The super elongation complex drives transcriptional addiction in MYCN-amplified neuroblastoma

Abstract

MYCN amplification in neuroblastoma leads to aberrant expression of MYCN oncoprotein, which binds active genes promoting transcriptional amplification. Yet, how MYCN coordinates transcription elongation to meet productive transcriptional amplification and which elongation machinery represents MYCN-driven vulnerability remain to be identified. We conducted a targeted screen of transcription elongation factors and identified the super elongation complex (SEC) as a unique vulnerability in MYCN-amplified neuroblastomas. MYCN directly binds EAF1 and recruits SEC to enhance processive transcription elongation. Depletion of EAF1 or AFF1/AFF4, another core subunit of SEC, leads to a global reduction in transcription elongation and elicits selective apoptosis of MYCN-amplified neuroblastoma cells. A combination screen reveals SEC inhibition synergistically potentiates the therapeutic efficacies of FDA-approved BCL-2 antagonist ABT-199, in part due to suppression of MCL1 expression, both in MYCN-amplified neuroblastoma cells and in patient-derived xenografts. These findings identify disruption of the MYCN-SEC regulatory axis as a promising therapeutic strategy in neuroblastoma.

Fig. 1.. Identification of the SEC as a dependency factor in MYCN-driven neuroblastoma.

(A) Immunoblots of nuclear MYCN in SHEP MYCN-ER cells upon 4-OHT (100 nM) treatment for the indicated time points. Histone H3 was used as a loading control. (B) Workflow of 4sU-labeled TT-seq. SHEP MYCN-ER cells were pretreated with DMSO or 4-OHT (100 nM) for 12 hours before being refed with fresh medium containing 400 μM 4sU for 15 min. After spike-in of 4sU-labeled S2 cells, RNA was extracted and subjected to stranded TT-seq. (C) Metaplot of TT-seq signals of the plus and minus strands in DMSO- and 4-OHT–treated cells. (D) Genome browser tracks of TT-seq signals at the NCL and NPM1 loci upon DMSO and 4-OHT treatment. (E) Targeted screen of elongation factor (EF) dependency in SHEP MYCN-ER cells (left). A total of sixteen elongation factors was individually depleted by the CRISPR-Cas9 with three single guide RNAs (sgRNAs) for each gene. Cell death was analyzed by propidium iodide (PI)–annexin V staining after 4-OHT or DMSO treatment for 72 hours (right). Heatmap visualization of the cell death upon depletion of each elongation factor and the ratio of cell death (4-OHT/DMSO) are shown on the right. (F) AFF1 and AFF4 were knocked down to fully disrupt SEC, and EAF1 were depleted by shRNA in Kelly and SHEP cells. Kelly and SHEP cells were infected for 2 days and selected by puromycin, before Western blotting to AFF1, AFF4, and EAF1 proteins. Cell death analysis was performed at day 4 after infection. One-way analysis of variance (ANOVA) followed by Bonferroni correction was performed in (F). ***P < 0.001. TSS, transcription start sites; TES, transcription end sites.

Fig. 2.. SEC is required for MYCN-dependent transcriptional programs.

(A) Experimental design for depletion of SEC in SHEP MYCN-ER cells followed by DMSO or 4-OHT treatment and RNA-seq. SHEP MYCN-ER cells were infected with lentiviruses expressing control (Ctrl) or AFF1 and AFF4 shRNA for 48 hours. Gene expression pattern of SHEP MYCN-ER cells with or without SEC knockdown were analyzed by RNA-seq 24 hours after 4-OHT (100 nM) or DMSO treatment. (B) MA-plots showing 4-OHT significantly changed genes (FDR < 0.05, log₂FC > 0.5) expression in control (left) and SEC knockdown group (right). Significant up-regulated and down-regulated genes are shown as red and blue dots, respectively. (C) Heatmap clustering of up-regulated genes in SHEP MYCN-ER cells with the indicated treatments. Rows show Z scores calculated for each vector. n indicates the number of genes. (D) Genome browser views of RNA-seq raw reads at the NCL and NPM1 loci. (E) Gene ontology analysis of genes in cluster 1. Gene ontology analysis was performed with Metascape (http://metascape.org/). rRNA, ribosomal RNA. (F and G) Quantitative reverse transcription polymerase chain reaction (qRT-PCR) validation of representative MYCN target genes in SHEP MYCN-ER cells (F) and Kelly cells (G) upon indicated treatment. SHEP MYCN-ER cells were treated with 4-OHT for 24 hours in (F). One-way ANOVA followed by Bonferroni correction was used in (F), and unpaired two-tailed Student’s t test was used in (G). **P < 0.01 and ***P < 0.001.

Fig. 3.. MYCN interacts with EAF1 and globally recruits SEC to target genes.

(A) ChIP-seq analysis of MYCN, MAX, AFF1, and AFF4 occupancies at the MYCN-binding promoter peaks (n = 9100) in SHEP MYCN-ER cells treated with 4-OHT or DMSO for 5 hours. Heatmaps show ChIP-seq signal in a window of ±1 kb centered at the peaks of MYCN. (B) Box plot analysis of occupancy signals showing the log₂ counts per million of MYCN, MAX, AFF1, and AFF4 ChIP-seq with or without 4-OHT treatment. P values were calculated by the Wilcoxon test. (C) 293T cells expressing MYCN-TurboID or Flag-MYCN were applied for TurboID or co-IP mass spectrometry analysis. The sum posterior error probability (PEP) score corresponds to the score calculated based on the PEP values of the peptide spectrum matches. (D) Purified recombinant GST-CDK9, CCNT1, or EAF1 proteins coupled to GST beads were used to pull down His-MYCN, followed by immunoblotting using the MYCN antibody. Coomassie blue staining of input proteins is shown at the bottom. IB, immunoblotting. (E) Co-IP with EAF1 antibody to detect MYCN in Kelly and NLF cell lysates. IgG, immunoglobulin G. (F and G) Heatmap (F) and metaplot (G) of MYCN, AFF1, and AFF4 occupancies at the MYCN-binding promoter peaks [n = 9100, as shown in (A)] in EAF1-depleted SHEP MYCN-ER cells treated with 4-OHT or DMSO for 5 hours. (H) ChIP-seq signals of MYCN, AFF1, and AFF4 at the NPM1 locus are shown. (I) Heatmap visualization of SEC component expression changes in SHEP MYCN-ER cells upon 4-OHT treatment. (J and K) qPCR analysis (J) and immunoblots (K) of SEC components in SHEP MYCN-ER cells upon 4-OHT treatment for 24 hours. Unpaired two-tailed Student’s t test was used in (J). *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 4.. SEC primarily mediates MYCN-dependent transcription elongation through EAF1.

(A) Metagene plots of Pol II S2P in MYCN-activated genes (n = 2116). SHEP MYCN-ER cells with or without EAF1 depletion were, respectively, treated with 4-OHT or DMSO for 5 hours. (B) Representative genome browser tracks of Pol II S2P ChIP-seq signals at the NPM1 loci with or without EAF1 depletion. (C) Workflow of 4sU-FP-seq–based measurement of transcription elongation in SHEP MYCN-ER cells. Cells with or without EAF1 knockout were treated with DMSO or 4-OHT for 24 hours before addition of FP (CDK9 inhibitor, 10 μM) to pause Pol II at the promoter. Flavopiridol was then washed out, followed by 4sU labeling for 10 min. The 4sU-labeled RNA was purified for RNA-seq. (D) Heatmap visualization of 4sU-FP-seq signals in DMSO- or 4-OHT–treated cells and changes of 4sU-FP-seq signal upon EAF1 knockout. All genes up-regulated by 4-OHT were plotted using the total 4sU-FP-seq signals. (E) Genome browser views of 4sU-FP-seq analysis at the NPM1 loci. (F) Heatmap analysis of 4sU-FP-seq in DMSO- or KL-2–treated SHEP MYCN-ER cells in the presence or absence of 4-OHT (24 hours). (G) Genome browser tracks of 4sU-FP-seq at the NPM1 locus with indicated treatments.

Fig. 5.. ABT-199 synergizes with KL-2 to induce MYCN-amplified neuroblastoma cell death.

(A) Analysis of KL-2–mediated cell death in SHEP MYCN-ER cells pretreated with or without 4-OHT (100 nm) for 24 hours. (B) Cell death analysis of neuroblastoma cells treated with KL-2 (10 μM). (C) Kelly xenograft tumor growth after 2-week treatment with vehicle (corn oil) and KL-2 (50 mg/kg). Six tumors were analyzed in each group. (D and E) High-throughput screening of clinically used drugs in Kelly cells to identify small molecules synergistic with KL-2 (D). Top hits are listed in (E). HDAC, histone deacetylase; c-RET, ret proto-oncogene; VEGFR, vascular endothelial growth factor receptor; HER2, human epidermal growth factor receptor 2. (F) Cell death analysis of neuroblastoma cell lines and primary human patient cells (with MYCN amplification) treated with KL-2 (5 μM) and/or ABT-199 (5 μM). (G) Combination index (CI) of KL-2 and ABT-199 was calculated using the CalcuSyn software (Biosoft). (H) Immunoblotting for several BCL-2 family proteins in Kelly cells upon KL-2 treatment. (I) Immunoblots of MCL1 and BCL-2 with or without KL-2 treatment for 24 hours in neuroblastoma cell lines. (J) Kelly cells expressing MCL1 were subjected to KL-2 (2.5 μM), ABT-199 (5 μM), or combination treatment, followed by cell death analysis. Unpaired two-tailed Student’s t test was used in (A) to (C). One-way ANOVA followed by Bonferroni correction was used in (F), and two-way ANOVA was used with Bonferroni correction in (J). ***P < 0.001.

Fig. 6.. KL-2 and ABT-199 synergistically suppress MYCN-mediated neuroblastoma in vivo.

(A and B) Tumor growth in Kelly xenografts treated with vehicle, KL-2 (25 mg/kg), ABT-199 (20 mg/kg), or in combination (n = 6 for each group) (A). Tumor images and weights at the 14th day after treatment are shown (B). (C and D) MYCN-amplified neuroblastoma patient-derived tumor samples were transplanted into NPG mice. Tumor growth of PDXs treated with vehicle, KL-2 (25 mg/kg), ABT-199 (20 mg/kg), or in combination for 2 weeks are shown in (C) (n = 5 for each group). Recipients were euthanized when the tumor size reached 2000 mm³, and Kaplan-Meier survival curves are shown in (D). (E and F) Representative images of immunohistochemical (IHC) staining of PCNA and cleaved caspase-3 (c-Cas-3) in PDX tumors at day 10 after starting treatment (E). Scale bar, 50 μm. Quantifications are shown in (F). One-way ANOVA test followed by Bonferroni correction was used in (A) to (C) and (F). Log-rank test was used in (D) for Kaplan-Meier survival analysis. *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 7.. Proposed mechanism for the action of SEC-mediated transcription elongation in MYCN-driven neuroblastoma cells and the mechanism-based targeted strategy.

In MYCN-amplified neuroblastoma cells, overexpressed MYCN recruits SEC to chromatin through EAF1. As a consequence, SEC-loaded Pol II produces enhanced processive transcription elongation (top). Pharmacological inhibition of SEC by KL-2 disrupts SEC and induces selective cell death. Drug library screen identifies ABT-199 as a synergistic drug in combination with KL-2, suggesting dual inhibition of SEC and BCL-2 as a promising therapeutic strategy for high-risk neuroblastoma (bottom).