Self-enforcing feedback activation between BCL6 and pre-B cell receptor signaling defines a distinct subtype of acute lymphoblastic leukemia

Abstract

Studying 830 pre-B ALL cases from four clinical trials, we found that human ALL can be divided into two fundamentally distinct subtypes based on pre-BCR function. While absent in the majority of ALL cases, tonic pre-BCR signaling was found in 112 cases (13.5%). In these cases, tonic pre-BCR signaling induced activation of BCL6, which in turn increased pre-BCR signaling output at the transcriptional level. Interestingly, inhibition of pre-BCR-related tyrosine kinases reduced constitutive BCL6 expression and selectively killed patient-derived pre-BCR(+) ALL cells. These findings identify a genetically and phenotypically distinct subset of human ALL that critically depends on tonic pre-BCR signaling. In vivo treatment studies suggested that pre-BCR tyrosine kinase inhibitors are useful for the treatment of patients with pre-BCR(+) ALL.

Figure 1. Expression and Activity of the pre-BCR Receptor in Subsets of pre-B ALL

(A) Flow cytometry staining for cell cytoplasmic μHC, (B) cell surface expression of the surrogate light chain components λ5 (IGLL1) and VpreB, and (C) Ca²⁺ mobilization in response to pre–BCR engagement using μHC-specific antibodies in different subtypes of ALL cases. (D) Patient-derived Pre-BCR⁺ ALL cells were treated with or without Dasatinib (25 μmol/l, 24 hr). Phosphorylation of SRC, BLK, SYK, BTK and PLCγ2 were measured by Western blot. (E) Protein expression and phosphorylation levels of LYN, SRC, PIK3R1 (p85), PIK3CD (p110δ) and AKT were measure by reverse phase protein arrays (RPPA) in pre-BCR⁺ vs. pre-BCR⁻ ALL patient samples (MDACC 1983–2007; n=208). Y axis shows log₂ expression values of RPPA data. P values were calculated from two-sided Wilcoxon test. See also Figure S1 and Table S1–S3.

Figure 2. Pre-BCR Signaling in ALL Is Associated with a Distinct Gene Expression and Signal Transduction Phenotype

(A–C) Gene expression microarray data was analyzed from 3 cohorts of pre-B ALL patient samples for children (St. Jude Research Hospital; COG P9906) and adults (ECOG E2993). In each dataset, the patient samples were ranked based on their average mRNA expression levels of pre-BCR molecules (IGLL1, IGLL3, VPREB1, VPREB3, IGHM, SYK and ZAP70). The top 15% and the bottom 25% cases were considered as the pre-BCR⁺ and pre-BCR⁻ and subject to the clustering analysis. Supervised analysis on pre-BCR⁺ vs. pre-BCR⁻ ALL revealed a 40-gene expression signature that is significantly up- or down-regulated across all 3 cohorts. The color scale bar represents the relative log₂ expression changes. (D) The configuration of the μ heavy chain locus (IGHM) was studied in human bone marrow pro-, pre- and immature B cells and (E) different subtypes of ALL patient samples (n=152). Y axis shows frequencies of normal B-cells or ALL clones with a functional or a non-functional IGHM gene rearrangement in these populations. (F) Frequency of different cytogenetic subtypes of pre-B ALL cases as pre-BCR⁺ or pre-BCR⁻ based on 4clinical trials (n=830, MDACC, St. Jude, COG and ECOG). See also Table S4–S6.

Figure 3. TCF3-PBX1 fusion protein binds to and upregulates genes encoding pre-BCR components

(A) ChIPseq tracks for TCF3^N, PBX1^C and p300 antibodies vs. input in a TCF3-PBX1 ALL line 697 and a primary sample ICN12 on IGLL1, VPREB1, VPREB3, CD79A and CD79B promoter regions and the IGHM enhancer regions (Eμ). Gene models were shown in UCSC genome browser hg18. B () QChIP validation using an HA antibody that is specific for the HA-Flag-tagged TCF3-PBX1 fusion or an EV as control. NCAPD2 serves as a negative control gene. Data represent means ± SEM (n=3). P values from t-test. (C) The scatter plot shows the average gene expression values of the TCF3-PBX1-rearranged (x-axis) vs. the non-TCF3-PBX1-rearranged ALL patient samples (y-axis) from the St. Jude dataset on the TCF3-PBX1 target genes (n=1,343). Blue and red highlighted are the down- (n=111) and up- (n=99) regulated genes (>1.5 fold change). (D) The heatmap representation for some up- or down-regulated genes. The color scale bar represents the log₂ expression changes. (E) Ingenuity pathway analysis for the up- and down-regulated genes. The color scale indicates p values calculated from Ingenuity. (F). Flow cytometry staining for cytoplasmic μHC, surface CD79B and CD19 in full-blown ALL populations that had developed after latencies of 90–180 days in the bone marrow from the MLL-AF4-, BCR-ABL1- or TCF3-PBX1-transgenic mice. See also Figure S2.

Figure 4. Pre-BCR Signaling in ALL Cells Drives Expression of BCL6

(A) BCL6, STAT5, and pSTAT5^Y694 Western blotting with β-actin as a loading control in human BCR-ABL1 or TCF3-PBX1 ALL samples. Asterisk denotes non-specific band. (B) Protein expression levels of BCL6, STAT5 and pSTAT5^Y694 by Western blot with β-actin as the loading control at different time points of tetracycline culture of Rag2^−/− tTA/μ-chain ALL cells (left panel), before and after pre-BCR signaling reconstitution by retroviral expression of Blnk-GFP or GFP EV in the Blnk^−/− ALL cells (middle panel), or CD8/μ-chain or CD8 EV in Ighm^−/− ALL cells (right panel). (C) BCL6 and STAT5 Western blot in the presence or absence of Cre-mediated deletion of Stat5 in Stat5^fl/fl pre-B ALL cells. (D) Immunohistochemistry double-stainings for BCL6 (nuclear, brown) and μHC (cytoplasmic/membrane, red) on the same slides of paraffin-embedded bone marrow samples from Pre-BCR⁻ and Pre-BCR⁺ ALL patients. A Burkitt’s lymphoma sample (12-926) carrying an IGH-BCL6 gene rearrangement was used as a positive control. Scale bars, 20 um. (E) Bone marrow pre-B cells transduced with BCR-ABL1 or TCF3-PBX1 vectors, or from full-blown ALL populations that had developed after latencies of 90–180 days from the BCR-ABL1- or TCF3-PBX1-transgenic mice were stained for intracellular Bcl6 protein. Cells were gated as live CD19+ cells. The Bcl6 expression threshold was set according to an isotype staining control. (F–G) BCL6 protein expression by Western blotting in presence or absence of SYK inhibitor (PRT062607, 10 μmol/l), BTK inhibitor (Ibrutinib, 10 μmol/l) for 24 hr in a TCF3-PBX1 ALL cell line (RCH-ACV) and a primary sample (ICN12), or (H) in presence or absence of Nilotinib (200 nmol/l) or Dasatinib (25 nmol/l) for 24 hr in primary BCR-ABL1 (ICN1) or TCF3-PBX1 (ICN12) ALL cells. β-actin was used as a loading control. (I) A schematic of pre-BCR, STAT5 and BCL6 regulation in BCR-ABL1 and TCF3-PBX1 ALL. See also Figure S3 and Table S7-S8.

Figure 5. BCL6 Is a Key Regulator of the Transcriptional Program in pre-BCR⁺ ALL Cells

(A) A genetic model for inducible ablation of BCL6 and a mCherry-based Bcl6 reporter (Bcl6^fl/fl-mCherry). Exons 5–10 of the Bcl6 locus were targeted for inducible Cre-mediated deletion. LoxP sites are indicated as black triangles. (B) PCR validation of the floxed, deleted and wild type Bcl6 allele from Bcl6^fl/fl-mCherry mouse bone marrow pre-B cells. (C) FACS analysis of EV^Neo or TCF3-PBX1^Neo retrovirally transduced Bcl6^fl/fl-mCherry pre-B cells that were transduced with either EV^GFP or Cre^GFP vector for Bcl6 deletion. Y-axis indicates mCherry expression and x-axis indicates GFP expression. (D) BCL6 ChIPseq binding tracks of target genes in patient-derived TCF3-PBX1 ALL cells (ICN12). Y axis represents the number of reads for peak summit normalized by the total number of reads per track. Gene models are shown in UCSC genome browser hg18. A control intragenic region and BCL6 were used as negative and positive controls. (E–F) A meta-analysis of BCL6 ChIPseq target genes (n=666) with gene expression microarray data for the EV vs. Cre transduced Bcl6^fl/fl TCF3-PBX1 ALL cells or (G–H) for the BCL6^High vs. BCL6^Low ALL patient samples from St. Jude. The scatter or heatmap plots showed up- and down-regulated BCL6 target genes in each data set. The color scale bar represents relative log₂ expression changes. (I) PBX1 and BCL6 ChIPseq binding tracks vs. input in ICN12 cells on pre-BCR-related genes (IGLL1, VPREB1, BLK, BANK1). See also Figure S4.

Figure 6. Pre-BCR⁺ ALL Cells Are Dependent on BCL6-Activity

(A–B) Patients were segregated into two groups based on higher or lower than the median expression of BCL6 in pre-BCR⁺ and pre-BCR⁻ ALL in 2 clinical trials: (A) COG P9906 and (B) ECOG E2993. Kaplan-Myer estimates were used to plot the survival probabilities. P values were calculated from the log-rank test. (C) Bcl6^fl/fl TCF3-PBX1 pre-B ALL cells were transduced with 4-OHT-inducible Cre (Cre-ER^T2-GFP) or EV control (ER^T2-GFP). Percentage of GFP-positive cells were measure by flow cytometry at different time points following 4-OHT treatment and time course data are depicted. (D) Patient-derive Pre-BCR⁺ ALL cells (ICN12) were transduced with 4-OHT-inducible dominant-negative BCL6 (^DNBCL6-ER^T2-GFP) or EV control (ER^T2-GFP). Percentages of GFP-positive cells were measured by the flow cytometry at different time points following 4-OHT treatment. (E) ICN12 cells were treated with vehicle or 5 μmol/l RI-BPI for 24 hr and then subjected to cell-cycle analysis (BrdU and 7-AAD staining). (F) ICN12 cells were exposed to vehicle, Vincristine (1 nmol/l), RI-BPI (5 μmol/l), or combination of Vincristine (1 nmol/l) and RI-BPI (5 μmol/l) for 3 days, followed by flow cytometry for Annexin V and 7-AAD staining. Data represent means ± SEM (n=3). P values from t-test. See also Figure S5.

Figure 7. Validation of Pharmacological Inhibition of BCL6-pre-BCR Signaling as Therapeutic Target in pre-BCR⁺ ALL Cells

(A) Cell viability was measured using CCK-8 in presence or absence of PRT062607 (SYK), Ibrutinib (BTK) or Dasatinib (ABL1/SRC/BTK) for 72 hr with gradients of concentrations as indicated in the X axis in pre-BCR⁺ ALL (n=7), pre-BCR⁻ ALL (n=8) and 2 Burkitt lymphoma (MN60, MHH-preB). Y axis shows the percentage of viable cells with the untreated cells as control (set to 100%). Data represent means ± SD (n=3). (B) Pre-BCR⁺ (n=8) and pre-BCR⁻ (n=9) patient-derived ALL samples were treated with a diverse panel of 51 kinase inhibitors as described previously (Tyner et al., 2013). The heatmap represents the IC₅₀ values for each sample relative to the observed median IC₅₀ value for over 400 primary leukemia samples interrogated by this assay at OHSU. Red or blue colors denote higher or lower than the median sensitivity of the pre-B ALL cells tested. Pre-BCR⁺ ALL: 07-112, 11-064, ICN12, 697, RCH-ACV, Kasumi-2, HPB-null, Nalm6. Pre-BCR⁻ ALL: BV173, SUPB15, BLQ5, LAX2, SEM, RS4;11, REH, LAX7R, SFO3. (C) Kinase dendrogram of pre-BCR⁺ and pre-BCR⁻ ALL based on experimentally measured sensitivities to individual inhibitors and their known inhibitory profile based on biochemical IC₅₀ values for individual kinase targets. TREESpots software was used (KINOMEscan, http://www.discoverx.com/). See also Figure S6.

Figure 8. Validation of Pharmacological Inhibition of BCL6-pre-BCR Signaling in Patient-Derived pre-BCR⁺ ALL Cells

(A) Patient-derived pre-BCR⁺ ALL cells (ICN12) were labelled with luciferase and 1,000,000 cells were injected intravenously to sublethally irradiated NOD/SCID mice. The mice were randomly separated into 4 groups and treated with vehicle, the SYK inhibitor PRT062607 (100 mg/kg), the BTK inhibitor Ibrutinib (75 mg/kg), or Dasatinib (40 mg/kg) respectively. Bioimages were taken at different time points for each of the groups. (B) Plot of Dasatinib IC₅₀’s from primary ALL patient samples (n=135, OHSU). ALL samples were cultured over 72 hr in the presence or absence of Dasatinib in concentrations ranging from 1 nmol/l to 1,000 nmol/l to calculate the IC₅₀ values. (C) Patient-derived pre-BCR⁺ ALL cells (n=4) were injected to sublethally irradiated NOD/SCID mice. Mice were treated with vehicle or Dasatinib (40–50 mg/kg). Kaplan-Meier estimates were used to plot the survival probabilities for each treatment group vs. vehicle control. P values were calculated from log-rank test. Peripheral blood was drawn weekly and analyzed for chimerism with either human CD19 or human CD45 vs. murine CD45 and plotted over time. Data represent means ± SD (n=4 or 5).