DREAM target reactivation by core transcriptional regulators supports neuroblastoma growth

ABSTRACT

Chromosome 17q gains are a common alteration in high-risk neuroblastomas with unknown functional significance. We identified a 17q super-enhancer regulated T-box Transcription Factor 2 (TBX2) as constituent of a core regulatory circuitry driving proliferation through enhancing V-myc myelocytomatosis viral-related oncogene, neuroblastoma derived (avian) (MYCN)/Forkhead box protein M1(FOXM1) reactivation of dimerization partner, RB-like, E2F and multi-vulval class B (DREAM) targets, which can be affected synergistically by combined cyclin-dependent kinase 7 and Bromo-domain inhibition.

Neuroblastoma (NB) is a pediatric cancer of the developing sympatho-adrenergic nervous system, predominantly affected by copy-number alterations. High-risk NBs represent half of all patients for which current multimodal intensive therapy is only successful in roughly half of these cases with survivors suffering from severe long-term side effects. V-myc myelocytomatosis viral-related oncogene, neuroblastoma derived (avian)) (MYCN) amplification is found in half of these high-risk patients and large 17q segmental gains occur in the majority of both MYCN amplified and non-amplified high-risk tumors.¹ The finding of recurrent gains of the syntenic human 17q region in MYCN driven NB mouse tumors further supports the putative functional importance of this genomic aberration in NB. Gene regulatory circuitries consisting of transcription factors (TFs) carefully control cell fate decisions and cell identity through execution of well-defined gene expression programs. Such TFs are typically associated by a super-enhancer (SE), the latter which is controlled by “master TF core regulatory circuits” driving the cell-type specific developmental transcriptional programs.² These SE-associated core-regulatory circuitries (CRCs) are well-known regulators in embryonal stem cell development and were recently described to be implicated in NB in the control of two divergent cell states, i.e. adrenergic and neural crest cell/mesenchymal cell like type of cells.^3,4 Of further note, CRCs consisting of SE-marked master TFs were recently shown to be dysregulated in NB through MYCN-dependent transcriptional amplification causing transcriptional addiction.⁵

We reveal the dosage-sensitive T-box TF 2 (TBX2) gene, located on the highly recurrent gained chromosome 17q, to be involved in the adrenergic CRC in NB, rendering a selective advantage to tumor cells exhibiting 17q gain⁶ (Figure 1). Using an integrative epigenomics search for dosage-sensitive TFs marked by H3K27ac defined SEs on 17q, we found TBX2 to be associated by an SE in 26 NB cell lines, further confirmed by promotor-super-enhancer interactions determined by 4C-seq analysis. In keeping with the lineage-specific characteristics of SEs, TBX2 is highly expressed in NB cell lines and primary tumors as compared to other tumor entities and strongly upregulated in mouse neural crest derived MYCN overexpressing NB. Of further interest, gains implicating the TBX2 locus were most frequently found in NB. Moreover, in a single MYCN-amplified case an additional focal 17q23.2 amplification encompassing the TF TBX2 was observed, similar to the previously reported focal high-level gain of a chromosome 17q23 segment in the NB cell line MP-N-TS. Furthermore, TBX2 expression was correlated with worse overall and progression-free survival in two different large NB tumor cohorts supporting the role of TBX2 in NB tumor aggressiveness.

Figure 1.

DREAM target reactivation by core transcriptional regulators supports neuroblastoma growth. We prioritized the chromosome 17q T-box Transcription Factor 2 (TBX2) as an important factor in neuroblastoma development based on (1) regulation by a super-enhancer, (2) correlation of its expression with worse survival, (3) high expression in neuroblastoma as compared to other entities and (4) identification of a focal 17q23.2 amplification encompassing the TBX2 locus (left panel). TBX2 is a constituent of a core regulatory circuitry together with ‘V-myc myelocytomatosis viral related oncogene, neuroblastoma derived (avian)’ (MYCN), and other genes known in neuroblastoma development, resulting in transcriptional addiction and rendering a selective advantage to tumor cells (middle panel). TBX2 repression of protein p21 enforces MYCN controlled p21 repression, and both MYCN and TBX2 activate Forkhead box protein M1 (FOXM1) expression and activity, resulting in DREAM complex composition exchange and cell cycle progression (right panel). Combination treatment of JQ1 with THZ1 to target the transcriptional addiction vulnerability induced by highly expressed CRC genes such as TBX2 results in strong synergistic effect on cell growth and apoptosis (bottom).

The recent findings of essential CRCs in NB prompted us to assess TBX2 DNA-occupancy. TBX2 ChIP-sequencing revealed strong enrichment and overlap of MYCN/GATA Binding Protein 3 (GATA3)/Heart- and neural crest derivatives-expressed protein 2 (HAND2)/Paired-like homeobox 2b (PHOX2B) motifs and ChIP-seq peaks at enhancer regions. This compelling evidence of TBX2 functioning as a CRC gene was further supported by binding of TBX2 to its own SE constituent which is co-occupied by GATA3/HAND2/PHOX2B and binding to the SE constituent of the latter CRC members. The overlap with MYCN is coherent with the recent report of Zeid et al. where invasion of the transcriptional amplifier MYCN in dose-dependent fashion was noted based on canonical and non-canonical MYCN recognition sites.⁵ Very recently, Durbin et al. confirmed TBX2 as an essential TF part of CRC maintaining cell state in MYCN amplified NB.⁷

So far, the functional significance of each of the individual CRC members has remained unexplored. Transcriptome analysis upon TBX2 knockdown in a MYCN amplified and high expressing MYCN NB cell line revealed enrichment for phosphorylate tumor protein p53 targets in the set of repressed genes, while activated genes were enriched for cell cycle hallmarks including G2/M checkpoint, E2 factor (E2F), MYC(N) targets, mitosis and DNA replication. Cyclin-Dependent Kinase Inhibitor 1A (CDKN1A) (also known as protein p21) is significantly upregulated upon TBX2 knockdown, in line with the described role of TBX2 in bypassing senescence, and the observed decrease in colony formation capacity, proliferation and G1 phase arrest. Of further interest, we identified motif enrichment for Forkhead box protein M1(FOXM1), E2F and E2F binding partners in nearly half of all downregulated genes, the latter which was also enriched at the TBX2 DNA-occupancy sites. We further scrutinized the functional TBX2-MYCN relationship, and found stronger decrease in cell proliferation, increased G1-phase arrest and a synergistic effect on expression levels of gene sets implicated in cell cycle and the “dimerization partner, RB-like, E2F and multi-vulval class B“ (DREAM)-E2F-FOXM1 complex upon combined TBX2 and MYCN knockdown. Based on these findings, we propose a model where TBX2 repression of CDKN1A enforces MYCN controlled CDKN1A repression, followed by exchange of DREAM complex composition and cell cycle progression.⁸ More directly, both MYCN and TBX2 activate FOXM1 expression and activity, promoting the switch of DREAM complex to an active form.

Our study also provided insight into a rational for targeting the transcriptional addiction vulnerability induced by highly expressed CRC genes such as TBX2. Based on this assumption, we combined cyclin-dependent kinase 7 (CDK7) inhibitor THZ1 with the bromo-domain inhibitor JQ1, both previously shown to affect transcription of SE-associated lineage-dependency genes in NB,^9,10 and observed strong synergistic effect on cell growth and apoptosis in eight NB cell lines and two primary tumor-derived cell lines (organoids) established prior to therapy. These observations were recently confirmed in vivo using a mouse xenograft approach.⁷ RNA-sequencing following THZ1 and JQ1 combination treatment induced synergistic downregulation of all predefined CRC genes, collapse of the CRC and loss of FOXM1 repression of E2F/DREAM complex core genes. Taken together, we propose that the MYCN-TBX2 CRC represents an important novel therapeutic vulnerability for high-risk NB, warranting future clinical trials to assess available bromo-domain and CDK7 inhibitors.

Funding Statement

This work was supported by the Belgian Foundation Against Cancer under Grant project 2015-146; Ghent University under Grant BOF10/GOA/019 and BOF16/GOA/23; the Belgian Program of Interuniversity Poles of Attraction under Grant IUAP Phase VII – P7/03; the Fund for Scientific Research Flanders under Grant Research projects G053012N, G050712N, G051516N, G021415N; the “Kom op tegen Kanker“ (Stand up to Cancer) the Flemish Cancer Society; the European Union H2020 under Grant OPTIMIZE-NB GOD9415N and TRANSCAN-ON THE TRAC GOD8815N; ‘Kinderkankerfonds’; Olivia Fund; and Villa Joep

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.