparisons post-test as indicated. A P-value of 0.05 or less was considered statistically significant. Except for the apoptosis-inducing agent drug screen and MI-2 72 h dose-responses, means were compared across a minimum of three biological replicates, with at least two technical replicates averaged where appropriate.
Data availability
RNA-seq fastq files are available on the BioProject database using ID: PRJNA1065630.
References
Matera AG, Wang Z. A day in the life of the spliceosome. Nat Rev Mol Cell Biol. 2014;15:108–21.
Lek M, Karczewski KJ, Minikel EV, Samocha KE, Banks E, Fennell T, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536:285–91.
Huber S, Haferlach T, Meggendorfer M, Hutter S, Hoermann G, Baer C, et al. SF3B1 mutations in AML are strongly associated with MECOM rearrangements and may be indicative of an MDS pre-phase. Leukemia. 2022;36:2927–2930.
Foy A, McMullin MF. Somatic SF3B1 mutations in myelodysplastic syndrome with ring sideroblasts and chronic lymphocytic leukaemia. J Clin Pathol. 2019;72:778–82.
Furney SJ, Pedersen M, Gentien D, Dumont AG, Rapinat A, Desjardins L, et al. SF3B1 mutations are associated with alternative splicing in uveal melanoma. Cancer Discov. 2013;3:1122–9.
Harbour JW, Roberson EDO, Anbunathan H, Onken MD, Worley LA, Bowcock AM. Recurrent mutations at codon 625 of the splicing factor SF3B1 in uveal melanoma. Nat Genet. 2013;45:133–5.
Kong Y, Krauthammer M, Halaban R. Rare SF3B1 R625 mutations in cutaneous melanoma. Melanoma Res. 2014;24:332–4.
Stephens PJ, Tarpey PS, Davies H, Van Loo P, Greenman C, Wedge DC, et al. It cuts both ways: reconciling the dual roles of caspase 8 in cell death and survival. Nat Rev Mol Cell Biol. 2011;12:757–763.
Darman RB, Seiler M, Agrawal AA, Lim KH, Peng S, Aird D, et al. Cancer-associated SF3B1 hotspot mutations induce cryptic 3’ splice site selection through use of a different branch point. Cell Rep. 2015;13:1033–45.
Alsafadi S, Houy A, Battistella A, Popova T, Wassef M, Henry E, et al. Cancer-associated SF3B1 mutations affect alternative splicing by promoting alternative branchpoint usage. Nat Commun. 2016;7:10615.
Liberante FG, Lappin K, Barros EM, Vohhodina J, Grebien F, Savage KI, et al. Altered splicing and cytoplasmic levels of tRNA synthetases in SF3B1-mutant myelodysplastic syndromes as a therapeutic vulnerability. Scientific Rep. 2019;9:2678.
Martinez-Ruiz G, Maldonado V, Ceballos-Cancino G, Grajeda JPR, Melendez-Zajgla J. Role of Smac/DIABLO in cancer progression. J Exp Clin Cancer Res. 2008;27:48.
Rajalingam K, Sharma M, Paland N, Hurwitz R, Thieck O, Oswald M, et al. IAP-IAP complexes required for apoptosis resistance of C. trachomatis-infected cells. PLoS Pathog. 2006;2:e114.
Petersen SL, Wang L, Yalcin-Chin A, Li L, Peyton M, Minna J, et al. Autocrine TNFalpha signaling renders human cancer cells susceptible to Smac-mimetic-induced apoptosis. Cancer Cell. 2007;12:445–456.
Varfolomeev E, Blankenship JW, Wayson SM, Fedorova AV, Kayagaki N, Garg P, et al. IAP antagonists induce autoubiquitination of c-IAPs, NF-kappaB activation, and TNFalpha-dependent apoptosis. Cell. 2007;131:669–681.
Vince JE, Wong WWL, Khan N, Feltham R, Chau D, Ahmed AU, et al. IAP antagonists target cIAP1 to induce TNFalpha-dependent apoptosis. Cell. 2007;131:682–693.
Probst BL, Liu L, Ramesh V, Li L, Sun H, Minna JD, et al. Smac mimetics increase cancer cell response to chemotherapeutics in a TNF-α-dependent manner. Cell Death Differ. 2010;17:1645–54.
Cazzola M, Rossi M, Malcovati L. Biologic and clinical significance of somatic mutations of SF3B1 in myeloid and lymphoid neoplasms. Blood. 2013;121:260–9.
Dolatshad H, Pellagatti A, Liberante FG, Llorian M, Repapi E, Steeples V, et al. Cryptic splicing events in the iron transporter ABCB7 and other key target genes in SF3B1-mutant myelodysplastic syndromes. Leukemia. 2016;30:2322–31.
Morgan MJ, Kim YS. Roles of RIPK3 in necroptosis, cell signaling, and disease. Exp Mol Med. 2022;54:1695–704.
Choudhary GS, Pellagatti A, Agianian B, Smith MA, Bhagat TD, Gordon-Mitchell S, et al. Activation of targetable inflammatory immune signaling is seen in myelodysplastic syndromes with SF3B1 mutations. eLife. 2022;11:e78136.
Seiler M, Peng S, Agrawal AA, Palacino J, Teng T, Zhu P, et al. Somatic mutational landscape of splicing factor genes and their functional consequences across 33 cancer types. Cell Rep. 2018;23:282–296.e4.
Bergot T, Lippert E, Douet-Guilbert N, Commet S, Corcos L, Bernard D. Human cancer-associated mutations of SF3B1 lead to a splicing modification of its own RNA. Cancers. 2020;12:652.
Taylor J, Lee SC, Stanley Lee CC, Sloan Kettering M. Mutations in spliceosome genes and therapeutic opportunities in myeloid malignancies. Genes Chromosomes Cancer. 2019;58:889–902.
Wright CW, Duckett CS. Reawakening the cellular death program in neoplasia through the therapeutic blockade of IAP function. J Clin Investig. 2005;115:2673–8.
Bertrand MJM, Milutinovic S, Dickson KM, Ho WC, Boudreault A, Durkin J, et al. cIAP1 and cIAP2 facilitate cancer cell survival by functioning as E3 ligases that promote RIP1 ubiquitination. Mol Cell. 2008;30:689–700.
Wang L, Du F, Wang X. TNF-alpha induces two distinct caspase-8 activation pathways. Cell. 2008;133:693–703.
Lin Y, Devin A, Rodriguez Y, Liu ZG. Cleavage of the death ___domain kinase RIP by caspase-8 prompts TNF-induced apoptosis. Genes; Dev. 1999;13:2514–26.
Brenner D, Blaser H, Mak TW. Regulation of tumour necrosis factor signalling: live or let die. Nat Rev Immunol. 2015;15:362–74.
Pop C, Oberst A, Drag M, Van Raam BJ, Riedl SJ, Green DR, et al. FLIPL induces caspase 8 activity in the absence of interdomain caspase 8 cleavage and alters substrate specificity. Biochem J. 2011;433:447–57.
Tsuchiya Y, Nakabayashi O, Nakano H. FLIP the switch: regulation of apoptosis and necroptosis by cFLIP. Int J Mol Sci. 2015;16:30321–41.
Feoktistova M, Geserick P, Kellert B, Dimitrova DP, Langlais C, Hupe M, et al. Prognostic and therapeutic relevance of FLIP and procaspase-8 overexpression in non-small cell lung cancer. Cell Death Dis. 2013;4:e951.
Hashimoto M, Saito Y, Nakagawa R, Ogahara I, Takagi S, Takata S, et al. CIAPs block ripoptosome formation, a RIP1/caspase-8 containing intracellular cell death complex differentially regulated by cFLIP isoforms. Mol Cell. 2011;43:449–463.
Eckelman BP, Salvesen GS, Scott FL. Human inhibitor of apoptosis proteins: why XIAP is the black sheep of the family. EMBO Rep. 2006;7:988–94.
Morgenstern JP, Land H. Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res. 1990;18:3587–96.
Carson R, Celtikci B, Fenning C, Javadi A, Crawford N, Perez-Carbonell L, et al. Multiplex screening for interacting compounds in paediatric acute myeloid leukaemia. Int J Mol Sci. 2021;22:10163.
Riley JS, Hutchinson R, McArt DG, Crawford N, Holohan C, Paul I, et al. HDAC inhibition overcomes acute resistance to MEK inhibition in BRAF-mutant colorectal cancer by downregulation of c-FLIPL. Clin Cancer Res. 2015;21:3230–3240
Cheng Z, Gong Y, Ma Y, Lu K, Lu X, Pierce LA, et al. Inhibition of BET bromodomain targets genetically diverse glioblastoma. Clin Cancer Res. 2013;19:1748–59.
Cairns LV, Lappin KM, Mutch A, Ali A, Matchett KB, Mills KI. A compound combination screening approach with potential to identify new treatment options for paediatric acute myeloid leukaemia. Sci Rep. 2020;10:18514.
Lappin KM, Davis L, Matchett KB, Ge Y, Mills KI, Blayney JK. EventPointer: an effective identification of alternative splicing events using junction arrays. BMC Genomics. 2016;17:467.
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29:15–21.
Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550.
Romero JP, Muniategui A, De Miguel FJ, Aramburu A, Montuenga L, Pio R, et al. EventPointer: an effective identification of alternative splicing events using junction arrays. BMC Genomics. 2016;17:467.
Funding
Leukaemia and Lymphoma Northern Ireland (R2739CNR). Leukaemia UK (R2569CNR). Cancer Research UK (A29302).
Author information
Authors and Affiliations
Contributions
KML, KIS, and KIM conceived the project. LR, KML, and KIS designed all experiments. LER, HPM, PJP, and AM performed experiments and analysed the data. JKB performed bioinformatic analysis. LR, KML, and KIS analysed and interpreted data. LER, KML, and KIS wrote the manuscript with input from all authors. KML, KIM, and KIS acquired funding.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Ethics statement
This study is a cell and molecular biology-based laboratory study and did not involve animal or human participants, and therefore did not require ethical approval within the UK.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Edited by Mauro Piacentini
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Roets, L.E., Blayney, J.K., McMillan, H.P. et al. The cancer-associated SF3B1K700E spliceosome mutation confers enhanced sensitivity to BV-6-induced cytotoxicity. Cell Death Dis 16, 476 (2025). https://doi.org/10.1038/s41419-025-07790-y
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41419-025-07790-y
