Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Nov;19(11):1818-1830.
doi: 10.1158/1541-7786.MCR-21-0259. Epub 2021 Jul 20.

Chemical Screen Identifies Diverse and Novel Histone Deacetylase Inhibitors as Repressors of NUT Function: Implications for NUT Carcinoma Pathogenesis and Treatment

Affiliations

Chemical Screen Identifies Diverse and Novel Histone Deacetylase Inhibitors as Repressors of NUT Function: Implications for NUT Carcinoma Pathogenesis and Treatment

Hitoshi Shiota et al. Mol Cancer Res. 2021 Nov.

Abstract

NUT carcinoma (NC), characterized most commonly by the BRD4-NUTM1 fusion, is a rare, aggressive variant of squamous carcinoma with no effective treatment. BRD4-NUT drives growth and maintains the poorly differentiated state of NC by activating pro-growth genes such as MYC, through the formation of massive, hyperacetylated, superenhancer-like domains termed megadomains. BRD4-NUT-mediated hyperacetylation of chromatin is facilitated by the chromatin-targeting tandem bromodomains of BRD4, combined with NUT, which recruits the histone acetyltransferase, p300. Here, we developed a high-throughput small-molecule screen to identify inhibitors of transcriptional activation by NUT. In this dCAS9-based GFP-reporter assay, the strongest hits were diverse histone deacetylase (HDAC) inhibitors. Two structurally unrelated HDAC inhibitors, panobinostat and the novel compound, IRBM6, both repressed growth and induced differentiation of NC cells in proportion to their inhibition of NUT transcriptional activity. These two compounds repressed transcription of megadomain-associated oncogenic genes, such as MYC and SOX2, while upregulating pro-differentiation, non-megadomain-associated genes, including JUN, FOS, and key cell-cycle regulators, such as CDKN1A. The transcriptional changes correlate with depletion of BRD4-NUT from megadomains, and redistribution of the p300/CBP-associated chromatin acetylation mark, H3K27ac, away from megadomains toward regular enhancer regions previously populated by H3K27ac. In NC xenograft models, we demonstrated that suppression of tumor growth by panobinostat was comparable with that of bromodomain inhibition, and when combined they improved both survival and growth suppression. IMPLICATIONS: The findings provide mechanistic and preclinical rationale for the use of HDAC inhibitors, alone or combined with other agents, in the treatment of NUT carcinoma.

PubMed Disclaimer

Conflict of interest statement

Competing interests

P.A. Cole has ownership interest (including patents) in Acylin Therapeutics and has consulted for Abbvie and Constellation Pharmaceuticals. C.A. French is a consultant at Boehringer-Ingelheim, and reports receiving a commercial research grant from Boehringer-Ingelheim, Glaxo-Smith-Kline, and Constellation Pharmaceuticals (Cambridge, MA). No potential conflicts of interest were disclosed by the other authors.

Figures

Figure 1.
Figure 1.
Chemical screen for small molecules that specifically inhibit the dCas9-NUT-dependent reporter gene activation. A. Schematic overview of the screen workflow. B. GFP expression in the presence (DMSO) or absence (ASV) of dCas9-NUT. ASV-induced dCas9-NUT degradation through a SMASh tag suppress GFP expression, confirming the specificity of the assay. C. Scatter plot of the Z-score of the GFP/mCherry fluorescent ratio. Weak hits (W) (−3 > Z > −4) are represented by magenta, medium hits (M) (−4 > Z > −5) are in green, and strong hits (S) (Z < −5) in red. Other data points shown in gray are non-hits (N/A).
Figure 2.
Figure 2.
Inhibition of NUT-dependent gene activation and growth of BRD4-NUT-positive NC cells with HDAC inhibitors. A. Inhibitory effect of entinostat (non-hit), nexturastat A (weak hit), apicidin (not included in screen), and panobinostat (strong hit) on dCas9-NUT-dependent reporter expression. Ratios of GFP/mCherry fluorescence were calculated. The results are from 3 biological replicates. B. IC50 and corresponding C. dose response of a series of indicated HDAC inhibitors measured on the NC cell line TC-797 by CellTiterGlo assay (Promega). Data are expressed as mean +/−SD.
Figure 3.
Figure 3.
Panobinostat induces growth arrest and differentiation through altering transcriptome in NC. A. Hemacolor staining, left, and cytokeratin (differentiation marker) immunoblots, right, of NC cells (TC-797, PER-403, 10–15) treated with panobinostat (15nM) for 72h. Scale bar, 20μm. B. Differentially expressed genes identified by RNA-seq of TC-797 and 10–15 cells treated with panobinostat (30nM, 4h). Venn diagram, left, shows the number of up- or down-regulated genes in TC-797, 10–15, or both NC cell lines treated with panobinostat compared with DMSO. The heatmap, right, shows differential expression of all affected genes for each treatment. C. Heatmap of row-scaled Normalized Enrichment Scores (NES) for Hallmark or Genome Ontology (GO) gene sets with family-wise error < 0.25 in comparisons by GSEA of pre-ranked log2 fold change in panobinostat treated cells relative to DMSO control.
Figure 4.
Figure 4.
Panobinostat treatment induces redistribution of BRD4-NUT and H3K27ac in NC cells. A. Images of immunofluorescence showing BRD4-NUT (green) and H3K27ac (red) localization in TC-797 cells treated with 30 nM of panobinostat for 1, 2, or 4 h. No further changes were seen at 8h (not shown). Control cells were treated with DMSO for 4 h. Scale bar, 10μm. B. Heatmaps of ChIP-seq performed on TC-797 cells treated with panobinostat (30nM) or DMSO for 4h. Top, H3K27Ac ChIPseq signal heatmaps of 4,741 H3K27Ac desert regions. Bottom, heatmaps of randomly sampled H3K27Ac peaks (n=2,754, 10% of 27,545 peaks). C-F. ChIP-seq signal profiles of BRD4-NUT and H3K27ac in TC-797 cells treated with panobinostat or DMSO for 4h obtained from ChIP-seq data as in B. ChIP-seq signal profiles within megadomain regions and regular peak regions (“regular”) as indicated are shown. Megadomains are defined by the inflection point obtained from the H3K27ac superset of peaks in DMSO-treated cells, for which the cutoff is 55kb. Regular domains are defined as non-megadomain peaks for the indicated antibodies used for ChIP. Number of peaks used for each profile are indicated. C. Enrichment profile in H3K27ac-associated megadomains. D. Enrichment profile in NUT-associated megadomains identified using the 55kb cutoff. E. Enrichment profile in H3K27ac-associated regular domains. F. Enrichment profile in NUT-associated regular domains. G. Integrated genome browser (IGV) track of regions around MYC/SOX2 and JUN/CDKN1A loci as examples of megadomain (MD) and non-megadomain regions, respectively. BRD4-NUT and H3K27ac ChIP-seq signals are colored rust and green, respectively. Each track shown for ChIP-seq and RNA-seq is from replicate 1 of two replicates. Black bars indicate domains (megadomain or regular, as indicated) identified for the provided conditions. Megadomains indicated by the black bars include only overlapping H3K27ac and NUT-associated megadomains. Regular domains, also indicated by black bars, are shown only for H3K27ac peaks. RNA-seq tracks are colored pink and blue. pano, panobinostat; reg, regular domain.
Figure 5.
Figure 5.
Improved in vivo growth repression and prolonged survival of mouse xenograft model of NC combining BET and HDAC inhibitors. A. Tumor growth in luciferase-expressing PER-403 NC cells engrafted into NOD-scid-GAMMA mice measured by bioluminescence imaging. Green and red triangles indicate treatment start and stop, respectively. B. Survival of PER-403 xenograft mice receiving various treatments (CPI-0610 (pelabresib): BETi, LBH589 (panobinostat)). The table summarizes median survival and p-values for each treatment. Green and red triangles are as described in A. C. Representative (4 of 8 mice per group) BLI images in prone mice. Color scale: minimum, 1 × 106; maximum, 5 × 107.

Similar articles

Cited by

References

    1. French CA, Miyoshi I, Kubonishi I, Grier HE, Perez-Atayde AR, Fletcher JA. Brd4-nut fusion oncogene: A novel mechanism in aggressive carcinoma. Cancer Res. 2003;63(2):304–7. - PubMed
    1. French CA, Kutok JL, Faquin WC, Toretsky JA, Antonescu CR, Griffin CA, et al. Midline carcinoma of children and young adults with nut rearrangement. J Clin Oncol. 2004;22(20):4135–9. Epub 2004/10/16. - PubMed
    1. French CA. Pathogenesis of nut midline carcinoma. Annu Rev Pathol. 2012;7:247–65. Epub 2011/10/25. - PubMed
    1. Bauer DE, Mitchell CM, Strait KM, Lathan CS, Stelow EB, Luer SC, et al. Clinicopathologic features and long-term outcomes of nut midline carcinoma. Clin Cancer Res. 2012;18(20):5773–9. Epub 2012/08/17. - PMC - PubMed
    1. Chau NG, Ma C, Danga K, Al-Sayegh H, Nardi V, Barrette R, et al. An anatomical site and genetic based prognostic model for patients with nut midline carcinoma: Analysis of 124 patients. JNCI Cancer Spectrum. 2019. - PMC - PubMed

Publication types

MeSH terms