K562 Nuclear Extracts Research Articles

The Role of p38 MAPK/CREB1 Signaling In γ-Globin Gene Regulation

AbstractAbstract 1012The clinical symptoms of sickle cell disease can be ameliorated by increased fetal hemoglobin (HbF) levels. Previous work from our laboratory demonstrated that Trichostatin A (TSA) and sodium butyrate (NaB) activate γ-globin expression via p38 MAPK signaling. In addition, cAMP response element binding protein 1 (CREB1) was shown to trans-activate the -1222 Gγ-globin cAMP response element (G-CRE) in a transient assay system. To study the role of p38 MAPK signaling in γ-globin regulation, loss of function siRNA studies were performed in K562 cells. siRNA-mediated knockdown of p38 MAPK resulted in 72% loss of γ-globin transcription. Furthermore, enforced stable expression of MKK3/6 increased the phosphorylated form of p38 MAPK by 70%, which in turn produced a 2- to 3-fold increase in γ-globin mRNA and HbF levels. Likewise, siCREB1 treatment reduced CREB1 levels by 62% and down regulated γ-globin expression 59%. In the same vein, stable expression of recombinant CREB1 activated HbF by 2-fold. These findings were subsequently confirmed in human primary erythroid cells grown in a two-phase liquid culture system. On day 11, we observed 50–70% γ-globin silencing after CREB1 and p38 MAPK siRNA knockdown with 60% target gene silencing. CREB1 enforced expression trans-activated γ-globin 4.5-fold which was accompanied by a 90% increase in HbF-FITC positive cells and HbF levels. Collectively, these data demonstrate that p38 MAPK and CREB1 are required for steady-state γ-globin gene transcription. To determine the role of the G-CRE in γ-globin regulation, the Gγ-globin promoter (-1500 to +36) was cloned into pGL4.17 Luc2/neo to produce pGγLuc2 (wild type) and mutant plasmids: -1225 G/A (m1), -1227 AC/TG (m2) and a scrambled G-CRE (m3s). Five K562 stable lines including KLuc2 (promoterless), KGγLuc2, KGγLuc2(m1), KGγLuc2(m2) and KGγLuc2(m3s) were established. Luciferase activity in KGγLuc2 was 1000-fold higher than in the control Kluc2 line; all mutations produced >90% loss of luciferase activity and a loss of γ-globin trans-activation by TSA and NaB. Next, siRNA studies were completed to determine if the G-CRE is required for γ-globin activation. A dose-dependent loss of promoter activity was observed after p38 MAPK and CREB1 siRNA knockdown of the KGγLuc2 cell line; however, promoter silencing was not observed in the mutant lines supporting a role for the G-CRE in p38 MAPK/CREB1 mediated γ-globin regulation. To study in vivo binding, chromatin immunoprecipitation (ChIP) assays were performed with CREB1 antibody in the KGγLuc2 stable line. We observed comparable 2- to 3-fold chromatin enrichment with CREB1 compared to the control IgG in the G-CRE regions of the pGγLuc2 plasmid and endogenous Gγ-globin promoter. To determine if an enhanceosome complex is bound to the G-CRE, we performed affinity column pull-down/mass spectrometry analysis. K562 nuclear extract was purified on a Heparin Sepharose column, following which fractions eluting at 0.6M NaCl showing peak gel shift binding activities with the G-CRE oligo were loaded into a size selecting Suprose 6 gel exclusion column. G-CRE eluting fractions were then identified by protein microsequencing (MS/MS). We identified CREB1, ATF2, c-Jun, BRG-1, hnRNPC1/C2, and the TCP-1 complex as major components. To determine protein co-localization, promoter pull-down assays were performed using biotinylated wild type and mutant (AC/TG) G-CRE probes and K562 nuclear extracts. We observed simultaneous CREB1, ATF-2 and cJun binding to the G-CRE which was abolished in the mutant probe. However, Brg1 was bound after NaB (2mM) induction. Subsequent co-IP studies showed interactions between ATF-2 and Brg1, CREB1, cJun, and hnRNPC1C2, which was further confirmed by co-elution profile of these molecules observed by sucrose gradient centrifugation, thus implying association as one complex. These data support complex protein-protein interactions in the G-CRE, which modulate γ-globin gene expression. Additional studies will be performed in primary erythroid cells using siRNA-based gene silencing and ChIP assays to determine novel mechanisms of γ-globin regulation and to define in vivo binding of proteins identified in the G-CRE enhanceosome complex. Disclosures:No relevant conflicts of interest to declare.

Transcriptional Regulation of γ-Globin Gene Expression by KLF4

AbstractAbstract 645Kruppel-like factors (KLFs) are a family of Cys2His2 zinc-finger DNA binding proteins that regulate gene expression through CACCC/GC/GT box binding in various gene promoters. The CACCC element is also critical for developmental regulation of the human γ-globin and β-globin genes; therefore studies to identify transcription factors that bind the CACCC element to alter gene expression are desirable. By microarray-based gene profiling, we identified two Kruppel-like factors, KLF4 and KLF12 whose expression levels decreased simultaneously with γ-globin silencing during in vitro erythroid maturation. Subsequent reverse transcription quantitative PCR (RT-qPCR) analysis confirmed KLF4 and KLF12 mRNA levels decreased 56-fold and 16-fold respectively in erythroid progenitors from day 7 to day 28 with over 90% γ-globin gene silencing. The effects of known fetal hemoglobin inducers hemin (50μM) and sodium butyrate (2mM) on KLF factor expression was tested in K562 cells. Hemin and sodium butyrate increased KLF4 3-fold (p<0.05) and 13-fold (p<0.01) respectively while KLF12 was only induced by butyrate. Likewise, hemin treatment of KU812 leukemia cells, which actively express γ-globin and β-globin, produced a 7-fold increase in KLF4 (p<0.05) while KLF12 levels were not changed suggesting KLF4 may be directly involved in γ-globin gene regulation. To characterize its role further siRNA-mediated loss of function studies were performed in K562 cells. A 60% knockdown of KLF4 expression produced 40% attenuation of γ-globin transcription (p<0.05). To confirm this effect, rescue experiments were performed as follows: K562 cells were treated with 100nM siKLF4 alone or in combination with the pMT3-KLF4 expression vector (10 and 20μg) for 48 hrs. The 40% knockdown of γ-globin expression produced by siKLF4 was rescued to baseline levels after enforced pMT3-KLF4 expression (p<0.05). To establish whether KLF4 directly stimulates γ-globin promoter activity, we performed co-transfection of pMT3-KLF4 and the Gγ-promoter (-1500 to +36) cloned into the pGL4.17 Luc2/neo vector; a dose-dependent increase in luciferase activity (2- to 5-fold; p<0.001) was observed. Furthermore, enforced expression of pMT3-KLF4 augmented endogenous γ-globin expression 2-fold (p<0.01). Collectively, these studies suggest that KLF4 acts as a trans-activator of γ-globin gene transcription. To address the physiological relevance of these findings, studies were extended to human primary erythroid cells grown in a two-phase liquid culture system. At day 11 when γ-globin gene expression was maximal, siKLF4 treatment produced a 60% decrease in γ/β-globin mRNA levels (p<0.001). By contrast, enforced pMT3-KLF4 expression enhanced γ/β-globin 1.5-fold at day 11 and day 28 (after γ-globin silencing); HbF levels were induced 1.5-fold (p<0.05) which was demonstrated by enzyme-linked immunosorbent assay. To gain insights into the molecular mechanism of KLF4-mediated γ-globin regulation, electrophoretic mobility shift assay (EMSA) and chromatin immunoprecipitation assay (ChIP) were completed. Since CREB binding protein (CBP) is known to function as a co-activator for KLF1, 4 and 13, we also tested its role in γ-globin gene regulation. EMSA performed with K562 nuclear extract and a [γ-32P] labeled γ-CACC probe (-155 to -132 relative to the γ-globin cap site) produced three DNA-protein complexes; the addition of KLF4 or CBP antibody resulted in a marked decrease in intensity of all complexes suggesting these factors bind the γ-CACC element. ChIP assay demonstrated 10-fold and 20-fold chromatin enrichment with KLF4 and CBP antibody respectively (p<0.001) confirming in vivo binding at the γ-CACC region. Lastly, co-immunoprecipitation established protein-protein interaction between KLF4 and CBP in K562 cells. Future studies will investigate the role of CBP in KLF4-mediated γ-globin regulation which will provide molecular targets for fetal hemoglobin induction and treatment of sickle cell anemia and β-thalassemia. Disclosures:No relevant conflicts of interest to declare.

Characterization of a Functional ZBP-89 Binding Site That Mediates Gata1 Gene Expression during Hematopoietic Development

GATA-1 is a lineage-restricted transcription factor that plays essential roles in hematopoietic development. The Gata1 gene hematopoietic enhancer allowed Gata1 reporter expression in erythroid cells and megakaryocytes of transgenic mice. The Gata1 hematopoietic enhancer activity is strictly dependent on a GATA site located in the 5' region of the enhancer. However, the importance of the GC-rich region adjacent to the 3'-end of this GATA site has been also suggested. In this study, we show that this GC-rich region contains five contiguous deoxyguanosine residues (G(5) string) that are bound by multiple nuclear proteins. Interestingly, deletion of one deoxyguanosine residue from the G(5) string (G(4) mutant) specifically eliminates binding to ZBP-89, a Krüppel-like transcription factor, but not to Sp3 and other binding factors. We demonstrate that GATA-1 and ZBP-89 occupy chromatin regions of the Gata1 enhancer and physically associate in vitro through zinc finger domains. Gel mobility shift assays and DNA affinity precipitation assays suggest that binding of ZBP-89 to this region is reduced in the absence of GATA-1 binding to the G1HE. Luciferase reporter assays demonstrate that ZBP-89 activates the Gata1 enhancer depending on the G(5) string sequence. Finally, transgenic mouse studies reveal that the G(4) mutation significantly reduced the reporter activity of the Gata1 hematopoietic regulatory domain encompassing an 8.5-kbp region of the Gata1 gene. These data provide compelling evidence that the G(5) string is necessary for Gata1 gene expression in vivo and ZBP-89 is the functional trans-acting factor for this cis-acting region.

Recruitment of Rad50 and MCM Complexes on the Human Beta Globin Locus Control Region (LCR): A Novel Role for LCR Hypersensitive Site 4 (HS4).

The human beta globin locus control region (LCR) aids in the establishment of an open chromatin configuration thereby facilitating the expression of linked downstream globin genes in a temporal and spatial manner along the developmental ontogeny. The LCR also functions as a classical enhancer of the beta like globin genes. The four DNAase-I hypersensitive regions (HS 1–4) on the LCR are believed to act as nucleation centers for the assembly of multi-protein transcriptional complexes that interact with the distant gene promoters. Present work is focused on characterizing the protein complex assemblies on the HS4 region and delineating their function in the context of goblin gene regulation. Our finding shows an assembly of a 2–5 MDa complex on the HS4 locus that is both erythroid restricted and sequence specific. Electrophoretic mobility shift assays (EMSA) were performed using K562 and HeLa nuclear extracts and ∼35mer double stranded oligonucleotides with 7 bp overlaps tiling the entire core HS4 sequence. We detected prominent K562 specific bands formed with an AT rich 39 bp HS4 sequence. Sequence specificity of DNA-protein complex formation on this AT rich sequence was determined by competitive EMSA assays. Gel supershift assays with a wide range of antibodies and subsequent partial purification of this multiprotein complex followed by the LC-MS/MS analysis reveal the presence of a mini chromosome maintenance (MCM) complex that is associated with the DNA replication, DNA repair associated Rad50-Mre-11-NBS-1 complex, histone acetyl transferase p300 and DNA binding transcription factors ILF2 (NF45) and ILF3 (NF90). The association of these proteins was determined by co-immunoprecipitation experiments. Chromatin immunoprecipitation (ChIP) experiments demonstrate the in vivo recruitment of these proteins on the HS4 region of the LCR. We are currently working on further characterization and functional studies of this complex and its significance on the role of the LCR in DNA replication and repair.

Characterization of Combinatorial Stat3/GATA-1 Complexes in the γ-Globin 5′UTR.

Regulation of human γ-globin expression depends on localized combinatorial interactions of trans-factors with cis-elements in the proximal promoter however, the full complement of proteins involved in γ-gene regulation remains obscure. We previously demonstrated the ability of Stat3 α to bind a +9 site in the Aγ-globin 5′untranslated region (Aγ5′UTR) to activate transcription. To expand on this observation we identified an overlapping Stat3/GATA-1 site (GAGATTATCAA) at nucleotide +21γ. Recently Stat3 and GATA-1 were shown to participate in protein-protein interactions to regulate cell differentiation. Therefore we hypothesized that GATA-1 and Stat3 may regulate γ-gene transcription through a similar mechanism. Initial chromatin immunoprecipitation (ChIP) assays were performed to demonstrate in vivo GATA-1 and Stat3 binding to the γ-promoter. ChIP assay for the Aγ-globin region between −122 to +65 produced 9-fold and 4-fold chromatin enrichment in K562 cells with GATA-1 and Stat3α antibodies respectively compared to no enrichment with IgG. Treatment with Interleukin-6 mediated a 12-fold increase in bound Stat3α and a concomitant 2-fold decrease in GATA-1 interactions. TFIID binding increased from 25-fold to 40-fold supporting transcriptional activation. To provide in vitro evidence for competitive binding in the +21γsite, electrophoretic mobility shift assays (EMSA) were performed with K562 nuclear extracts and 32P- labeled γ41 (+1 to +41), +21γ, and Stat3 and GATA-1 consensus probes. Four specific DNA-protein complexes were established with γ41. Stat3 and GATA-1 antibody studies with the four probes demonstrated that Stat3 dimers and GATA-1 dimers were bound in the Band 1 and Band 4 complexes respectively; in contrast Stat3/GATA-1 heterodimers were present in the Band 3 complex. Studies with the +21γ probe confirmed our findings with γ41. The data are consistent with our hypothesis that GATA-1 and Stat3 compete for binding to the +21γ cis-element. To further our understanding of the in vivo consequences of GATA-1/Stat3 interactions, preliminary transient transfections were performed in K562 cells. The wild type -210γLuc reporter containing the −210 to +36 γ-promoter (10μg) and this mutant −201γ(+21mStat3)Luc reporter were co-transfected individually with β-galactosidase. The luciferase activity of −201γLuc increased 110-fold after enforced pZeoStat3α (30μg) expression which dropped 75-fold when tested with the mutant Stat3 reporter. Residual activity was attributed to the intact +9γ Stat3 site. Enforced pGFP-GATA-1 expression activated the −201γLuc reporter 35-fold which increased further to 70-fold for the mutant reporter. Additional cotransfection studies will be completed to determine whether competitive interactions of Stat3 and GATA-1 occur to regulate γ-globin transcription in vivo. To confirm and verify these observations, vitro competitive binding utilizing purified Stat3 and GATA-1 proteins will be performed. 2-Dimensional EMSA coupled with mass spectroscopy will be used for comprehensive identification of the protein components of complexes bound to the Aγ5′UTR.

G125A Single Nucleotide Polymorphism (SNP) in 5′-UTR of the Tumor Suppressor Gene Bax Affects Its Transcriptional Regulation.

Lower expression of Bax protein in various human malignancies is associated with poor response to treatment and shorter disease-free survival. We (Cancer Lett 2002; 187:199–205) and others (J Clin Oncol ; 23:1514–21) have shown the association of a single nucleotide polymorphism (SNP) in the BAX promoter (G125A) with reduced protein expression and treatment resistance in chronic lymphocytic leukemia (CLL). Using luciferase reporter gene assay we demonstrated that this SNP significantly reduced BAX promoter activity (Oncogene 2005; 24:2042–9). Our aim was to determine the effect of this polymorphism on the binding of transcription factors. For electrophoretic mobility shift (EMSA), HeLa and K562 nuclear extracts, and for chromatin immunoprecipitation (ChIP), K562 cells were used to study their ability to bind to a radiolabeled DNA probe corresponding to the BAX promoter region with G nucleotide at the position 125 or bearing G125A SNP. Super-shift assay was performed to determine the transcription factor involved in binding. Competition assays were performed to determine differences in the binding ability of the two probes. A panel of antibodies was tested by super-shift and ChIP assays, non-specific antibodies served as negative controls. Two major band shifts were detected by EMSA. The mobility of the detected complexes was different from those observed with GC1 probe, specific for Sp1/Sp3, suggesting the involvement of transcription factors other than Sp1/ Sp3. This was confirmed in super-shift assay and ChIP assay by incubating DNA probes with Sp1 or/ and Sp3 antibodies. We also found in the competition assays that the cold probes competed differently for binding. The findings provide evidence of the ability of G125A SNP to influence transcription factor binding in vitro (as shown in EMSA experiments) and in vivo (in ChIP experiments).

GATA-1 Binding to Exon 1 Directs High Level Erythroid-Specific Expression of the Human Protoporphyrinogen Oxidase Gene.

Protoporphyrinogen oxidase (PPOX), the penultimate enzyme in the heme biosynthetic pathway, catalyzes the six-electron oxidation of protoporphyrinogen IX to protoporphyrin IX. Like other heme biosynthetic proteins, PPOX is involved in synthesizing heme for red cells (erythroid-specific expression) and as a cofactor for the respiratory cytochromes (housekeeping expression). To date, little is known about transcriptional regulation of the human PPOX gene (PPOX). We established the molecular basis for erythroid-specific expression of PPOX. Using transient in vitro transfection assays in human erythroleukemic K562 cells we studied tissue-specific expression of PPOX. We found that reporter constructs lacking exon 1 showed a 75% reduction in promoter strength in K562 cells (Figure, no 1 and 2). Hence, in vitro high-level erythroid-specific expression of PPOX is dependent on the presence of exon 1. Examination of erythroid-specific regulatory elements in exon 1 revealed two GATA-1 sites, one consensus (A/T)GATA(A/G) site (GATA-1_II AGATAA) and one non-consensus site, deviating at the first nucleotide (GATA-1_I, CGATAG). To study the relative contribution of these two GATA-1 sites to erythroid-specific transcriptional regulation, we performed in vitro transfections of wild-type and mutant (GATA → GTTA) reporter plasmids in K562 cells. We found that the highest level of transcription depended on the integrity of both sites (Figure, no 5). The consensus GATA-1_II site contributed the most to promoter strength (Figure, no 4). Subsequent electrophoretic mobility shift assay and supershift experiments using K562 nuclear extracts demonstrated that both GATA sites were able to bind GATA-1 in vitro. Our experiments showed that exon 1 was dispensable for PPOX promoter activity in human hepatoma HepG2 cells. Interestingly, in HeLa human cervical carcinoma cells the presence of exon 1 decreased promoter activity. Conclusively, exon 1 of the human PPOX gene contains two GATA-1 binding sites, which are required for high level erythroid-specific expression of PPOX and, in addition, bind GATA-1 in vitro. Our results contribute to a better understanding of the molecular mechanisms involved in differential regulation of the human PPOX promoter in erythroid and non-erythroid cells. [Display omitted]

In Vivo and In Vitro Analysis of an Aγ-Globin Gene Silencer in Human β-Yac Transgenic Mice.

Autonomous silencing of gene expression is one mechanism operative in the control of human β-like globin gene switching and is best exemplified by the ε-globin gene. Experiments using variously truncated Aγ-globin genes linked to LCR sequences suggested that a region of the Aγ-globin gene between −730 to −378 relative to the mRNA CAP site may function as an adult stage-specific silencer element. A 5.4 Kb marked Aγ-globin gene (Aγm) inserted between LCR 5′HS1 and the ε-globin gene in a β-YAC (Aγm 5′ ε β-YAC) was silenced in transgenic mice during adult definitive erythropoiesis, even in the absence of an adult β-globin gene. In contrast, when a marked β-globin gene (βm) was inserted in this same location in another β-YAC (βm 5′ ε β-YAC), the βm-globin gene was expressed throughout ontogeny. From these data we concluded that: 1) any gene located near the LCR will be strongly expressed throughout ontogeny, unless some gene-specific silencing mechanism exists, 2) competition between the γ- and β-globin genes for interaction with the LCR is not the exclusive mechanism controlling γ- to β-globin gene switching, and 3) that the Aγm-globin gene was autonomously silenced. A -730 to -378 deletion of the Aγm-globin gene was introduced into the Aγm 5′ ε β-YAC via homologous recombination to produce a Δ1s Aγm 5′ ε β-YAC. This YAC was microinjected and six founders were obtained. Four transgenic lines were established carrying at least one full-length β-globin locus and two were established that lacked the adult β-globin gene. All founders containing an intact β-globin gene expressed the Δ1s Aγm-globin during adult erythropoiesis (45% – 122% relative to human β-globin expression). In one line examined in detail, the Δ1s Aγm-globin gene was expressed in the embryonic yolk sac, fetal liver, and adult blood. ε-globin gene expression was not detected in the embryonic yolk sac and expression of the normally located γ-globin genes was not observed at any developmental stage. β-globin gene expression was observed in the fetal liver and adult blood, although its expression was decreased. To further delineate the function of the Δ1s fragment, transient transfection assays to test silencer function and protein-DNA interaction assays were performed. Silencer activity of the352 bp Δ1s fragment was examined using a series of pGL2 luciferase reporter plasmids that were synthesized to include the Δ1s fragment; these were electroporated into various cells. Electrophoretic mobility shift assays (EMSAs) and DNAse I footprinting were employed to begin assessment of protein binding within the Δ1s fragment. A 50 bp DNA fragment spanning −713 to −664 of the Δ1s element was used in EMSAs; DNA binding activity was observed in K562 nuclear extracts. These preliminary data suggest that the −730 to −378 γ-globin gene silencer binds a repressor protein complex.

The -Globin mRNA Stabilization Complex Is Assembled in the Nucleus Supporting CP Integration of Nuclear and Cytoplasmic Controls over -Globin Gene Expression. ααα.

Stability of human -globin mRNA in erythroid cells is critically dependent on the function of a 3UTR RNA-protein (RNP) complex ( complex ). This complex is assembled by the high-affinity binding of a polyC-specific RNA binding protein (CP) to a defined C-rich segment in the 3UTR. Recently we observed that certain CP protein isoforms contain nuclear localization signals, co-localize with nuclear speckles, and shuttle between nuclear and cytoplasmic compartments. These data suggested that the -complex may initially assemble on nascent nuclear transcripts and may have nuclear as well as cytoplasmic functions. To test this model, we immuno-enriched CP-RNP complexes from K562 nuclear extracts and probed these complexes for -globin transcripts. We found that these CP-containing nuclear RNP complexes were enriched for intron-containing -globin transcripts as well as mature, spliced -globin mRNAs. The association of mature -globin mRNA in the complex was entirely dependent on the presence of the 3UTR CP binding site. In contrast, the unspliced -globin transcripts remained associated with CP complexes in the absence of the 3 UTR binding site. These data suggested that one or more previously unrecognized CP binding sites are present within and specific to unspliced -globin transcripts. Visual inspection of -globin intronic sequences revealed two prominent C-rich elements. These sequences are in an open secondary structure and conform to established parameters for high-affinity CP binding sites. RNA EMSA confirmed two high-affinity CP binding sites within intron 1; the stronger site was coincident with the pyrimidine-rich tract of the splice-acceptor site. In vitro splicing assays were used to explore nuclear functions of the a complexes. Incubation of full-length -globin transcripts with HeLa cell nuclear extract resulted in the efficient removal of both introns. Elimination of the 3UTR complex slowed removal of intron 1 but accelerated intron 2 excision. The impact of the intron-1 complexes on splicing was also studied. Incubation of an -globin transcript segment extending from exon 1 through intron 1 and into exon 2 resulted in efficient intron excision. Affinity-depletion of CP from the HeLa extract enhanced intron 1 excision from this substrate. This splicing enhancement was reversed by the addition of recombinant CP to the depleted extract. In summary, we report that assembly of the complexes on -globin transcripts takes place in the nucleus, that these RNP complexes assemble at sites within intron 1 as well as at the previously described 3UTR binding site, and that the both sets of a complexes have the potential to alter the efficiency of -globin intron excision. The fully processed -globin mRNA, containing the remaining 3UTR complex, is then exported to the cytoplasm in a pre-packaged form where it is stabilized by the UTR complex and is robustly translated. We propose that CP mediates and integrates nuclear as well as cytoplasmic post-transcriptional controls over human -globin gene expression.α′'α' αα′αααααααα′ααα′αααααα′ααααααα′αα′α“”′ααα

Disruption of a novel regulatory element in the erythroid-specific promoter of the human PKLR gene causes severe pyruvate kinase deficiency

We established the molecular basis for pyruvate kinase (PK) deficiency in a white male patient with severe nonspherocytic hemolytic anemia. The paternal allele exhibited the common PKLR cDNA sequence (c.) 1529G>A mutation, known to be associated with PK deficiency. On the maternal allele, 3 in cis mutations were identified in the erythroid-specific promoter region of the gene: one deletion of thymine -248 and 2 single nucleotide substitutions, nucleotide (nt) -324T>A and nt -83G>C. Analysis of the patient's RNA demonstrated the presence of only the 1529A allele, indicating severely reduced transcription from the allele linked to the mutated promoter region. Transfection of promoter constructs into erythroleukemic K562 cells showed that the most upstream -324T>A and -248delT mutations were nonfunctional polymorphisms. In contrast, the -83G>C mutation strongly reduced promoter activity. Site-directed mutagenesis of the promoter region revealed the presence of a putative regulatory element (PKR-RE1) whose core binding motif, CTCTG, is located between nt -87 and nt -83. Electrophoretic mobility shift assay using K562 nuclear extracts indicated binding of an as-yet-unidentified trans-acting factor. This novel element mediates the effects of factors necessary for regulation of pyruvate kinase gene expression during red cell differentiation and maturation.

Cooperation and competition between the binding of COUP-TFII and NF-Y on human epsilon- and gamma-globin gene promoters.

The nuclear receptor COUP-TFII was recently shown to bind to the promoter of the epsilon- and gamma-globin genes and was identified as the nuclear factor NF-E3. Transgenic experiments and genetic evidence from humans affected with hereditary persistence of fetal hemoglobin suggest that NF-E3 may be a repressor of adult epsilon and gamma expression. We show that, on the epsilon-promoter, recombinant COUP-TFII binds to two sites, the more downstream of which overlaps with an NF-Y binding CCAAT box. Binding occurs efficiently to either the 5' or the 3' COUP-TFII site but not to both sites simultaneously. However, adding recombinant NF-Y induces the formation of a stable COUP-TFII.NF-Y-promoter complex at concentrations of COUP-TFII that would not give significant binding in the absence of NF-Y. Mutations of the promoter indicate that COUP-TFII cooperates with NF-Y when bound to the 5' site, whereas binding at the 3' site is mutually exclusive. Likewise, in the gamma-promoter, COUP-TFII binds to a site overlapping the distal member of a duplicated CCAAT box, competing with NF-Y binding. Transfections in K562 cells show that both the mutation of the 5' COUP-TFII or of the NF-Y site on the epsilon-promoter decrease the activity of a luciferase reporter; the mutation of the 3' COUP-TFII site has little effect. These results, together with transgenic experiments suggesting a repressive activity of COUP-TFII on the epsilon-promoter and the observation that, on the 3' site, COUP-TFII and NF-Y binding is mutually exclusive, suggest that COUP-TFII may exert different effects on epsilon transcription depending on whether it binds to the 5' or to the 3' site. At the 5' site, COUP-TFII might cooperate with NF-Y, forming a stable complex, and stimulate transcription; at the 3' site, COUP-TFII might compete for binding with NF-Y and, directly or indirectly, decrease gene activity.

CrkL functions as a nuclear adaptor and transcriptional activator in Bcr-Abl–expressing cells

Objective To identify tyrosine phosphorylated proteins that interact with CrkL in Bcr-Abl–expressing cells and analyze the function of that association. Materials and Methods Immunoprecipitation of CrkL was performed on lysates from parental cells (Rat-1, MO7e, or 32D) or Bcr-Abl–expressing cells (Rat-1p185, MO7p210, 32Dp210, K562) followed by immunoblotting for pTyr, Stat5, or CrkL. Interactions were confirmed in vitro using GST-CrkL fusion proteins. Electrophoretic mobility shift assays were performed on K562 nuclear extracts using a β-casein promoter-derived probe. Supershift analysis was performed with CrkL, Stat5, Stat1, Grb2, and peptide-blocked CrkL and Stat5 antibodies. CrkL localization in Rat-1 and Rat-1p185 cells was detected with indirect immunofluorescence. Transcriptional activation was analyzed in COS7 cells transfected with a Stat-responsive luciferase reporter construct and Bcr-Abl, kinase-defective Bcr-Abl, CrkL, or Grb2. Results We show that, in Bcr-Abl–expressing cells, CrkL interacts with tyrosine phosphorylated Stat5. Additionally, in the presence of Bcr-Abl, CrkL is found in the nucleus, can be detected in a Stat5/DNA complex, and increases transcriptional activation from a Stat-responsive reporter construct. Conclusion This suggests a novel role for CrkL, functioning as a nuclear adaptor protein that can associate with and activate Stat proteins in Bcr-Abl–expressing cells.

DNA-Protein Interactions in the Proximal ζ-Globin Promoter: Identification of Novel CCACCC- and CCAAT-binding Proteins

ABSTRACT: The human ζ-globin gene is expressed in a tissue- and developmental-specific pattern, with expression confined to primitive erythroid cells of the embryonic yolk sac blood islands. Transgenic mouse studies have shown that the proximal ζ-globin promoter contains sequences that contribute to the stage-specificity of expression, but no systematic functional studies of theciselements in the proximal ζ-globin promoter have been reported. In this paper, we show that a number of conserved sequence elements in the ζ-globin promoter are important for promoter activity in transiently transfected K562 erythroleukemia cells, which constitutively express ζ-globin. These include a GATA site at -105, a CCACC site at -93, a CCAAT box at -65, and a TATA box at -29. A highly conserved CCTCC sequence at -78 is not important for ζ-globin promoter activity in this system. Mutations at these sites do not result in increased promoter activity in OCIM1 cells, an erythroid line that does not express ζ-globin, suggesting none of these sites is a developmental silencer. Electrophoretic mobility shift assays show that K562 and OCIM1 nuclear extracts contain DNA-binding activities that interact with the -105 GATA, -65 CCAAT, and -29 TATA sites. In addition K562 cells, but not OCIM1 cells, have an activity that binds the -93 CCACC site. GATA-1 interacts with the GATA site. The K562 CCACC-binding protein is distinct from Sp1, Sp2, Sp3, Sp4, EKLF, and BKLF. A specific -65 CCAAT-binding activity is present in K562 and OCIM1 nuclear extracts that is distinct from other CCAAT-binding proteins including CBF/NF-Y, C/EBP, NF-1, and CP2. Thus, we have identified two novel factors that may contribute to the tissue or developmental stage-specific expression of ζ-globin.

Transactivation of an Intronic Hematopoietic-specific Enhancer of the Human Wilms' Tumor 1 Gene by GATA-1 and c-Myb

The Wilms' tumor 1 gene (WT1) encodes a zinc-finger transcription factor which is expressed in a tissue-specific manner. Our studies indicate that in addition to the promoter, other regulatory elements are required for tissue-specific expression of this gene. A 258-base pair hematopoietic specific enhancer in intron 3 of the WT1 gene increased the transcriptional activity of the WT1 promoter by 8-10-fold in K562 and HL60 cells. Sequence analysis revealed both a GATA and a c-Myb motif in the enhancer fragment. Mutation of the GATA motif decreased the enhancer activity by 60% in K562 cells. Electrophoretic mobility shift assays showed that the GATA-1 protein in K562 nuclear extracts binds to this motif. Cotransfection of the enhancer containing reporter construct with a GATA-1 expression vector showed that GATA-1 transactivated this enhancer, increasing the CAT reporter activity 10-15-fold. Similar analysis of the c-Myb motif by cotransfection with the enhancer CAT reporter construct and a c-Myb expression vector showed that c-Myb transactivated the enhancer by 5-fold. A DNase I-hypersensitive site has also been mapped in the 258-base pair enhancer region. These data suggest that GATA-1 and c-Myb are responsible for the activity of this enhancer in hematopoietic cells and may bind to the enhancer in vivo.

Conserved E Boxes Function as Part of the Enhancer in Hypersensitive Site 2 of the β-Globin Locus Control Region: ROLE OF BASIC HELIX-LOOP-HELIX PROTEINS

The human beta-globin gene cluster is regulated in part by a distal locus control region that is required for opening a chromatin domain in erythroid cells and enhancing expression of the beta-like globin genes at the correct developmental stages. One part of the locus control region, called hypersensitive site 2 (HS2), functions as a strong enhancer. Matches to the consensus binding sites for basic helix-loop-helix (bHLH) proteins (E boxes) are well conserved within the HS2 core. We show that mutations of the HS2 core that alter an invariant E box cause a 3.5-fold reduction in enhancement of expression of an epsilon-globin reporter gene in transiently transfected K562 cells, both before and after induction. Mutations of the HS2 core that alter a less-highly conserved E box cause a more modest reduction in enhancement. Footprint analysis shows binding of erythroid nuclear proteins in vitro to the invariant E box as well as an adjacent CAC/GTG box. Probes containing the E box regions form sequence-specific complexes with proteins from both K562 and MEL nuclear extracts; these are disrupted by the same mutations that decrease enhancement. Some of these latter complexes contain known bHLH proteins, as revealed by specific loss of individual complexes when treated with antibodies against TAL1 and USF. Interaction between the E boxes and the bHLH proteins, as well as other binding proteins, could account for the role of these sites in enhancement by HS2.

The transcriptional regulator YY1 binds to the 5'-terminal region of B19 parvovirus and regulates P6 promoter activity

We performed a systematic study to identify cellular factors that bound to the terminal repeat region of B19 parvovirus. Using electrophoretic mobility shift assays, we detected one cellular factor which prominently bound to the repeat region. The factor was purified from K562 nuclear extract by specific DNA affinity column chromatography and identified as YY1, a multifunctional transcription factor. Of multiple possible YY1 binding sites in the upstream region of the P6 promoter, three showed specific strong binding. Transcription enhancement by YY1 was demonstrated in vitro by transient transfection assays. In studies using truncated and mutated versions of this region, YY1 activity was diminished by the alteration of any two of these three sites and abolished by the alteration of all three sites. Our results suggest that YY1 is a positive regulator of B19 parvovirus transcription.

The 87-kD Aγ-Globin Enhancer-Binding Protein Is a Product of the HOXB2(HOX2H) Locus

AbstractDevelopmental regulation of globin gene expression may be controlled by developmental stage-specific nuclear proteins that influence interactions between the locus control region and local regulatory sequences near individual globin genes. We previously isolated an 87-kD nuclear protein from K562 cells that bound to DNA sequences in the β-globin locus control region, γ-globin promoter, and Aγ-globin enhancer. The presence of this protein in fetal globin-expressing cells and its absence in adult globin-expressing cells suggested that it may be a developmental stage-specific factor. A λ gt11 K562 cDNA clone encoding a portion of the HOXB2 (formerly H0X2H) homeobox gene was isolated on the basis of the ability of its β-galactosidase fusion protein to bind to the same DNA sequences as the 87-kD K562 protein. Because no other relationship had been established between the 87-kD K562 protein and the HOXB2 protein other than their ability to bind to the same DNA sequences, we have investigated whether the two proteins are related antigenically. Our data show that antisera produced against the HOXB2-β-gal fusion protein and a synthetic HOXB2 decapeptide react specifically with an 87-kD protein from K562 nuclear extract, showing that the 87-kD K562 nuclear protein is a product of the HOXB2 locus, and is the first demonstration of cellular H0XB2 protein.

The 87-kD A gamma-globin enhancer-binding protein is a product of the HOXB2(HOX2H) locus

Developmental regulation of globin gene expression may be controlled by developmental stage-specific nuclear proteins that influence interactions between the locus control region and local regulatory sequences near individual globin genes. We previously isolated an 87-kD nuclear protein from K562 cells that bound to DNA sequences in the beta- globin locus control region, gamma-globin promoter, and A gamma-globin enhancer. The presence of this protein in fetal globin-expressing cells and its absence in adult globin-expressing cells suggested that it may be a developmental stage-specific factor. A lambda gt11 K562 cDNA clone encoding a portion of the HOXB2 (formerly HOX2H) homeobox gene was isolated on the basis of the ability of its beta-galactosidase fusion protein to bind to the same DNA sequences as the 87-kD K562 protein. Because no other relationship had been established between the 87-kD K562 protein and the HOXB2 protein other than their ability to bind ot the same DNA sequences, we have investigated whether the two proteins are related antigenically. Our data show that antisera produced against the HOXB2-beta-gal fusion protein and a synthetic HOXB2 decapeptide react specifically with an 87-kD protein from K562 nuclear extract, showing that the 87-kD K562 nuclear protein is a product of the HOXB2 locus, and is the first demonstration of cellular HOXB2 protein.

In Vitro Transcription From the Human Aγ-Globin Gene Promoter

We report enhanced transcription from the human A gamma-globin gene promoter in nuclear extracts (NE) of erythroleukemia (K562) cells compared with that in HeLa NE. We do not observe differences in transcription levels in the two extracts with nonglobin promoter templates. Our findings, indicating preferential recognition of the globin gene promoter by nuclear factors in K562 cells, are consistent with results of studies previously reported by ourselves and others. A novel finding described here is that the addition of a double-stranded octamer motif oligonucleotide to K562 NE increases the level of transcription from the A gamma-globin gene promoter, suggesting a potential role for an octamer motif-binding factor in the repression of A gamma-globin gene transcription. A cosmid construct containing extensive human gamma- and beta-globin gene promoter and structural sequences as well as upstream control sequences also exhibits higher levels of globin gene transcription in K562 NE than in HeLa NE. Our demonstration of the feasibility of efficient, globin promoter-specific in vitro transcription of this complex template offers a novel approach for the systematic analysis of the effects of putative regulatory factors on globin gene expression in vitro in the context of a genetic environment approximating that found in vivo.

In vitro transcription from the human A gamma-globin gene promoter

We report enhanced transcription from the human A gamma-globin gene promoter in nuclear extracts (NE) of erythroleukemia (K562) cells compared with that in HeLa NE. We do not observe differences in transcription levels in the two extracts with nonglobin promoter templates. Our findings, indicating preferential recognition of the globin gene promoter by nuclear factors in K562 cells, are consistent with results of studies previously reported by ourselves and others. A novel finding described here is that the addition of a double-stranded octamer motif oligonucleotide to K562 NE increases the level of transcription from the A gamma-globin gene promoter, suggesting a potential role for an octamer motif-binding factor in the repression of A gamma-globin gene transcription. A cosmid construct containing extensive human gamma- and beta-globin gene promoter and structural sequences as well as upstream control sequences also exhibits higher levels of globin gene transcription in K562 NE than in HeLa NE. Our demonstration of the feasibility of efficient, globin promoter-specific in vitro transcription of this complex template offers a novel approach for the systematic analysis of the effects of putative regulatory factors on globin gene expression in vitro in the context of a genetic environment approximating that found in vivo.