ETV4-Dependent Transcriptional Plasticity Maintains MYC Expression and Results in IMiD Resistance in Multiple Myeloma

Abstract

Immunomodulatory drugs (IMiD) are a backbone therapy for multiple myeloma (MM). Despite their efficacy, most patients develop resistance, and the mechanisms are not fully defined. Here, we show that IMiD responses are directed by IMiD-dependent degradation of IKZF1 and IKZF3 that bind to enhancers necessary to sustain the expression of MYC and other myeloma oncogenes. IMiD treatment universally depleted chromatin-bound IKZF1, but eviction of P300 and BRD4 coactivators only occurred in IMiD-sensitive cells. IKZF1-bound enhancers overlapped other transcription factor binding motifs, including ETV4. Chromatin immunoprecipitation sequencing showed that ETV4 bound to the same enhancers as IKZF1, and ETV4 CRISPR/Cas9-mediated ablation resulted in sensitization of IMiD-resistant MM. ETV4 expression is associated with IMiD resistance in cell lines, poor prognosis in patients, and is upregulated at relapse. These data indicate that ETV4 alleviates IKZF1 and IKZF3 dependency in MM by maintaining oncogenic enhancer activity and identify transcriptional plasticity as a previously unrecognized mechanism of IMiD resistance.

Significance: We show that IKZF1-bound enhancers are critical for IMiD efficacy and that the factor ETV4 can bind the same enhancers and substitute for IKZF1 and mediate IMiD resistance by maintaining MYC and other oncogenes. These data implicate transcription factor redundancy as a previously unrecognized mode of IMiD resistance in MM. See related article by Welsh, Barwick, et al., p. 34. See related commentary by Yun and Cleveland, p. 5. This article is featured in Selected Articles from This Issue, p. 4.

Figure 1.

Lenalidomide (Len) downregulates MYC in Len-sensitive MM cells. A, MTT assay of MM cells treated with Len (10 μmol/L) at the indicated time points. The line shows the threshold separating sensitive (Len sen) from resistant (Len res) cell lines. B, Immunoblot analysis of IKZF3, MYC, CRBN, and GAPDH in 7 MM cell lines treated with lenalidomide (10 μmol/L) at the indicated time points grouped by lenalidomide response. C, Scatter plot of RNA-seq gene-expression changes in MM1S (top) and OPM2 (bottom) treated with lenalidomide for 12 and 24 hours, respectively. Differentially expressed genes (DEG; FDR ≤0.05) are shown in red (upregulated) and blue (downregulated). Ctrl, control; FPKM, Fragments Per Kilobase Per Million reads. D, Heat map of lenalidomide DEGs in MM1S (top) and OPM2. Samples are represented in columns and genes by rows. E, Gene set enrichment analysis (GSEA) enrichment scores for the gene sets: HALLMARK_INTERFERON_ALPHA_RESPONSE (top) and MYC_TARGET_V1 (bottom) in both MM1S (blue) and OPM2 (green). Lenalidomide-induced gene-expression changes are sorted from most upregulated (left) to most downregulated (right) and the overlap with gene set genes are shown in blue and green for MM1S and OPM2, respectively. F, Expression of selected DEGs measured by RNA-seq. Data presented are the mean ± SEM (A) or ± SD (F) with N ≥ 3 (A) N = 2 (F).

Figure 2.

IKZF1 binds MM enhancers and super-enhancers. A, Heat map of IKZF1 of ChIP-seq peaks (left) and the input control (right) in MM1S cells. The scale is expressed in reads per million (RPM). B, IKZF1 binding at the MYC (red) locus compared with the input control. Translocation breakpoints (Tr, black) and the frequency of copy-number variant gains (CNV; gray) from CoMMpass samples are shown (bottom). C, P300 (blue) super-enhancer analysis in MM1S (top) and RPMI8226 (bottom). Binding of IKZF1 (orange), H3K27ac (gray), and BRD4 (green) to each P300-defined region is shown using the normalized RPM. D, Heat maps of P300, H3K27ac, BRD4, and IKZF1 in P300-defined enhancers from MM1S (top) and RPMI8226 (bottom). All heat maps were sorted by the size of the P300-defined super-enhancer (SE; annotated left). E, Overlap of H3K27ac, BRD4, and IKZF1 with P300 enhancers (blue) and super-enhancers (black). F, Genome plots of CCND2 (left) and BCL2L1 (right) loci showing P300 (blue), H3K27ac (gray), BRD4 (green), and IKZF1 (orange) binding in MM1S (top) and RPMI8226 (bottom). Scale is RPM.

Figure 3.

Lenalidomide-induced loss of enhancer coactivators in IMiD-sensitive cells. A–C, Scatter plots of IKZF1 (A), P300 (B), and BRD4 (C) ChIP-seq signal in MM1S control (ctrl) and lenalidomide (Len)-treated cells (10 μmol/L, 24 hours). D–F, Scatter plots of IKZF1 (D), P300 (E), and BRD4 (F) in RPMI8226 ctrl and Len-treated cells. The signal at each region is shown to the right of the scatter plots for control and lenalidomide treatment. G, Scatter plot of lenalidomide-induced fold changes at regions cobound with IKZF1 and P300 (left), as well as a binned analysis of fold changes ranked by lenalidomide-induce IKZF1 fold change (right) in MM1S. Correlation is shown (Spearman ρ). H, Scatter plot and binned analysis of IKZF1 and BRD4 cobound regions for MM1S as in part G. I, Scatter plot and binned analysis of IKZF1 and P300 cobound regions for RPMI8226 as in part G. J, Scatter plot and binned analysis of IKZF1 and BRD4 cobound regions for RPMI8226 as in G. K, Genome plot of DUSP22 and IRF4 showing IKZF1 (orange), P300 (blue), and BRD4 (green) ChIP-seq in MM1S (left) and RPMI8226 (right). The change in lenalidomide-treated samples is shown in darker tones, and the scale is reads per million (RPM).

Figure 4.

IKZF1 shares similar motifs with ETS factors that correlate with IMiD resistance. A, Correlation matrix of motifs at IKZF1-bound regions in RPMI8226 (left) with select transcription factor families annotated in color (key far left). The frequency of overlap with IKZF1-bound regions is shown (right). Only motifs for expressed (≥1 FPKM) transcription factors in RPMI8226 significantly (FDR ≤10⁻¹⁰, odds ratio ≥1.5) enriched at IKZF1-bound regions are shown. B, Logo plots of select motifs enriched at IKZF1-bound regions in RPMI8226. The frequency of overlap with IKZF1-bound regions is shown in parenthesis. C, Pearson correlation (R) of transcription factor expression as measured by RNA-seq with lenalidomide (Len; 10 μmol/L)-induced change in growth index on Day 6 as measured by MTT assay in 12 MM cell lines from Fig. 1A. D, Progression-free survival (PFS) and hazard ratio (HR) for the expression of each transcription factor shown in C and observed in IMiD-treated patients from CoMMpass. E, Correlation of ETV4 expression as determined by RNA-seq with Len-induced change in growth index in 12 MM cell lines as determined by MTT assay on Day 6 (from Fig. 1A).

Figure 5.

ETV4 is cobound at IKZF1 enhancers. A, Heat map of ETV4 ChIP-seq in RPMI8226 cells. B, Scatter plot of genomic binding occupancy for IKZF1 and ETV4 in RPMI8226 cells with color denoting if the region is bound by IKZF1 only, ETV4 only, or both (key right). C, Venn diagram of IKZF1 (orange) and ETV4 (red) bound regions in RPMI8226 cells. Note: the 11,847 ETV4-bound regions overlap 12,419 IKZF1-bound regions. D, Overlap of ETV4 sites with IKZF1, H3K27ac, P300, and BRD4 enriched regions in RPMI8226. E, Genomic plot of P300, H3K27ac, BRD4, IKZF1, and ETV4 at the IGH (left) and PIM1 (right) super-enhancers.

Figure 6.

ETV4 mediates lenalidomide resistance in MM cells. A, Western blot of ETV4 and GAPDH expression in MM1S, RPMI8226 with sgRNAs targeting ETV4 or a nontargeting (NT) control, and AMO1 MM cells. B,ETV4 RNA expression by RT-qPCR in RPMI8226 control cells (Cas9) and ETV4KO (sgETV4-1, sgETV-2, and sgETV4-3). C, Western blot for MYC, IKZF3, and GAPDH in RPMI8226 nontargeting control or sgRNA ETV4-2 and ETV4-3 cells treated with lenalidomide (Len; 10 μmol/L). D, F, H, Change in growth index measured by MTT assay at the indicated time points in RPMI8226 (D), L363 (F) and ARD (H) cell lines transduced with Cas9 control or sgRNA ETV4-2 and ETV4-3 cells after treatment with lenalidomide (Len; 10 μmol/L). E, G, Western blot analysis for NIK/MAP3K14, NFkB subunit (p100), ETV4, and MYC in L363 (E) and ARD cell lines (G) transduced with Cas9 control and ETV4KO cells treated with 10 μmol/L Len. Note the downregulation of NIK/MAP3K14 and lack of p100 degradation in ETV4KO cells in response to lenalidomide. F, Viability of L363 Cas9 and ETV4KO after Len treatment. The data presented are mean ± SEM.

Figure 7.

ETV4 expression is associated with poor outcome and a proliferation gene-expression program. A, Kaplan–Meier PFS (left) and OS (right) of CoMMpass patients treated with IMiD as part of their first-line therapy (N = 567) stratified by ETV4 expression of 1 FPKM. B, PFS (left) and OS (right) analysis as in A applied to POLLUX patients treated with daratumumab, lenalidomide, and dexamethasone. C, Expression of ETV4, IKZF1, IKZF3, CRBN, IRF4, and MYC in paired samples collected at diagnosis (ND) and relapse (RR; N = 103) from CoMMpass patients (N = 47). D, Expression analysis of paired samples (N = 28) from POLLUX patients (N = 14) as in C. E, GSEA of gene sets associated with ETV4 expression in NDMM and RRMM samples from CoMMpass. Only gene sets with an FDR ≤0.05 in both analyses are shown. P values were determined using a Cox proportional hazards model Wald test (A and B) or a linear regression with a covariate for patients (C and D).