Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Oct 12.
Published in final edited form as: Cancer Cell. 2015 Sep 24;28(4):486–499. doi: 10.1016/j.ccell.2015.09.001

hnRNP K is a haploinsufficient tumor suppressor that regulates proliferation and differentiation programs in hematologic malignancies

Miguel Gallardo 1, Hun Ju Lee 2, Xiaorui Zhang 1, Carlos Bueso-Ramos 3, Laura R Pageon 4, Mark McArthur 4, Asha Multani 5, Aziz Nazha 1, Taghi Manshouri 1, Jan Parker-Thornburg 5, Inmaculada Rapado 6, Alfonso Quintas-Cardama 1, Steven M Kornblau 1, Joaquin Martinez-Lopez 6, Sean M Post 1
PMCID: PMC4652598  NIHMSID: NIHMS734287  PMID: 26412324

Summary

hnRNP K regulates cellular programs and changes in its expression and mutational status have been implicated in neoplastic malignancies. To directly examine its role in tumorigenesis, we generated a mouse model harboring an Hnrnpk knock-out allele (Hnrnpk+/−). Hnrnpk haploinsufficiency resulted in reduced survival, increased tumor formation, genomic instability, and the development of transplantable hematopoietic neoplasms with myeloproliferation. Reduced hnRNP K expression attenuated p21 activation, downregulated C/EBP levels, and activated STAT3 signaling. Additionally, analysis of samples from primary acute myeloid leukemia patients harboring a partial deletion of chromosome 9 revealed a significant decrease in HNRNPK expression. Together, these data implicate hnRNP K in the development of hematological disorders and suggest hnRNP K acts as a tumor suppressor.

Introduction

Cytogenetic alterations and chromosomal deletions are often observed in cancers. Recurrent genetic losses are thought to contribute to disease processes through deregulating critical cellular programs that govern proliferation and differentiation (Burnett et al., 2011; Mrozek et al., 2007; 2008). However, even though these genetic alterations can be clinically identified, we often do not fully understand which gene(s) are responsible for driving a particular malignancy. For example, it is well established that 9q deletions are recurrent abnormalities in myeloid neoplasias and are present in ∼2% of acute myeloid leukemias (AMLs); however, which gene(s) are responsible for the pathogenesis of this disease remains unclear (Mecucci et al., 1984). In a subset of AMLs harboring 9q deletions, the 9q21.32 locus is specifically lost, suggesting a tumor suppressor resides at this locus and its haploinsufficiency contributes to disease progression (Dayyani et al., 2008; Kronke et al., 2013; Sweetser et al., 2005). A recent study mapped six possible candidate genes to the minimally deleted region (MDR) of chromosome 9q21.32 (Kronke et al., 2013). Of these six genes, the DNA and RNA binding protein, Heterogeneous Nuclear Ribonuclear Protein K (HNRNPK), is an intriguing potential tumor suppressor.

Mutational analyses of AML patient samples by The Cancer Genome Atlas revealed that hnRNP K mutations may drive AML progression, implicating the potential involvement of hnRNP K in leukemogenesis (2013). Furthermore, in vitro studies suggest that altered hnRNP K expression may contribute to cancer phenotypes through its transcriptional and translational regulation of genes that control proliferation and differentiation programs (Liu and Szaro, 2011; Notari et al., 2006; van Domselaar et al., 2012; Wang et al., 2011).

It is currently unclear how aberrant hnRNP K expression contributes to tumor phenotypes in vivo, as underexpression, overexpression, and mutations have all been have been reported in patient samples. On one hand, several studies suggest increased hnRNP K expression results in a potential oncogenic effect through regulation of c-Myc (Notari et al., 2006; Roychoudhury and Chaudhuri, 2007). However, much of our understanding of hnRNP K's potential role in tumorigenesis stems from biochemical and cell line studies demonstrating that hnRNP K acts as a transcriptional co-activator of p53-dependent p21 expression, suggests a potential role in tumor suppression (Enge et al., 2009; Moumen et al., 2005). Furthermore, in vitro studies have shown that hnRNP K controls programs that influence differentiation potential in hematopoietic stem cells (HSCs) which may ultimately impact leukemogenesis (Miau et al., 1998; Ostareck-Lederer and Ostareck, 2012). Additionally, other studies have shown that hnRNP K represses C/EBPβ-mediated differentiation in the liver and that it regulates the expression of genes required for hematopoietic maturation (Miau et al., 1998; Ostareck et al., 1997). Together, these studies suggest hnRNP K plays a critical role in proliferation and differentiation and that deregulation of hnRNP K expression may promote tumorigenesis (Liu and Szaro, 2011; Mikula et al., 2013; van Domselaar et al., 2012; Wang et al., 2011).

To understand how hnRNP K functionally influences hematopoiesis and tumorigenesis in vivo, we generated an Hnrnpk haploinsufficient mouse model (Hnrnpk+/−) that potentially mimics the putative haploinsufficiency observed in AML patients with 9q21.32 deletions.

Results

HNRNPK expression is reduced in patients with AML that harbor 9q deletions

Alterations in the long arm of chromosome 9 (9q) are recurrent genetic anomalies observed in acute myeloid leukemia (AML). In a subset of individuals patients with AML, the 9q21.32 genomic section containing the Heterogeneous Nuclear Ribonuclear Protein K (HNRNPK) gene is specifically lost, suggesting a tumor suppressor may reside at this locus (Kronke et al., 2013). To evaluate a potential link between HNRNPK expression and AML, we examined HNRNPK expression levels in patients with 9q21.32 deletions. Analysis of CD34+ primary bone marrow samples from these patients (n = 12) revealed HNRNPK expression is significantly reduced when compared to CD34+ bone marrow cells from healthy donors (n = 8) (Figure 1, p=0.0001), suggesting reduced HNRNPK expression may contribute to the etiology of myeloid malignancies.

Figure 1. HNRNPK expression is reduced in patients with AML that harbor a 9q deletion.

Figure 1

Open in a new tab

Quantitative RT-PCR analysis of HNRNPK levels from bone marrow aspirates of AML patients who carry 9q deletions (n = 12) and healthy donors controls (n = 8) (p = 0.0001). Data are represented as mean ± SEM determined from triplicate samples after normalization to geometric mean of β-ACTIN and GAPDH. ***p < 0.005. P-values were calculated using the Mann-Whitney test.

Generation of Hnrnpk haploinsufficient mice

Biochemical and cell based studies have previously been employed to examine the role of hnRNP K in regulating cellular processes (Gao et al., 2009; Notari et al., 2006; van Domselaar et al., 2012; Wang et al., 2011). However, in vivo evaluation of hnRNP K has been limited due to the lack of animal models. Thus, to directly examine the biological functions of hnRNP K, we used mouse embryonic stem cells (mESCs) generated by the KOMP consortium that harbor an Hnrnpk knockout allele (Hnrnpktm1(KOMP)Wtsi) to generate Hnrnpk+/− chimeric mice (Figure 2A). However, after screening more than 400 potential offspring on the C57Bl/6 background, only one Hnrnpk+/− mouse was generated (data not shown). This single mouse (male) rapidly developed a lymphoma and became moribund; however, prior to death, sperm was isolated and used for in vitro fertilization to generate Hnrnpk+/− mice on the Balb/C background. Germline transmission, validation of targeted recombination, evaluation of recombination sites, and verification of the absence of mutations were confirmed by PCR analyses using internal and external primers followed by DNA sequencing (Figure 2B and Figure S1A and B). Cre-mediated deletion of the neomycin cassette and Flp-mediated deletion of the LacZ and neomycin cassettes were confirmed by PCR analyses (Figure S1C).

Figure 2. Generation of Hnrnpk haploinsufficient mice.

Figure 2

Open in a new tab

(A) Schema of the Hnrnpk locus and the KOMP targeting vector (Hnrnpk1(KOMP)Wtsi). The targeting vector replaced exons 3 through 6 and with a neomycin selectable marker flanked by LoxP sites (ovals) and LacZ cassette flanked by Flp sites (arrowheads). Black numbered boxes denote Hnrnpk exons. (B) PCR analysis of targeted recombination in Hnrnpk+/− mice. The 5.9-kb amplicon (dark blue arrow heads) represents the recombined Hnrnpk allele using external and internal primers for the 5′ arm. Wild-type DNA serves as a negative control. The 10.1-kb (black arrow heads), 9.1-kb (red arrow heads), 8.1-kb (green arrow heads), and 5.5-kb (light blue arrow heads) amplicons represent the recombined Hnrnpk allele using external and internal primers for the 3′ arm. See also: Figure S1.

Biallelic loss of Hnrnpk results in embryonic lethality while haploinsufficiency results in developmental defects

Hnrnpk+/− mice were born at reduced Mendelian ratio and a significant portion of Hnrnpk+/− mice that were born died prior to weaning (31 of 103, 30%), typically within the first several days of life (Figure 3A). Hnrnpk+/− mice that did survive past weaning (21 days), displayed a pronounced retardation in growth compared to wild-type littermates (Figure 3B and 3C). This reduced growth persisted throughout the life of these mice, as both male and female Hnrnpk+/− mice were significantly smaller than their wild-type counterparts at six months (female: 25.74 ± 0.97 grams versus 20.57 ± 0.81 grams (p = 0.011) and male: 32.38 ± 2.49 grams versus 19.72 ± 1.89 grams (p = 0.004); respectively (Figure 3D). We next sought to generate a cohort of Hnrnpk-null mice (Hnrnpk−/−); however, we were unable to obtain mice harboring biallelic deletion of Hnrnpk (Figure 3E). These results suggest loss of Hnrnpk is incompatible with embryonic survival while reduced hnRNP K expression causes pleiotropic effects that impact neonatal survival and development.

Figure 3. Hnrnpk+/− mice have developmental defects and reduced hnRNP K expression.

Figure 3

Open in a new tab

(A) Observed and expected number of Hnrnpk+/− and wild type mice at weaning (21 days). (B) Growth retardation in Hnrnpk+/− mice at day 3 (Left panel) and 10 days (Right panel). (C) Weight (in grams) of Hnrnpk+/− (n = 10) and wild type littermates (n = 8) over the first four weeks of life (p = 0.0001). (D) Weight analysis (in grams) of adult Hnrnpk+/− and wild type mice stratified by sex (Female, p = 0.011) and (Male, p = 0.003). (E) Observed and expected ratio of Hnrnpk−/−, Hnrnpk+/−, and wild type mice from 12 separate hnRNP K+/− matings at weaning (21 days). (F) Quantitative RT-PCR analysis of Hnrnpk expression in the bone marrow of wild type (n = 4) and Hnrnpk+/− (n = 4) mice (p = 0.0396). (G) Western blot analyses of hnRNP K levels in lysates from wild type and Hnrnpk+/− mice, (upper-left panel: MEFs; upper-right panel: whole bone marrow; bottom-left panel: spleen; bottom-right panel: liver). Data are represented as mean ± SEM. *p < 0.05 and ***p < 0.005. p-values were calculated using Mann-Whitney test.

Given the Hnrnpk-null lethality, our attention was focused on investigating the potential impact of Hnrnpk haploinsufficiency. We evaluated expression changes resulting from deletion of one Hnrnpk allele by examining hnRNP K levels in tissues from Hnrnpk+/− and wild-type mice. Western blot and qRT-PCR analysis of Hnrnpk+/− spleens, bone marrow, livers, and mouse embryo fibroblasts (MEFs) revealed a significant decrease in hnRNP K levels compared to their wild type littermates (Figure 3F and 3G). Together, these results demonstrate that loss of one Hnrnpk allele significantly diminished hnRNP K levels and identified reduced hnRNP K expression as a casual factor in the Hnrnpk haploinsufficient phenotypes.

The p53/p21 pathway is altered in Hnrnpk +/− mice

Previous biochemical and pharmacological analyses have implicated hnRNP K in directly regulating p21 expression through the p53 pathway (Enge et al., 2009; Moumen et al., 2005). To precisely evaluate a link between hnRNP K and p21 expression in a genetically defined model, we attempted to generate Hnrnpk−/− MEFs to examine potential hnRNP K-mediated proliferation changes. We were unable to obtain Hnrnpk-null MEFs as Hnrnpk−/− embryos either fail to form or died in utero prior to day 13.5 (data not shown). We were, however, successful in generating Hnrnpk+/− MEFs and used these MEFs to evaluate the impact that reduced hnRNP K expression has on proliferation potential. Hnrnpk+/− MEFs had a growth advantage compared to wild type MEFs (Figure 4A and Figure S2). Given the proposed role of hnRNP K as a regulator of p53-dependent p21 expression, we next evaluated p21 levels in wild type and Hnrnpk+/− MEFs following γ-irradiation and observed Hnrnpk+/− MEFs did not fully transcriptionally activate p21 expression (Figure 4B), resulting in reduced p21 levels (Figure 4C). These data suggest that hnRNP K directly contributes to regulating the antiproliferative effects of p53 activities at the p21 promoter.

Figure 4. Reduced hnRNP K expression results in increased proliferation and a dampened p21 response following DNA damage.

Figure 4

Open in a new tab

(A) Increased cell proliferation in Hnrnpk+/− MEFs. Three low passage (P2) MEF cell lines per genotype were plated for 24, 48 or 72 hr, and then assayed by WST-8 (p = 0.0070). Each data point represents the mean ± SEM for three separate MEF lines per genotype (B) Quantitative RT-PCR analysis of p21 levels before and after irradiation from WT and Hnrnpk+/− MEFs. (C) Western blot analysis of hnRNP K and p21 expression in lysates from γ-IR treated wild type and Hnrnpk+/− MEFs. β-actin serves as an internal control. Data are represented as mean ± SEM. ***p < 0.005. P-values were calculated using unpaired t-tests and Mann-Whitney tests. See also, Figure S2.

Decreased hnRNP K expression results in myeloid hyperplasia

Hematopoietic cells undergo constant cycles of self-renewal and differentiation, making them extremely sensitive to genetic alterations that disrupt proliferation and differentiation programs (Seita and Weissman, 2010). Given that hnRNP K has been implicated in regulating differentiation potential, we next evaluated whether reduced hnRNP K expression impacts hematopoietic processes by evaluating complete blood counts from the peripheral blood of Hnrnpk+/− mice. Hnrnpk+/− mice had a significant increase in the number and percentage of neutrophils (29.37×103/μl ± 3.96 ×103/μl versus 12.10 ×103/μl ± 2.24 ×103/μl; p = 0.0035), basophils (0.19 ×103/μl ± 0.046 ×103/μl versus 0.083 ×103/μl ± 0.019 ×103/μl; p = 0.0477), and platelets (1858 ×103/μl ± 126.8 ×103/μl versus 992.3 ×103/μl ± 87.05 ×103/μl; p = 0.0095) compared to wild type mice (Figure 5A, Figure S3). Additionally, flow cytometry revealed that reduced hnRNP K expression resulted in a significant increase in CD11b/LyG6 double positive cells (markers of mature granulocytes) in the peripheral blood of Hnrnpk+/− mice compared to wild type mice (25.42% ± 15.72% versus 12.93% ± 11.06%; respectively, p = 0.0354) and bone marrow (53.76% ± 8.61% versus 30.17% ± 13.98%; respectively, p = 0.008) (Figure 5B). These findings indicate that Hnrnpk haploinsufficiency promotes a myeloproliferative phenotype.

Figure 5. Hnrnpk+/− mice develop myeloid hyperplasia.

Figure 5

Open in a new tab

(A) Cell blood count (CBC) of peripheral blood comparing neutrophils (p = 0.0006) and basophils (p = 0.0477) from wild type (n = 8) and Hnrnpk+/− (n = 11) mice and platelets (p = 0.0095) of wild type (n = 6) and hnRNP K+/− (n = 6) mice. (B) Flow cytometry analysis of double positive Ly6G/CD11b myeloid populations (50,000 gated cells) in the peripheral blood of wild type (n = 6) and hnRNP K+/− (n =6) (upper panel, p = 0.0354) and bone marrow of wild type (n = 7) and hnRNP K+/− (n = 13) (bottom panel, p = 0.0008) mice. Percentages were compared and analyzed using Mann-Whitney test. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.005. See also, Figure S3.

Hnrnpk+/− mice have reduced survival and are tumor prone

To evaluate the long-term consequence of reduced hnRNP K expression, a cohort of Hnrnpk+/− mice were generated and monitored for changes in survival. To minimize the impact that the Hnrnpk-dependent neonatal lethal phenotype would have on our survival analyses, only Hnrnpk+/− mice that survived past 21 days (weaning) were included in our survival comparisons. Hnrnpk+/− mice showed a significant decrease in overall survival compared to wild type littermates (p = 0.0001) (Figure 6A). The median overall survival for the Hnrnpk+/− mice was 474 days while wild type mice have not reached the 50 percent survival by that time point (hazard ratio = 5.45; confidence interval = 3.30 to 9.12).

Figure 6. Hnrnpk+/− mice have reduced survival and are tumor prone.

Figure 6

Open in a new tab

(A) Kaplan-Meier curves indicating survival of Hnrnpk+/− (n = 76) and wild type (n = 37) mice. Statistical significance was determined by log rank test (p = 0.0001). (B) Hnrnpk haploinsufficiency results in tumor phenotypes (n = 55; 62% myeloproliferation, 31% lymphoma and 4% hepatocellular carcinoma). (C) Hematoxylin and Eosin (H&E) staining of paraffin-embedded bone marrow sections and peripheral blood (PB) smears from wild type Hnrnpk+/− mice diagnosed with myeloid hyperplasias. The scale bar for H&E staining represents 50 μm and 25 μm for PB smears. (D) H&E staining of paraffin-embedded splenic sections and PB smears from wild type and lymphoma burdened Hnrnpk+/− mice. (E) H&E and immunohistochemical staining of malignant T-cell (CD3 positive, Bottom Panel) and B-cell (B220 positive, Top Panel) lymphomas in the liver of Hnrnpk+/− mice. (F) H&E staining of paraffin-embedded liver sections from wild type and Hnrnpk+/− mice diagnosed with hepatocellular carcinoma. See also Figure S4.

To explore the causes of this reduced survival, moribund Hnrnpk+/− mice were sacrificed and tissues were harvested for evaluation. Pathological analyses revealed that Hnrnpk+/− mice had a proclivity to develop hematologic malignances, as the majority of Hnrnpk+/− mice pathologically analyzed (34 out of 55; 62%) displayed a significant myeloproliferative phenotype (Figures 6B and 6C). To investigate the basis of these myeloproliferative phenotypes, we performed karyotype analyses and observed that hnRNP K expression resulted in significant genomic instability as cells from Hnrnpk+/− bone marrows had frequent structural aberrations compared to wild type control bone marrows (Figure S4A). Hnrnpk+/− bone marrows exhibited at least one genomic alteration, with the majority being fusions (93%) and the remaining seven percent breaks. Additionally, 29% of the total bone marrow cells analyzed contained polyploidy cells. These observations suggest that Hnrnpk loss contributes to tumorigenesis through its regulation of genomic stability. Even though we routinely observed Hnrnpk+/− bone marrows packed with myeloid cells that harbored chromosomal abnormalities, this did not often result in a significant impact in splenic size. This lack of difference in splenic volume was attributed to the significant overall reduction in the size of the Hnrnpk+/− compared to wild type littermates (Figure 3C). Accordingly, analyses of organ weights revealed that Hnrnpk+/− mice had significantly smaller spleens (0.045 ± 0.024 grams, n = 7) than age-matched wild type control mice (0.14 ± 0.07 grams, n = 3; p = 0.0357).

In addition to myeloproliferative phenotypes, Hnrnpk+/− mice also frequently harbored other hematologic malignancies. Lymphomas were observed in 31% (17 out of 55) of Hnrnpk+/− mice (Figure 6B, D, and E). To determine the cell of origin of these hnRNP K+/−-dependent lymphomas, we evaluated the expression of lineage-specific lymphoid markers. Lymphomas from these Hnrnpk+/− mice expressed either T-cell (CD3) or B-cell (B220) surface markers, indicating that reduced hnRNP K expression promotes both T- and B-cell lymphomas (Figure 6E, right panels). Furthermore, we observed a thrombocytosis-like phenotype in some Hnrnpk+/− mice, as reduced hnRNP K expression often resulted in increased megakaryopoesies (Figure S4B). In addition to the pronounced hematopoietic phenotypes, we observed that two Hnrnpk+/− mice (2 of 55; 4%) developed hepatocellular carcinomas (Figure 6F). These findings indicate that hnRNP K plays a significant role in preventing tumor formation and further suggests that hnRNP K may function as a previously uncharacterized tumor suppressor.

Reduced hnRNP K expression results in altered cytokine signaling and increased proliferation and differentiation potential

To gain an understanding of the molecular pathways altered by diminished hnRNP K expression, we evaluated cytokine levels in wild type (n = 10) and Hnrnpk+/− (n = 9) mouse sera. Interestingly, Hnrnpk+/− mice had a significantly higher serum concentration of interleukin-3 (IL-3) and IL-6 which are well known to drive myeloproliferation and differentiation (Figure 7A). Additionally, we found that serum levels of granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte-colony stimulating factor (G-CSF), which are required for propagation of the myeloid lineage, were also significantly increased in Hnrnpk+/− mice (Figure 7A). Since myeloproliferation can occur as a result of an acute inflammatory process (Gabay and Kushner, 1999), we evaluated global inflammation by examining expression of C-reactive protein in the livers and bone marrows of wild type and Hnrnpk+/− mice. We observed no significant difference in the levels of C-reactive protein in Hnrnpk+/− mice and wild type littermates (Figure S5), indicating an acute phase response is not the primary source of these myeloproliferative phenotypes.

Figure 7. hnRNP K-mediated activation of cytokines contributes to proliferation and differentiation phenotypes.

Figure 7

Open in a new tab

(A) Cytokines array analysis of IL-3, IL-6, G-CSF and GM-CSF levels in the peripheral blood of wild type (n = 10) and Hnrnpk+/− mice (n = 9) diagnosed with myeloid hyperplasia (IL-3: p = 0.0002, IL-6: p = 0.0004, G-CSF: p = 0.0002, and GM-CSF: p = 0.0002). (B) Analyses of colony formation assays after 14 days using Lin-CD117+ hematopoietic stem cells from wild type (n = 8) and Hnrnpk+/− (n = 17) mice (number of colonies: p = 0.0078 and number of cells: p = 0.0001). All experiments were performed in triplicate. Data are represented as mean ± SEM. **p < 0.01 and ***p < 0.005. P-values were calculated using Mann-Whitney test. (C) Replating analysis of Lin-CD117+ hematopoietic stem cells from wild type (n = 5) and Hnrnpk+/− (n = 7) mice. Each passage was analyzed every 14 days. All experiments were performed in quadruplicate. Data are represented as mean ± SEM. (D) Flow cytometry analysis of CD45.1+ (host) and CD45.2+ (donor) in the transplanted host mice and B220+ and Ly6G+ cells within the CD45.2+ population. (E) CBC analysis of untransplanted NSG (control) mice and NSG mice transplanted with wild type or Hnrnpk+/− HSCs. (F) Wright's staining of peripheral blood (PB) smears from NSG mice transplanted with wild type or Hnrnpk+/− HSCs harboring malignant lymphoid cells. See also, Figure S5.

Given the significant changes in cytokine levels observed in Hnrnpk+/− mice, we next evaluated the impact that these specific cytokines have on myeloproliferative potential in Hnrnpk+/− hematopoietic stem cells (HSC) using in vitro colony formation assays. Lin-CD117+ HSCs from Hnrnpk+/− mice exhibited a significant increase in proliferation and differentiation potential when compared to wild type Lin-CD117+ HSCs grown in MethoCult (Figure 7B). To further explore the replicative potential of Hnrnpk+/− HSCs, we performed replating in vitro colony formation assays. Hnrnpk+/− HSCs displayed higher replating potential than wild type HSCs, as Hnrnpk+/− HSCs had a significant increase in the number of colonies at each replating and sustained their proliferation and differentiation potential for more passages (Figure 7C). Together, these findings suggest Hnrnpk haploinsufficiency and cytokine signaling may cooperate to disrupt hematopoietic homeostasis leading to proliferative and differentiation defects.

Hnrnpk+/− hematopoietic stem cells develop myeloid proliferative and malignant lymphoid phenotypes in transplanted mice

To more fully understand the replicative potential of Hnrnpk+/− HSCs and determine whether transplantation of these cells could give rise to malignancies, we isolated Lin-CD117+ HSCs from three separate CD45.2+ Hnrnpk+/− and wild type mice and injected them into irradiated CD45.1+ NOD.Cg-Prkdc(scid)Il2rg(tm1Wjl)/SzJ (NSG) recipient mice. While we observed engraftment of CD45.2+ HSCs isolated from both Hnrnpk+/− and wild type mice in the recipient CD45.1+ NSG mice; two out of the three NSG mice transplanted with Hnrnpk+/− HSCs had a significant expansion of the myeloid compartment with a hypereosinophilic phenotype compared to NSG mice transplanted with wild type HSCs (Figure 7D and E). Interestingly, the remaining NSG mouse transplanted with the Hnrnpk+/− HSCs did not develop a myeloproliferative phenotype but instead developed a malignant lymphoma phenotype marked by expansion of aberrant lymphocytes (Figure 7F). These mice displayed medium to large lymphoma cells in the peripheral blood with irregular nuclear contours, small nucleoli, and basophilic cytoplasm. Together, these transplantation data indicate the cell autonomous ability of Hnrnpk+/− HSCs to drive these malignant phenotypes independent of the Hnrnpk+/− microenvironment.

Diminished hnRNP K expression results in deregulation of proliferation and differentiation pathways

In order to examine the molecular mechanisms that contribute to the observed hematopoietic alterations, we isolated tissues from Hnrnpk+/− and wild type mice and evaluated expression of genes known to influence myeloproliferative neoplasms. Hnrnpk+/− mice had significantly reduced expression of the differentiation factors C/EBPα and β in the bone marrow and liver (Figures 8A and 8B). Interestingly, only one isoform of C/EBPα— C/EBPα p42— appeared to be reduced in an hnRNP K-dependent manner (upper band, denoted by arrow head, Figure 8A). In addition, reduced hnRNP K levels were associated with reduced expression of its transcriptional target p21 (Figure 8C) and activation of STAT3 as determined by phosphorylation of tyrosine 705 (Figure 8D). Given the role of hnRNP K in transcriptional regulation and the fact that p21, C/EBPα, and C/EBPβ contain putative hnRNP K DNA binding sites (Lynch et al., 2005), we next performed hnRNP K-specific chromatin immunoprecipitation (ChIP) assays. These experiments revealed that hnRNP K directly interacted with the p21, C/EBPα, and C/EBPβ genes (Figure 8E and Figure S6A). Importantly, hnRNP K binding to these targets was significantly diminished in Hnrnpk+/− tissues. Our observation that the p42 isoform of C/EBPα was reduced in Hnrnpk+/− tissues, coupled with hnRNP K's putative translational activity, suggested hnRNP K may also interact with the C/EBPα transcript. To examine this, we performed RNA immunoprecipitation assays and observed a significant interaction between hnRNP K and the C/EBPα transcript in both wild type murine tissues and in the human leukemic cell line HL-60 (Figure 6B and Figure S6B and C). However, in Hnrnpk+/− samples the interaction between hnRNP K and C/EBPα was reduced. Together, these data suggest that hnRNP K may directly regulate cellular differentiation through the C/EBP pathways and proliferation through the p53/p21 and JAK/STAT pathways.

Figure 8. Hnrnpk haploinsufficiency alters expression of genes controlling proliferation and differentiation.

Figure 8

Open in a new tab

(A) Western blot analysis of C/EBPα and C/EBPβ from lysates of wild type and Hnrnpk+/− mice. Upper arrow denotes the p42 isoform of C/EBPα and the lower arrow marks the p30 isoform. β-actin serves as a loading control. (B) Quantitative RT-PCR analysis of C/EBPα and C/EBPβ levels from whole bone marrows of wild type (n = 8) and Hnrnpk+/− (n = 10) mice (C/EBPα: p = 0.042 and C/EBPβ: p = 0.020). (C) Quantitative RT-PCR analysis of p21 levels in wild type (n = 8) and Hnrnpk+/− (n = 10) bone marrows (p = 0.0044). (D) Immunohistochemical analysis of phos-Y705Stat3 levels in the bone marrow wild type and Hnrnpk+/− mice diagnosed with myeloid hyperplasia. The scale bar represents 50 μm. (E) ChIP analysis of hnRNP K interacting with the C/EBPα, C/EBPβ and p21 genes in wild type or Hnrnpk+/− tissues. Each rectangle represents the corresponding gene. The closed black box denoted the region harboring a putative hnRNP K binding sequence that is amplified by the corresponding primer sets. The “check-mark” denotes positive hnRNP K binding and the “X” denotes a lack of interaction. (F) Quantitative RT-PCR analysis of TNFα levels in spleens from wild type (n = 7) and tumor burdened Hnrnpk+/− (n = 5) mice (p = 0.0177). (G) H&E staining and immunohistochemical analyses of B220, TNFα, and phos-Y705STAT3 expression in Hnrnpk+/− mice with malignant lymphoid infiltration in the liver. Data are represented as mean ± SEM. *p < 0.05 and ***p < 0.005. P-values were calculated using unpaired t and Mann-Whitney tests. See also Figure S6.

We next examined expression changes in genes known to be deregulated specifically in lymphomas and thought to be regulated by hnRNP K. Lymphomas from Hnrnpk+/− had significantly elevated levels of TNF-α, an activator of cell proliferation, and increased phos-Stat3 levels compared to wild type mice (Figures 8F and 8G, bottom panels). Similar to the increased proliferation potential observed in the myeloid compartment, we also observed malignant phenotypes arising in lymphoid lineages. Interestingly, in the sera of these mice, we also observed an increase in the pro-lymphoid cytokines IL-2, IL-10 and TNFα (Figure S6D). These findings suggest the impact of reduced hnRNP K expression may occur at an early hematopoietic precursor stage and that hnRNP K-mediated alterations in specific targets may dictate phenotypic manifestations.

Discussion

Over the past decade, in vitro and biochemical studies have shown that hnRNP K impacts cellular programs that influence cancer development (Notari et al., 2006). As such, clinical association studies have implicated that change in hnRNP K expression associate with disease phenotypes. (2013; Enge et al., 2009; Kronke et al., 2013; Moumen et al., 2005; Roychoudhury and Chaudhuri, 2007). In this study, we examined how reduced hnRNP K expression influences malignant phenotypes in the hematopoietic compartment. Herein, we demonstrate that HNRNPK expression is significantly reduced in patients with AML that carry a 9q21.32 deletion. To fully examine the role of hnRNP K in vivo, we generated an Hnrnpk+/− mouse model, which potentially mimics the genetic loss observed in AML patients with 9q21.32 deletions. Here, we report that hnRNP K is critical for embryogenesis and development, as biallelic deletion of Hnrnpk results in embryonic lethality prior to day 13.5 in utero while haploinsufficiency results in a partial neonatal lethal phenotype and developmental defects. Hnrnpk+/− mice displayed a significant reduction in survival and a highly penetrant cancer phenotype, as defined by lymphomas, myeloproliferation, and hepatocellular carcinomas. Reduced hnRNP K expression contributed to these phenotypes by disrupting differentiation and proliferation pathways and through decreased genomic stability. These results indicate that hnRNP K is a tumor suppressor that may influence proliferation and differentiation activities.

The role of hnRNP K in tumor formation has been difficult to address due to the number of cellular processes it regulates. On one hand, hnRNP K overexpression has been reported in some solid tumors and associates with poor clinical status (Matta et al., 2009; Roychoudhury and Chaudhuri, 2007). In studies using CML cell lines, the oncogenic potential of the BCR-ABL fusion gene appears to be dependent on hnRNP K, suggesting hnRNP K may play a role in the BCR-ABL-mediated oncogenic addiction (Notari et al., 2006). In fact, shRNA knock-down of hnRNP K in these cells resulted in decreased differentiation potential and uncovered the fact that hnRNPK's translational functions, but not its transcription functions, contribute to the oncogenic potential of BCR-ABL. Furthermore, experiments in pancreatic cancer cells revealed that siRNA-mediated knock down of hnRNP K resulted in an inhibition of proliferation (Zhou et al., 2010). When hnRNP K is overexpressed in cell lines, this cytoplasmic portion became elevated where it was shown to influence the activity of c-Src (Ostareck-Lederer et al., 2002). Taken as a whole, these observations suggest that when hnRNP K is overexpressed, it may possess oncogenic functions through its translational and cytoplasmic functions.

In contrast, there is also biochemical and clinical data that suggest hnRNP K loss may influence tumorigenesis as well. hnRNP K has been shown to influence the tumor suppressive activities of p53 (Enge et al., 2009; Moumen et al., 2005) and in AML, hnRNP K is deleted in a subset of AML patients (Kronke et al., 2013). Furthermore, mutations in hnRNP K have also been implicated in driving AML progression (2013). These findings indicate that hnRNP K potentially functions as a tumor suppressor through the p53 pathway, and that its loss may directly result in malignant phenotypes. This notion of hnRNP K as having a role in tumor suppression gained significant traction following biochemical and cell line studies that revealed hnRNP K was required activation of p21 in a p53-dependent manner following DNA damage or treatment with preclinical compounds that reactivate mutant p53 functions. To examine how hnRNP K impacts p21 activation in a genetically defined system, we treated Hnrnpk+/− MEFs with ionizing radiation and observed a significant hnRNP K-dependent reduction in p21 activation. Using an in vivo approach, we determined that p21 levels were significantly decreased in tissues from Hnrnpk+/− mice. Furthermore, this in vivo regulation appears to be a consequence of a physical interaction between hnRNP K and the p21 gene. Together, these in vitro and in vivo results indicate that hnRNP K is indeed a direct regulator of p21 activation and loss of this hnRNP K-mediated activity partially explains the observed proliferative phenotypes observed in the MEFs and tissues of Hnrnpk+/− mice.

In addition to influencing proliferative programs, hnRNP K has also been implicated in regulating cellular differentiation. In our study, we observed that C/EBPα and β were significantly deregulated in tissues from Hnrnpk+/− mice. While recent ChIP-Seq studies in cell lines suggest hnRNP K may directly modulate C/EBPβ expression by specifically interacting with its promoter (Mikula et al., 2013), less is known regarding how hnRNP K regulates C/EBPα levels. Examination of the potential interaction between hnRNP K and the C/EBPα and C/EBPβ genes revealed that hnRNP K directly interacts with both genes. Similar to our observation at the p21 promoter, reduced hnRNP K expression also resulted in decreased binding at these sites. Interestingly, we observed that the C/EBPα p42 isoform, but not p30, was significantly downregulated in Hnrnpk+/− mice, suggesting that changes in hnRNP K expression altered expression of the C/EBPα p42 isoform only. The observed changes in the expression ratio of C/EBPα isoforms (e.g.; p42 to p30) may have a significant impact on the myeloproliferative phenotypes observed in the Hnrnpk+/− mice, as deregulation of the p42 isoform significantly alters myelogenesis (Schuster et al., 2013) and the p30 isoform does not possess the anti-mitotic activity of the p42 isoform (Nerlov, 2004). As such, the p42 isoform has been identified as a myeloid tumor suppressor, as its absence allows for alterations in lineage commitment and unchecked myeloid proliferation (Kirstetter et al., 2008). Due to leaky ribosomal scanning, both the p42 and p30 isoforms of C/EBPα are translated from the same mRNA (Hsieh et al., 1998). Given the known translational activity of hnRNP K, and our observation that hnRNP K interacts with C/EBPα mRNA, it is plausible that hnRNP K may also influence the expression of C/EBPα isoforms. Taken together, these findings suggest perturbations in hnRNP K levels deregulate C/EBP expression leading to a failure in maintaining a homeostatic balance in hematopoietic differentiation that culminates in the observed myeloproliferative phenotypes.

In contrast to the C/EBPα p42 mouse model (Kirstetter et al., 2008) that produces a myeloid hyperplasia phenotype with high levels of CD11b+/Ly6G- cells, Hnrnpk+/− mice displayed high levels of CD11b+/Ly6G+ cells, suggesting reduced hnRNP K expression likely contributes to alterations in differentiation potential through additional mechanisms outside of the C/EBP pathway. As such, we observed a significant increase in cytokines that regulate proliferation and differentiation potential (e.g.; IL-3, IL-6, GM-CSF, and G-CSF) and that are required for transitioning HSCs to the myeloid lineage. Together, these additional factors may be a necessary and sufficient secondary input required for HSCs to develop into myeloid malignancies. To examine the interplay between HSC maturation and myeloid proliferation and differentiation defects, we used in vitro cytokine dependent colony formation assays. These experiments demonstrated that Hnrnpk haploinsufficiency led to increased differentiation and proliferation potential in hematopoietic stem cells (HSCs). Furthermore, replating in vitro cytokine dependent colony formation assays demonstrated the ability of Hnrnpk+/− HSCs to be serially propagated, indicating the increased long-term replicative potential of the Hnrnpk+/− HSCs. Given that cytokines necessary for HSCs to transition into myeloid and lymphoid lineages (IL-2, IL-10 and TNFα; Figure S6D) were significantly upregulated in Hnrnpk+/− mice, suggests these in vivo events may simply be a consequence of aberrant cytokine production specific to the Hnrnpk+/− bone marrow niche. To test the possibility, we transplanted Lin-CD117+ Hnrnpk+/− cells into recipient NSG mice. Critically, Hnrnpk+/− HSCs had the capacity to trigger myeloproliferative and lymphoma phenotypes independent of the Hnrnpk+/− microenvironment. Together, these results indicate the Hnrnpk+/− HSCs, by themselves, have the capacity to induce cancer processes and stimulate these proliferative phenotypes.

In myeloproliferative neoplasms, as well as other cancers, the JAK/STAT pathway is a critical hub that links cytokine signaling with transcriptional programs controlling proliferation and differentiation potential (Quintas-Cardama and Verstovsek, 2013). Consistent with our findings of increased IL-3 and IL-6 in Hnrnpk+/− mice, we observed a significant increase in STAT3 activation in the myeloproliferative bone marrows. In patients with myeloproliferative neoplasms (MPN), activation of the JAK2/Stat pathway commonly results from activating mutations in JAK2 (V617F) (James et al., 2005). However, our data suggest that reduced hnRNP K expression also results in a similar activation of the STAT pathway but without the need for JAK2 mutations. Thus, reduced hnRNP K expression may serve as an initial event to stimulate proliferation and force cancer phenotypes.

The relationship between hnRNP K and STAT3 may not be exclusively related to the granulocytic compartment, as JAK2 mutations leading to STAT3/5 activation are commonly observed in patients with essential thrombocythemia (Quintas-Cardama and Verstovsek, 2013). Likewise, we observed a substantial number of Hnrnpk+/− bone marrows that displayed megakaryopoietic phenotypes (Figure S4B). These data support the idea that hnRNP K is a critical factor that regulates the transition of committed myeloid progenitors to terminally differentiated hematopoietic lineages through activation of cytokine- and Stat3-dependent pathways. In addition to controlling myeloid hyperplasia, activation of STAT3 also promotes B cell proliferation and associated with poor overall survival in patients with diffuse large B cell lymphoma (Huang et al., 2013). Similar to these findings, we observed that hnRNP K mediated STAT3 activation also contributed to lymphomogenesis in Hnrnpk+/− mice. Together, these results indicate that hnRNP K loss significantly influences STAT3 activation which directly contributes to the differentiation and proliferation defects in the hematologic compartment of Hnrnpk+/− mice.

In this study, we present in vivo evidence that hnRNP K is directly involved in cancer development. Clinical studies have revealed that the HRNPK gene resides in the minimally deleted region of the 9q21.32 locus which is lost in a subset of patients with AML (Dayyani et al., 2008; Kronke et al., 2013; Sweetser et al., 2005), suggesting alterations in hnRNP K expression associate with leukemogenesis in these patients. Given that hnRNP K expression has not been examined in primary AML patient samples that harbor a 9q deletion, we evaluated whether the remaining HNRNPK allele could compensate for the deleted gene in maintaining wild type hnRNP K expression levels. In our 9q deleted AML patient samples, we observed a dramatic decrease in HNRNPK levels, suggesting this loss contributed to the disease process (Figure 1). This finding may have significant translational applications, as there are currently no targeted therapies for patients with AML that harbor a 9q21.32 deletion. Thus, a better understanding of how hnRNP K influences the p53/p21- and STAT3- pathways could lead to management of this disease and the development of tailored therapeutic regimens for these individuals. Therapies that reactivate the p53 pathway (e.g.; nutlin-3) or disrupt the STAT3 pathway (ruxolitinib or AZD9150, a STAT3 inhibitor currently in clinical trial) could be therapeutically beneficial and used to enhance current treatment strategies.

Based on data from our Hnrnpk+/− mouse model, we demonstrate that hnRNP K behaves as a tumor suppressor through its regulation of p21 and the p42 isoform of C/EBPα. However, there are also clinical association data and cell line studies suggesting that hnRNP K may also have a role in oncogenesis. These two dichotomous results suggest that a homeostatic balance in hnRNP K expression must be maintained for proper cellular regulation and to prevent tumorigenesis. Thus, any disequilibrium in expression (reduced or increased) could result in dire consequences. Similar to the importance of our Hnrnpk+/− mouse model in determining that hnRNP K has tumor suppressive functions, the generation of mouse models that overexpresses hnRNP K will be critical to determining whether hnRNP K has oncogenic potential and if its overexpression directly impacts tumor formation in vivo. These overexpression models could be employed to examine a causal relationship between increased hnRNP K levels and changes in its cellular localization, its direct influences on c-Myc expression, and its interaction with kinases such as c-Src.

In summary, our findings demonstrate that Hnrnpk haploinsufficiency results in highly penetrant and transplantable cancer phenotypes, particularly in the hematological compartment with expansion of myeloid lineages, and a significant reduction in survival. Reduced hnRNP K levels contributed to these phenotypes by deregulating proliferation and differentiation programs through its control of the p53/p21 and C/EBP pathways, deregulation of cytokines, and STAT3 activation. Our clinical data suggest changes in hnRNP K levels associate with myeloid malignancies as hnRNP K expression is reduced in patients with AML that carry 9q21.32 deletions. Together, these findings suggest that hnRNP K is a tumor suppressor gene involved in hematologic malignancies.

Experimental Procedures

Analysis of HNRNPK expression in AML patient samples

All twelve human AML patient samples were obtained from the AML tissue bank at MD Anderson Cancer Center following protocol PA14-0478, which was approved by the Institutional Review Board at MD Anderson. Patient consent was obtained at time of collection. All CD34+ cell examined harbored a deletion of the 9q21.32 locus in greater than 50% of metaphases analyzed. In addition to the 9q deletions, three out of twelve samples contained several additional chromosomal aberrations and one out of twelve harbored a t(8;21) translocation. RNA from Ficoll-enriched mononuclear cells of AML patients (n = 12) and wild type donors (n = 8) was isolated and reverse transcribed to generate cDNA. cDNAs were used in TaqMan assays to determine expression of HNRNPK and two housekeeping control genes (GAPDH and β-ACTIN). All assays were performed in triplicate and changes in expression were determined by comparing HNRNPK expression to housekeeping controls (Pfaffl, 2001).

Generation of Hnrnpk+/− mice

Mice carrying a knockout allele (Hnrnpktm1(KOMP)Wtsi) were generated using KOMP targeted JM8A1.N3 mESCs (a C57BL/6N strain) mouse embryonic stem (mES) cells. Quality control of the mES cells was performed in the MD Anderson Genetically Engineered Mouse Facility (GEMF). mES cells were injected into C57/Bl6 blastocysts using standard procedures as previously performed in GEMF to generated chimeric Hnrnpk+/− mice (Post et al., 2010). Due to a highly penetrant haploinsufficient lethal phenotype, only one viable Hnrnpk+/− mouse (a male) was generated on the C57Bl/6 mouse background. Sperm from this mouse was isolated and used in IVF procedures to generate mice on a Balb/C background. Balb/C;C57BL/6 hybrid mice were subsequently backcrossed for three generations with wild type Balb/C mice. These Hnrnpk+/− Balb/C mice were used to generate the Hnrnpk+/− cohort. The neomycin cassette was removed by crossing the Hnrnpk+/− mice with ZP-3 Cre expressing mice. Females from these crosses were then bred to wild type males. The LacZ and neomycin cassette was removed by crossing Hnrnpk+/− mice with mice expressing Flp recombinase. All mouse studies were conducted with approval from the Institutional Animal Care and Use Committee at MD Anderson Cancer Center under protocol 0000787-RN01.

Genotyping and Validation of the Hnrnpk Allele

PCR based strategies using primer sets that were external and internal to both 5′ and 3′ arms were initially performed to confirm homologous recombination and germline transmission. The 5′ arm was verified by PCR amplification and visualization using external primer 5′-GCATAATAGCAACATATAAAGTTAC-3′ (forward) and internal primers 5′-AACCTTCCCCAACACCGAAACCAATG-3′ (forward) and 5′-CACAACGGGTTCTTCTGTTAGTCC -3′ (reverse) primers. The 3′ arm was verified by external primers 5′-CAGCATCAGCCACTTCCATAACC-3′ (reverse) and 5′-TGCCAATGTATTTTGTCTATGTCC-3′ (reverse) and internal primers 5′-CAGTCTCAACTCGCAATCAAAGTCAC-3′ (forward) and 5′-ATACTGTCGTCGTCCCCTCAAAC-3′ (forward); 5′-GTTGCAGTGCACGGCAGATACACTTGCTGA-3′ (forward), 5′-TTCGGCTATGACTGGGCACAACAG-3′(forward) and 5′-CAACTGCCTAAGATGGAAGTAACCT-3′ (reverse) primers. In addition to visualization of the PCR amplicon following gel electrophoresis, PCR amplicons were purified and evaluated by Sanger sequencing to confirm the sequence of the targeted allele. Additionally, the lacZ reporter, neomycin cassette, Flp-FRT sites and loxP sites were evaluated by PCR amplification and Sanger sequencing using primers 5′-AACCTTCCCCAACACCGAAACCAATG-3′ (Forward), GTGGCTTCAATTACCCTTTCTTGG (Forward), 5′-AGTCCCAACCCCTTCCTCCTAC-3′(Reverse), 5′-CGACTCCTGAGCCCGTCAGTATC-3′(Forward) and 5′-CCAAACTCATCAATGTATCTTATC-3′ (Reverse).

qRT-PCR

RNA from wild type and Hnrnpk+/− mice bone marrows, HSCs, spleens, and livers were extracted and purified using Tri Reagent (Sigma Aldrich). Purified RNA was reverse-transcribed using the iScript cDNA Synthesis kit (BioRad) to generate cDNA. qRT-PCR was performed with iTaq UniveSYBR Green SMX as per instructions on an ABI StepOnePlus Real Time PCR System. Details are described in Supplemental Experimental Procedures.

ChIP-qPCR

Chromatin immunoprecipitation assays and quantification of chromatin interaction by real-time PCR were performed as described (Bochkis et al., 2008). p21, C/EBPα and C/EBPβ primers selection was determined by in silico analyses of putative hnRNP K binding motifs (TCCCCA) near the p21, C/EBPα and C/EBPβ loci (Lynch et al., 2005). Details are described in Supplemental Experimental Procedures.

Western Blot

Western Blot analyses were performed using protein lysates isolated from the bone marrows, HSCs, spleens, and livers of wild type and Hnrnpk+/− mice. Primary antibodies used were hnRNP K (D-6, SantaCruz), C/EBPα (D56F10, Cell Signaling), C/EBPβ (LAP, Cell Signaling) and p21 (F-5, SantaCruz). Details are described in Supplemental Experimental Procedures.

Mouse Embryo Fibroblasts

Mouse embryo fibroblasts (MEFs) from Hnrnpk+/− and wild type mice were prepared from embryonic day 13.5 embryos and cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37°C and 5% CO2. For analysis of proliferation, MEFs were equally plated, collected at indicated time points, and counted using trypan blue exclusion analysis or analyzed by WST-8 assay (Abnova) according to the manufacture's protocol. Prior to WST-8 analysis, cells were incubated with reconstituted WST-8 for two hr, and absorbance was read at a wavelength of 450 nm.

DNA Damage

For DNA damage experiments, MEFs were irradiated with 3 Gy ionizing radiation and collected 3 hr post treatment.

Flow Cytometry

Flow cytometry was performed using cells from colony forming units, bone marrows, and peripheral blood. Details are described in Supplemental Experimental Procedures.

Survival Analysis

Survival analysis was performed using the Kaplan-Meier method. Differences between survival distributions were analyzed using the log-rank test. Hazard ratio and confidence interval was obtained by Mantel-Haenszel analysis. Statistical computations were performed using GraphPad Prism 6.0.

Pathological Analysis and Immunohistochemistry

Moribund mice were sacrificed according to IACUC and MD Anderson guidelines under protocol 0000787-RN01. Details are described in Supplemental Experimental Procedures.

Cytokine multi-plex assay

Sera from the peripheral blood of wild type and Hnrnpk+/− mice were isolated following high speed centrifugation. Cytokine levels were analyzed using the Bio-Plex Pro™ Mouse Cytokine Multi-Plex Assay (BioRad) according to the manufacturing protocol. Results were processed on the Bio-Plex® MAGPIX™ Multiplex Reader (BioRad) and analyzed using the Bio-Plex manager software with 5PL curve fitting.

Colony Formation Assay

Hematopoietic stem cells (Lin-CD117+) were isolated from the bone marrows of Hnrnpk+/− (n = 17) and wild type mice (n = 8) using CD117+ microbead kits (#130-091-224, Miltenyi Biotech). 50,000 fractionated HSCs were cultured in triplicate wells of a P24 plate in methylcellulose medium with cytokines (IL-3 20 ng/ml, IL-6 20 ng/ml, EPO 3units and SCF 50 ng/ml), MethoCult™ GF M3434 (Stem Cell Technologies). Colony formation assays using Hnrnpk+/− and wild type Lin-CD117+ HSCs were performed in triplicate for each sample. Results were read after 14 days by counting the number of colonies and extracting the cells to perform viability assay (trypan blue) and flow cytometry.

Colony Formation Replating Assay

Fourteen days after the initial colony formation assay, Hnrnpk+/− and wild type cells were harvested and resuspended in PBS. Viable cells were counted using trypan blue and 50,000 total cells were re-seeded in quadruplicate cultures in methylcellulose medium supplemented with cytokines (IL-3 20 ng/ml, IL-6 20 ng/ml, EPO 3 units and SCF 50 ng/ml) (MethoCult™ GF M3434, Stem Cell Technologies). This process was repeated on a 14-day cycle.

Transplant Assays

500,000 Lin-CD117+ cells from Hnrnpk+/−(n = 3) and WT (n = 3) were injected into the tail veins of irradiated NOD.Cg-Prkdc(scid)Il2rg(tm1Wjl)/SzJ (NSG) mice. NSG mice were subjected to a non-lethal dose of γ-irradiation (2.5Gy) 24 hr before injection of cells. Three months post injection, mice were sacrificed and evaluated for alterations in the hematological compartment between the Hnrnpk+/− and wild type mice by CBC and pathologic and flow cytometry analyses. Engraftment was determined by flow cytometry using CD45.1-FITC and CD45.2-PE antibodies (eBiosciences). Corresponding isotypes were used as controls. Subpopulations of CD45.2+ cells were determined as above using FlowJo software (http://www.flowjo.com/).

Statistical Analysis

Statistical analyses were performed using unpaired t and Mann-Whitney tests. P-values less than 0.05 were considered statistically significant. Survival curve used the Kaplan-Meier method. Differences between survival distributions were analyzed using the log-rank test. Hazard ratios and confidence intervals were obtained by Mantel-Haenszel analysis. Statistical computations were performed using GraphPad Prism 6.0.

Supplementary Material

1
2

Significance.

Deletion of the 9q21.32 locus is observed in some patients with AML, suggesting an tumor suppressor resides in this region. HNRNPK is an attractive candidate, as it is one of six genes mapped to this region and is thought to drive AML progression when mutated. Here, we show that AML patients harboring this deletion have reduced HNRNPK expression, and demonstrate through an animal model that Hnrnpk haploinsufficiency results in significant hematologic and malignant phenotypes, aberrant p21 and C/EBP expression, and Stat3 activation. These data provide evidence that hnRNP K acts as a haploinsufficient tumor suppressor and may underpin the pathogenesis of a subset of AML, which may open new frontiers in investigating potential hnRNP K-mediated tumor suppressor pathways.

Acknowledgments

We thank members of the Post laboratory for helpful discussions. We appreciate the technical advice and support from members of Dr. Micheal Andreeff's laboratory. We thank the AML tissue bank at MD Anderson Cancer Center for AML patient samples. The AML Tissue Bank, Veterinary and Pathology Core Facilities, DNA Sequencing Core, Flow Cytometry Core, and Genetically Engineered Mouse Facility were supported by an NCI Cancer Center Support Grant CA16672. This study has been supported by an NIH Career Development Award (P50 CA1000632-09), Leukemia Research Foundation Award, Center for Genetics and Genomics Award, and MDACC start-up funds to (S.M.P) and (J. M-L.) and (I. R.) have been supported by the Cancer Research Innovation Spain.

Authorship: M.G. designed the research, performed experiments, analyzed and interpreted data, and wrote the manuscript. H.J.L. performed experiments, analyzed and interpreted data, and critically revised the manuscript. X.Z. performed experiments, supported the experimental procedures, and critically revised the manuscript. L.R.P., M.M., and C.B.-R., performed and analyzed pathology studies. A.N. performed MEFs studies. A.M. performed karyotyping. T.M. supported experimental procedures. J.P.T. supported the development of the mouse model. I.R. collected samples and performed and analyzed gene expression assays. A.Q.C. supported the analysis and interpretation of data, and critically revised the manuscript. S.M.K. collected samples and clinical data. J.M.L collected samples and clinical data. S.M.P. designed and supervised research and experiments, performed experiments, analyzed and interpreted data, and wrote the manuscript.

Footnotes

Conflicts of interests: The authors declare no conflicts.

Disclosures: All authors reviewed and accepted the manuscript.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bibliography

  1. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 2013;368:2059–2074. doi: 10.1056/NEJMoa1301689. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bochkis IM, Rubins NE, White P, Furth EE, Friedman JR, Kaestner KH. Hepatocyte-specific ablation of Foxa2 alters bile acid homeostasis and results in endoplasmic reticulum stress. Nat Med. 2008;14:828–836. doi: 10.1038/nm.1853. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Burnett A, Wetzler M, Lowenberg B. Therapeutic advances in acute myeloid leukemia. J Clin Oncol. 2011;29:487–494. doi: 10.1200/JCO.2010.30.1820. [DOI] [PubMed] [Google Scholar]
  4. Dayyani F, Wang J, Yeh JR, Ahn EY, Tobey E, Zhang DE, Bernstein ID, Peterson RT, Sweetser DA. Loss of TLE1 and TLE4 from the del(9q) commonly deleted region in AML cooperates with AML1-ETO to affect myeloid cell proliferation and survival. Blood. 2008;111:4338–4347. doi: 10.1182/blood-2007-07-103291. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Enge M, Bao W, Hedstrom E, Jackson SP, Moumen A, Selivanova G. MDM2-dependent downregulation of p21 and hnRNP K provides a switch between apoptosis and growth arrest induced by pharmacologically activated p53. Cancer Cell. 2009;15:171–183. doi: 10.1016/j.ccr.2009.01.019. [DOI] [PubMed] [Google Scholar]
  6. Gabay C, Kushner I. Acute-phase proteins and other systemic responses to inflammation. N Engl J Med. 1999;340:448–454. doi: 10.1056/NEJM199902113400607. [DOI] [PubMed] [Google Scholar]
  7. Gao FH, Wu YL, Zhao M, Liu CX, Wang LS, Chen GQ. Protein kinase C-delta mediates down-regulation of heterogeneous nuclear ribonucleoprotein K protein: involvement in apoptosis induction. Exp Cell Res. 2009;315:3250–3258. doi: 10.1016/j.yexcr.2009.09.005. [DOI] [PubMed] [Google Scholar]
  8. Hsieh CC, Xiong W, Xie Q, Rabek JP, Scott SG, An MR, Reisner PD, Kuninger DT, Papaconstantinou J. Effects of age on the posttranscriptional regulation of CCAAT/enhancer binding protein alpha and CCAAT/enhancer binding protein beta isoform synthesis in control and LPS-treated livers. Mol Biol Cell. 1998;9:1479–1494. doi: 10.1091/mbc.9.6.1479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Huang X, Meng B, Iqbal J, Ding BB, Perry AM, Cao W, Smith LM, Bi C, Jiang C, Greiner TC, et al. Activation of the STAT3 signaling pathway is associated with poor survival in diffuse large B-cell lymphoma treated with R-CHOP. J Clin Oncol. 2013;31:4520–4528. doi: 10.1200/JCO.2012.45.6004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. James C, Ugo V, Le Couedic JP, Staerk J, Delhommeau F, Lacout C, Garcon L, Raslova H, Berger R, Bennaceur-Griscelli A, et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera. Nature. 2005;434:1144–1148. doi: 10.1038/nature03546. [DOI] [PubMed] [Google Scholar]
  11. Kirstetter P, Schuster MB, Bereshchenko O, Moore S, Dvinge H, Kurz E, Theilgaard-Monch K, Mansson R, Pedersen TA, Pabst T, et al. Modeling of C/EBPαlpha mutant acute myeloid leukemia reveals a common expression signature of committed myeloid leukemia-initiating cells. Cancer Cell. 2008;13:299–310. doi: 10.1016/j.ccr.2008.02.008. [DOI] [PubMed] [Google Scholar]
  12. Kronke J, Bullinger L, Teleanu V, Tschurtz F, Gaidzik VI, Kuhn MW, Rucker FG, Holzmann K, Paschka P, Kapp-Schworer S, et al. Clonal evolution in relapsed NPM1-mutated acute myeloid leukemia. Blood. 2013;122:100–108. doi: 10.1182/blood-2013-01-479188. [DOI] [PubMed] [Google Scholar]
  13. Liu Y, Szaro BG. hnRNP K post-transcriptionally co-regulates multiple cytoskeletal genes needed for axonogenesis. Development. 2011;138:3079–3090. doi: 10.1242/dev.066993. [DOI] [PubMed] [Google Scholar]
  14. Lynch M, Chen L, Ravitz MJ, Mehtani S, Korenblat K, Pazin MJ, Schmidt EV. hnRNP K binds a core polypyrimidine element in the eukaryotic translation initiation factor 4E (eIF4E) promoter, and its regulation of eIF4E contributes to neoplastic transformation. Mol Cell Biol. 2005;25:6436–6453. doi: 10.1128/MCB.25.15.6436-6453.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Matta A, Tripathi SC, DeSouza LV, Grigull J, Kaur J, Chauhan SS, Srivastava A, Thakar A, Shukla NK, Duggal R, et al. Heterogeneous ribonucleoprotein K is a marker of oral leukoplakia and correlates with poor prognosis of squamous cell carcinoma. International Journal of Cancer. 2009;125:1398–1406. doi: 10.1002/ijc.24517. [DOI] [PubMed] [Google Scholar]
  16. Mecucci C, Vermaelen K, Kulling G, Michaux JL, Noens L, Van Hove W, Tricot G, Louwagie A, Van den Berghe H. Interstitial 9q- deletions in hematologic malignancies. Cancer Genet Cytogenet. 1984;12:309–319. doi: 10.1016/0165-4608(84)90064-5. [DOI] [PubMed] [Google Scholar]
  17. Miau LH, Chang CJ, Shen BJ, Tsai WH, Lee SC. Identification of heterogeneous nuclear ribonucleoprotein K (hnRNP K) as a repressor of C/EBPbeta-mediated gene activation. J Biol Chem. 1998;273:10784–10791. doi: 10.1074/jbc.273.17.10784. [DOI] [PubMed] [Google Scholar]
  18. Mikula M, Bomsztyk K, Goryca K, Chojnowski K, Ostrowski J. Heterogeneous nuclear ribonucleoprotein (HnRNP) K genome-wide binding survey reveals its role in regulating 3′-end RNA processing and transcription termination at the early growth response 1 (EGR1) gene through XRN2 exonuclease. J Biol Chem. 2013;288:24788–24798. doi: 10.1074/jbc.M113.496679. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Moumen A, Masterson P, O'Connor MJ, Jackson SP. hnRNP K: an HDM2 target and transcriptional coactivator of p53 in response to DNA damage. Cell. 2005;123:1065–1078. doi: 10.1016/j.cell.2005.09.032. [DOI] [PubMed] [Google Scholar]
  20. Mrozek K, Marcucci G, Paschka P, Whitman SP, Bloomfield CD. Clinical relevance of mutations and gene-expression changes in adult acute myeloid leukemia with normal cytogenetics: are we ready for a prognostically prioritized molecular classification? Blood. 2007;109:431–448. doi: 10.1182/blood-2006-06-001149. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Nerlov C. C/EBPαlpha mutations in acute myeloid leukaemias. Nat Rev Cancer. 2004;4:394–400. doi: 10.1038/nrc1363. [DOI] [PubMed] [Google Scholar]
  22. Notari M, Neviani P, Santhanam R, Blaser BW, Chang JS, Galietta A, Willis AE, Roy DC, Caligiuri MA, Marcucci G, Perrotti D. A MAPK/HNRPK pathway controls BCR/ABL oncogenic potential by regulating MYC mRNA translation. Blood. 2006;107:2507–2516. doi: 10.1182/blood-2005-09-3732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Ostareck-Lederer A, Ostareck DH. Precision mechanics with multifunctional tools: how hnRNP K and hnRNPs E1/E2 contribute to post-transcriptional control of gene expression in hematopoiesis. Curr Protein Pept Sci. 2012;13:391–400. doi: 10.2174/138920312801619484. [DOI] [PubMed] [Google Scholar]
  24. Ostareck-Lederer A, Ostareck DH, Cans C, Neubauer G, Bomsztyk K, Superti-Furga G, Hentze MW. c-Src-Mediated Phosphorylation of hnRNP K Drives Translational Activation of Specifically Silenced mRNAs. Molecular and Cellular Biology. 2002;22:4535–4543. doi: 10.1128/MCB.22.13.4535-4543.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Ostareck DH, Ostareck-Lederer A, Wilm M, Thiele BJ, Mann M, Hentze MW. mRNA silencing in erythroid differentiation: hnRNP K and hnRNP E1 regulate 15-lipoxygenase translation from the 3′ end. Cell. 1997;89:597–606. doi: 10.1016/s0092-8674(00)80241-x. [DOI] [PubMed] [Google Scholar]
  26. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001;29:e45. doi: 10.1093/nar/29.9.e45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Post SM, Quintas-Cardama A, Pant V, Iwakuma T, Hamir A, Jackson JG, Maccio DR, Bond GL, Johnson DG, Levine AJ, Lozano G. A high-frequency regulatory polymorphism in the p53 pathway accelerates tumor development. Cancer Cell. 2010;18:220–230. doi: 10.1016/j.ccr.2010.07.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Quintas-Cardama A, Verstovsek S. Molecular pathways: Jak/STAT pathway: mutations, inhibitors, and resistance. Clin Cancer Res. 2013;19:1933–1940. doi: 10.1158/1078-0432.CCR-12-0284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Renneville A, Roumier C, Biggio V, Nibourel O, Boissel N, Fenaux P, Preudhomme C. Cooperating gene mutations in acute myeloid leukemia: a review of the literature. Leukemia. 2008;22:915–931. doi: 10.1038/leu.2008.19. [DOI] [PubMed] [Google Scholar]
  30. Roychoudhury P, Chaudhuri K. Evidence for heterogeneous nuclear ribonucleoprotein K overexpression in oral squamous cell carcinoma. Br J Cancer. 2007;97:574–575. doi: 10.1038/sj.bjc.6603911. author reply 576. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Schuster MB, Frank AK, Bagger FO, Rapin N, Vikesaa J, Porse BT. Lack of the p42 form of C/EBPalpha leads to spontaneous immortalization and lineage infidelity of committed myeloid progenitors. Exp Hematol. 2013;41:882–893. e816. doi: 10.1016/j.exphem.2013.06.003. [DOI] [PubMed] [Google Scholar]
  32. Seita J, Weissman IL. Hematopoietic stem cell: self-renewal versus differentiation. Wiley Interdiscip Rev Syst Biol Med. 2010;2:640–653. doi: 10.1002/wsbm.86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sweetser DA, Peniket AJ, Haaland C, Blomberg AA, Zhang Y, Zaidi ST, Dayyani F, Zhao Z, Heerema NA, Boultwood J, et al. Delineation of the minimal commonly deleted segment and identification of candidate tumor-suppressor genes in del(9q) acute myeloid leukemia. Genes Chromosomes Cancer. 2005;44:279–291. doi: 10.1002/gcc.20236. [DOI] [PubMed] [Google Scholar]
  34. van Domselaar R, Quadir R, van der Made AM, Broekhuizen R, Bovenschen N. All human granzymes target hnRNP K that is essential for tumor cell viability. J Biol Chem. 2012;287:22854–22864. doi: 10.1074/jbc.M112.365692. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Wang N, Zhang P, Guo X, Zhou Z, Sha J. Hnrnpk, a protein differentially expressed in immature rat ovarian development, is required for normal primordial follicle assembly and development. Endocrinology. 2011;152:1024–1035. doi: 10.1210/en.2010-0797. [DOI] [PubMed] [Google Scholar]
  36. Zhou R, Shanas R, Nelson MA, Bhattacharyya A, Shi J. Increased expression of the heterogeneous nuclear ribonucleoprotein K in pancreatic cancer and its association with the mutant p53. International Journal of Cancer. 2010;126:395–404. doi: 10.1002/ijc.24744. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS734287-supplement-1.docx (12.8KB, docx)
2
NIHMS734287-supplement-2.pdf (8.7MB, pdf)

RESOURCES