Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 11 (1), 740

A Combination Strategy Targeting Enhancer Plasticity Exerts Synergistic Lethality Against BETi-resistant Leukemia Cells

Affiliations

A Combination Strategy Targeting Enhancer Plasticity Exerts Synergistic Lethality Against BETi-resistant Leukemia Cells

Lei Guo et al. Nat Commun.

Abstract

Primary and acquired drug resistance imposes a major threat to achieving optimized clinical outcomes during cancer treatment. Aberrant changes in epigenetic modifications are closely involved in drug resistance of tumor cells. Using BET inhibitor (BETi) resistant leukemia cells as a model system, we demonstrated herein that genome-wide enhancer remodeling played a pivotal role in driving therapeutic resistance via compensational re-expression of pro-survival genes. Capitalizing on the CRISPR interference technology, we identified the second intron of IncRNA, PVT1, as a unique bona fide gained enhancer that restored MYC transcription independent of BRD4 recruitment in leukemia. A combined BETi and CDK7 inhibitor treatment abolished MYC transcription by impeding RNAPII loading without affecting PVT1-mediated chromatin looping at the MYC locus in BETi-resistant leukemia cells. Together, our findings have established the feasibility of targeting enhancer plasticity to overcome drug resistance associated with epigenetic therapies.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1. BETi-resistant leukemia cells undergo BRD4-independent enhancer remodeling.
a Heatmap representing the IC50 values of the indicated leukemia cell lines to a BRD4 inhibitor JQ1. The IC50 values were obtained from the published database PHARMACODB. b The principal component analysis (PCA) of enhancer region distributions in the indicated BETi-sensitive and BETi-resistant leukemia cells. c Smoothed scatter plot representation of H3K27ac-enriched regions in paired BETi-sensitive (parental, x-axis) and BETi-resistant (y-axis) murine AF9 AML cells. Color code represents the dot density. d Smoothed scatter plot representation of differential H3K27ac-enriched regions and Brd4 enriched regions in paired BETi-sensitive (parental) and BETi-resistant murine AF9 AML cells. Red highlighted region: genomic regions displayed increased H3K27ac enrichment with no change or decreased Brd4 binding in BETi-resistant murine AF9 AML cells. All the publicly available H3K27ac and BRD4 ChIP-seq datasets used in Fig. 1b–d were summarized in Supplementary Table 1. Color code represents the dot density. e The histogram showing the distribution of distance between the genomic regions highlighted within the red box in Fig. 1d and their closest genes. f Genomic Regions Enrichment of Annotations Tool (GREAT) analysis of genomic regions highlighted in the red box of Fig. 1d.
Fig. 2
Fig. 2. Dual inhibition of CDK7 and BRD4 synergistically suppresses the growth of BETi-resistant leukemia cells.
a Heatmap representation of Fraction Affected (FA) and Bliss interaction index across five-point dose range of a BET inhibitor (I-BET151) and a CDK7 inhibitor (THZ1) in K562 and Jurkat cells. Mean values of triple biological experiments were shown. (b and c) Proliferation analysis of K562 and Jurkat cells (b) or murine AF9 AML cells (c) treated with DMSO (black), THZ1 (green), I-BET151 (blue), and the combination of THZ1 + I-BET151 (red). The concentrations of I-BET151 and THZ1 were 2.5 μM and 12.5 nM for K562, 2.5 μM and 3.125 nM for Jurkat, 2.5 μM and 50 nM for murine AF9 AML cells, respectively. Data were shown as mean ± S.D; n = 6 from 3 independent assays, **P = 0.00002 (K562), 0.00009 (Jurkat) and 0.000006 (AF9 resistant), by two-tailed Student’s t test. d, e Immunoblot analysis on apoptosis-related marker PARP and cleaved caspase 3 (C/Caspase3) in K562 (d, top), Jurkat (d, bottom) and murine AF9 AML cells (e) treated with DMSO, THZ1, I-BET151, and the combination of THZ1 + I-BET151. The inhibitor concentrations were the same as shown in Fig. 2b, c. Three independent assays were performed. f, g Quantification of proliferation of K562, Jurkat cells (f) and murine AF9 AML cells (g) transduced with shRNAs targeting CDK7 and/or BRD4. Data were shown as mean ± S.D; n = 8 from three independent assays, **P = 0.00003 (K562), 0.00001 (Jurkat) and 0.000002 (AF9 resistant), by two-tailed Student’s t test.
Fig. 3
Fig. 3. Synthetic lethality imposed by dual inhibition of CDK7 and BRD4 toward BETi-resistant leukemia cells in vivo.
a Schematic of the experimental design for CDK7 and / or BRD4 inhibitor treatment in vivo. b Kaplan–Meier survival curves of recipient mice transferred with BETi-resistant murine AF9 AML cells and then treated with DMSO (black), THZ1 (green, 10 mg/kg), I-BET151 (blue, 10 mg/kg), or a combination of THZ1 with I-BET151 (red). n = 5 mice. *P = 0.0112, by log rank Mantel-Cox test. c Representative images and weights of spleens collected from recipient mice transferred with BETi-resistant murine AF9 AML cells at 20 days after DMSO, THZ1, I-BET151, or combination treatment. n = 5 mice, P = 0.0049, by ANOVA with Dunnett’s post-hoc correction. d Percentage of YFP+ murine BETi-resistant AF9 AML cells collected from the spleen and bone marrow of the mice treated with DMSO (black), THZ1 (green), I-BET151 (blue), or a combination of THZ1 with I-BET151 (red). Data were shown as mean ± S.D; n = 3 mice, **P = 0.0044, by two-tailed Student’s t test. e Representative Giemsa staining of the peripheral blood collected from recipient mice transferred with BETi-resistant murine AF9 AML cells at 20 days after DMSO, THZ1, I-BET151, or combination treatment of THZ1 and I-BET151. n = 5 mice. Scale bar, 10 μm. f Representative Hematoxylin and eosin (HE) staining of liver, spleen, and bone marrow tissues collected from the recipient mice transferred with BETi-resistant murine AF9 AML cells at 20 days after DMSO, THZ1, I-BET151, or the combination treatment. n = 5 mice. Scale bar, 100 μm.
Fig. 4
Fig. 4. MYC identified as a critical target that mediates the synergistic lethality of I-BET151 and THZ1 in BETi-resistant leukemia cells.
a The PCA analysis of RNA-seq results in K562, Jurkat and BETi-resistant murine AF9 AML cells treated with DMSO (black), I-BET151 (blue), THZ1 (green), and the combination of I-BET151 + THZ1 (red) for 24 h. b Heatmap representation of differentially expressed genes (DEGs) identified between the DMSO, I-BET151, THZ1, and the combination treatment in K562 and murine AF9 BETi-resistant AML cells. DEGs were defined as q-value < = 0.05. Red and blue color stand for up- and down-regulated genes, respectively. (c) Gene Set Enrichment Analysis (GSEA) of DEGs identified from Fig. 4b. d GSEA presentation of MYC or E2F targeted genes in the identified DEGs. Genes were ranked by fold changes. e Real-time qPCR analysis of super-enhancer related genes in K562 and murine BETi-resistant AF9 AML cells after DMSO (black), I-BET151 (blue), THZ1 (green), and the combination treatment (red) for 24 h. Data were shown as mean ± S.D; n = 4 from four independent assays. **P = 0.002 (MYC), 0.0015 (MYB), 0.01 (MEIS1) and 0.0006 (LMO2), by two-tailed Student’s t test. f Representative Western blotting showing the protein levels of MYC, CDK7, BRD4, phosphorylated and total RNAPII in K562, murine BETi-sensitive or resistant AF9 AML cells after DMSO, I-BET151, THZ1, and the combination treatment for 24 h. GAPDH was used as control. n = 3 from 3 independent assays. g Immunoblot analysis of MYC expression in K562 and BETi-resistant murine AF9 AML cells transduced with the empty vector (EV) or a lentivirus encoding MYC. Cells were treated with DMSO or combination (Combo) for 24 h. Three independent assays were performed. h Proliferation rate of K562 and BETi-resistant murine AF9 AML cells expressing the empty vector (EV) or MYC after DMSO or the combination treatment (Combo) for 3 days. Data were shown as mean ± S.D; n = 8 from 4 independent assays. Drug concentrations used in Fig. 4 were the same as in Fig. 2b, c for each cell line.
Fig. 5
Fig. 5. A de novo remodeled PVT1 enhancer in BETi-resistant leukemia cells.
a Differential enrichment of H3K27ac at all identified MYC enhancers between parental (sensitive) and BETi-resistant murine AF9 AML cells. Red arrow: The genomic region within Pvt1 locus at the top differential enriched regions between parental and BETi-resistant murine AF9 AML cells. b Genome browser views of H3K27ac enrichment in the PVT1 locus (left) and BENC enhancers (E1–E5, right) of MYC in the indicated leukemia cell lines with relatively high (red) or low (blue) IC50 values of BETi. Green highlighted regions represented MYC enhancers. c Genome browser view of H3K27ac enrichment at Myc Pvt1 locus (top) and BENC enhancers (E1–E5, bottom) in BETi-sensitive (Vehicle, blue) and resistant (red) murine AF9 AML cells. Green highlighted regions were identified MYC enhancers. d Genome browser views of H3K27ac and ATAC-seq signals of MYC PVT1 enhancers in K562 (red) and THP1 (blue) cells. Green highlighted regions were identified MYC enhancers.
Fig. 6
Fig. 6. PVT1 enhancer positively regulates MYC expression in BETi-resistant leukemia cells.
a Schematic depicting the use of dCas9-KRAB-based CRISRPi to target MYC-associated PVT1 or BENC enhancers. b ChIP-qPCR analysis of H3K27ac enrichment in K562 cells transduced with a scrambled sgRNA or sgRNA targeted to the PVT1 and BENC loci. Data were shown as mean ± S.D; n = 4 from four independent assays, **P = 0.01 (E1), 0.0007 (E2), 0.0006 (E3), 0.0001 (E4), and 0.01 (PVT1), by two-tailed Student’s t test. ce Transcription (c), protein (d) levels and cell viability (e) of K562 cells expressing dCas9-KRAB and two independent sgRNAs targeted to PVT1 after BETi treatment. K562 cells transduced with dCas9-KRAB and scrambled sgRNAs were used as control. Data were shown as mean ± S.D; n = 3 from three independent assays, **P = 0.0002 (c) and 0.0004 (e), by two-tailed Student’s t test. f Schematics depicting the use of the dCas9-p300Core-based CRISRPa system to target MYC-associated PVT1 enhancers. g ChIP-qPCR analysis of H3K27ac enrichment in THP1, MOLM-13 and murine MLL-AF9 parental cells transduced with a scrambled sgRNA or sgRNA targeted to the PVT1 locus. Data were shown as mean ± S.D; n = 3 from three independent assays, **P = 0.028 (AF9), 0.012 (MOLM-13) and 0.011 (THP1) by two-tailed Student’s t test. hj Comparison of gene transcription (h), protein expression (i) levels and dose response curves (j) of THP1 cells expressing dCas9-p300Core and two independent sgRNAs targeted to PVT1 after BETi treatment. THP1 cells transduced with dCas9-p300Core and scrambled sgRNAs were used as control. Data were shown as mean ± S.D; n = 3 from three independent assays, **P = 0.0001 (sg-scram) and 0.0005 (sg-PVT1), by two-tailed Student’s t test.
Fig. 7
Fig. 7. THZ1 inhibits MYC transcription by disrupting RNAPII loading at a BETi-resistant specific enhancer-promoter loop in BETi-resistant leukemia cells.
a Heatmap showing HiC signals for the K562 (top) and THP1 (bottom) cell lines at the MYC locus (arrow) with flanking regions (chr8: 128,700–129,000 kb). Circled region indicated the PVT1 enhancer. b Genome browser views of H3K27ac HiChIP and RNAPII ChIP-seq signals within the MYC and PVT1 loci. The heatmaps were generated using Juicebox. Circled regions indicated the PTV1 enhancer. c 4C-seq of K562 cells treated with DMSO (black), THZ1 (green), I-BET151 (blue), and a combination of THZ1 and I-BET151 (red) for 24 h. Viewpoints were selected at the MYC promoter. Red arrow indicated the PVT1 locus (left). The normalized 4C reads per fragment at the PVT1 locus were shown as boxplots (right). n = 3 biological samples; NS: Not significant, by One-way ANOVA with Dunnett’s post hoc correction.For the boxplots, bounds of the box spans from 25 to 75% percentile, center line within box represents median. Whiskers represent median ± 1.5 times interquartile range. Dots represent outliers. d ChIP-qPCR analysis of RNAPII enrichment (red) at transcription starting site (TSS) and genebody of MYC in K562 cells treated with DMSO, THZ1, I-BET151, and a combination of THZ1 and I-BET151 for 24 h. Chromatin pulled down with IgG was used as control (black). Data were shown as mean ± S.D; n = 3 from 3 independent assays. **P = 0.014 and 0.026 (K562, TSS and gene body); p = 0.01 and 0.007 (AF9, TSS and gene body), by two-tailed Student’s t test. Drug concentrations used in Fig. 6c, d were the same as in Fig. 2b, c for each cell line. e Genome browser views showing the enrichment of different transcription factors (TFs) at the PVT1 locus. Hi-C, H3K27ac HiChIP and ChIP-seq data of the corresponding TFs and RNAPII were obtained from publicly available data listed in Supplementary Table 1. f H3K27ac HiChIP intensities at the PVT1 and MYC loci in BETi-resistant murine MLL-AF9 AML cells treated with DMSO (blue) or LSD1i (red).
Fig. 8
Fig. 8. A tentative enhancer remodeling model to explain the synergistic effect of BETi and THZ1 on BETi-resistant leukemia.
MYC expression is regulated by the classic super-enhancer BENC (E1-E5) in BETi-sensitive leukemia cells, which is mediated by BRD4 binding. BETi blocks the BRD4 binding to its genomic targets and subsequently inhibits the expression of MYC and cell growth. Long-term drug treatment or primary resistance may restore MYC expression by enhancer remodeling: PVT1 acts as a de novo BRD4 binding-independent enhancer, which can recruit other transcription factors and RNA Polymerase II (Pol II) to the MYC promoter and initiate MYC expression. THZ1 treatment reduces the Pol II occupancy to suppress the re-activated MYC transcription.

Similar articles

See all similar articles

References

    1. Mansoori B, Mohammadi A, Davudian S, Shirjang S, Baradaran B. The different mechanisms of cancer drug resistance: a brief review. Adv. Pharm. Bull. 2017;7:339–348. doi: 10.15171/apb.2017.041. - DOI - PMC - PubMed
    1. Konieczkowski DJ, Johannessen CM, Garraway LA. A convergence-based framework for cancer drug resistance. Cancer Cell. 2018;33:801–815. doi: 10.1016/j.ccell.2018.03.025. - DOI - PMC - PubMed
    1. Holohan C, Van Schaeybroeck S, Longley DB, Johnston PG. Cancer drug resistance: an evolving paradigm. Nat. Rev. Cancer. 2013;13:714–726. doi: 10.1038/nrc3599. - DOI - PubMed
    1. Zanconato F, et al. Transcriptional addiction in cancer cells is mediated by YAP/TAZ through BRD4. Nat. Med. 2018;24:1599–1610. doi: 10.1038/s41591-018-0158-8. - DOI - PMC - PubMed
    1. Takeda DY, et al. A somatically acquired enhancer of the androgen receptor is a noncoding driver in advanced prostate. Cancer Cell. 2018;174:422–432 e413. - PMC - PubMed

LinkOut - more resources

Feedback