Burkholderia Lethal Factor 1, a Novel Anti-Cancer Toxin, Demonstrates Selective Cytotoxicity in MYCN-Amplified Neuroblastoma Cells

Abstract

Immunotoxins are being investigated as anti-cancer therapies and consist of a cytotoxic enzyme fused to a cancer targeting antibody. All currently used toxins function via the inhibition of protein synthesis, making them highly potent in both healthy and transformed cells. This non-specific cell killing mechanism causes dose-limiting side effects that can severely limit the potential of immunotoxin therapy. In this study, the recently characterised bacterial toxin Burkholderia lethal factor 1 (BLF1) is investigated as a possible alternative payload for targeted toxin therapy in the treatment of neuroblastoma. BLF1 inhibits translation initiation by inactivation of eukaryotic initiation translation factor 4A (eIF4A), a putative anti-cancer target that has been shown to regulate a number of oncogenic proteins at the translational level. We show that cellular delivery of BLF1 selectively induces apoptosis in neuroblastoma cells that display MYCN amplification but has little effect on non-transformed cells. Future immunotoxins based on this enzyme may therefore have higher specificity towards MYCN-amplified cancer cells than more conventional ribosome-inactivating proteins, leading to an increased therapeutic window and decreased side effects.

Keywords: BLF1; MYCN; apoptosis; cancer; eIF4A; neuroblastoma; ribosome-inactivating protein.

Figure 1

BLF1 causes growth inhibition preferentially in MYCN-amplified cell lines. (A) Immunoblot showing protein levels of MYCN and GAPDH in the MYCN-amplified cell lines IMR-32 and SK-N-BE(2), and the non-MYCN-amplified cell lines SH-SY5Y and LA-N-6; (B) Alamarblue assay following 72 h titration with BLF1 or saporin in the presence of LF3000. MYCN-amplified cell lines are shown in red and non-MYCN-amplified cell lines are shown in blue (n ≥ 3, ± SEM); (C) Table showing calculated GI50s and 95% confidence intervals.

Figure 2

BLF1 induces apoptosis in MYCN-amplified neuroblastoma cell lines. (A) Fluorescence microscopy using Hoechst (blue) and propidium iodide (red) staining shows viability of neuroblastoma cell lines following 72 h incubation with 300 nM BLF1 or 300 nM saporin in the presence of LF3000; (B) Quantification of A shows that BLF1 exhibits similar cytotoxicity to saporin in MYCN-amplified cell lines but is not significantly cytotoxic to non-MYCN-amplified cells when compared to the untreated control (LF3000 alone), whereas saporin still exhibits high levels of cytotoxicity (n = 3, ± SEM). Data was analysed by two-way ANOVA and Tukey’s multiple comparisons test with *** p < 0.001 and **** p < 0.0001; (C) CellEvent Caspase 3/7 activation assay shows that 300 nM BLF1 induces similar levels of apoptosis to saporin (300 nM) following a 72-h incubation in the presence of LF3000 (n = 3, ± SEM). Data was analysed by two-way ANOVA and Tukey’s multiple comparisons test with **** p < 0.0001; ns: non significant.

Figure 3

BLF1 exhibits increased potency in SHEP-21N cells expressing MYCN. (A) Immunoblot showing protein levels of MYCN, eIF4A and a loading control α-tubulin in SHEP-21N cells following 72 h of incubation in the presence or absence of tetracycline; (B) AlamarBlue assay following 72-h titrations with BLF1 in the presence of LF3000 demonstrates increased growth inhibition in SHEP-21N cells expressing MYCN (n = 3, ± SEM); (C) Fluorescence microscopy using Hoechst (blue) and propidium iodide (red) staining shows viability of SHEP-21N cells following 72-h incubation with 300 nM BLF1 in the presence of LF3000; (D) Quantification of the panel C results shows that BLF1 is significantly cytotoxic to SHEP-21N cells expressing MYCN but has no significant effect on cells not expressing MYCN when compared to the untreated control (LF3000 alone) (n = 3, ± SEM). Data was analysed by multiple unpaired t-tests with **** p < 0.0001; ns: non significant. (E) Fluorescence microscopy using Hoechst and propidium iodide staining shows viability of SHEP-21N cells following 72 h incubation with 1 μM eIF4A or eIF4A-Q339E dominant-negative mutant in the presence of LF3000.

Figure 4

RocA exhibits increased potency in SHEP-21N cells expressing MYCN. (A) AlamarBlue cell assay shows more potent growth inhibition in SHEP-21N cells expressing MYCN, following 72 h titrations with RocA (n = 3, ± SEM); (B) Fluorescence micrographs showing cell viability following 72 h of incubation with 100 nM RocA. All nuclei are stained with Hoechst (blue) whereas the nuclei of dead cells are stained with propidium iodide (red); (C) Quantification of the panel C results shows that RocA is significantly cytotoxic to SHEP-21N cells expressing MYCN but has no effect on viability of cells not expressing MYCN when compared to the untreated control (DMSO alone) (n = 3, ± SEM). Data was analysed by multiple unpaired t-tests with ** p < 0.01; ns: non significant.

Figure 5

BLF1 down-regulates protein levels of MYCN and other eIF4A sensitive transcripts. (A) Schematic showing eIF4A sensitive (red) and non-eIF4A sensitive (blue) mRNAs. eIF4A sensitive mRNAs contain long and highly structured 5′ UTRs that require the RNA helicase of eIF4A to unwind secondary structures such as G-quadruplexes for scanning by the 43S pre-initiation translation complex. These coding transcripts usually encode for oncogenes and pro-survival proteins [10]; (B) Immunoblot showing protein levels of MYCN, CDK4, c-Jun, RhoA, α-tubulin and GAPDH in IMR-32 cells following a 24-h incubation with 30 nM BLF1 or saporin in the presence of LF3000; (C) Quantification of the panel B results shows the high level of down-regulation of MYCN and CDK4 following treatment with either toxin, whereas only saporin down-regulates the eIF4A insensitive transcripts c-Jun and RhoA. No effect on housekeeping genes GAPDH or α-tubulin was seen. Protein levels are shown as a ratio of the untreated control (LF3000 alone) (n ≥ 3, ± SEM). Data was analysed by multiple one-sample t-tests to a normalised control of 1 with ** p < 0.01 and *** p < 0.001; ns: non significant.

Figure 6

BLF1 has no effect on viability of primary mouse fibroblasts. (A) AlamarBlue assay following titrations of BLF1 and saporin in the presence of LF3000 in quiescent mouse fibroblasts shows that BLF1 has no effect on cell numbers whereas saporin causes a large reduction (n = 3, ± SEM); (B) Fluorescence microscopy images showing Hoechst 33342 (blue) and propidium iodide (red) staining following treatment of mouse fibroblasts with 300 nM BLF1 or saporin in the presence of LF3000; (C) Quantification of B shows that BLF1 has no effect on cells whereas saporin causes a significant amount of cell death (n = 3, ± SEM). Data was analysed by one-way ANOVA and Dunnet’s multiple comparisons test with **** p < 0.0001; ns: non significant.