Max deletion destabilizes MYC protein and abrogates Eµ- Myc lymphomagenesis

Abstract

Although MAX is regarded as an obligate dimerization partner for MYC, its function in normal development and neoplasia is poorly defined. We show that B-cell-specific deletion of Max has a modest effect on B-cell development but completely abrogates Eµ-Myc-driven lymphomagenesis. While Max loss affects only a few hundred genes in normal B cells, it leads to the global down-regulation of Myc-activated genes in premalignant Eµ-Myc cells. We show that the balance between MYC-MAX and MNT-MAX interactions in B cells shifts in premalignant B cells toward a MYC-driven transcriptional program. Moreover, we found that MAX loss leads to a significant reduction in MYC protein levels and down-regulation of direct transcriptional targets, including regulators of MYC stability. This phenomenon is also observed in multiple cell lines treated with MYC-MAX dimerization inhibitors. Our work uncovers a layer of Myc autoregulation critical for lymphomagenesis yet partly dispensable for normal development.

Keywords: B-cell development; MAX; MYC stability; lymphomagenesis; transcription.

Figure 1.

Conditional deletion of Max in the B-cell lineage. (A) Schematic depicting the location of loxP sites at the Max locus. (B) Representative immunoblots for MAX in B220⁺ splenocytes from Max^fl/fl (wild-type [WT]) and Max^fl/fl mb1-cre (knockout [KO]) animals. (C) Representative flow plots showing B220⁺ and IgM⁺ populations in CD45 gated bone marrow (BM) cells. (D) Quantification of B lymphocyte precursor populations in Max WT (n = 5) and knockout (n = 6) BM. (E) Dual immunofluorescence (IF) for MAX and B220 in spleens. Quantification of MAX⁺ B220⁺ cells from MAX knockout spleens. n = 3. Yellow arrowheads indicate MAX⁺ B220⁺ cells in Max knockout. Total number of splenocytes (F) and CD19⁺ B220⁺ cells (G) in Max WT and knockout mice. WT n = 8; knockout n = 9. (H) IF staining for B220 and proliferation marker Ki67 in Max WT and knockout spleens. (I) IF staining of germinal centers (PNA) in Max WT and knockout spleens. Representative image. n = 3 animals per genotype. Scale bars, 100 µM. All error bars represent SEM.

Figure 2.

Requirement for Max in activated B cells and Eµ-Myc-induced lymphomagenesis. (A) Cell size as determined by forward scatter in WT unstimulated and LPS-activated Max WT and Max knockout (KO) B220⁺ cells. (B) Cell number 72 h after treatment in LPS-treated Max WT and knockout (n = 10 from five WT and knockout mice). (C,D) Cell viability (C) and apoptosis (D) in LPS-activated B cells (n = 9 from three WT and knockout mice) assessed using luciferase-based Cell Titer Glo and Caspase Glo assays. (E) Kaplan-Meier curve showing survival analysis of Max WT (n = 23) and knockout (n = 14) Eµ-Myc animals up to 180 d. P-value was calculated using log-rank (Mantel-Cox) test. (F) Analysis of B-cell precursor populations in Eµ-Myc BM. (G) Representative spleens from normal and Eµ-Myc mice. (H) Histogram of cell size of Eµ-Myc Max WT and knockout mice. (I,J) Total splenocyte number (I) and CD19⁺ B220⁺ cell number (J) in Eµ-Myc WT and knockout mice. n = 3 for each. (K) Proportion on mature IgD-positive B cells in Eµ-Myc spleens. n = 3. All error bars represent SEM.

Figure 3.

Gene expression profiling and genomic occupancy of MAX in B cells. (A,B) Hallmark gene set enrichment for pathways up-regulated (A) and down-regulated (B) in Max knockout (KO) B cells relative to WT B cells. (C) Next-generation sequencing (NGS) plots depicting genomic occupancy of MYC, MAX, MNT, and E2F1 in Max WT and knockout B cells ranked on expression changes. (D) Representative peaks for MAX, MYC, and MNT at E2F target Cbx5. (E) Motifs significantly enriched at MAX-bound genes. (F) Gene set enrichment for pathways enriched in MNT–MAX–MYC-bound genes. (G) Overlap of MAX-bound genes with genes that are differentially expressed in Max knockout B cells (false discovery rate <0.05 cutoff for differential expression).

Figure 4.

MYC stability upon MAX loss in normal and premalignant B cells. (A,B) mRNA (A) and protein (B) levels of Myc in WT and knockout (KO) B cells. n = 4 for WT and knockout. (C,D) Immunoblot (C) and quantification (D) of MYC levels following 2 h of MG132 treatment of Max WT and knockout B cells. n = 3 for WT and knockout. (E,F) Representative micrographs (E) and mean fluorescence intensity quantification (F) of MYC staining in sorted B220⁺ splenocytes from Max WT and knockout mice. n = 54 WT cells; n = 61 knockout cells. Scale bar, 100 µm. (G,H) Myc mRNA levels (G) and MYC protein levels (H) in Eµ-Myc Max WT and knockout cells. All error bars represent SEM.

Figure 5.

Max loss leads to a global down-regulation of the Myc signature in Eµ-Myc premaligant cells. (A) Summary of total differentially expressed genes in Max knockout normal and premalignant B cells. (B) Principal component analysis of all four genotypes. (EWT) Eu-Myc Max WT; (EKO) Eu-Myc Max knockout; (WT) Max WT; (KO) Max knockout. (C,D) Hallmark gene set enrichment for pathways up-regulated (C) and down-regulated (D) in Eu-Myc premalignant Max knockout B cells. (E) Heat map representation of global transcriptional changes in Eu-Myc premalignant Max WT and Max knockout cells. Representative genes from important categories are labeled. (F,G) Venn diagram (F) and volcano plot (G) of differentially expressed genes that are directly bound by MYC. (H,I) Volcano plots depicting the proportion of differentially expressed inflammation-related genes (H) and E2F target genes (I) that are directly bound by MYC in Eu-Myc premalignant cells.

Figure 6.

Factors mediating MYC degradation in the absence of MAX. (A) Normalized expression values of MYC-stabilizing genes in WT and Eµ-Myc B220⁺ cells. (B) Volcano plot showing expression changes for MYC stability genes and ChIP binding data for MYC in premalignant Eµ-Myc B220⁺. (C) qPCR for BTRC, CIP2A, and SET in DMSO- and Myci (10058-F4)-treated Daudi cells. (D) Western blot for pMYC S62 and CIP2A levels in Myci-treated Daudi cells. (E) Growth curves for HCT116 cells at different concentrations of Myci. (F) Western blot showing MYC levels in Myci-treated HCT116 cells. (G) MYC RNA levels in Myci-treated HCT116 cells. (H,I) Representative immunoblot of MYC levels (H) and MYC mRNA levels (I) in Omomyc- versus GFP-expressing HCT116 cells. All error bars represent SEM.

Figure 7.

MYC degradation in FBW7^−/− and MYC phospho-mutant-expressing cells. (A) Representative blot for MYC levels in HCT116 and HCT116 FBW7^−/− cells following a cycloheximide chase. (B) Determination of MYC half-life in Myci-treated control and FBW7^−/− cells. n = 3 experiments. (C) Half-life of MYC under each of the four conditions. (D) qPCR for CIP2A levels in HCT116 cells. n = 3. P = 0.048 for Myci versus control in WT. (E) Immunoblot of MYC levels following treatment with 75 µM Myci in MYC- or MYC^T58A-overexpressing HCT116 cells. (F) Quantification of MYC levels following MYCI treatment in MYC- versus MYC^T58A-expressing HCT116. n = 3 for each condition. All error bars represent SEM. (G) Model depicting proposed network dynamics in normal B cells and premalignant Eµ Myc cells and the consequences of Max deletion in each context. In normal B cells, MNT–MAX activity largely balances MYC–MAX activity, leading to the activation of only a subset of MYC target genes. Upon Max loss, alleviation of MNT–MAX repression and E2F activation of target genes partially compensates for loss of MYC–MAX activity. In premalignant cells, MYC–MAX heterodimers show increased activity and activate MYC-stabilizing genes such as Cip2a and Set. Disruption of this circuit via Max deletion leads to destabilization of MYC protein and loss of the MYC signature expression. Hence, no tumors arise in knockout mice.