Knockdown of ZNF268, which is transcriptionally downregulated by GATA-1, promotes proliferation of K562 cells

Abstract

The human ZNF268 gene encodes a typical KRAB-C2H2 zinc finger protein that may participate in hematopoiesis and leukemogenesis. A recent microarray study revealed that ZNF268 expression continuously decreases during erythropoiesis. However, the molecular mechanisms underlying regulation of ZNF268 during hematopoiesis are not well understood. Here we found that GATA-1, a master regulator of erythropoiesis, repressed the promoter activity and transcription of ZNF268. Electrophoretic mobility shift assays and chromatin immunoprecipitation assays showed that GATA-1 directly bound to a GATA binding site in the ZNF268 promoter in vitro and in vivo. Knockdown of ZNF268 in K562 erythroleukemia cells with specific siRNA accelerated cellular proliferation, suppressed apoptosis, and reduced expression of erythroid-specific developmental markers. It also promoted growth of subcutaneous K562-derived tumors in nude mice. These results suggest that ZNF268 is a crucial downstream target and effector of GATA-1. They also suggest the downregulation of ZNF268 by GATA-1 is important in promoting the growth and suppressing the differentiation of K562 erythroleukemia cells.

Figure 1. GATA-1 represses ZNF268 promoter activity and transcription.

(A) Western blot analysis of exogenous GATA-1 and GFP expression in transfected HEK293 and HeLa cells using anti-Flag antibody. (B) Luciferase assays in HEK293 and HeLa cells co-transfected with GATA-1 expression plasmid (0.2 µg in 48-well plates) and a luciferase reporter under the control of the ZNF268 promoter. GFP expression plasmid served as a control. (C) Quantitative real-time PCR analysis of ZNF268 mRNA in HEK293 and HeLa cells transfected with plasmid expressing GATA-1 or GFP. GAPDH mRNA was used to normalize ZNF268 expression. Data (mean ± SD) are derived from an average of three independent experiments. *p<0.05 and **p<0.01 (standard t test).

Figure 2. GATA-1 selectively binds to the GATA binding site in the ZNF268 promoter in vitro.

(A) Schematic diagram of the 11 GATA sites (G1–G11) in the ZNF268 promoter. (B) EMSAs using K562 nuclear extract and biotin-labeled probes corresponding to the GATA binding sites in the human ZNF268 promoter. Nuclear extract was omitted from the binding reaction as a negative control. (C) Competitive EMSAs and supershift assays showing the binding of a GATA-1 complex to the G1 site (−1412 to −1388). Labeled wild type G1 probe or labeled mutant probe was added to the reaction (lanes 2 and 3). Unlabeled competitors were added prior to G1 probe addition (lanes 4 and 5). For supershift experiments, anti-GATA-1 antibody was incubated with nuclear extracts before addition to the reaction mixture (lane 6).

Figure 3. GATA-1 binds to the ZNF268 promoter in vivo.

ChIP assays were performed with K562 cells using the indicated antibodies, with IgG serving as a negative control. The precipitated DNA was amplified by PCR, electrophoresed, and stained with ethidium bromide. For all antibodies, primers C-s/C-a (−1166 to −962) served as a negative control. Input lanes show products after PCR amplification and before immunoprecipitation. (A) PCR amplification of DNA precipitated with anti-GATA-1 or anti-FOG antibodies using the primers G1-s/G1-a, which flank the GATA-binding sites contained within −1406 to −1266. (B) PCR amplification of DNA precipitated with anti-RNA polymerase II (pol II) or anti-TFIID antibodies using primers flanking the transcription start site (PES1/PE12). (C) As a positive control, ChIP assays were conducted using anti-CREB-2 antibody and primers flanking the CRE binding site in the ZNF268 promoter (+594 to +925).

Figure 4. Stable silencing of ZNF268 accelerates the proliferation of K562 cells.

K562 cells were transfected with recombinant lentiviral particles containing ZNF268 short hairpin RNA (shRNA; shh-268) or control lentiviral vector (shh-control). Transfected cells were then sorted according to GFP expression to generate stably transfected cell lines. (A) Quantitative real-time PCR analysis of ZNF268 mRNA in ZNF268-silenced and control cells. (B) Cellular proliferation, as determined by cell counting. Values are derived from an average of three independent experiments. (C, D) Cell cycle profiles, as assessed by DNA content in PI-stained cells. (E, F) EdU labeling showing proliferation of ZNF268-silenced and control cells. The percentage of positive cells was derived from triplicate samples. (G) Western blot analysis of c-myc, p53, and cyclin D1. ZNF268 and β-actin levels were also analyzed. *p<0.05 and **p<0.01 (standard t test).

Figure 5. Stable silencing of ZNF268 suppresses apoptosis and promotes tumor formation in nude mice.

(A) Apoptosis in of ZNF268-silenced (shh-268) and control (shh-c) cells, as assessed by FACS analysis of PI or Annexin V staining. **p<0.01 (standard t test). (B, C). ZNF268-silenced or control K562 cells (∼1×10⁷ cells) were subcutaneously implanted into male athymic nude mice. Tumor-bearing mice were then sacrificed 30 days later. Representative mice and excised tumors are shown (B), along with a comparison of tumor weight between the groups (C).

Figure 6. Stable silencing of ZNF268 suppresses erythroid differentiation of K562 cells.

(A, B) Surface expression of the erythroid markers CD71 and glycophorin A in ZNF268-silenced (shh-268) and control (shh-c) cells. Expression was analyzed using PE-conjugated antibodies against human CD71 or glycophorin A. PE-conjugated control IgG served as a negative control. (C) Quantitative real-time PCR analysis of γ-hemoglobin (γ-globin) mRNA in ZNF268-silenced and control cells. Expression of γ-globin was normalized to GAPDH. Data were derived from triplicate samples. **p<0.01 (standard t test).