ASCL1 is a MYCN- and LMO1-dependent member of the adrenergic neuroblastoma core regulatory circuitry

A heritable polymorphism within regulatory sequences of the LMO1 gene is associated with its elevated expression and increased susceptibility to develop neuroblastoma, but the oncogenic pathways downstream of the LMO1 transcriptional co-regulatory protein are unknown. Our ChIP-seq and RNA-seq analyses reveal that a key gene directly regulated by LMO1 and MYCN is ASCL1, which encodes a basic helix-loop-helix transcription factor. Regulatory elements controlling ASCL1 expression are bound by LMO1, MYCN and the transcription factors GATA3, HAND2, PHOX2B, TBX2 and ISL1—all members of the adrenergic (ADRN) neuroblastoma core regulatory circuitry (CRC). ASCL1 is required for neuroblastoma cell growth and arrest of differentiation. ASCL1 and LMO1 directly regulate the expression of CRC genes, indicating that ASCL1 is a member and LMO1 is a coregulator of the ADRN neuroblastoma CRC.

, 500 bps in total were retrieved from the peak summits (the highest point of each peak). Enriched motifs were analyzed using the MEME-chip package from MEME-Suite. (c, d) 293T cell were co-transfected with a plasmid encoding a C-terminus FLAG-HA-tagged LMO1 (pOZC-LMO1) or an empty vector (pOZC-EV) and plasmids encoding LDB1 (pCS2-LDB1) and GATA3 (pCS2-GATA3) (c). Kelly cells were transduced with pOZC-EV or pOZC-LMO1 by retrovirus infection (d). Immunoprecipitation (IP) experiments were then performed using anti-FLAG antibody. Antibodies against LMO1, LDB1 and GATA3 were used for immunoblotting. (e) The LMO1-bound gene loci were first selected in Jurkat cells. Density plots show the distribution of GATA3, H3K27ac and H3K4me1 signals at the LMO1-bound regions (± 5 kb from binding sites) in Jurkat and Kelly cells. Metagene plots show distribution of each transcription factor or histone mark signals at the LMO1-bound regions (± 5 kb from binding sites) in Jurkat and Kelly cells. The color scale shows the intensity of the distribution signal.  (P<1e-4), 500 bps in total were retrieved from the peak summits (the highest point of each peak). Enriched motifs were analyzed using the MEME-chip package from MEME-Suite. (d) The single cell RNA-seq data was reported by Furlan et al. 3 for mouse embryo at stage E12.5 and E13.5. Heatmap was generated using global normalization to Log2 TPM. Heatmap image represents gene expression of Ascl1 along with transcription factor genes that are dominantly expressed in the ADRN subtype of neuroblastoma (Ascl1, Hand2, Phox2b, Mycn, Tbx2, Isl1, Lmo1 and Gata3). Singles cells were ordered based on expression of Ascl1 from high to low. The color scale represents row z-scores. (i) Cell viability was measured after 3, 5 and 7 days of lentiviral transduction of shRNA. The growth rate (fold-change) over 7 days compared to day 3 was indicated (n=3 per group). (j) Cell viability on day 7 were shown in bar-chart. Data are presented as the fold-change of the shGFP control (n=3 per group). Values represent means ± SD for technical triplicates in g, i and j. The p-value by two-way ANOVA (repeated measurements) followed by Tukey's multiple comparisons post hoc test are indicated in g, i and j. ****, p-value<0.0001.

Supplementary Methods Cell cycle analysis
Approximately 1 million cells were counted, washed with PBS and fixed with 70% ethanol at 4°C for 1 hour. The cells were washed with PBS, treated with 50uL RNase H (final concentration at 2mg/mL) for 30 min at 37°C, and incubated with 200uL of 50ug/mL propidium iodide (PI). The cells were then subjected to flow cytometry analysis using the BD LSRII (BD Biosciences). The percentage of cells in each cell cycle phase was measured.

Motif analysis
Based on the ChIP-seq peaks called by MACS14, 500 bps in total were retrieved from the peak summits. Enriched motifs were analyzed using the MEME-chip package from MEME-Suite 4 using the Homo sapiens Comprehensive Model Collection database (HOCOMOCO) v11 and the TRANSFAC motif databases.
Gene set enrichment analysis (GSEA) GSEA 5 analysis of the normalized genes was performed using the log2 ratio of classes metric to rank genes for the comparison between control and LMO1 knockdown samples in SH-SY5Y cells. The direct target genes regulated by LMO1 in Kelly cells were first defined by ChIP-seq based on peak calling and by RNA-seq based on a significant gene expression change upon shRNA knockdown (absolute log2-fold-change ≥ or ≤1; p-value <0.05). The lists of upregulated genes and downregulated genes were used as gene sets.
Lamda phosphatase treatment 1X10 6 cells were lysed with RIPA buffer in the presence of protease inhibitor (Roche). A total of 1uL lambda phosphatase (NEB, P0753S) and 1X NEB buffer (final concentration) for protein metallophosphatases (PMP) and 1mM MnCl2 (final concentration) were added to lysate to make a total reaction volume of 50uL and incubated at 30 degree for 30 minutes. Lysate was then mixed with Laemmli sample buffer (Bio-Rad) with 10% betamercaptoethanol and boiled for 10 min at 95 °C for western blot analysis.

Calyculin A treatment
A total of 800,000 cells were seeded one day before in 6-well plates. The cells were then treated with Calyculin A (Cell Signaling, #9902) at different concentration for 30 min and then collected for Western blot analysis.

Statistics and Reproducibility
All the statistical analyses were done in GraphPad Prism software. A p-value less than 0.05 was considered statistically significant. The details of methods used can be found in each supplementary figure legend. Experiments in Supplementary Fig 1a, 2c, 2d, 3d, 4b, 4d, 4e, 4f and 7a were repeated three or more times. Experiments in Supplementary Fig 3c, 4c, 5a, 5b and 7e were repeated two times.