ANP32B suppresses B‐cell acute lymphoblastic leukemia through activation of PU.1 in mice

Abstract

ANP32B, a member of the acidic leucine‐rich nuclear phosphoprotein 32 kDa (ANP32) family of proteins, is critical for normal development because its constitutive knockout mice are perinatal lethal. It is also shown that ANP32B acts as a tumor‐promoting gene in some kinds of cancer such as breast cancer and chronic myelogenous leukemia. Herein, we observe that ANP32B is lowly expressed in B‐cell acute lymphoblastic leukemia (B‐ALL) patients, which correlates with poor prognosis. Furthermore, we utilized the N‐myc or BCR‐ABL^p190‐induced B‐ALL mouse model to investigate the role of ANP32B in B‐ALL development. Intriguingly, conditional deletion of Anp32b in hematopoietic cells significantly promotes leukemogenesis in two B‐ALL mouse models. Mechanistically, ANP32B interacts with purine rich box‐1 (PU.1) and enhances the transcriptional activity of PU.1 in B‐ALL cells. Overexpression of PU.1 dramatically suppresses B‐ALL progression, and highly expressed PU.1 significantly reverses the accelerated leukemogenesis in Anp32b‐deficient mice. Collectively, our findings identify ANP32B as a suppressor gene and provide novel insight into B‐ALL pathogenesis.

Loss of Anp32b promotes B‐cell acute lymphoblastic leukemia (B‐ALL) development in mice. ANP32B enhances PU.1 activity to regulate B‐ALL leukemogenesis.

Abbreviations

B‐ALL

B‐cell acute lymphoblastic leukemia

BCL6

B cell lymphoma 6

CCND1

recombinant cyclin d1

CCND2

recombinant cyclin d2

CML

chronic myelogenous leukemia

HDAC1

histone deacetylase 1

IRES

internal ribosome entry site

IRF7

interferon regulatory factor 7

KLF5

krüppel‐like factor 5

PEST

polypeptide sequences enriched in proline, glutamate, serine, and threonine

PU.1

purine rich box‐1

Q‐pcr

quantitative real‐time polymerase chain reaction

SMARCA5

SWI/SNF‐related, matrix‐associated, actin‐dependent regulator of chromatin, subfamily A member 5

TAD

topologically associating domain

TSS

transcriptional start sites

1. INTRODUCTION

B‐cell acute lymphoblastic leukemia (B‐ALL) is a group of hematological malignancies caused by the clonal proliferation of lymphoid progenitor cells combined with a blockage of B‐cell differentiation, which commonly occurs in children and also in adult populations. ¹ The oncogenesis of B‐ALL is often associated with various genetic lesions, including oncogenic fusions derived from chromosomal translocations such as ETV6‐RUNX1, BCR‐ABL1, or TCF3‐PBX1. ² These chromosomal aberrations are important in leukemia initiation, but they alone are insufficient to generate a full leukemic phenotype. ³ It has been discovered that the mutations or abnormal expressions of related genes involved in B‐cell development, signal transduction, and epigenetic regulation play important roles in the development of B‐ALL. ³ , ⁴ , ⁵ Despite advances in the treatment of B‐ALL, including chemotherapy, bone marrow (BM) transplantation, chimeric antigen receptor T cell (CAR‐T) immunotherapy, or combinations of these treatments, ⁶ , ⁷ , ⁸ 20% of child B‐ALL patients still have treatment failure. The prognosis is even worse in adult B‐ALL patients, as only 30% of them achieve long‐term survival, ⁹ , ¹⁰ therefore novel therapeutic targets are urgently needed to treat B‐ALL more effectively.

ANP32B belongs to the highly conserved acidic leucine‐rich nuclear phosphoprotein 32 kDa (ANP32) family, whose members including ANP32A, B, and E, which is characterized by an N‐terminal leucine‐rich repeat domain and a C‐terminal low‐complexity acidic region. ¹¹ Although the ANP32 proteins functionally overlap in a broad array of physiological processes, they have been reported to have diverse roles in cancer progression. ¹² ANP32A is a putative tumor suppressor based on studies that it could inhibit cell transformation and has reduced expression in prostate and breast cancer. ¹³ , ¹⁴ , ¹⁵ Furthermore, ANP32A is a positive prognostic marker in non‐small‐cell lung cancer. ¹⁶ However, ANP32A was also shown to be upregulated in primary acute myeloid leukemia cells and promotes leukemogenesis. ¹⁷ ANP32E similarly shows enhanced expression in gastric cancer and is a negative prognostic marker in myeloma. ¹⁸ , ¹⁹ However, it is also reported that higher expression of ANP32E was associated with extended survival in follicular lymphoma. ²⁰

ANP32B is the most critical gene for normal development by comparing the effects of Anp32b deficiency to those of Anp32a or Anp32e deficiency in mice. ²¹ Previously, we showed that ANP32B acts as a negative regulator of leukemic cell apoptosis and a master enforcer of cell proliferation in breast cancer cells. ²² , ²³ Recently, we also demonstrated that ANP32B‐mediated repression of p53 maintains the function of chronic myelogenous leukemia (CML) stem cells and promotes CML progression. ²⁴ Although it is ranked among the highest candidates in a tumor‐suppressor‐rich genome‐wide search for recessive cancer genes, ²⁵ the potential tumor‐suppressor role of ANP32B remains largely unknown. Herein we investigate the function of ANP32B to B‐ALL development using hematopoietic‐specific Anp32b knockout mice and demonstrate that ANP32B suppresses B‐ALL leukemogenesis by enhancing PU.1 activity.

2. METHODS

2.1. Mice

To delete the Anp32b gene specifically in the hematopoietic system, Anp32b ^fl/fl mice established at the Shanghai Model Organisms Center were crossed with Scl‐Cre transgenic mice. All these strains were maintained on a C57BL/6 background. Six‐week‐old Scl‐Cre mice were injected intraperitoneally daily with tamoxifen (10 mg/mL in corn oil; Sigma) at 50 μg/g body weight for 21 days to induce the Scl‐Cre transgene. Genotyping primers are listed in Table S1. All of the animal experiments were conducted according to the Guideline for Animal Care at Shanghai Jiao Tong University School of Medicine.

2.2. B‐ALL mice model

To establish an N‐myc‐induced or BCR‐ABL^p190‐induced murine B‐ALL, B220⁺ cells were sorted from the bone marrow of 10–12‐week‐old donor mice, and infected with BCR‐ABL^p190‐GFP or N‐myc‐GFP retrovirus with polybrene (4 μg/mL). For transplantation, 1 × 10⁵ B220⁺ infected cells mixed with 2 × 10⁵ competitor BM cells were directly injected through the tail vein into lethally irradiated C57BL/6 mice. For secondary transplantation, 2 × 10⁵ GFP⁺ cells were isolated and transplanted into irradiated C57BL/6 mice with 2 × 10⁵ competitor BM cells.

2.3. Flow cytometry

For flow cytometry analyses, BM cells were filtered through a 40‐μm strainer to obtain a single‐cell suspension. Peripheral blood (PB) cells were treated with ammonium chloride potassium lysis buffer to remove red blood cells. The cells were stained with indicated fluorochrome‐conjugated antibodies following the manufacturer's instructions. The antibodies and dyes used are listed in Table S1.

2.4. Immunoprecipitation and mass spectrometry analysis

Immunoprecipitation of endogenous proteins or Flag‐tagged proteins and nano liquid chromatography with tandem mass spectrometry (LC–MS/MS) used to identify interacting proteins were performed as previously described. ²⁶ , ²⁷ All the antibodies used for immunoprecipitation are listed in Table S1. ANP32B‐interacting proteins identified by mass spectrometry are listed in Table S3.

2.5. Immunofluorescence

The details of immunofluorescence have been described previously. ²⁸ The primary antibodies used are shown in Table S1.

2.6. ChIP‐qRT‐PCR

Chromatin immunoprecipitation (ChIP) experiments were performed similarly to those described previously ²⁴ and 2 × 10⁷ Anp32b ^+/+ and Anp32b ^−/− GFP⁺ cells were performed per reaction. The purified ChIP DNA was quantified by qRT‐PCR. The primers used are listed in Table S1.

2.7. Quantification and statistical analysis

For comparison between two experimental groups or a specific pair in a multigroup, two‐tailed unpaired Student's t‐test was used, where error bars denote mean ± SEM. For comparison of cell growth curves, two‐way ANOVA was used. For analysis of survival rates, a log‐rank (Mantel‐Cox) test was performed. All the differences were considered to be statistically significant if P < 0.05.

3. RESULTS

3.1. ANP32B is lowly expressed and predicts a better prognosis in human B‐ALL

To investigate the expression of ANP32B in B‐ALL, we first analyzed the mRNA expression level of ANP32B in three B‐ALL patient cohorts (GSE13159, GSE28497, and GSE33315). Interestingly, we observed that ANP32B was expressed at a significantly lower level in B‐ALL patients compared with normal controls (Figure 1A). To support this, we further analyzed ANP32B mRNA and protein levels in mononuclear cells from bone marrow of eight cases of primary adult patients with B‐ALL together with samples from 10 cases of nonleukemic individuals as control, and the results demonstrated that both ANP32B mRNA and protein levels were also aberrantly downregulated in B‐ALL patients (Figure 1B,C and Table S2). Notably, higher ANP32B expression was associated with longer overall survival (OS) in B‐ALL patients according to the COG P9906 childhood B‐ALL cohort (Figure 1D). Together, these data suggest that ANP32B might play an important role in B‐ALL.

FIGURE 1.

ANP32B is lowly expressed and predicts a better prognosis in B‐ALL. (A) ANP32B mRNA expression level in B‐ALL and normal bone marrow samples was analyzed in three published datasets. (B and C) Relative mRNA and protein expression levels of ANP32B in bone marrow mononuclear cells (BMMCs) of B‐ALL and nonleukemic individuals were analyzed. (D) Overall survival of pediatric B‐ALL patients from COG P9906 relative to ANP32B mRNA expression level.

3.2. Loss of Anp32b promotes N‐myc‐induced B‐ALL development

To investigate the potential role of ANP32B in B‐ALL development, we established the N‐myc‐induced B‐ALL mouse model, which can be achieved by retrovirally introducing N‐myc‐IRES‐GFP into B220⁺ BM cells, followed by transplantation into lethally irradiated mice (Figure S1A). Three weeks after transplantation, the number of peripheral blood mononuclear cells (PBMC) and GFP⁺ leukemic cells in PB and BM in mice transplanted with N‐myc‐transduced B220⁺ BM cells were significantly higher than mice transplanted with empty vector cells (Figure S1B–D). All GFP⁺ leukemic cells in BM mainly expressed B‐cell markers (B220⁺CD19⁺) but not myeloid markers (Mac‐1⁺Gr‐1⁺), and these B cells were mainly undifferentiated B progenitor cells (B220⁺CD43⁺IgM⁻) (Figure S1E). In addition, tumor‐bearing mice also had enlarged lymph nodes and exhibited hepatosplenomegaly (Figure S1F,G).

B220⁺ BM cells of tamoxifen treated Scl‐Cre ⁻; Anp32b ^fl/fl and Scl‐Cre ⁺; Anp32b ^fl/fl mice, which we previously reported and referred to as Anp32b ^+/+ and Anp32b ^−/− mice, ²⁴ were infected with N‐myc‐GFP retrovirus and then transplanted into lethally irradiated recipients. The results showed that Anp32b deficiency significantly increased the engraftment of GFP⁺B220⁺ B lymphoid leukemia cells in PB and BM of recipient mice (Figure 2A,B). Giemsa–Wright staining displayed a higher frequency of immature blast cells in BM (Figure 2C), together with more massive splenomegaly and enlarged lymph nodes (Figure 2D,E), and significantly increased leukemia infiltration in spleen and lymph nodes in recipients of Anp32b ^−/− cells (Figure 2F). In line with this, recipients of Anp32b ^−/− cells showed remarkably reduced survival times during primary and secondary transplantation (Figure 2G). To exclude the possibility that the enhanced homing ability contributes to the effects of Anp32b loss, GFP⁺ B‐ALL cells sorted from Anp32b ^+/+ and Anp32b ^−/− recipients were injected into lethally irradiated mice, followed by detection of homed GFP⁺ B‐ALL cells in PB, spleen, and BM. No significant difference was detected between two groups 18 h after injection (Figure 2H). Taken together, our data indicate that Anp32b loss significantly promotes N‐myc‐induced B‐ALL development.

FIGURE 2.

Loss of Anp32b promotes N‐myc‐induced B‐ALL development. (A) Flow cytometry plots (left) and statistics (right) of the percentage of GFP⁺B220⁺ cells in PB from recipients transplanted with N‐myc‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells (n = 7). (B) Flow cytometry plots (left) and statistics (right) of the percentage of GFP⁺B220⁺ cells in BM from recipients transplanted with N‐myc‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells (n = 5). (C) Representative images of Giemsa–Wright staining (left) for Anp32b ^+/+ and Anp32b ^−/− BM cells on transplantation. Quantification of the percentage of blast cells is shown on the right (n = 5). Representative blast cells are indicated with red arrows. (D–F) Gross pathology of lymph nodes, spleens (D) and relative weights of the spleens (E), and hematoxylin–eosin staining of the spleens and lymph nodes (F) from recipients (n = 5). (G) Survival curves for recipients transplanted with N‐myc‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells on the first (n = 5) and second transplantation (n = 4). (H) Quantification of the frequencies of homed GFP⁺ cells in PB, BM, and spleens in recipient mice receiving B‐ALL cells 18 h after transplantation (n = 5). Error bars denote mean ± SEM. Statistical significance was determined by two‐tailed unpaired t‐test (A–C, E and H) or log‐rank test (G) and the P values are shown. All animal experiments were repeated at least twice with similar results, and the results from one representative experiment are shown.

3.3. Loss of Anp32b promotes BCR‐ABL^p190 ‐induced B‐ALL development

To further confirm the role of ANP32B in B‐ALL development, we established another B‐ALL mouse model generated by transplantation of BCR‐ABL^p190‐transduced B220⁺ BM cells into lethally irradiated recipient mice, ²⁹ which reflects the pathology of human disease since BCR‐ABL^p190 fusion gene accounts for 90% in pediatric Ph⁺ B‐ALL and 50%–80% in adult Ph⁺ B‐ALL (Figure S2A). Leukemic development was characterized by expansion of B lymphoid leukemic cells in PB and BM, frequent hind leg paralysis, moderate hepatosplenomegaly, and enlarged lymphoid organs of recipients (Figure S2B–G). Then B220⁺ BM cells of Anp32b ^+/+ and Anp32b ^−/− mice were infected with BCR‐ABL^p190 retrovirus and transplanted into lethally irradiated recipients. Similarly, recipients of Anp32b ^−/− BM cells developed B‐ALL significantly faster than recipients of Anp32b ^+/+ BM cells. The GFP⁺ leukemic cells were largely increased in PB and BM of Anp32b ^−/− B‐ALL mice (Figure 3A–C). Anp32b deficiency also exhibited more frequent hind leg paralysis, massive splenomegaly, and enlarged lymph nodes with more severe infiltration of leukemic cells compared with their wild‐type counterparts (Figure 3D–F). Consistently, recipients of Anp32b ^−/− cells had markedly decreased survival during primary and secondary transplantation (Figure 3G). Furthermore, the number of colonies formed by Anp32b ^−/− B‐ALL cells was significantly increased compared with Anp32b ^+/+ controls in the first plating, and this enhancement was even more impressive in the second plating (Figure 3H). Collectively, these data suggest that Anp32b deficiency significantly promotes BCR‐ABL^p190‐induced B‐ALL development.

FIGURE 3.

Loss of Anp32b promotes BCR‐ABL^p190‐induced B‐ALL development. (A) Flow cytometry plots (left) and statistics (right) of the percentage of GFP⁺B220⁺ cells in PB from recipients receiving BCR‐ABL^p190‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells (n = 5). (B) Flow cytometry plots (left) and statistics (right) of the percentage of GFP⁺ cells in BM from recipients receiving BCR‐ABL^p190‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells (n = 4). (C) Representative images of Giemsa–Wright staining (left) and frequencies of blast cells (right) for Anp32b ^+/+ and Anp32b ^−/− BM cells on transplantation (n = 5). Representative blast cells are indicated with red arrows. (D) Gross appearance of Anp32b ^+/+ and Anp32b ^−/− recipients. (E and F) Gross pathology of spleens, lymph nodes (E), and hematoxylin–eosin staining of lymph nodes (F) from recipients. (G) Survival curves for recipients transplanted with BCR‐ABL^p190‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells on the first (n = 5) and second transplantation (n = 6). (H) Primary and secondary colonies of GFP⁺ cells sorted from the recipients receiving BCR‐ABL^p190‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells on first transplantation (n = 3). Error bars denote mean ± SEM. Statistical significance was determined by two‐tailed unpaired t‐test (A–C and H) or log‐rank test (G) and the P values are shown. All animal experiments were repeated at least twice with similar results, and the results from one representative experiment are shown.

3.4. ANP32B directly interacts with PU.1

ANP32B has been found to bind transcription factors (TFs) and modulate their activity. ²⁴ , ³⁰ To explore the potential ANP32B‐interacting proteins in B cells, BaF3 cells (a murine pro‐B‐cell line) were transfected with empty vector or Flag‐ANP32B, followed by affinity purification using anti‐Flag antibody, and the precipitates were analyzed by LC–MS/MS. In total, we identified 259 ANP32B‐interacting proteins in B cells including p53 (Figure S3A and Table S3). Our recent study reveals that ANP32B interacts with p53 to regulate hematopoiesis and CML leukemogenesis. ²⁴ To determine whether ANP32B inhibits B‐ALL development through regulating p53 activity, we used Anp32b ^+/+ p53 ^+/+, Anp32b ^+/+ p53 ^+/−, Anp32b ^−/− p53 ^+/+, and Anp32b ^−/− p53 ^+/− mice in our pervious study to induce B‐ALL. Survival analysis showed that although heterozygous p53 loss in Anp32b ^+/+ cells accelerated BCR‐ABL^p190‐induced leukemogenesis, Anp32b ^−/− p53 ^+/− mice presented similar survival compared with Anp32b ^−/− p53 ^+/+ mice (Figure S3B), suggesting that Anp32b‐deficiency promotes B‐ALL development in a p53‐independent manner.

In parallel, we performed RNA‐seq analysis and revealed 393 significantly differentially expressed genes (DEGs) in Anp32b ^−/− B‐ALL cells compared with the control ones. To further screen ANP32B‐interacting transcriptional regulators, we compared the top 200 extracted upstream regulators through Ingenuity Pathway Analysis (IPA) with 259 ANP32B‐interacting proteins and found an overlap of seven candidate genes (Figure S3A). Among these genes, only PU.1, a key regulator of B‐cell fate specification, ³¹ was marked as a significantly inhibited regulator (z‐score = −2.608) in the Anp32b ^−/− B‐ALL group (Figure S3C), suggesting that ANP32B might interact with PU.1 and enhance its transcriptional activity.

We continued to investigate the relationship between ANP32B and PU.1. As shown in Figure 4A and Figure S3D, Flag‐tagged ANP32B could immunoprecipitate endogenous PU.1 in mouse BaF3 and BaF3/BCR‐ABL^p190‐expression cells. Endogenous ANP32B‐PU.1 interaction was also validated in human B‐ALL cell line SEM and Nalm6 (Figure 4B,C). To define the domains of ANP32B and PU.1 required for their interaction, Flag‐tagged full‐length PU.1 (FL) and its two fragments, PU.1 (1‐169aa, TAD+PEST), PU.1 (117‐270aa, ETS + PEST) (Figure 4D), were transfected in 293T cells, followed by co‐immunoprecipitation (IP) with ANP32B antibody. The results showed that ANP32B pulled down FL and the 117–270 fragment, suggesting that ETS (DNA‐binding domain) of PU.1 is essential for its interaction with ANP32B (Figure 4E). On the other hand, the N‐terminal (1‐163aa) of ANP32B did not interact with PU.1 (Figure 4F). These data suggest that the DNA‐binding domain of PU.1 is required for its interaction with the C‐terminal acidic domain of ANP32B. Furthermore, in vitro GST‐pull down assay showed that either GST‐PU.1 (FL) or GST‐PU.1 (170‐270aa) pulled down ANP32B (Figure 4G), supporting a direct interaction of ANP32B with PU.1 protein.

FIGURE 4.

ANP32B interacts with PU.1 and enhances the transcriptional activity of PU.1. (A) Western blot analysis of indicated proteins in the inputs and immunoprecipitates of Flag‐tagged ANP32B‐transfected BaF3 cells. Empty vector (EV) serves as negative control. (B) Western blot analysis of indicated proteins in the inputs and immunoprecipitates of endogenous ANP32B in SEM cells. (C) Immunofluorescent staining of endogenous ANP32B, PU.1 together with re‐staining of DAPI in Nalm6 cells followed by imaging with confocal microscopy. (D) Structure illustrations of full‐length (FL) and fragments of PU.1 and ANP32B. (E) Western blot analysis of indicated proteins in the inputs and immunoprecipitates of anti‐ANP32B antibody in 293T cells transfected with Flag‐PU.1 full‐length and its two fragments. (F) Western blot analysis of indicated proteins in the inputs and immunoprecipitates of anti‐FLAG M2 beads in 293T cells transfected with Flag‐ANP32B full‐length and N163 segments. (G) Bacterially expressed ANP32B was incubated with GST or GST‐tagged PU.1, GST‐tagged PU.1 (170‐270aa) followed by GST‐tag pull down and Western blot analysis of indicated proteins. (H) GSEA analysis of RNA‐seq data from recipients receiving N‐myc‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells using the PU.1‐regulated gene set (GSE13125). (I and J) Clonally derived 293T cell lines depleted of ANP32B (gANP32B) or not (gNS), empty vector (EV), or Flag‐ANP32B‐infected 293T cells were co‐transfected with PLVX‐PU.1, luciferase reporter for PU.1 transcription (PU.1‐luc), and Renilla luciferase reporter, and the relative luciferase activities were determined. (K) Relative mRNA expression levels of indicated genes in BM GFP⁺ cells sorted from recipients receiving N‐myc‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells on first transplantation. (L) ChIP‐quantitative RT‐PCR of IgG and PU.1 on the promoters of the indicated genes in BM GFP⁺ cells from recipients receiving N‐myc‐transduced Anp32b ^+/+ and Anp32b ^−/− B220⁺ BM cells. Error bars denote mean ± SEM. Statistical significance was determined by two‐tailed unpaired t‐test (I–L) and the P values are shown. The experiments in (A–G) and (I–L) were repeated three times independently with similar results.

3.5. ANP32B enhances the transcriptional activity of PU.1 in B‐ALL

The above observations prompted us to investigate how ANP32B regulates PU.1 function. After ruling out that Anp32b deficiency or overexpression did not change the mRNA and protein level of PU.1 in B‐ALL cells (Figure S3E–G), we performed a gene set enrichment analysis (GSEA) to gain a global view of the transcriptome profile regulated by ANP32B. In line with our previous hypothesis, PU.1 target genes were significantly enriched in the transcriptome of Anp32b ^+/+ B‐ALL cells, suggesting that ANP32B may positively regulate the transcriptional activity of PU.1 (Figure 4H). To support this, a specific PU.1 responsive element (RE)‐driven luciferase assay showed that the transcriptional activity of PU.1 was ANP32B dose dependently enhanced on ANP32B overexpression and reduced on ANP32B knockout in 293T cells (Figure 4I,J). Accordingly, the mRNA levels of PU.1‐activated genes Irf7 and p21 ³² , ³³ , ³⁴ were downregulated, while PU.1‐inhibited genes Ccnd1 and Ccnd2 ³⁴ , ³⁵ , ³⁶ were upregulated in Anp32b ^−/− B‐ALL cells (Figure 4K). Because ANP32B antibody is not suitable for ChIP assay, we performed ChIP‐seq analyses using Flag antibody in BaF3 cells transfected with Flag‐ANP32B and compared the results with PU.1 ChIP‐seq data from GSE22178 in hematopoietic progenitor cells. We obtained 41,223 and 25,855 binding sites for ANP32B and PU.1, respectively. ANP32B showed a similar peak distribution pattern to PU.1, most frequently in the promoter regions, followed by intergenic regions and introns (Figure S4A), especially around the regions of the TSS (Figure S4B). Significantly, 9638 (38.6%) of the PU.1 peaks overlapped with ANP32B peaks (Figure S4C), including PU.1‐activated p21, Bcl6, Rel and PU.1‐inhibited Ccnd1,Ccnd2 (Figure S4D–H). ³⁴ , ³⁶ , ³⁷ The common peak annotations and gene list are listed in Table S4. Homer known motif analysis showed that the PU.1 binding sequence enriched in ANP32B, PU.1 and ANP32B/PU.1 common binding sites (Figure S4I). To assess whether ANP32B affects the chromatin occupancy of PU.1 in B‐ALL cells, we conducted ChIP‐qRT‐PCR assays in Anp32b ^+/+ and Anp32b ^−/− B‐ALL cells to monitor whether PU.1 is recruited to the promoters of PU.1 targeted genes including p21, Ccnd1, Ccnd2, Bcl6, and Rel. Intriguingly, Anp32b depletion significantly abolished the binding of PU.1 to these target genes (Figure 4L). All these data indicated that ANP32B co‐localizes with PU.1 on the genome and promotes the binding of PU.1 to target genes.

3.6. PU.1 suppresses B‐ALL progression

Several lines of evidence have demonstrated that PU.1 is a potent tumor suppressor for B cell malignancies including multiple myeloma (MM) and classical Hodgkin lymphoma (cHL). ³² , ³³ , ³⁸ Conditional double deletion of PU.1/Spi‐B or PU.1/IRF4 or PU.1/IRF8 developed pre‐B‐ALL at high frequency. ³⁹ , ⁴⁰ In line with this, PU.1 exhibited a significantly lower expression level in B‐ALL patients compared with normal controls in GSE13159 dataset (Figure 5A). To investigate the functional role of elevated PU.1 expression in B‐ALL cells, we transfected Flag‐PU.1 into Nalm6 and BaF3/BCR‐ABL^p190 cells and found that PU.1 overexpression significantly inhibited cell proliferation and clonogenicity in these two cells (Figure 5B–E). Furthermore, we introduced Flag‐PU.1 by lentiviral Flag‐PU.1‐IRES‐RFP plasmids in BM GFP⁺ cells collected from N‐myc‐induced B‐ALL mice, followed by transplantation into irradiated recipient mice. Then, the GFP⁺RFP⁺ B‐ALL cells were sorted and injected into lethally irradiated mice (Figure 5F). Western blot analysis showed that PU.1 was highly expressed in GFP⁺RFP⁺ B‐ALL cells (Figure 5G). Indeed, PU.1 overexpression obviously decreased the frequencies of GFP⁺RFP⁺ B‐ALL cells in PB of recipient mice (Figure 5H). Meanwhile, we observed decreased spleen, liver, and lymph node size (Figure 5I), and extended survival times in PU.1 overexpression recipient mice (Figure 5J). Consequently, the mRNA levels of PU.1‐activated genes Irf7 and p21 were upregulated, while PU.1‐inhibited genes Ccnd1 and Ccnd2 were downregulated in PU.1 overexpression B‐ALL cells (Figure 5K). Collectively, these data indicate that overexpression of PU.1 inhibits B‐ALL progression.

FIGURE 5.

Overexpression of PU.1 suppresses B‐ALL progression. (A) PU.1 mRNA expression level in B‐ALL and normal bone marrow samples was analyzed in GSE13159. (B) Western blot analysis of PU.1 expression in Nalm6 cells infected with EV and Flag‐PU.1 (left). Cell numbers were counted on the indicated days (right, n = 3). (C) Colony‐forming assay for Nalm6 cells infected with EV or Flag‐PU.1. Colony numbers were evaluated at day 7 (n = 3). (D) Western blot analysis of BCR‐ABL^p190 and PU.1 expression in BaF3/BCR‐ABL^p190 cells infected with EV and Flag‐PU.1 (left). Cell numbers were counted at the indicated days (right, n = 3). (E) Colony‐forming assay for BaF3/BCR‐ABL^p190 cells infected with EV or Flag‐PU.1. Colony numbers were evaluated at day 5 (n = 3). (F) Schematic diagram evaluating the effect of PU.1 in the N‐myc‐induced B‐ALL mice model. (G) Western blot analysis of indicated proteins in GFP⁺RFP⁺ cells sorted from secondary recipients. (H) Flow cytometry plots (left) and statistics (right) of the percentage of GFP⁺RFP⁺ cells in peripheral blood on secondary transplantation (n = 5). (I) Gross pathology of the livers, spleens, and lymph nodes from the secondary recipients. (J) Survival curves for recipients receiving EV and Flag‐PU.1 GFP⁺RFP⁺ cells on secondary transplantation (n = 6). (K) Relative mRNA expression levels of indicated genes in GFP⁺RFP⁺ BM cells sorted from EV/Flag‐PU.1 mice BM cells on secondary transplantation. Error bars denote mean ± SEM. Statistical significance was determined by two‐tailed unpaired t‐test (A, C, E, H, K), two‐way ANOVA (B and D) or log‐rank test (J) and the P values are shown. All animal experiments were repeated at least twice with similar results, and the results from one representative experiment are shown.

3.7. PU.1 signaling rescues Anp32b‐deficiency B‐ALL phenotype

To confirm the requirement for PU.1 signaling for ANP32B‐mediated B‐ALL progression in vivo, we transfected XZ201‐RFP‐EV and XZ201‐RFP‐PU.1 into N‐myc‐induced Anp32b ^+/+ and Anp32b ^−/− B‐ALL cells, followed by transplantation into irradiated recipient mice. Then, the same number of GFP⁺RFP⁺ B‐ALL cells were sorted and injected into lethally irradiated mice. The mice transplanted with PU.1‐overexpressed Anp32b ^−/− B‐ALL cells presented extended survival compared with Anp32b ^−/− control mice, which was slightly inferior to Anp32b ^+/+ mice (Figure 6A), suggesting PU.1 overexpression in Anp32b ^−/− B‐ALL cells partially but significantly reversed accelerated leukemogenesis in Anp32b ^−/− mice. Accordingly, the higher percentages of blast cells in BM, increased sizes of spleen/lymph nodes, and more severe tissue infiltration with leukemic cells in Anp32b ^−/− B‐ALL mice were greatly rescued in PU.1‐overexpressed Anp32b ^−/− B‐ALL (Figure 6B–D). We sorted GFP⁺RFP⁺ BM cells from Anp32b ^+/+ /EV, Anp32b ^−/− /EV, and Anp32b ^−/− /Flag‐PU.1 mice and determined the mRNA level of PU.1 target genes by quantitative real‐time polymerase chain reaction (Q‐PCR). The mRNA levels of downregulated PU.1‐activated genes Irf7, p21 and upregulated PU.1‐inhibited genes Ccnd1, Ccnd2 in the Anp32b ^−/− B‐ALL mice were rescued in PU.1 overexpressing Anp32b KO transformed cells (Figure 6E). All these data indicate that ANP32B enhances PU.1 activity to suppress B‐ALL progression.

FIGURE 6.

PU.1 signaling rescues the Anp32b‐deficiency B‐ALL phenotype. (A) Survival curves from secondary recipients injected with Anp32b ^+/+ /EV, Anp32b ^−/− /EV, Anp32b ^−/− /Flag‐PU.1 B‐ALL cells (n = 5). (B) Representative images of Giemsa–Wright staining for Anp32b ^+/+ /EV, Anp32b ^−/− /EV, Anp32b ^−/− /Flag‐PU.1 BM cells on second transplantation. Quantification of the frequencies of blast cells is shown on the right (n = 5). Representative blast cells are indicated with red arrows. Gross pathology (C) and hematoxylin–eosin staining (D) of the spleens and lymph nodes from the secondary recipients. (E) Relative mRNA expression levels of indicated genes in GFP⁺RFP⁺ BM cells sorted from Anp32b ^+/+ /EV, Anp32b ^−/− /EV, Anp32b ^−/− /Flag‐PU.1 mice BM cells on secondary transplantation. Error bars denote mean ± SEM. Statistical significance was determined by two‐tailed unpaired t‐test (B and E) or log‐rank test (A), and the P values are shown. All animal experiments were repeated at least twice with similar results, and the results of one representative experiment are shown.

4. DISCUSSION

The roles of ANP32B in tumorigenesis are due to its cellular and genetic context. Although ANP32B has been reported to serve as a tumor‐promoting gene in breast cancer and CML, ²³ , ²⁴ to our knowledge this study provides the first evidence for the tumor‐suppressive role of ANP32B in B‐ALL. In contrast to high expression of ANP32B in CML patients compared with nonleukemic controls, ANP32B expression is inhibited in B‐ALL patients through database analysis and clinical patient samples, and patients with low expression of ANP32B have poor prognostic outcome. DNA hypermethylation and post‐translational regulation are common mechanisms for deregulation of tumor suppressor genes, ⁴¹ , ⁴² so the underlying mechanisms of down‐regulation of ANP32B in B‐ALL patients need to be further studied.

Substantial effort has been made to identify multiple key signaling pathways in B‐ALL. ² , ⁴³ Mouse models provide invaluable tools for such studies, in part because they allow genetic manipulation of leukemic cells that is difficult to achieve using human cell lines or leukemia cells from patients. Here, we used two well‐characterized mouse models to study the effect of Anp32b deficiency in B‐ALL initiation and development. N‐myc overexpression in committed progenitor B cells is able to induce pre‐B‐ALL/lymphoma. ⁴⁴ The BCR‐ABL^p190‐induced mouse model is a human‐relevant model reflecting the pathology of the human disease since human B‐ALL cells usually contain a BCR‐ABL^p190 fusion gene. ⁴⁵ Our results demonstrated that Anp32b deficiency significantly promotes B‐ALL development in both models, suggesting that ANP32B has the same effect on different types of B‐ALL driven by different gene mutants.

As a histone chaperone protein, ANP32B has been reported to form a repressive complex with p53 and KLF5, and thus inhibit their transcriptional activity. ²⁴ , ³⁰ After ruling out p53 signaling through which ANP32B regulates B‐ALL progression, we further used combined proteomics and transcriptomics analysis to explore the potential ANP32B‐interacting TFs in B cells. Interestingly, we found that ANP32B physically interacts with PU.1 and is recruited to the promotor of PU.1‐targeted genes to enhance its transcription in B‐ALL cells. Of note, ANP32B binds to the DNA‐binding domain of PU.1, whereas binds to the C‐terminal domain (CTD) of p53. The mechanisms of ANP32B acting as a coactivator for PU.1 or a corepressor for p53 need to be further addressed.

PU.1 plays crucial roles in the determination and differentiation of hematopoietic lineages, ⁴⁶ , ⁴⁷ and is associated with the occurrence of erythrocyte leukemia, pre‐B‐ALL, acute myeloid leukemia, and other diseases. ⁴⁸ , ⁴⁹ , ⁵⁰ Here, we used cell lines and mouse models to demonstrate that PU.1 overexpression dramatically suppresses B‐ALL progression, which is consistent with the tumor suppressor function in B‐ALL revealed by PU.1 and Spi‐B double deletion. In particular, PU.1 overexpression in Anp32b‐deficient B‐ALL cells partially but significantly reversed accelerated leukemogenesis in Anp32b knockout mice, indicating that PU.1 is involved in the regulation of B‐ALL progression by ANP32B. ANP32B is a well‐known histone chaperone responsible for chromatin remodeling and epigenetic modification. ³⁰ , ⁵¹ In this work, we identified seven candidate ANP32B‐interacting proteins through LC–MS/MS combined IPA analysis. In addition, PU.1, histone deacetylase HDAC1, and chromatin remodeling gene SMARCA5 also play important roles in hematologic malignancies. ⁴ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ Whether or not these genes are involved in the accelerated B‐ALL leukemogenesis caused by Anp32b deficiency deserves to be further explored. Given the cancer‐promoting function of ANP32B in CML together with its tumor‐suppressing role in B‐ALL, we hypothesize that ANP32B is a “double‐edged sword” in cancer progression. The diverse roles of ANP32B are likely due to the selective regulation of its binding TFs in different contexts. In summary, our results provide first evidence that ANP32B interacts with PU.1 and enhances its transcriptional activity, thereby suppressing B‐ALL development in mice. Notably, ANP32B is lowly expressed in B‐ALL patients, thus highlighting upregulation of ANP32B as a very promising therapeutic strategy for the treatment of B‐ALL.

AUTHOR CONTRIBUTIONS

Q.Y., H.‐R.L., and S.Y. performed most experiments. Y.‐S.W., X.‐N.Z., and Z.Z. conducted partial experiments. D.Z. provided clinical samples. Y.Y. and G.‐Q.C. designed and supervised the entire project and prepared the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

ETHICAL APPROVAL

Approval of the Research Protocol by an Institutional Reviewer Board: N/A.

Informed Consent: N/A.

Registry and the Registration No. of the Study/Trial: N/A.

Animal Studies: All the animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at Shanghai Jiao Tong University School of Medicine.

Supporting information

DATA S1 Figures S1–S4.

TABLE S1 Reagents, primers, and plasmids used

TABLE S2 Patient information

TABLE S3 Mass spectrometry data

TABLE S4 ChIP‐seq data

Yang Q, Liu H‐R, Yang S, et al. ANP32B suppresses B‐cell acute lymphoblastic leukemia through activation of PU.1 in mice. Cancer Sci. 2023;114:2882‐2894. doi: 10.1111/cas.15822