ZNF300 Knockdown Inhibits Forced Megakaryocytic Differentiation by Phorbol and Erythrocytic Differentiation by Arabinofuranosyl Cytidine in K562 Cells

Previously, we reported that ZNF300 might play a role in leukemogenesis. In this study, we further investigated the function of ZNF300 in K562 cells undergoing differentiation. We found that ZNF300 upregulation in K562 cells coincided with megakaryocytic differentiation induced by phorbol-12-myristate-13-acetate (PMA) or erythrocytic differentiation induced by cytosine arabinoside (Ara-C), respectively. To further test whether ZNF300 upregulation promoted differentiation, we knocked down ZNF300 and found that ZNF300 knockdown effectively abolished PMA-induced megakaryocytic differentiation, evidenced by decreased CD61 expression. Furthermore, Ara-C-induced erythrocytic differentiation was also suppressed in ZNF300 knockdown cells with decreased γ-globin expression and CD235a expression. These observations suggest that ZNF300 may be a critical factor controlling distinct aspects of K562 cells. Indeed, ZNF300 knockdown led to increased cell proliferation. Consistently, ZNF300 knockdown cells exhibited an increased percentage of cells at S phase accompanied by decreased percentage of cells at G0/G1 and G2/M phase. Increased cell proliferation was further supported by the increased expression of cell proliferation marker PCNA and the decreased expression of cell cycle regulator p15 and p27. In addition, MAPK/ERK signaling was significantly suppressed by ZNF300 knockdown. These findings suggest a potential mechanism by which ZNF300 knockdown may impair megakaryocytic and erythrocytic differentiation.


Introduction
Krüppel-associated box (KRAB)-containing zinc finger proteins (ZFPs) comprise a large family of transcription regulators in mammals. KRAB-ZFPs typically bear an N-terminal KRAB (Krüppel-associated box) domain that functions to suppress transcription by recruiting KRAB domain-associated protein 1 (KAP-1). KAP1 subsequently recruits histone deacetylase and histone methyltransferase machinery to mediate heterochromatinization and gene silencing [1][2][3][4][5][6][7]. Based on the structure of the KRAB domain, the KRAB-ZFPs can be further classified into three subfamilies: KRAB (AB) with a classical A-box and a B-box, KRAB (A) with a classical A-box only, and KRAB (Ab) with a classical A-box and a highly divergent B-box [8]. The A-box is highly conserved and plays a key role in the repression of target genes while the B-box is less conserved and plays an auxiliary role [9]. It's been reported that the KRAB-ZFPs are only found in the tetrapod vertebrate, suggesting an important function of KRAB-ZFPs in the evolution process of the higher organisms [10,11].
ZNF300 is a typical member of KRAB-ZNFs. It was originally isolated from the human embryos based on the enrichment of C2H2-specific mRNA and primarily expressed in heart, skeletal muscle, and brain. It encodes a KRAB domain and 12 C2H2 type zinc finger domains as a nuclear protein. The KRAB domain of the ZNF300 protein exhibits typical transcription repressor activity [12] while the zinc finger domain binds the consensus sequence C(t/a)GGGGG(g/c)G that are found in the promoter regions of multiple genes such as IL2, IL2RB, CD44, TP53, tumor necrosis factor-a (TNFa), and TNF-a receptor associated factor 2 (TRAF2) [13]. Indeed, ZNF300 was shown to activate IL-2Rb promoter activity [13]. Recently, inflammation was shown to upregulate ZNF300 expression, which further increased NF-kB activity by up-regulating TRAF2 and interacting with IKKb [14]. ZNF300 upregulation also induced the expression of IL6 and IL8, which may lead to the exacerbation of inflammation and tumor metastasis [14]. In addition, ZNF300 was downregulated during embryonic stem cell differentiation in vitro [15] and associated with 5q-syndrome, a distinct subtype of primary myelodysplastic syndrome (MDS) defined by interstitial deletion of chromosome 5q31-33 [16,17]. Our previous studies also showed that ZNF300 was associated with myeloid differentiation [18]. Although these data suggested that ZNF300 is likely to play an important role in leukemogenesis and hematopoiesis, the exact role of ZNF300 remains unknown.
In this study, we aimed to reveal the potential role of ZNF300 in blood cell differentiation by using a K562 cell model. K562 is a human erythroleukemia cell line, approximates to megakaryocyte-erythrocyte progenitor stage, and has the bipotency to differentiate into megakaryocytes or erythrocytes induced by phorbol-12-myristate-13-acetate (PMA) or cytosine arabinoside (Ara-C), respectively [19]. We demonstrated that ZNF300 was upregulated in K562 cells undergoing megakaryocytic differentiation induced by PMA or erythrocytic differentiation induced by Ara-C, respectively. Furthermore, ZNF300 knockdown potently abolished K562 cell differentiation under both conditions. The loss of differentiation capacity in ZNF300 knockdown cells coincided with increased proliferation evidenced by increased cell percentage at S phase, upregulation of PCNA, and decreased expression of cell cycle regulators p15 and p27. In addition, MAPK/ERK signaling was quenched by ZNF300 knockdown. These observations (B) The resultant cells were also stained with PE-conjugated GPIIIa (CD61)-specific antibody. The samples were analyzed using flow cytometer. Data was analyzed with Flowjo and presented as histogram graph. (C, D) The mRNA level of ITGB3 (CD61) and ITGA2B (CD41a) in the resultant cells was measured by quantitative RT-PCR. Data was normalized to GAPDH and presented as bar graph. (E) The mRNA level of ZNF300 in the resultant cells was measured by quantitative RT-PCR and represented as the relative expression (mean¡SD). Data were representative results of 3 independent experiments with similar results. *** indicates p,0.001. (F) The protein expression level of ZNF300 in resultant cells was measured by western blot and quantified by densitometry. Numbers indicate the densitometry of ZNF300 protein normalized by that of HSC70, which is further normalized to that of untreated cells. Result was the representative blot from 3 experiments with similar result.  suggest that the increased proliferation and impaired MAPK/ERK may contribute to the loss of differentiation capacity in K562 cells.
To establish stable cell line with ZNF300 knockdown, we transfected K562 cells with shZNF300 constructs or control vector by electroporation. Briefly, the K562 were washed twice with PBS and resuspended in electroporation buffer at the concentration of 2610 7 cells/ml. Four mg of plasmid DNA was mixed with 100 ml of cell suspension. The DNA-cell mixture was subjected to electroporation in a 2 mm cuvette using a Nucleofector II electroporator (Amaxa). The electroporated cells were selected with puromycin (2 mg/ml) for one week. The expression of ZNF300 was measured by western blot analysis and quantitative RT-PCR analysis.

FACS analysis
Megakaryocytic or erythrocytic differentiation was measured by flow cytometry. Approximately, 1610 5 cells were collected and washed with PBS containing 1% BSA and 0.1% sodium azide followed by incubation with PE-conjugated anti-CD61 (GPIIIa) (BD Biosciences) or PE-conjugated anti-CD235a (Biolegend) at 4˚C for half an hour. The expression of CD61 and CD235a was measured by flow cytometry on a Beckman CyAn. Data were further analyzed using FlowJo software (Tree Star Inc).
For cell cycle profile analysis, cells were fixed with 2% PFA overnight at 4˚C, stained with 1 mg/ml DAPI in the presence of saponin (PBS with 0.1% BSA and 0.05% Saponin) for 2 hrs. The DNA content was measured by flow cytometry. Data were analyzed using ModFit LT (Verity Software House). Quantitative RT-PCR analysis Total RNA was isolated with TRIzol reagent and 1 mg of RNA was used for firststrand cDNA synthesis using RevertAid First Strand cDNA Synthesis kit (Thermo scientific, Lithuania, EU). SYBR Green Bestar Real-time PCR Master Mix (DBI Bioscience) was used and the PCR reactions were run on an ABI 7500 real-time PCR system (Applied Biosystems, Foster City, CA, USA). The PCR amplification conditions were: Denaturation at 95˚C for 5 min followed by 95˚C 30 sec, 60˚C 30 sec, 72˚C 30 sec for 40 cycles. Each PCR reaction was performed in triplicates and GAPDH was used as an endogenous control for normalization. The relative quantitation of real-time PCR product was measured using the comparative DDC T method [20,21] and presented as bar graph. Primers used for quantitative RT-PCR are as following: ZNF300-S 59-GGATGTGGCTGTGGATTT-39, ZNF300-A 59-ATGGCTCTTCTCCTTGTT-39; CD41-S 5-9GATGAGACCCGAAATGTAG-GC-39, CD41-A 59-GTCTTTTCTAGGACGTTCCAGTG-39; CD61-S 59-GTGAC-CTGAAGGAGAATCTGC-39, CD61-A 59-CCGGAGTGCAATCCTCTGG-39. Primers for c-Globin and GAPDH were derived from references [13,22].

Wright-Giemsa staining and benzidine staining
Wright-Giemsa staining was performed following the manual from the supplier (Sigma). Cell morphology was observed under a light microscopy. Hemoglobin-containing cells were identified by benzidine staining as described [25]. In brief, cells were collected and washed twice with the cold phosphate-buffered saline and then stained with benzidine solution. Benzidine dihydrochloride (2 mg/ml) was prepared in 0.5 M (3%) acetic acid solution and H 2 O 2 (1%) was added immediately before use. The cell suspensions were mixed with the benzidine solution in a 1:1 ratio and incubated for 5 min. Cells with blue-brown-stained cytoplasm were counted as benzidine-staining positive cells and at least 1, 000 cells were counted per sample. The experiments were repeated three times.

Statistical analysis
All data were the statistics of three independent experiments and presented as mean ¡ standard deviation. A Student's t test (two-tail, unpaired) was used to test the difference in two experiment groups. A p value less than 0.05 was considered significance.

ZNF300 is upregulated in K562 cells undergoing megakaryocytic differentiation
Previously, we reported that the ZNF300 protein expression levels correlated to differential stages of leukemic blasts [18]. In addition, ZNF300 expression was upregulated in HL-60 cells undergoing differentiation induced by DMSO. These results suggest that ZNF300 likely plays a role in the pathogenesis of leukemia or blood cell differentiation. To address the potential role of ZNF300 in blood cell differentiation, we chose K562 cells as a model. PMA treatment effectively induced megakaryocytic differentiation in K562 cells. These cells showed typical characters of megakaryocytic differentiation with a marked increase in cell size, extensive multinuclearity, and the presence of vacuoles (Fig. 1A). Megakaryocytic differentiation was also evidenced by a significant increase of CD61 (ITGB3) expression, the differentiation surface marker of megakaryocytes, determined by flow cytometry and quantitative RT-PCR (Fig. 1B,C). The mRNA expression level of CD41 (ITGA2B), another differentiation surface marker of megakaryocytes, was also upregulated (Fig. 1D). More importantly, PMA treatment also significantly upregulated ZNF300 expression at both mRNA and protein levels as shown in Fig. 1E and Fig. 1F compared to the untreated control. These observations suggest that ZNF300 upregulation correlate to megakaryocytic differentiation in K562 cells.

ZNF300 is upregulated in K562 cells undergoing erythrocytic differentiation
To determine whether ZNF300 expression is altered in K562 cells undergoing erythrocytic differentiation, we treated K562 cells with Ara-C as previously reported [26,27]. As shown in Fig. 2A, the K562 cells treated with Ara-C exhibited condensed nuclei and high proportion of nucleus contraction and fragmentation in contrast to untreated control cells. Erythrocytic differentiation was also evidenced by an increase of CD235a, a differentiation surface maker for erythrocytes, measured by flow cytometry (Fig. 2B). In addition, Ara-C treatment also significantly increased the percentage of benzidine-staining positive cells, which measured hemoglobin expression as an endogenous erythrocytic differentiation marker in K562 cells [22,28] (Fig. 2C). The c-globin expression was confirmed at mRNA level (Fig. 2D). Interestingly, we observed upregulation of ZNF300 at both mRNA and protein levels (Fig. 2E, F). These observations suggest that ZNF300 upregulation correlate to erythrocytic differentiation in K562 cells.

ZNF300 knockdown abolishes PMA-induced megakaryocytic differentiation and Ara-C-induced erythrocytic differentiation in K562 cells
To establish the causal-effective relationship between upregulation of ZNF300 and megakaryocyte differentiation, we inhibited ZNF300 expression in K562 cells by short hairpin RNA (shRNA) technique. We designed 5 different shRNAs and subcloned into pLKO.1 vector to make shRNA-expressing vectors (shZNF300). K562 cells were transfected with shZNF300 or control constructs (Ctrl) and selected with puromycin. As shown in S1 Figure, three out of 5 constructs (shZNF300 #2, 3, 4) significantly knocked down ZNF300 expression at both mRNA and protein levels. In addition, ZNF300 knockdown did not significantly alter CD61 or CD235a expression determined by flow cytometry at basal level (data not shown). Thus three cell lines derived from K562 cells stably transfected with three shZNF300 constructs (shZNF300#2, #3, #4) were used for further experiments.
To examine the effect of ZNF300 knockdown on megakaryocytic differentiation in K562 cells, control (Ctrl) or ZNF300 knockdown cells (shZNF300) were treated with PMA. As previously described, control cells exhibited typical characters of megakaryocytic differentiation after PMA induction. In contrast, shZNF300 cells appeared to be insensitive to PMA treatment: these cells looked bright and remained in round shape, suggesting that ZNF300 knockdown diminishes megakaryocyte differentiation (Fig. 3A). Indeed, PMA treatment failed to upregulate CD61 (ITGB3) expression in shZNF300 cells (shZNF300#2, 3, 4) compared to control cells (Ctrl) (Fig. 3B, C). Upregulation of CD41 (ITGA2B) mRNA expression was also eliminated by ZNF300 knockdown (Fig. 3D). Consistently, the shRNAs that failed to efficiently knock down ZNF300 (shZNF300#1 and 5) lost the ability to abrogate megakaryocyte differentiation (data not shown). In addition, PMA also failed to promote megakaryocytic differentiation in HEL cells with ZNF300 knockdown (data not shown). These results suggest that ZNF300 knockdown abrogate megakaryocyte differentiation induced by PMA treatment.
To assess the role of ZNF300 in erythrocytic differentiation induced by Ara-C, control (Ctrl) or ZNF300 knockdown cells (shZNF300) were treated with Ara-C.
As shown in Fig. 4A, Ara-C treatment led to high percentage of benzidinestaining positive cells in control cells (Ctrl). In contrast, benzidine-staining positive cells in ZNF300 knockdown cells (shZNF300) were barely observed (Fig. 4A, B), suggesting that ZNF300 knockdown abrogates erythrocytic differentiation induced by Ara-C. The diminished erythrocytic differentiation in shZNF300 cells was also confirmed by failure to upregulate CD235a and c-globin expression compared to that of control (Ctrl) (Fig. 4C, D). In addition, we measured the cleaved caspase 3 (Fig. 4E). As expected, we barely detected any cleaved caspase 3 in control cells (Ctrl) or ZNF300 knockdown cells without Ara-C treatment unless we overexposed the film as shown in Fig. 4E. With Ara-C treatment, only slight upregulation of cleaved caspase 3 was observed in control cells but not in ZNF300 knockdown cells. These results were consistent to previous reports showing that Ara-C treatment did not induce significant apoptosis [2,29]. These observations suggest that ZNF300 knockdown block erythrocytic differentiation induced by Ara-C without affecting apoptosis.

ZNF300 knockdown promotes cell proliferation in K562 cells
Failure to undergo differentiation frequently accompanies increased proliferation in blood cells. Thus we investigated the effect of ZNF300 knockdown on cell proliferation. We measured cell proliferation by two means. One was to count viable cells and the other was to detect dehydrogenase activity with CCK-8. In two days, the number of viable shZNF300 cells (shZNF300#2, 3, 4) significantly exceeded that of control cells (Ctrl) and the discrepancy was dramatically amplified over time (Fig. 5A). Consistently, the relative absorbance of ZNF300 knockdown cells (shZNF300#2, 3, 4) was higher than that of control cells (Ctrl) (Fig. 5B). In contrast, cells stably transfected with shZNF300#1 and 5 that failed to knock down ZNF300 proliferated normally comparable to that of control cells (data not shown). These observations suggest that ZNF300 knockdown promote cell proliferation in K562 cells. To support this, cell cycle profile analysis demonstrated that shZNF300 cells exhibited increased percentage of cells at S phase. As shown in Fig. 5C and 5D, the percentage of cells at S phase in shZNF300 cells were 40.5%, 40.2%, and 41.4% respectively compared to 20.3% in control cells and the difference was significant. Consistently, cell cycle regulator p15 and p27 was downregulated in shZNF300 cells and the proliferation marker PCNA was upregulated (Fig. 5E). These results suggest that ZNF300 somehow affect cell cycle progress and ZNF300 downregulation lead to increased proliferation.
Sustained MAPK/ERK signaling is essential for megakaryocyte differentiation in K562 cells. We thus examined the phosphorylation of ERK in ZNF300 knockdown cells. We found that the phosphorylation of ERK (P-ERK) was significantly reduced in ZNF300 knockdown cells compared to that in control cells (Fig. 5E). This result was consistent to the phenotype that shZNF300 failed to undergo megakaryocytic differentiation.
Previously, ZNF300 was shown to localize in both cytosol and nucleus [30]. To test whether alteration of ZNF300 subcellular distribution may contribute to the phenotype, we measured the protein level of ZNF300 in both cytosol and nucleus. We found that ZNF300 dominantly localized in cytosol (Cyto) and PMA treatment did not alter the distribution (Fig. 5F).
Taken together, the increased proliferation and impaired MAPK/ERK signaling may contribute to the effect of ZNF300 knockdown on proliferation and differentiation in K562 cells.

Discussion
Previously, ZNF300 was shown to correlate with Crohn's disease and 5qsyndrome [16,31]. Further studies suggest that ZNF300 may play a role in cell proliferation, apoptosis and immune response [12,13]. In this study, we discovered that ZNF300 downregulation abolished forced differentiation in K562 in response to PMA or Ara-C treatment. Our study suggests a novel function of ZNF300 in megakaryocytic and erythrocytic differentiation.
ZNF300 function study has been impeded in part due to its lack of orthologous in mice. In order to study its function, we tried to overexpress ZNF300 in K562 by lentiviral transduction. We failed to obtain any transductants that stably expressed full length ZNF300 (data not shown). This is similar to another research on ZNF268 showing that no transfectants expressing full length ZNF268 could be established in HEK293 cells [32]. Thus knockdown of ZNF300 is the only choice. These observations suggest that KRA-ZFPs may play important roles and have to be tightly regulated. However, how KRAB-ZFPs are regulated is largely unknown. Recent ChIP-Seq data of KRAB-associated protein 1 (KAP1), the most important partner of KRA-ZFPs, showed that KAP1-binding was significantly enriched in the zinc finger region of KRAB-ZFPs [33]. These observations suggest that KRAB-ZFPs may negatively regulate themselves and mediate long-range heterochromatinization. This may partially explain the reason why ZNF300 could not be overexpressed. Further study on the regulation of ZNF300 will significantly help us understand how ZNF300 exerts its function.
ZNF300 may play multiple functions as transcription factor and signaling molecule. As a typical KRAB-ZFPs, ZNF300 protein bears 12 zinc finger motifs. Interestingly, ZNF300 localizes in both cytoplasm and nucleus [34]. In HeLa cells, ZNF300 enhanced NF-kB signaling and promoted tumorigenesis in a xenograft nude mice model [14]. In contrast, ZNF300 knockdown promoted cell proliferation in K562 cells in this study. We speculate that two possibilities may explain the apparent inconsistency. On one hand, the same signaling molecule affected by ZNF300 may play completely opposite functions in different cell types. For instance, MAPK/ERK signaling is activated in various types of carcinoma and supposed to be one of critical signaling pathways for carcinogenesis [35,36]. However, MAPK/ERK is critical for megakaryocyte differentiation in K562 cells. Therefore, the impaired MAPK/ERK may explain the failure to undergo megakaryocyte differentiation in ZNF300 knockdown cells. Comparison of signaling pathway affected by ZNF300 in carcinoma cells and leukemic cells may provide more information. On the other hand, the target genes regulated by ZNF300 may be different in these cells. Although the potential ZNF300 DNAbinding consensus sequence was determined, very few target genes were identified. Further study using microarray or ChIP sequencing may significantly promote study on ZNF300 function.
The increased proliferation may contribute to impaired differentiation phenotype in ZNF300 knockdown cells. ZNF300 knockdown cells showed impaired erythrocytic differentiation by Ara-C and increased proliferation ( Fig. 2 and Fig. 5). Our findings supported a previous study showing that hemoglobin induction by Ara-C in K562 cells was cell-cycle dependent [26]. Our study also support a previous report showing that nuclear receptor co-repressor N-CoR was required for Ara-C-induced erythrocyte differentiation in K562 cells using similar knockdown technique [37]. However, N-CoR seemed not to be required for PMA-induced megakaryocytic differentiation of K562 cells. Given that both KRAB-ZFPs and N-CoR can mediate gene repression at chromosomal level, it is possible that change in chromatin structure may be a common feature of erythrocyte differentiation. Indeed, a recent work showed that developmentally silenced globin genes could be reactivated by forced chromatin looping [38]. Further study on this may facilitate us to understand the erythrocyte differentiation at chromosomal level.
Taken together, we have discovered that ZNF300 may control expression of cell cycle regulators as well as signal transduction in K562 cells that subsequently lead to increased proliferation and resistance to PMA-or Ara-C-induced cell differentiation.