C/EBPβ Is a Transcriptional Regulator of Wee1 at the G2/M Phase of the Cell Cycle

The CCAAT/enhancer-binding protein β (C/EBPβ) is a transcription factor that regulates cellular proliferation, differentiation, apoptosis and tumorigenesis. Although the pro-oncogenic roles of C/EBPβ have been implicated in various human cancers, how it contributes to tumorigenesis or tumor progression has not been determined. Immunohistochemistry with human non-small cell lung cancer (NSCLC) tissues revealed that higher levels of C/EBPβ protein were expressed compared to normal lung tissues. Knockdown of C/EBPβ by siRNA reduced the proliferative capacity of NSCLC cells by delaying the G2/M transition in the cell cycle. In C/EBPβ-knockdown cells, a prolonged increase in phosphorylation of cyclin dependent kinase 1 at tyrosine 15 (Y15-pCDK1) was displayed with simultaneously increased Wee1 and decreased Cdc25B expression. Chromatin immunoprecipitation (ChIP) analysis showed that C/EBPβ bound to distal promoter regions of WEE1 and repressed WEE1 transcription through its interaction with histone deacetylase 2. Treatment of C/EBPβ-knockdown cells with a Wee1 inhibitor induced a decrease in Y15-pCDK1 and recovered cells from G2/M arrest. In the xenograft tumors, the depletion of C/EBPβ significantly reduced tumor growth. Taken together, these results indicate that Wee1 is a novel transcription target of C/EBPβ that is required for the G2/M phase of cell cycle progression, ultimately regulating proliferation of NSCLC cells.

C/EBPβ was shown to be required for carcinogen-induced mouse skin tumorigenesis by regulating p53-induced apoptosis upon carcinogen treatment [12,13], even though it induced

Cell Cycle Analysis
For thymidine double block [36], cells were seeded at 2 × 10 5 cells per well in 6-well culture plates and treated with thymidine at 2 mM for 14 h. Cells were then washed, supplemented with normal media for 12 h, and treated with 2 mM thymidine for another 14 h. The cells were harvested at 2 h intervals up to 16 h and 26 h after release. Afterward, cells were harvested and fixed in 75% (v/v) cold ethanol at −20 • C for at least 2 h. The fixed cells were collected by centrifugation and resuspended in propidium iodide (PI) Staining Buffer (Sigma, St. Louis, MO, USA) to stain DNA and analyzed for DNA content on a flow cytometry (FACSCaliber; Becton Dickinson, Franklin Lakes, NJ, USA).

Live Cell Fluorescence Imaging for Analysis of Motisis
A549 cells were seeded in 12-well culture plates and transfected with negative control siNC or siC/EBPβ. After 24 h, cells were stained with NucBlue ® live reagent (Hoechst 33342 dye, Thermo Scientific, Rockford, IL, USA) for 20 min following matufcturer's protocol, then replaced with fresh culture medium. Images were taken by Operetta High Content Screening (HCS) System (PerkinElmer, Waltham, MA, USA) every 10 min for 48 h under 200× magnification.

Luciferase Reporter Assay
The constructs of Wee1 promoter region were cloned into pGL3-promoter firefly luciferase vector (Promega, Madison, WI, USA), which were named R3 (−4932 to −4679), and R2 (−4543 to −4380), C/EBPβ and HDAC2 cDNA clones purchased from OriGene were sub-cloned into pcDNA3.1 vector (Invitrogen, Waltham, MA, USA). Cells were seeded in 24-well plates and co-transfected with reporter vectors and pcDNA3.1 or pcDNA3.1-C/EBPβ and/or pcDNA3.1-HDAC2 as indicated using Lipofectamine 2000 (Invitrogen). After 48 h, cells were harvested. Firefly luciferase activities were determined using the Luciferase assay system kit (Promega, Madison, WI, USA), as described by the manufacturer, with a luminescence plate reader (VICTOR™ X, PerkinElmer, Waltham, MA, USA). The firefly luciferase activity was normalized for transfection efficiency with protein measurement using a BCA protein assay. Data are expressed as relative luciferase activity/µg protein.

Immunoprecipitation
Cells were lysed in cell lysis buffer (20 mM Tris-HCl pH8.0, 150 mM NaCl, 1 mM Na 2 EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium pyrophosphate, and 1 mM β-glycerophosphate). Each cell lysate (1 mg) was incubated with C/EBPβ monoclonal antibody (Santacruz Biotechnology, Santa Cruz, CA, USA) overnight at 4 • C. Following incubation, protein was immunoprecipitated using protein G agarose beads (GE Healthcare, Chicago, IL, USA) for 2 h at 4 • C with gentle rotation. The immunoprecipitates were washed three times with lysis buffer and boiled in 20 µL of 1× SDS sample buffer for 5 min at 95 • C. After centrifugation, the supernatant was analyzed using Western blot.

Xenograft Mouse Model and siRNA Delivery
A549 (5 × 10 6 ) cells were suspended in 100 µL PBS and mixed with 50 µL Matrigel (Corning Inc.). The mixtures were implanted subcutaneously into 6-week-old athymic nude mice. When the tumor size reached 60 to 80 mm 3 , the dilute siRNA solution in sterile PBS (50 µL) was directly injected into the xenograft tumor via electroporation using NEPA21 Super Electroporator (Nepa gene Co., Chiba, Japan). The tumor size was monitored every 7 days up to 7 weeks. Tumor diameters were measured twice a week and the volume was calculated with the following formula: V (mm 3 ) = longest diameter × shortest diameter 2 /2.

Immunohistochemical Staining for Lung Cancer Tissue Microarray
Lung tissue arrays [CCN5, Human, Normal lung (59 adjacent normal lung tissues matching CC5, 1 carbon); CC5, Human, Lung cancer (59 NSCLC tissues, 1 carbon); CCA4 Human, Lung cancer-metastasis-normal (36 NSCLCs, 1 missing NSCLC, 2 small cell lung cancers (SCLCs), 1 malignant mesothelioma; 9 metastatic tissues matching 9 among 36 NSCLCs, 1 metastatic tissue matching 1 among 2 SCLCs; 9 normal lung tissues matching 9 among 36 NSCLC, 1 carbon] were purchased from Superbiochips Laboratories (Seoul, Korea) [37]. Total number of tissues on 3 microarrays was 68 for adjacent normal lung tissues, 95 for NSCLC tissues and 9 for metastatic tissues from 95 patients. Each array contained 59 sections of 4 µm thickness obtained by surgical resection and one carbon for orientation. The sections were used for immunohistochemical staining performed with the Ventana BenchMark XT Staining systems (Ventana Medical Systems, Inc.) using C/EBPβ antibody (1:30, sc-150 Santacruz Biotechnology, Santa Cruz, CA, USA) and the UltraView Universal DAB detection kit (Ventana Medical Systems, Inc.). Parallel sections incubated with normal IgG antibody instead of primary antibodies were used for negative controls. The stainings were scored from 0 to 4 based on the intensity and proportion of positive staining in a tissue field. Stained tissue arrays were reviewed by two experienced medical pathologists. To obtain representative images, slides were scanned by the Aperio ScanScope scanner (Aperio Technologies, Vista, CA, USA) and images were captured using Aperio ImageScope software.

Statistical Analysis
All data points represent average values and standard deviation (SD) or standard error (SE) obtained from three independent experiments performed in triplicate. Comparison between two groups was performed using Student's t-test. The relationship between C/EBPβ expression and clinicopathologic characteristics was analyzed using the Student's t-test. Statistical significance was defined as a p value < 0.05. For survival analysis, Kaplan-Meier Plotter [38] was used to investigate the association of C/EBPβ mRNA expression in overall survival and post-progression survival in lung cancer patients.

Levels of C/EBPβ Protein in Human Lung Cancer Tissues Were Elevated
To explore the clinical significance of C/EBPβ in lung cancer, we examined the expression of C/EBPβ in patient-derived lung cancer tissue microarrays. Lung cancer tissues and adjacent normal lung tissue obtained from 95 patients were immunohistochemically stained for C/EBPβ. Representative pictures of C/EBPβ staining results with varying intensities (0 to 4) are shown in Figure 1A. As shown in Figure 1B and Table 1, the proportion of normal lung tissues with positive C/EBPβ staining was only 30.9%, whereas positive C/EBPβ staining increased up to 67.3% in the primary lung cancer tissues, including 29.4% of strong staining intensity (score 2-4). Scoring of C/EBPβ staining in each NSCLC subtype is listed in Figure S1A. Tumor samples from metastatic patients tended to display strong staining for C/EBPβ (score 2-3, 55.5%), and positive C/EBPβ staining of metastatic tissues was significantly higher than that of normal lung tissues ( Figure 1B). Immunohistochemical analyses of primary NSCLC tissues showed that C/EBPβ expression was already enhanced from the early stages of lung cancer but there were no significant differences among different clinical stages, tumor, node, metastasis (TNM) stage, age, or sex ( Table 2). A significant increase in C/EBPβ positive staining was observed in squamous cell carcinoma and other types of NSCLCs compared with lung adenocarcinoma. Using a public database, we also analyzed gene alteration status of C/EBPB in lung adenocarcinoma and lung squamous cell carcinoma, the most common subtypes of non-small cell lung cancers (NSCLCs). The C/EBPB gene was amplified by 1.22% and 1.57% and deleted by 0.1% and 0.17% in lung adenocarcinoma and lung squamous cell carcinoma, respectively, indicating that alteration of the C/EBPB gene is not frequent ( Figure S1B). An association between clinical outcomes of lung cancer patients and C/EBPβ expression was examined using the Kaplan-Meier Plotter [38]. This revealed that the levels of C/EBPβ mRNA were inversely correlated with the overall survival of patients with lung cancers, adenocarcinomas, and squamous cell carcinomas ( Figure 1C and Figure  S2A). Additionally, increased expression of C/EBPβ mRNA was associated with poor post-progression survival in all lung cancer and adenocarcinoma patients ( Figure S2A,B). All these data indicate that the expression of C/EBPβ is upregulated at the protein levels, which is a functional moiety, in human lung cancer and possibly correlated with clinical outcome of patients. Patients-derived lung cancer tissue arrays were examined for C/EBPβ expression using the immunoperoxidase method. Staining results were graded according to the intensity and proportion of positive area. Images were captured at a magnification of 200X by using the Aperio ImageScope software. Scale bars: 200 μm. (B) The histogram represents the percentage of the immunohistochemistry (IHC) score for C/EBPβ in 68 normal tissues, 95 primary, and 9 metastatic tumor tissues. The statistical significance was determined using the t-test, p < 0.05. (C) The association between C/EBPβ mRNA expression and overall survival of adenocarcinoma (whole dataset) and squamous cell carcinoma patients (GSE37745) was analyzed using the Kaplan-Meier Plotter. Hazard ratio (HR) significance was found with log-rank tests.  The association between C/EBPβ mRNA expression and overall survival of adenocarcinoma (whole dataset) and squamous cell carcinoma patients (GSE37745) was analyzed using the Kaplan-Meier Plotter. Hazard ratio (HR) significance was found with log-rank tests.

Knockdown of C/EBPβ Inhibits Cell Proliferation Rates of NSCLC Cells
In this research, we measured the levels of C/EBPβ protein in normal human bronchiolar epithelial cells (NHBE), immortalized human bronchial epithelial cells (BEAS-2B), and various subtypes of NSCLC cell lines. Different subtypes of lung cancers are known to be distinct in origin and in disease progression [39]. NHBE cells expressed C/EBPβ protein in marginal amounts compared with BEAS-2B cells and NSCLC cell lines tested ( Figure 2A). There was no significant association between C/EBPβ protein levels and specific NSCLC subtype.
To examine the function of C/EBPβ in lung cancer cells, we knocked down C/EBPβ using two different siRNAs in NSCLC cell lines. Transfection of each siC/EBPβ dramatically suppressed proliferation of A549 cells with C/EBPβ down-regulation ( Figure 2B). We also employed doxycycline-inducible shRNA targeting another sequence of C/EBPβ and observed that doxycycline treatment decreased C/EBPβ protein and cell proliferation ( Figure 2C).
To exclude the possibility that retarded cell proliferation induced by C/EBPβ-knockdown was limited to A549 cells, we knocked down C/EBPβ in several NSCLC cell lines with decent levels of C/EBPβ protein ( Figure 2D). Consistent with results shown in Figure 2B,C, cell proliferation rates were significantly slower in siC/EBPβ-transfected NSCLC cells than in negative control-siRNA (siNC)-transfected cells ( Figure 2E). Neither histological subtypes nor EGFR and K-Ras mutation status of the cell lines produced significant differences in the growth arrest induced by C/EBPβ-knockdown ( Figure 2E). NSCLC subtypes and mutational status of K-RAS and EGFR in each cell line are listed in Table S1. Although C/EBPβ-knockdown attenuated cell proliferation in BEAS-2B cells, the degree of inhibition appeared minimal compared with NSCLC cell lines. These observations indicate that C/EBPβ is important for cell proliferation in NSCLC cells. doxycycline-inducible shRNA targeting another sequence of C/EBPβ and observed that doxycycline treatment decreased C/EBPβ protein and cell proliferation ( Figure 2C). To exclude the possibility that retarded cell proliferation induced by C/EBPβ-knockdown was limited to A549 cells, we knocked down C/EBPβ in several NSCLC cell lines with decent levels of C/EBPβ protein ( Figure 2D). Consistent with results shown in Figure 2B,C, cell proliferation rates were significantly slower in siC/EBPβ-transfected NSCLC cells than in negative control-siRNA (siNC)-transfected cells ( Figure 2E). Neither histological subtypes nor EGFR and K-Ras mutation status of the cell lines produced significant differences in the growth arrest induced by C/EBPβ-knockdown ( Figure 2E). NSCLC subtypes and mutational status of K-RAS and EGFR in each cell line are listed in Table S1. Although C/EBPβ-knockdown attenuated cell proliferation in BEAS-2B cells, the degree of inhibition appeared minimal compared with NSCLC cell lines. These observations indicate that C/EBPβ is important for cell proliferation in NSCLC cells. Aenocarcinoma, A549, NCI-H1975, NCI-H23, NCI-H1703, NCI-H522, A427, Calu-3, NCI-H358; adeno-squamous cell carcinoma, HCC2279; squamous cell carcinoma, HCC95, HCC1588; large cell carcinoma, H460, NCI-H1299; anaplastic carcinoma, Calu-6. (B) Live cells of A549 transfected with si-Negative Control (siNC), siC/EBPβ #1, or siC/EBPβ #2 were counted with trypan blue staining at indicated times after transfection. Data are presented as fold increase. (C) Doxycycline-inducible shC/EBPβ cells using A549 were generated and treated with or without doxycycline (100 ng/mL). Using the IncuCyte live cell imaging system, proliferation cells was monitored and quantified by the percentage of cell confluence. (D) The protein levels of C/EBPβ were detected by Western blotting to check C/EBPβ-knockdown in each cell line. (E) Cell number of lung cancer cell lines transfected with siNC or siC/EBPβ (#1 + #2) was counted using a Coulter counter at intervals of 24 h up to 120 h after siRNA transfection. K-RAS: mutant K-RAS/wild-type EGFR, EGFR: wild-type K-RAS/mutant EGFR, WT: wild-type K-RAS/wild-type EGFR. Data are presented as mean ± standard deviation (SD). The statistical significance was determined using t-tests, * p < 0.05, ** p < 0.01. Using the IncuCyte live cell imaging system, proliferation cells was monitored and quantified by the percentage of cell confluence. (D) The protein levels of C/EBPβ were detected by Western blotting to check C/EBPβ-knockdown in each cell line. (E) Cell number of lung cancer cell lines transfected with siNC or siC/EBPβ (#1 + #2) was counted using a Coulter counter at intervals of 24 h up to 120 h after siRNA transfection. K-RAS: mutant K-RAS/wild-type EGFR, EGFR: wild-type K-RAS/mutant EGFR, WT: wild-type K-RAS/wild-type EGFR. Data are presented as mean ± standard deviation (SD). The statistical significance was determined using t-tests, * p < 0.05, ** p < 0.01.

C/EBPβ-Knockdown Delayed G 2 /M Phase of the Cell Cycle Progression with Elevated Levels of Inhibitory Phosphorylation of CDK1
Next, we investigated whether inhibition of cell proliferation was due to increased cell death or inhibited cell cycle progression in the C/EBPβ-knockdown cells. There was little difference in cell death between control and C/EBPβ-knockdown cells ( Figure 3A), indicating cell death was not involved in C/EBPβ-knockdown-induced inhibition of cell proliferation. We then examined the role of C/EBPβ in cell cycle regulation. We analyzed the cell cycle of C/EBPβ-knockdown cells by flow cytometry, which showed a significant decrease in the population of cells in the G 0 /G 1 and a simultaneous increase in the G 2 /M phase of the cell cycle compared with the control ( Figure 3B). G 2 /M progression of the cell cycle is mainly regulated by cyclin B/CDK1 [45]. Activity of CDK1 is regulated by its phophorylation. The Wee1 and Myt1 kinases phosphorylate CDK1 at tyrosine 15 (Tyr15), thereby inhibiting CDK1. Dephosphorylation of Tyr15 by cell division cycle 25 (Cdc25) phosphatase is required for the activation of CDK1 and further entry into mitosis [46][47][48][49][50]. Knockdown of C/EBPβ increased phosphorylation of CDK1 at the tyrosine 15 residue (Y15-pCDK1) with a simultaneous increase in the levels of Wee1 and cyclin B1 and a decrease in the levels of Cdc25B ( Figure 3C). The expression of Cdc25A and Cdc25C remained unchanged between the groups. C/EBPβ-knockdown did not appeared to affect the protein levels of Myt1 and mitotic arrest deficient 2 (Mad2), mitotic spindle checkpoint component [51]. Taken together, these data indicate that C/EBPβ-knockdown cells undergo a delay in the G 2 /M phase of the cell cycle, displaying elevated Y15-pCDK1 possibly due to increased Wee1 and decreased Cdc25B.

C/EBPβ-Knockdown Delayed G2/M Phase of the Cell Cycle Progression with Elevated Levels of Inhibitory Phosphorylation of CDK1
Next, we investigated whether inhibition of cell proliferation was due to increased cell death or inhibited cell cycle progression in the C/EBPβ-knockdown cells. There was little difference in cell death between control and C/EBPβ-knockdown cells ( Figure 3A), indicating cell death was not involved in C/EBPβ-knockdown-induced inhibition of cell proliferation. We then examined the role of C/EBPβ in cell cycle regulation. We analyzed the cell cycle of C/EBPβ-knockdown cells by flow cytometry, which showed a significant decrease in the population of cells in the G0/G1 and a simultaneous increase in the G2/M phase of the cell cycle compared with the control (Figure 3B). G2/M progression of the cell cycle is mainly regulated by cyclin B/CDK1 [45]. Activity of CDK1 is regulated by its phophorylation. The Wee1 and Myt1 kinases phosphorylate CDK1 at tyrosine 15 (Tyr15), thereby inhibiting CDK1. Dephosphorylation of Tyr15 by cell division cycle 25 (Cdc25) phosphatase is required for the activation of CDK1 and further entry into mitosis [46][47][48][49][50]. Knockdown of C/EBPβ increased phosphorylation of CDK1 at the tyrosine 15 residue (Y15-pCDK1) with a simultaneous increase in the levels of Wee1 and cyclin B1 and a decrease in the levels of Cdc25B ( Figure 3C). The expression of Cdc25A and Cdc25C remained unchanged between the groups. C/EBPβ-knockdown did not appeared to affect the protein levels of Myt1 and mitotic arrest deficient 2 (Mad2), mitotic spindle checkpoint component [51]. Taken together, these data indicate that C/EBPβ-knockdown cells undergo a delay in the G2/M phase of the cell cycle, displaying elevated Y15-pCDK1 possibly due to increased Wee1 and decreased Cdc25B.  In order to observe a temporal progression in the G 2 /M phase of the cell cycle, we synchronized cells at the G 1 /S boundary using thymidine double block methods [36], and released cells into the cell cycle by removing thymidine. Control cells progressed into the cell cycle and accumulated in the G 2 /M phase 6 h after release at the highest levels, and subsequently moved into the G 1 phase in just 2-4 h ( Figure 4A,B). In contrast, C/EBPβ-knockdown cells reached the G 2 /M peak 8 h after release into the cell cycle, and stayed in the G 2 /M phase for much longer up to 16 h. Cell populations in the G 2 /M phase in an asynchronized condition (26 h) were significantly higher in C/EBPβ-knockdown cells compared with controls, consistent with results shown in Figure 3B. To determine the effect of C/EBPβ-knockdown on mitotic progression, we analyzed mitotic duration using time-lapse live cell imaging with nuclear Hoechst 33342 staining. As shown in Figure 4D, mitotic duration from nuclear envelope breakdown to anaphse onset [52] was not significantly different between control and C/EBPβ-knockdown cells, suggesting that a delay in the G 2 /M phase of the cell cycle progression displayed in C/EBPβ-knockdown cells was not due to a delay in the mitotic phase, but was rather due to prolonged stay in the G 2 phase of the cell cycle. To confirm a delay in the G 2 -M transition of C/EBPβ-knockdown cells, we analyzed mitotic cell populations using flow cytometry along with phospho-histone H3 (Ser10) and PI staining. Phosphorylation of histone H3 (Ser10) is essential for mitotic condensation and thus considered a mitotic marker [53]. As shown in Figure S3, more C/EBPβ-knockdown cells were accumulated in the G 2 /M phase than control cells, consistent with results in Figure 3B, but mitotic cells with positive phospho-histone H3 (Ser10) staining decreased significatly in C/EBPβ-knockdown cells. Combined with the live cell imaging results, C/EBPβ seem to be important in the G 2 /M transition of the cell cycle.
In the control cells, levels of Y15-pCDK1 started to decrease within 6 h and suddenly dropped 10 h after release when the most cells returned to the G 1 phase ( Figure 4C). A timely and transient increase in Cdc25B protein levels was correlated with a decrease in Y15-pCDK1, which is an important event for mitotic entry. However, Y15-pCDK1 stayed at higher levels for a longer time with an increase in Wee1 and a delayed and attenuated Cdc25B induction in the C/EBPβ-knockdown cells compared with control cells ( Figure 4C). Taken together, our data demonstrate that C/EBPβ is an important mediator at the G 2 /M phase of the cell cycle progression by regulating the expression of Wee1 and Cdc25B, which critically modulates Y15-pCDK1. (D) Mitotic duration of control or C/EBPβ-knockdown A549 cells stained with NucBlue ® was monitored by an Operetta High Content Screening (HCS) System. Representative images of siNC and siC/EBPβ-transfected cells from time lapse series were shown. Images were acquired every 10 min. Data are presented as mean ± SD of mitotic duration in 30 cells in each group. Mitotic duration was measured from nuclear envelope breakdown (prometaphase) to anaphase onsets [52]. Statistical significance was determined using the t-test; n.s., not significant. Cell images were taken at a magnification of 200X using an Operetta High Content Screening (HCS) System. Scale bar: 20 μm.

C/EBPβ Regulates Wee1 Expression at the Transcription Levels
Next, we investigated if either WEE1 or CDC25B was the transcription target of C/EBPβ playing a role in cell cycle regulation. As shown in Figure 5A, in C/EBPβ-knockdown cells, mRNA levels of Wee1 and Cdc25B increased and decreased, respectively, correlated with their protein levels ( Figure  5A). With our C/EBPβ-ChIP on chip data (unpublished data) and using a web-based tool for searching transcription factor binding sites, TFSEARCH [54], we predicted four putative C/EBPβ binding regions on the WEE1 distal promoter, as shown in Figure 5B. We performed a C/EBPβ-ChIP assay and real-time PCR to verify that C/EBPβ bound to −4.7 to −4.9 kB (R3) and −4.4 to −4.5 kB (R2) upstream of the Wee1 transcription start site ( Figure 5C,F). However, we did not find any significant binding of C/EBPβ to the CD25B promoter based on our C/EBPβ-ChIP on chip data (unpublished data).
C/EBPβ has been reported to interact with histone deacetylase1 (HDAC1) and repress the activation of the C/EBPα promoter [55]. Our data showed that C/EBPβ interacts with HDAC2, not with HDAC1 in A549 cells ( Figure 5D). In addition, the HDAC2-ChIP assay revealed that HDAC2 shown. Images were acquired every 10 min. Data are presented as mean ± SD of mitotic duration in 30 cells in each group. Mitotic duration was measured from nuclear envelope breakdown (prometaphase) to anaphase onsets [52]. Statistical significance was determined using the t-test; n.s., not significant. Cell images were taken at a magnification of 200X using an Operetta High Content Screening (HCS) System. Scale bar: 20 µm.

C/EBPβ Regulates Wee1 Expression at the Transcription Levels
Next, we investigated if either WEE1 or CDC25B was the transcription target of C/EBPβ playing a role in cell cycle regulation. As shown in Figure 5A, in C/EBPβ-knockdown cells, mRNA levels of Wee1 and Cdc25B increased and decreased, respectively, correlated with their protein levels ( Figure 5A). With our C/EBPβ-ChIP on chip data (unpublished data) and using a web-based tool for searching transcription factor binding sites, TFSEARCH [54], we predicted four putative C/EBPβ binding regions on the WEE1 distal promoter, as shown in Figure 5B. We performed a C/EBPβ-ChIP assay and real-time PCR to verify that C/EBPβ bound to −4.7 to −4.9 kB (R3) and −4.4 to −4.5 kB (R2) upstream of the Wee1 transcription start site ( Figure 5C,F). However, we did not find any significant binding of C/EBPβ to the CD25B promoter based on our C/EBPβ-ChIP on chip data (unpublished data).
promoter activity, we constructed a promoter-reporter vector containing either the R2 or R3 region of the distal WEE1 promoter. Enforced expression of C/EBPβ or HDAC2 alone reduced WEE1 promoter-reporter activity and C/EBPβ along with HDAC2 further decreased transcriptional activity of the WEE1 promoter, indicating that C/EBPβ recruits HDAC2 to the WEE1 promoter, negatively regulating WEE1 transcription ( Figure 5G). Taken together, we identified Wee1, a mitosis inhibitor, as a novel transcription target gene of C/EBPβ. and TFSEARCH based binding sites are indicated as a rectangle and an oval, respectively. The prediction of WEE1 promoter regions was based on NCBI accession number (NC_000011.10, GRCh37.p11) (C) The C/EBPβ-ChIP assay followed by qRT-PCR on putative C/EBPβ binding regions on the WEE1 promoter was performed to determine endogenous C/EBPβ occupancy at the specified region. The fold enrichment of C/EBPβ occupancy over GAPDH exon (negative control) is shown. Data are presented as mean ± SE. (D) A549 cell lysates were immunoprecipitated using anti-C/EBPβ antibodies. Immunocomplexes were analyzed by Western blot with either anti-HDAC1 or -HDAC2 antibodies. IgG was used as a negative control. (E) HDAC2-ChIP assay followed by qRT-PCR on putative C/EBPβ binding regions at the WEE1 promoter was performed. The fold enrichment of HDAC2 occupancy over GAPDH exon (negative control) is shown. Data are presented as mean ± SE. (F) The C/EBPβ-ChIP or HDAC2-ChIP assay followed by PCR on putative C/EBPβ binding regions at the WEE1 promoter was performed. The PCR products resolved on 2% agarose gel were visualized. (G) A549 cells were co-transfected with a WEE1 promoter-luciferase construct containing R2, or R3 along with C/EBPβ and/or HDAC2, as indicated, for 48 h, and then luciferase activities were measured. Data are expressed as relative luciferase activity/ug protein standardized by a control pGL3-promoter vector. Data are presented as mean ± SD. Statistical significance was determined using the t-test, * p < 0.05. C/EBPβ-ChIP on chip data and TFSEARCH based binding sites are indicated as a rectangle and an oval, respectively. The prediction of WEE1 promoter regions was based on NCBI accession number (NC_000011.10, GRCh37.p11) (C) The C/EBPβ-ChIP assay followed by qRT-PCR on putative C/EBPβ binding regions on the WEE1 promoter was performed to determine endogenous C/EBPβ occupancy at the specified region. The fold enrichment of C/EBPβ occupancy over GAPDH exon (negative control) is shown. Data are presented as mean ± SE. (D) A549 cell lysates were immunoprecipitated using anti-C/EBPβ antibodies. Immunocomplexes were analyzed by Western blot with either anti-HDAC1 or -HDAC2 antibodies. IgG was used as a negative control. (E) HDAC2-ChIP assay followed by qRT-PCR on putative C/EBPβ binding regions at the WEE1 promoter was performed. The fold enrichment of HDAC2 occupancy over GAPDH exon (negative control) is shown. Data are presented as mean ± SE. (F) The C/EBPβ-ChIP or HDAC2-ChIP assay followed by PCR on putative C/EBPβ binding regions at the WEE1 promoter was performed. The PCR products resolved on 2% agarose gel were visualized. (G) A549 cells were co-transfected with a WEE1 promoter-luciferase construct containing R2, or R3 along with C/EBPβ and/or HDAC2, as indicated, for 48 h, and then luciferase activities were measured. Data are expressed as relative luciferase activity/ug protein standardized by a control pGL3-promoter vector. Data are presented as mean ± SD. Statistical significance was determined using the t-test, * p < 0.05. C/EBPβ has been reported to interact with histone deacetylase1 (HDAC1) and repress the activation of the C/EBPα promoter [55]. Our data showed that C/EBPβ interacts with HDAC2, not with HDAC1 in A549 cells ( Figure 5D). In addition, the HDAC2-ChIP assay revealed that HDAC2 also bound to the R2 and R3 regions ( Figure 5E,F), suggesting that C/EBPβ recruits HDAC2 at the WEE1 distal promoter. To determine if C/EBPβ in conjunction with HDAC2 repressed WEE1 promoter activity, we constructed a promoter-reporter vector containing either the R2 or R3 region of the distal WEE1 promoter. Enforced expression of C/EBPβ or HDAC2 alone reduced WEE1 promoter-reporter activity and C/EBPβ along with HDAC2 further decreased transcriptional activity of the WEE1 promoter, indicating that C/EBPβ recruits HDAC2 to the WEE1 promoter, negatively regulating WEE1 transcription ( Figure 5G). Taken together, we identified Wee1, a mitosis inhibitor, as a novel transcription target gene of C/EBPβ.

C/EBPβ-Knockdown Cells Treated with MK1775 Were Recovered from G 2 /M Delay
To confirm Wee1 plays an important role in a delay in the G 2 /M cell cycle shown in C/EBPβ deficiency, we examined whether the inhibition of Wee1 could release C/EBPβ-knockdown cells from the delay in the G 2 /M. Four hours after release into the cell cycle at the completion of the thymidine double block, cells were treated with a Wee1 inhibitor, MK1775, or DMSO. MK1775 treatment of control cells had little effect on cell cycle progression ( Figure 6A,B). However, Wee1 inhibition of C/EBPβ-knockdown cells diminished accumulation of cell populations in the G 2 /M phase induced by C/EBPβ-deficiency ( Figure 6A,B). More specifically, C/EBPβ-knockdown cells accumulated in the G 2 /M phase up to 10 h (62.2%) because of a delay in the G 2 /M phase and decreased to 43. 6% at 12 h after release. However, upon MK-1775 treatment of C/EBPβ-knockdown cells, cell populations in the G 2 /M phase maximally reached up to 57.8% at 8 h, further progressed into the G 0 /G 1 and S phase, and then decreased to 44.9% and 22% at 10 and 12 h after release, respectively. Consistent with the results in Figures 3 and 4, compared with control cells, Y15-pCDK1 increased in the C/EBPβ-knockdown cells ( Figure 6C). Treatment with MK1775 induced a rapid decrease in Y15-pCDK1 with little changes in total levels of CDK1. As MK1775 treatment rescued C/EBPβ-knockdown cells from the delay in the G 2 /M phase, we argue that increased Y15-pCDK1 by Wee1 is responsible for the delay in the G 2 /M phase with defective C/EBPβ function.  Figure 6C). Treatment with MK1775 induced a rapid decrease in Y15-pCDK1 with little changes in total levels of CDK1. As MK1775 treatment rescued C/EBPβ-knockdown cells from the delay in the G2/M phase, we argue that increased Y15-pCDK1 by Wee1 is responsible for the delay in the G2/M phase with defective C/EBPβ function.

C/EBPβ-Knockdown Inhibits Tumor Growth In Vivo
Since C/EBPβ-knockdown inhibits cell proliferation inducing a delay in the G 2 /M phase of the cell cycle, we tested if C/EBPβ-knockdown reduces the growth of xenograft tumors in vivo. Tumors were produced by injecting A549 cells subcutaneously into the dorsal area of athymic nude mice, and siNC or siC/EBPβ RNA was delivered into the tumors via electroporation. Tumor growth was monitored for seven weeks. As shown in Figure 7A, siC/EBPβ treatment markedly suppressed tumor growth by at least 50% compared with the siNC treatment. Consistent with in vitro results, immunohistochemistry analysis revealed that treatment with siC/EBPβ increased the expression of Y15-pCDK1 and Wee1 while decreasing Cdc25B expression ( Figure 7B). C/EBPβ-knockdown tumors displayed lower Ki67 and higher cleaved caspase-3 expression compared with the control. These data indicate that C/EBPβ is important for tumor growth in vivo.

C/EBPβ-Knockdown Inhibits Tumor Growth In Vivo
Since C/EBPβ-knockdown inhibits cell proliferation inducing a delay in the G2/M phase of the cell cycle, we tested if C/EBPβ-knockdown reduces the growth of xenograft tumors in vivo. Tumors were produced by injecting A549 cells subcutaneously into the dorsal area of athymic nude mice, and siNC or siC/EBPβ RNA was delivered into the tumors via electroporation. Tumor growth was monitored for seven weeks. As shown in Figure 7A, siC/EBPβ treatment markedly suppressed tumor growth by at least 50% compared with the siNC treatment. Consistent with in vitro results, immunohistochemistry analysis revealed that treatment with siC/EBPβ increased the expression of Y15-pCDK1 and Wee1 while decreasing Cdc25B expression ( Figure 7B). C/EBPβ-knockdown tumors displayed lower Ki67 and higher cleaved caspase-3 expression compared with the control. These data indicate that C/EBPβ is important for tumor growth in vivo. were implanted subcutaneously into athymic nude mice. When tumor size reached 60 to 80 mm 3 , siNC or siC/EBPβ were delivered into the tumors via electroporation once a week for seven weeks. Tumors were measured at the indicated time and tumor volume was calculated as described in Section 2. Photos from siNC-or siC/EBPβ-treated tumors are shown. Similar results were observed in three independent experiments. Data are presented as mean ± SD. The statistical significance was determined using the t-test, * p < 0.05, ** p < 0.01, significantly different from siC/EBPβ-treated tumor volume. (B) Immunohistochemical staining for C/EBPβ, Y15-pCDK1, Wee1, Cdc25B, Ki67, and cleaved caspase-3 (CC-3) was conducted with paraformaldehyde-fixed, paraffin-embedded xenograft tumors. Images were captured at a magnification of 400X by using the Aperio ImageScope software. Scale bars: 100 μm.

Discussion
Mitosis is regulated by cyclin B/CDK1 [56]. Wee1 kinase phosphorylates CDK1 at Tyr15, and dephosphorylation of this site by Cdc25 is required for the activation of CDK1 and further entry into mitosis. We showed that C/EBPβ is important for the proliferation of NSCLC cells by mediating the G2/M transition of cell cycle. C/EBPβ regulated the inhibitory phosphorylation at the Tyr15 residue of CDK1 (Figures 3 and 4). This resulted from the function of C/EBPβ, which increased the expression of Cdc25B phosphatase but inhibited the expression of Wee1 kinase at the mRNA and protein levels (Figures 3, 4 and 5A). The mechanism of regulating Cdc25B expression by C/EBPβ Figure 7. C/EBPβ-knockdown inhibits tumor growth. (A) A549 cells (5 × 10 6 ) were implanted subcutaneously into athymic nude mice. When tumor size reached 60 to 80 mm 3 , siNC or siC/EBPβ were delivered into the tumors via electroporation once a week for seven weeks. Tumors were measured at the indicated time and tumor volume was calculated as described in Section 2. Photos from siNC-or siC/EBPβ-treated tumors are shown. Similar results were observed in three independent experiments. Data are presented as mean ± SD. The statistical significance was determined using the t-test, * p < 0.05, ** p < 0.01, significantly different from siC/EBPβ-treated tumor volume. (B) Immunohistochemical staining for C/EBPβ, Y15-pCDK1, Wee1, Cdc25B, Ki67, and cleaved caspase-3 (CC-3) was conducted with paraformaldehyde-fixed, paraffin-embedded xenograft tumors. Images were captured at a magnification of 400X by using the Aperio ImageScope software. Scale bars: 100 µm.

Discussion
Mitosis is regulated by cyclin B/CDK1 [56]. Wee1 kinase phosphorylates CDK1 at Tyr15, and dephosphorylation of this site by Cdc25 is required for the activation of CDK1 and further entry into mitosis. We showed that C/EBPβ is important for the proliferation of NSCLC cells by mediating the G 2 /M transition of cell cycle. C/EBPβ regulated the inhibitory phosphorylation at the Tyr15 residue of CDK1 (Figures 3 and 4). This resulted from the function of C/EBPβ, which increased the expression of Cdc25B phosphatase but inhibited the expression of Wee1 kinase at the mRNA and protein levels ( Figures 3, 4 and 5A). The mechanism of regulating Cdc25B expression by C/EBPβ does not seem to be mediated by direct transcriptional activation and is yet to be determined. C/EBPβ bound to two WEE1 distal promoter regions, −4.7 to −4.9 kB and −4.4 to −4.5 kB, upstream of the transcription start site and repressed transcription through recruiting HDAC2 ( Figure 5). These findings are summarized as a schematic diagram in Figure 8. To the best of our knowledge, this is the first report that Wee1, a key regulator of G 2 /M progression, is a transcriptional target of C/EBPβ. Finally, we showed that a Wee1 inhibitor, MK-1775, significantly recovered C/EBPβ-knockdown cells from the delay in the G 2 /M phase of the cell cycle ( Figure 6). Taken together, these results indicate that C/EBPβ is a transcriptional regulator of Wee1, mitosis inhibitor protein, ultimately regulating the G 2/ M phase of the cell cycle progression.
Cell cycle proteins need to be tightly controlled temporally and spatially. Wee1 increases during the S and G 2 phase to block premature mitotic entry and then decreases during the M phase of the cell cycle [57]. Regulation of the Wee1 protein has been extensively studied and several mechanisms of modulating its kinase activity and protein levels have been demonstrated [58]. For example, Wee1 kinase activity is regulated by Akt-mediated phosphorylation on Ser642 and phosphorylated Wee1 binds to 14-3-3θ and translocates to the cytoplasm, resulting in G 2 /M cell cycle progression [59]. In late G 2 phase, phosphorylation of Wee1 by CDK1 and polo-like kinase 1 creates a phosphodegron which targets Wee1 for SCF-β-TrCP ubiquitin ligase-mediated proteasomal degradation [60]. However, transcriptional regulation of WEE1 is not well known. WEE1 transcription was shown to be repressed by direct binding of kruppel-like factor 2 (KLF2) and chromodomain helicase DNA binding protein 5 (CHD5) and activated by c-Fos/activator protein-1 (AP-1) [61][62][63]. Our results also demonstrated that Wee1 is downregulated by C/EBPβ via direct binding to the distal promoter and it is required for the G 2 /M cell cycle progression in lung cancer cells.
C/EBPβ positively regulated the proliferation of various NSCLC cells regardless of NSCLC subtypes or gene mutation status of EGFR or K-RAS, whereas it did not have a substantial effect on immortalized lung epithelial cells (Figure 2). Even if we identified Wee1 and Cdc25B as downstream players of C/EBPβ in the cell cycle regulatory role, the upstream cue, possibly altered in NSCLC, remains to be determined.
In our study, control cells after release from the thymidine double block progressed to the S phase rapidly reaching the peak levels in 2 h, and majority of them moved into the G 2 /M phase in another 4 h ( Figure 4A,B and Figure 6A,B). However, in C/EBPβ-knockdown cells, G 1 /S transition was delayed and significant population of cells remained in the G 0 /G 1 phase ( Figure 4A,B), suggesting that G 1 /S cell cycle arrest is also induced in the absence of C/EBPβ. Consistent with our observation, C/EBPβ-deficient human endometrial stromal cells underwent G 1 /S arrest shown by the reduction of BrdU incorporation after release from thymidine double block. In those cells, cyclin E was downregulated and thereby cyclin E-CDK2 was not functional for G 1 /S phase progression [64]. C/EBPβ has been shown to cooperate with E2F in activating transcription of E2F target genes involved in the G 1 /S transition by recruiting CBP/p300 coactivator [65].
C/EBPs are expressed in human lung epithelium and they play roles in lung development and differentiation [66]. C/EBPα-deficient mice displayed hyper-proliferative alveolar type II cells and a defect in alveolar architecture and suffered from respiratory distress [67,68], whereas lung phenotypes of C/EBPβ knockout mice were not specified [66]. Even if C/EBPβ is involved in the lung-specific gene expression in lung epithelial cells [69,70], it does not seem to be essential in basal lung development or differentiation, possibly due to functional redundancy with other C/EBP family members, as described in liver and mouse skin [71,72]. Rather acute lung injury is induced by lipopolysaccharide (LPS)-induced C/EBPβ mRNA [73], suggesting C/EBPβ might play a distinct role in pathological status.
Loss in the function mutations were found in acute myeloid leukemia, suggesting C/EBPα as potential human tumor suppressor [74]. C/EBPα has been reported to be down-regulated in more than 40% of human primary lung cancers [70,75], and it seems to be associated with increased DNA methylation of C/EBPA promoter [76]. More recently, the oncogenic role of C/EBPβ has been suggested in human cancers, but how it contributes to tumorigenesis or tumor progression needs to be determined. We found that C/EBPβ protein is up-regulated in NSCLCs (Figure 1). Additionally, the delivery of siRNA against C/EBPβ into xenografted mouse tumors effectively inhibited tumor growth (Figure 7). Considering the notion that the function of C/EBPβ is not vital in the lung, C/EBPβ could be an attractive target for cancer therapy alone or in combination. be determined. We found that C/EBPβ protein is up-regulated in NSCLCs ( Figure 1). Additionally, the delivery of siRNA against C/EBPβ into xenografted mouse tumors effectively inhibited tumor growth (Figure 7). Considering the notion that the function of C/EBPβ is not vital in the lung, C/EBPβ could be an attractive target for cancer therapy alone or in combination. Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1. Figure S1: Expression and gene alteration of C/EBPβ in human lung cancers. Figure S2: Overall survival (OS) and post-progression survival (PPS) of lung cancer patients. Table S1: Histological subtypes and gene mutation status of NSCLC cell lines. Figure S3: FACS analysis of mitotic cells using mitotic marker, phospho-Histone H3 (Ser10)  Figure S1: Expression and gene alteration of C/EBPβ in human lung cancers. Figure S2: Overall survival (OS) and post-progression survival (PPS) of lung cancer patients. Table S1: Histological subtypes and gene mutation status of NSCLC cell lines. Figure S3: FACS analysis of mitotic cells using mitotic marker, phospho-Histone H3 (Ser10).