The transcription factors signal transducer and activator of transcription 5A (STAT5A) and STAT5B negatively regulate cell proliferation through the activation of cyclin-dependent kinase inhibitor 2b (Cdkn2b) and Cdkn1a expression

Authors

  • Ji Hoon Yu,

    1. Laboratory of Genetics and Physiology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD
    Search for more papers by this author
  • Bing-Mei Zhu,

    1. Laboratory of Genetics and Physiology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD
    Search for more papers by this author
  • Mark Wickre,

    1. Laboratory of Genetics and Physiology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD
    Search for more papers by this author
  • Gregory Riedlinger,

    1. Laboratory of Genetics and Physiology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD
    Search for more papers by this author
  • Weiping Chen,

    1. Genomics Core Laboratory, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD
    Search for more papers by this author
  • Atsushi Hosui,

    1. Department of Gastroenterology and Hepatology, Osaka University Graduate School of Medicine, Suita, Osaka, Japan
    Search for more papers by this author
  • Gertraud W. Robinson,

    1. Laboratory of Genetics and Physiology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD
    Search for more papers by this author
  • Lothar Hennighausen

    Corresponding author
    1. Laboratory of Genetics and Physiology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD
    • Laboratory of Genetics and Physiology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, 8 Center Drive, Room 101, Bethesda, MD 20892-0822
    Search for more papers by this author
    • fax: 301-480-7312


Abstract

Although the cytokine-inducible transcription factor signal transducer and activator of transcription 5 (STAT5) promotes proliferation of a wide range of cell types, there are cell-specific and context-specific cases in which loss of STAT5 results in enhanced cell proliferation. Here, we report that loss of STAT5 from mouse embryonic fibroblasts (MEFs) leads to enhanced proliferation, which was linked to reduced levels of the cell cycle inhibitors p15INK4B and p21CIP1. We further demonstrate that growth hormone, through the transcription factor STAT5, enhances expression of the Cdkn2b (cyclin-dependent kinase inhibitor 2B) gene and that STAT5A binds to interferon-gamma–activated sequence sites within the promoter. We recently demonstrated that ablation of STAT5 from liver results in hepatocellular carcinoma upon CCl4 treatment. We now establish that STAT5, like in MEFs, activates expression of the Cdkn2b gene in liver tissue. Loss of STAT5 led to diminished p15INK4B and increased hepatocyte proliferation. Conclusion: This study for the first time demonstrates that cytokines, through STAT5, induce the expression of a key cell cycle inhibitor. These experiments therefore shed mechanistic light on the context-specific role of STAT5 as tumor suppressor. (HEPATOLOGY 2010;52:1808-1818)

Signal transducer and activator of transcription 5A (STAT5A) and STAT5B are latent transcription factors that are induced by a plethora of cytokines, including growth hormone (GH), prolactin, and several interleukins.1 Gene knockout experiments have revealed context-specific functions of STAT5, that range from the specification, proliferation and survival to differentiation of normal cells.1 Aberrant activation of STAT5 has been detected in a majority of leukemias and many solid tumors, suggesting a critical role in the initiation/progression of these tumors. Notably, the deletion of STAT5 from breakpoint cluster region/Abelson murine leukemia (BCR-ABL)–induced leukemic cells results in their regression,2 providing evidence that STAT5 is a critical transcription factor in the progression of leukemias. Moreover, transgenic experiments using constitutively active STAT5 mutants have supported this concept.3-5 In addition to its oncogenic role, a context-specific tumor suppressor function has been associated with STAT5, such as inhibiting expression of nucleophosmin/anaplastic lymphoma kinase (NPM1-ALK)6 and suppressing STAT3 and transforming growth factor-β activity in liver.71

Table 1. Expression of a Selected Set of Genes in STAT5+/+ and STAT5−/− MEFs
Gene SymbolGene TitleStat5+/+ MEFsStat5−/− MEFsFold Change (WT/KO)
  1. Expressed genes were identified by Affymetrix Mouse Genome 430 version 2.0 array GeneChips Microarray-based gene expression analysis. Microarray analysis was performed using three samples from each group. Gene expression ratio showing at least a two-fold change in STAT5+/+ MEFs relative to STAT5−/− MEFs. Values indicated are the difference between the mean value of three samples.

STAT5aSignal transducer and activator of transcription 5A173.513.612.7
STAT5bSignal transducer and activator of transcription 5B398.111.036.1
Socs2Suppressor of cytokine signaling 2563.6265.62.1
Cdkn2bCyclin-dependent kinase inhibitor 2B (p15INK4B)4876.22120.72.3
Cdkn1aCyclin-dependent kinase inhibitor 1A (P21CIP1)9942.6438.922.7

Mice from which the Stat5 locus has been deleted specifically in liver tissue displayed altered metabolic pathways and developed fatty liver (nonalcoholic steatohepatitis).8, 9 Treatment of these mice with CCl4 led to liver fibrosis and tumors, suggesting that STAT5 is a tumor suppressor in the context of hepatocytes.7 Aberrant activation of the transforming growth factor-β and STAT3 pathways in these mice appears to contribute to the CCl4-induced fibrosis and hepatocellular carcinoma (HCC).7

Evidence is emerging that cytokines, through STAT5, regulate the cell cycle by facilitating progression through the G1 phase.10-13 This opens the possibility that STAT5 directly controls cell cycle genes and thereby can function either as an oncogene or a tumor suppressor. We have now addressed this question by studying mouse embryonic fibroblasts (MEFs) and liver tissue devoid of STAT5. In particular, STAT5-null MEFs were chosen because they exhibit an increased proliferation rate, suggesting that in this context STAT5 is a cell cycle suppressor.

Abbreviations

Cdkn, cyclin-dependent kinase inhibitor; ChIP, chromatin immunoprecipitation; DAPI, 4′,6-diamidino-2-phenylindole; DMEM, Dulbecco's modified Eagle medium; EDTA, ethylene diamine tetraacetic acid; FACS, fluorescence-activated cell sorting; GAS, interferon-gamma–activated sequence; GH, growth hormone; HCC, hepatocellular carcinoma; HSC, hematopoietic stem cell; i.p., intraperitoneal; MEF, mouse embryonic fibroblast; PCNA, proliferating cell nuclear antigen; PCR, polymerase chain reaction; SD, standard deviation; SEM, standard error of the mean; SOCS2, suppressor of cytokine signaling 2; STAT, signal transducer and activator of transcription.

Materials and Methods

Mice Breeding.

We generated Stat5f/f;Alb-Cre mice by breeding Stat5f/f mice with albumin/cyclic adenosine monophosphate response element (Alb-Cre) transgenic mice.14Stat5f/f and Alb-Cre transgenic mice were on a mixed background. Only 10- to 18-week-old male mice were used in the experiments unless otherwise indicated. We treated the animals humanely and performed procedures according to the protocol approved by the Animal Use and Care Committee at the National Institute of Diabetes and Digestive and Kidney Diseases.

Liver Fibrosis Induced by CCl4.

Hepatic fibrosis in mice was induced by intraperitoneal (i.p.) injection with 2 mL/kg body weight of 10% CCl4 (Sigma, St. Louis, MO) dissolved in olive oil (Sigma, St. Louis, MO), three times a week for 12 weeks. For GH stimulation, mice were i.p. injected with 2 μg/g body weight GH. They were sacrificed 4 hours after injection and livers were harvested for analyses.

Isolation of Primary MEF Cells.

Primary MEFs were isolated from embryonic day 14.5 (E14.5) Stat5+/+ and Stat5−/− embryos by first mincing the embryos, then digesting in 0.05% trypsin/0.02% ethylene diamine tetraacetic acid (EDTA) for 30 minutes at 37°C, pelleting the tissue, and resuspending in growth medium consisting of Dulbecco's modified Eagle medium (DMEM) with 10% fetal bovine serum. MEFs were maintained in high-glucose DMEM supplemented with 15% fetal bovine serum, 50 μg/mL streptomycin sulfate, 50 U/mL penicillin G sodium, β-mercaptoethanol, and nonessential amino acid in an atmosphere of 5% CO2 at 37°C.

Retrovirus Infection.

The retroviral-expression vector carrying a wild-type Stat5A15 gene was based on an MSCV-IRES-GFP (murine stem cell virus–internal ribosome entry site–green fluorescent protein) backbone (gift from Richard Moriggl, Ludwig-Boltzmann Institute, Vienna, Austria). The 293T cells were transfected with the plasmid using FuGENE (Roche, Indianapolis, IN). Supernatants were collected for 48-72 hours after transfection and passed through a 0.45-μm filter before freezing at −80°C. For the infection, 106Stat5−/− MEFs were seeded on a 10-cm culture dish and infected the next day with retrovirus in the presence of 8 μg/mL polybrene. After infection, nonfluorescent cells and GFP-expressing cells were isolated using the FACS Vantage (Becton Dickinson, San Jose, CA) fluorescence-activated cell sorter (FACS) and sorted directly into phosphate-buffered saline (PBS). Sorted MEFs were maintained in DMEM supplemented as described above.

Affymetrix Microarray Analysis.

Primary MEFs derived from Stat5+/+ or Stat5−/− embryos were cultured to passage 13. After starvation for 5 hours, MEFs were stimulated with GH (1 μg/mL) for 2 hours. Unstimulated samples were used as control. Total RNAs were prepared by using a RNeasy Plus Mini Kit (Qiagen, Valencia, CA). RNA quality was verified using an Agilent Bioanalyzer (Agilent Technologies, Palo Alto, CA). Microarray analyses were performed using Affymetrix Mouse Genome 430 version 2.0 array GeneChips (Affymetrix, Santa Clara, CA). Microarray signals were analyzed using the Affymetrix RMA algorithm. Up-regulated and down-regulated genes were selected based on P < 0.05 and fold-changes of more than 1.5 or less than 1.5 as assessed by analysis of variance using Partek Pro software (Partek, St. Louis, MO). The microarray analyses were performed with three independent biological sample sets. Microarray data have been deposited in Gene Expression Omnibus (GEO; accession number: GSE21861).

Analysis of Cell Proliferation.

The proliferation of cells was determined by a Trypan blue dye exclusion assay. In brief, primary Stat5+/+ and Stat5−/− MEFs (1 × 105 cells/well) were seeded on tissue culture plates and cultured in high-glucose DMEM. The number of viable cells was counted after 2, 4, and 6 days. MEFs were harvested with trypsin-EDTA. The cell suspension was loaded onto a hemocytometer (1:1) with the dye Trypan blue, which is taken up by dead cells. Both viable and dead cells were counted, from which both the percentage of dead cells and total cell number were calculated.

Cell-Cycle Analysis.

Primary Stat5+/+ and Stat5−/− MEFs were washed twice with PBS and fixed for 30 minutes at −20°C in 70% ethanol. Total DNA was stained with propidium iodide (5 μg/mL in PBS containing 50 μg/mL ribonuclease A). Cell cycle distribution was determined by FACS analysis using a FACSCalibur (Becton Dickinson, San Jose, CA). Data are presented as a percentage of viable cells remaining in the respective cell cycle phase.

Antibodies, Immunoblotting, and Immunostaining.

In brief, primary MEFs were lysed by adding NuPAGE LDS sample buffer (Invitrogen, Carlsbad, CA). Cell lysates were heat-denatured for 10 minutes at 70°C and loaded on a NuPAGE 10% Bis-Tris polyacrylamide gel. After electrophoresis in NuPAGE sodium dodecyl sulfate running buffer using the Xcell SureLock Mini-Cell, proteins were transferred to a polyvinylidene fluoride membrane according to the manufacturer's instructions (Invitrogen, Carlsbad, CA). The rabbit polyclonal anti-STAT5 (N-20), anti-STAT5 (C-17), anti-p15 (K-18), anti-p21 (C-19), anti–cyclin D1 (HD-11), anti–cyclin A (H-432), anti–cyclin B1 (H-433), anti-Cdk4 (C-22), and anti–β-actin antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) were used for probing western blots. Immunohistochemistry was performed using standard procedures. In short, liver tissues were removed and fixed in 10% neutral buffered formalin and embedded in paraffin wax. Sections (5 μm) were prepared for hematoxylin and eosin (H&E) staining and immunofluorescence analyses. After deparaffinization, antigen unmasking was performed in a Decloaking chamber (Biocare Medical, San Diego, CA) using BORG Decloaker Solution (Biocare Medical, San Diego, CA) for 5 minutes at 125°C. The sections were blocked for 30 minutes in Tris-buffered saline with Tween-20 containing 3% goat serum. Primary antibodies used in this study included rabbit anti-STAT5 (N-20), rabbit anti–phospho-STAT5 (Tyr694) (Cell Signaling Technology, Beverly, MA), rabbit anti-p15 (K-18) (Santa Cruz Biotechnology, Santa Cruz, CA), rabbit anti–phospho-histone H3 (Ser-10) (Upstate Biotechnology, Lake Placid, NY), in addition to mouse anti-PCNA (DAKO Cytomation, Carpinteria, CA) and mouse anti–β-catenin (BD Transduction Laboratories, San Jose, CA). For double-labeling immunofluorescence analyses, sections exposed to a pair of primary antibodies were incubated in a 1:400 dilution of goat anti-rabbit immunoglobulin G (IgG) conjugated with a red fluorophore (AlexaFluor 594; Molecular Probes, Eugene, OR) and goat anti-mouse IgG conjugated with a green fluorophore (AlexaFluor 488; Molecular Probes, Eugene, OR) for 30 minutes at room temperature. Images were obtained with a Retiga Exi camera on a Olympus BX51 microscope (Olympus America, Center Valley, PA) using Image-Pro 5.1 software. For quantitation, three images taken with the 40× objective were counted per mouse. Three mice from each experimental group were evaluated.

Immunoprecipitation Analysis.

Cellular proteins were extracted from primary Stat5+/+ and Stat5−/− MEFs. Protein fractions were incubated overnight with an anti-CDK4 antibody and protein A–sepharose beads at 4°C in radioimmunoprecipitation assay buffer containing 25 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 0.25% Nonidet P-40, and 0.5% sodium deoxycholate. The protein A–sepharose antibody–antigen complex was collected and washed three times with ice-cold radioimmunoprecipitation assay buffer. The final pellet was resuspended with sodium dodecyl sulfate sample buffer and boiled for 5 minutes. This preparation was subjected to western blot analysis with the anti-p15 or anti-p21 antibody (Santa Cruz Biotechnology, Santa Cruz, CA).

Chromatin Immunoprecipitation Assay.

Chromatin immunoprecipitation (ChIP) assay was performed as described.16 In brief, after starvation for 5 hours, primary Stat5+/+ MEFs were stimulated with GH for 45 minutes. Unstimulated samples were used as controls. MEFs were cross-linked in 1.5% formaldehyde for 15 minutes at 37°C. Cells were then harvested and sonicated using the Misonix Sonicator 3000 (Misonix, Farmingdale, NY). Immunoprecipitation was carried out in Tris-EDTA buffer containing protease inhibitors (Sigma, St. Louis, MO). Chromatin was incubated with protein A Dynabeads (Invitrogen, Carlsbad, CA), which were preincubated with STAT5A or IgG antibody (R&D Systems, Minneapolis, MN). Immunoprecipitated DNA was eluted and amplified by real-time polymerase chain reaction (PCR) using a 7900 HT fast real-time PCR system (Applied Biosystems, Foster City, CA) and analyzed using SDS2.3 Software (Applied Biosystems, Foster City, CA). Sequence-specific primers used for amplification of the putative STAT5 binding sites (interferon-gamma–activated sequence [GAS] sites) within the Socs2 (suppressor of cytokine signaling 2) and Cdkn2b (cyclin-dependent kinase inhibitor 2b) genes were as following: For the Socs2 GAS, forward primer 5′-GGAGGGCGGAGT CGCAGGC-3′, reverse primer 5′-GACTTGGCAAGA GTTAACCGTC-3′; the primer sets for Cdkn2b gene were: GAS1, forward 5′-GTTTTGCCGTGATGTCC TTG-3′, reverse 5′-ATCGCACTGCTTCGTGTAAC-3′; GAS2, forward 5′-GACAGGCATTGTCCAAGACA-3′, reverse 5′-GTGCCACATTCTCCCACTTT-3′.

RNA Isolation and Quantitative Real-Time PCR Analysis.

Total RNA was isolated from primary Stat5+/+ MEFs, Stat5−/− MEFs, and liver tissues using Trizol reagent (Invitrogen, Carlsbad, CA). One μg amounts of RNA were reverse transcribed (complementary DNA reverse transcription kit; Applied Biosystems, Foster City, CA). Real-time quantification of messenger RNA (mRNA) transcript levels was performed using the SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. Real-time PCR was carried out using an ABI-Prism 7900HT (Applied Biosystems, Foster City, CA). Individual PCRs were performed in triplicate on samples using glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a housekeeping gene. The primers used were: Cdkn2b, forward 5′-CCCTGCCA CCCTTACCAGA-3′, reverse 5′-CAGATACCTCGCA ATGTCACG-3′, yielding a 169–base pair (bp) PCR product; Cdkn1a, forward 5′-GTGGCCTTGTCGCT GTCTT-3′, reverse 5′-GCGCTTGGAGTGATAGAAA TCTG-3′, yielding a 126-bp PCR product; GAPDH, forward 5′-AACGACCCCTTCATTGAC-3′, reverse 5′ TCCACGACATACTCAGCAC-3′, yielding a 191-bp PCR product.

Statistics.

All statistical analyses were performed using the Student t test (two-tailed, unpaired). A P value of 0.05 or less was considered significant.

Results

Increased Proliferation Rate of STAT5-Deficient MEFs.

Fibroblasts were isolated from control and Stat5−/− fetuses and their growth curves were established (Fig. 1). Although Stat5−/− MEFs from passage 3 displayed a small but significantly increased proliferation rate when compared to matched controls, the difference was more profound with cells from passage 8 (Fig. 1A) and later passages (Supporting Information Fig. 1). Notably, after 6 days in culture, the number of passage 8 Stat5−/− MEFs had increased five-fold, whereas control MEFs had doubled in number. DNA content profiles demonstrated that the percentage of Stat5−/− MEFs in G0/G1 and S phases was decreased compared to controls, but percentages in the G2/M phase were increased (Fig. 1B).

Figure 1.

Cell proliferation in primary Stat5+/+ and Stat5−/− MEFs. (A) Cell growth curve of primary MEFs. Cell viability was determined in Stat5+/+ and Stat5−/− MEFs at the third (p3) and eighth (p8) passage. All values represent means ± standard deviation (SD) from three independent experiments. (B) Cell cycle analysis of control and Stat5−/− MEFs. (C,D) mRNA expression of Cdkn2b and Cdkn1a in primary MEFs. Levels of Cdkn2b and Cdkn1a mRNA were analyzed by quantitative real-time PCR in Stat5+/+ and Stat5−/− MEFs at the third (p3) and eighth (p8) passage. The values were expressed as means ± SD. *P < 0.05; **P < 0.01 compared with corresponding controls.

To further explore the molecular basis of this accelerated growth, we performed global gene expression analyses of control and Stat5−/− MEFs (GEO accession number: GSE21861). In addition to the reduced expression of bona fide STAT5 target genes, such as Socs2, we observed a two-fold and 22-fold reduced expression of Cdkn2b (p15INK4B) and Cdkn1a (p21CIP1), respectively (Table 1) (Janus kinase [JAK]-STAT Prospector at http://jak-stat.nih.gov). To establish a link between reduced expression of Cdkn2b and Cdkn1a in MEFs and their growth behavior, we analyzed p15INK4B and p21CIP1 mRNA and protein levels in MEFs at passages 3 and 8. Although Cdkn2b mRNA levels were reduced in the absence of STAT5 by approximately 80% and 60% at passages 3 and 8, respectively (Fig. 1C), Cdkn1a mRNA levels were reduced by 75% and 50% (Fig. 1D), respectively. Western blot analyses confirmed the sharp reduction of p15INK4B and to a lesser extent of p21CIP1 (Fig. 2A). CDK4 levels were slightly increased in the Stat5−/− MEFs. Notably, levels of cyclin D1 and cyclin A, and in particular cyclin B1, were elevated in the absence of STAT5 (Fig. 2B). Absence of a STAT5 signal corroborated the genotype of the cells used. The p15INK4B forms an inhibitory complex with CDK4, and immunoprecipitation analyses demonstrated that this complex was greatly reduced in the absence of STAT5 (Fig. 2C), which is in agreement with the reduced p15INK4B levels (Fig. 2A).

Figure 2.

Expression of p15INK4B, p21CIP1, cyclin D1, cyclin A, and cyclin B1 in primary MEFs. (A,B) Level of p15INK4B, p21CIP1, cyclin D1, cyclin A, and cyclin B1 in Stat5+/+ and Stat5−/− MEFs at third (p3) and eighth (p8) passages. Expression of p15INK4B, p21CIP1, cyclin D1, cyclin A, and cyclin B1 was determined by western blotting. (C) The p15INK4B-CDK4 complex was analyzed by coimmunoprecipitation in Stat5+/+ and Stat5−/− MEFs. The p15INK4B was detected in anti-CDK4 immunoprecipitation (IP) derived from protein fraction of MEFs by immunoblotting.

Expression of Cdkn2b Is Under STAT5 Control In Vitro.

To determine whether expression of Cdkn2b is under direct GH/STAT5 regulation, control and Stat5−/− MEFs were stimulated with GH for 4 hours and RNA was analyzed. Although Cdkn2b expression was induced two-fold in control MEFs, no induction was observed in experimental cells, which also exhibited lower baseline levels (Fig. 3A). These experiments demonstrate the STAT5 dependency of Cdkn2b expression. Moreover, we demonstrated that Cdkn1a expression was also under STAT5-dependent GH control (Fig. 3B) supporting and extending earlier findings.17

Figure 3.

GH-induced expression and STAT5 binding to conserved GAS sites in the Cdkn2b (p15INK4B) gene promoter. (A,B) mRNA expression of Cdkn2b and Cdkn1a in primary MEFs treated with GH. MEFs were treated with GH for 4 hours. Levels of Cdkn2b and Cdkn1a mRNA were analyzed by quantitative real-time PCR in Stat5+/+ and Stat5−/− MEFs. Values are shown as means ± standard error of the mean (SEM). (C) Schematic of the Cdkn2b (p15INK4B) gene. Vertical boxes indicate translated exons. Locations of the conserved GAS sequences are indicated. Sites are shown aligned with sequences from other species. (D) ChIP analysis of STAT5 binding to the putative GAS sites. Stat5+/+ MEFs were treated with GH. Binding to GAS sites was analyzed by quantitative real-time PCR. DNA was amplified from STAT5-precipitated complexes using specific primers for known (Socs2) and suspected (Cdkn2b) GAS regions. All values represent means ± SEM from three independent experiments performed in triplicate. *P < 0.05; compared with corresponding controls.

To determine whether STAT5 directly controls the Cdkn2b gene, we scanned the promoter region for GAS motifs. Two conserved GAS motifs were identified at positions −2584, and −1976 (Fig. 3C). ChIP analyses in MEFs confirmed GH-induced STAT5 binding to both sites (Fig. 3D). Binding to the Socs2 gene promoter served as a positive control. STAT5 binding was also detected at the Cdkn1a gene promoter (data not shown).

Restored Growth and Cdkn2b Expression Upon Reestablishing STAT5 Expression.

Impaired proliferation of Stat5−/− MEFs coincided with reduced expression of the Cdkn1a and Cdkn2b genes suggesting a causal relationship. To establish the link between STAT5A, Cdkn2b expression and cell proliferation kinetics, we reintroduced STAT5A into Stat5−/− MEFs. Stat5−/− MEFs that were complemented with retrovirally expressed STAT5A displayed reduced cell proliferation as compared to parent Stat5−/− MEFs (Fig. 4A) and exhibited growth kinetics reminiscent of wild-type MEFs. Moreover, the GH-induced expression of Cdkn1a and Cdkn2b was reinstated (Fig. 4B,C) as was Socs2 expression (Fig. 4D). This was also reflected on the protein level (Fig. 4E). Although p15INK4B and p21CIP1 levels did not change in Stat5−/− MEFs carrying an empty control retrovirus, ectopic expression of STAT5A resulted in higher levels (Fig. 4E).

Figure 4.

Stat5−/− MEFs complemented with transgenic STAT5A regain suppressed cell proliferation. (A) Growth curves of Stat5−/− MEFs infected with an empty (control) retrovirus or with a STAT5A-expressing retrovirus. Cell viability was determined at passage 9. Note, Stat5−/− MEFs in the absence or presence of the control vector display identical proliferation curves. Introduction of STAT5A results in reduced cell growth. (B,C,D) mRNA expression of Cdkn2b, Cdkn1a, and Socs2 in primary Stat5−/− MEFs carrying either the control (Con) or STAT5A-expressing retrovirus. Levels of Cdkn2b, Cdkn1a, and Socs2 were analyzed by quantitative real-time PCR in MEFs at passage 9. The values were expressed as the means ± SD. (E) p15INK4B and p21CIP1 levels in Stat5−/− MEFs carrying either the control (Con) or STAT5A-expressing retrovirus at passage 9. Expression of p15INK4B and p21CIP1 was determined by western blotting. All values represent means ± SD from three independent experiments. The values were expressed as the means ± SD. *P < 0.05; **P < 0.01 compared with corresponding controls.

STAT5-Dependent Regulation of Cdkn1a and Cdkn2b in Liver Tissue.

We discovered earlier that hepatocyte-specific loss of STAT5 led to hepatosteatosis and HCC upon CCl4 treatment.7 H&E staining revealed that hepatosteatosis with accumulation of fat droplets progressed in liver-specific STAT5-null mice (Stat5f/f;Alb-Cre) upon CCl4 treatment (Supporting Information Fig. 2). To test whether the STAT5-regulated expression of Cdkn1a and Cdkn2b was also occurring in liver tissue, we analyzed the expression of Cdkn1a and Cdkn2b mRNA and the corresponding protein levels in control and liver-specific STAT5-null mice. In mice at 2 months of age, p15INK4B levels were greatly reduced in experimental liver tissue (Fig. 5A), supporting the notion that STAT5 controls expression of this cell cycle regulator. Levels of p21CIP1 were marginally reduced. Similar to the case in MEFs, cyclin D1 levels were elevated in STAT5-null livers (Fig. 5A). Further evidence for this was derived from in vivo experiments in which control and experimental mice were injected with GH followed by mRNA analyses. Although GH treatment of control mice induced Cdkn2b mRNA levels, no such increase was observed in experimental mice (Fig. 5B). The p21CIP1 levels were only slightly reduced in experimental liver tissue. Immunoprecipitation analyses demonstrated an absence of CDK4-p15INK4B complexes whereas CDK4-p21CIP1 complexes were reduced to a lesser extent (Fig. 5C).

Figure 5.

Expression of p15INK4B and p21CIP1 in liver tissue from Stat5f/f and Stat5f/f;Alb-Cre mice. (A) Protein expression of p15INK4B, p21CIP1, and cyclin D1 was determined by western blotting. (B) mRNA expression of Cdkn2b in Stat5f/f and Stat5f/f;Alb-Cre mice injected with GH. Mice were treated with GH for 4 hours. Expression of Cdkn2b was analyzed by quantitative real-time PCR in liver tissues from Stat5f/f and Stat5f/f;Alb-Cre mice. All values represent means ± SEM. *P < 0.05; compared with corresponding controls. (C) p15INK4B-CDK4 and p21CIP1-CDK4 complex were analyzed by coimmunoprecipitation in Stat5f/f and Stat5f/f;Alb-Cre mice injected with GH. p15INK4B and p21CIP1 were detected in anti-CDK4 immunoprecipitates (IP) derived from liver tissues of Stat5f/f and Stat5f/f;Alb-Cre mice by immunoblotting.

Immunohistochemistry was used as an independent means to corroborate the dependence of p15INK4B levels on STAT5. Nuclear p15INK4B was observed in liver tissue of control mice and the levels increased after GH treatment (Fig. 6A,C). In contrast, low expression was observed in the liver-specific STAT5-null mice before and after GH injection (Fig. 6A,C). Strong nuclear STAT5 staining was observed after GH only in the control but not the experimental mice (Fig. 6B). Moreover, the number of nuclei positive for p15INK4B decreased in liver-specific STAT5-null mice (Fig. 6C). Nuclear p15INK4B was observed in liver tissue of control mice in the absence and presence of CCl4 (Fig. 7). In contrast, low expression was observed in the liver-specific STAT5-null mice (Fig. 7). There were no differences in p15INK4B levels in the absence or presence of CCl4 (Fig. 7).

Figure 6.

Immunostaining of p15INK4B and STAT5 in Stat5f/f and Stat5f/f;Alb-Cre mice injected with GH. (A) Livers from Stat5f/f and Stat5f/f;Alb-Cre mice were harvested after GH injection and analyzed for p15INK4B expression using immunofluorescence staining with anti-p15INK4B (red), anti–β-catenin (green) antibodies and 4′,6-diamidino-2-phenylindole (DAPI; blue). (B) Livers from Stat5f/f and Stat5f/f;Alb-Cre mice were harvested after GH injection and analyzed for STAT5 expression using immunofluorescence staining with anti-STAT5 (red), anti–β-catenin (green) antibodies, and DAPI (blue). (C) Expression of p15INK4B was analyzed by immunostaining. The p15INK4B-positive cells in liver tissues were counted from three different areas and the number per area is shown. All values represent means ± SD from 3-4 mice in each group, **P < 0.01; compared with corresponding controls.

Figure 7.

Immunostaining of p15INK4B in Stat5f/f and Stat5f/f;Alb-Cre mice injected with CCl4. (A) Livers from Stat5f/f and Stat5f/f;Alb-Cre mice were harvested after CCl4 injection and analyzed for p15INK4b expression using immunofluorescence staining with anti-p15INK4b (red), anti-β-catenin (green) antibodies, and DAPI (blue). (B) p15INK4B-positive cells in liver tissues were counted from three different areas and the number per area is shown. All values represent means ± SD from 3-4 mice in each group, **P < 0.01; compared with corresponding controls.

CCl4 Treatment Results in Increased Cell Proliferation.

We recently demonstrated that liver-specific STAT5-null mice develop hepatosteatosis and HCC upon treatment with CCl4.7 A second cohort of mice confirmed that approximately 25% of the CCl4-treated STAT5-null mice developed HCC, whereas treated control mice remained free of tumors (Supporting Information Fig. 3). To establish whether loss of STAT5 and the sharply reduced levels of p15INK4B correlated with increased cell proliferation, we stained tissue sections for phosphorylated histone H3 as a measure of cell proliferation (Fig. 8A). The number of phospho-histone H3–positive nuclei in experimental mice was three times higher than in control mice (Fig. 8B). Moreover, CCl4 treatment resulted in a further increase only in the absence of STAT5 (Fig. 8B). In support of this, the number of nuclei positive for proliferating cell nuclear antigen (PCNA) was increased in liver-specific STAT5-null mice (Fig. 9).

Figure 8.

Immunostaining of phospho-histone H3 in Stat5f/f and Stat5f/f;Alb-Cre mice injected with CCl4. (A) Livers from Stat5f/f and Stat5f/f;Alb-Cre mice were harvested after 12 weeks of CCl4 injection and analyzed for phospho-histone H3 using immunofluorescence staining with anti–phospho-histone H3 (red), anti–β-catenin (green) antibodies, and DAPI (blue). (B) Phospho-histone H3–positive cells in liver tissues were counted from three different areas and the number/area is shown. All values represent means ± SD from 3-4 mice in each group, *P < 0.05; compared with corresponding controls.

Figure 9.

Immunostaining of PCNA in Stat5f/f and Stat5f/f;Alb-Cre mice injected with CCl4. (A) Livers from Stat5f/f and Stat5f/f;Alb-Cre mice were harvested after 12 weeks of CCl4 injection and analyzed for PCNA expression using immunofluorescence staining with anti-PCNA (red) and DAPI (blue). (B) PCNA-positive cells in liver tissues were counted from three different areas and the number/area is shown. All values represent means ± SD from 3-4 mice in each group, **P < 0.01; compared with corresponding controls.

Discussion

The contribution of STAT5 to cell proliferation and survival has been firmly established.1 For example, upon ablation of STAT5 in mice, mammary luminal epithelial cells fail to proliferate and form alveoli18 and cells from various hematopoietic lineages are absent.19-21 Conditional ablation of STAT5 upon the establishment of leukemia results in a complete remission,2 further suggesting a role of STAT5 for tumor maintenance. Loss of STAT5A also protects mice from oncogene-induced mammary tumorigenesis.22-24 Lastly, ectopic expression of a constitutively active STAT5A in mice, cells, or transplanted bone marrow cells can induce myeloproliferative disorders3, 5, 25 and breast cancer.4 These experimental models are in agreement with the findings that many solid tumors and probably all leukemias display constitutively active STAT5. Although the complete set of STAT5 target genes that might control cell proliferation and survival remains elusive, genes encoding cyclin D1, bcl-xL (B cell lymphoma extra large), and Akt have been shown to be under STAT5 control.26, 27

This study for the first time establishes that STAT5 activates the cell cycle suppressor gene Cdkn2b, and thereby suppresses cell proliferation. This was not only observed in MEFs, but also in liver tissue. The ability of STAT5 to activate expression of the Cdkn2b gene is likely cell- and context-dependent. We performed global gene expression analyses of several cell lineages and tissues in the presence and absence of STAT5, and a STAT5-regulated expression of the Cdkn2b gene was only observed in MEFs and liver tissue. Expression of the Cdkn2b gene is not only controlled by STAT5 but also by other transcription factors, including Forkhead box and Octamer 1,28, 29 suggesting that the context specificity is the result of a combinatorial effect of several transcription factors. Importantly, re-expression of STAT5A in Stat5−/− MEFs led to the reactivation and GH-induced expression of the Cdkn2b and Cdkn1a genes and a concomitant reduced cell proliferation. Although this study demonstrates for the first time that expression of the Cdkn2b gene is under the control of STAT5A, other studies have linked STATs to expression of the Cdkn1a gene.21, 25, 30

The sharp reduction of p15INK4B in STAT5-null liver tissue corresponds with increased phospho-histone H3 and PCNA staining, indicating increased cell proliferation. Loss of STAT5 in liver tissue combined with a CCl4 insult results in liver tumors and we suggest that the reduction of appropriate p15INK4B levels contributes to these events. In support of this hypothesis, loss of p15INK4B, due to methylation of the Cdkn2b promoter, has been detected in patients with HCC.31

We propose that STAT5 exhibits two distinct inroads into the cell cycle: a generic one observed in many cancer cells and a cell- and/or context-specific one. Findings described in this study highlight context-specific contributions of STAT5 to cell cycle control, in particular, the potential role of STAT5 as a brake. For example, in mammary tissue, STAT5 has distinct roles depending on the particular developmental stage. Although STAT5 controls progenitor populations during puberty, the presence during early stages of pregnancy is critical for cell proliferation.4, 18, 32 Upon completion of the proliferation program, STAT5 is critical for the differentiation of mammary alveolar cells leading to the production of milk.33 Clearly in the context of mammary epithelium, cell proliferation and differentiation occur in different, albeit overlapping, time windows and STAT5 assumes different roles depending on the cell type and hormonal status. Similarly, STAT5 exerts several distinct and unique roles in hematopoietic cell lineages that point to STAT5 as a modulator of cell proliferation and differentiation. Similar to the case in MEFs, STAT5 can also be a negative regulator of cell proliferation in hematopoietic stem cells (HSCs). A recent study by Wang and colleagues21 demonstrated increased cycling of STAT5-null HSCs, followed by reduced survival and a depleted long-term HSC pool. Similar to our study, loss of STAT5 from HSCs also resulted in an increased pool of cells in the S/G2/M phase. Although Cdkn1a levels were only slightly reduced in the absence of STAT5A, a more profound reduction of Cdkn1c was observed, suggesting transcriptional regulation by STAT5. Notably, we have identified STAT5A binding to the Cdkn1c gene using ChIP-seq technology, and expression of the Cdkn1c gene in MEFs was reduced by more than 90% in Stat5−/− MEFs (unpublished data). Lastly, STAT5 also appear to contribute to and maintain different physiological states in the liver. Most obvious is the contribution of STAT5 to the differentiation of hepatocytes as measured by the GH-dependent expression of liver-specific genes.9 In hepatocytes, the presence of STAT5 is also required to curtail GH-induced activation of STAT3, which in itself is a transforming stimulus.7, 8 Finally, in hepatocytes, STAT5 also activates the Cdkn2b gene whose product p15INK4B negatively controls the cell cycle. Experiments provided in this study therefore shed light on yet another mechanism exploited by STAT5 to evoke unique cellular responses.

Acknowledgements

The authors contributed to this work as follows: JHY, experimental design, performed experiments, data analysis, wrote manuscript; BMZ, ChIP assay (Figure 3D); MW and GR, generated STAT5A overexpressing MEFs (Figure 4); WC, analyzed microarray experiments (Table 1); AH, performed microarray experiments (Table 1); GWR, contributed embryos, wrote manuscript; LH, experimental design, data analysis, wrote manuscript.

We thank Richard Moriggl for providing us with the STAT5a expressing retrovirus.

Ancillary