Langdon S D, Kaufman R E
Departments of Medicine and Biochemistry, Duke University Medical Center, Durham, NC 27710, USA.
Blood. 1998 Jan 1;91(1):309-18.
Normal expression of the human beta-globin domain genes is dependent on at least three types of regulatory elements located within the beta-globin domain: the locus control region (LCR), globin enhancer elements (3'beta and 3'Agamma), and the individual globin gene promoter and upstream regions. It has been postulated that regulation occurs through physical interactions between factors bound to these elements, which are located at considerable distances from each other. To identify the elements required for promoter-enhancer interactions from a distance, we have investigated the expression of the wild-type, truncated, and mutated gamma-globin promoters linked to the 5'HS2 enhancer. We show that in K562 cells, 5'HS2 increases activity approximately 20-fold from both a wild-type and truncated (-135 --> +25) gamma promoter and that truncation or site-directed mutagenesis of the tandem CCAAT boxes eliminated the enhancement by 5'HS2. Mutation of the gamma-globin gene promoter GATA-1 binding sites did not decrease either promoter strength or enhancement of activity by 5'HS2. To determine if enhanced expression of gamma-globin gene promoters carrying mutations associated with hereditary persistence of fetal hemoglobin (HPFH) was due to greater interactions with enhancers, we linked these HPFH gamma-globin gene promoters to 5'HS2 and demonstrated a twofold to threefold higher expression than the corresponding wild-type promoter plus enhancer in MEL cells. Addition of the Agamma-globin gene 3' enhancer to a plasmid containing the gamma-globin gene promoter and 5'HS2 did not further enhance promoter strength. Furthermore, we have demonstrated that the previously identified core 5'HS2 enhancer (46-bp tandem AP-1/NF-E2 sites) increased expression only when located 5', but not 3', to the gamma-globin-luciferase reporter gene, suggesting that its enhancer effect is not by DNA looping. Our results suggest that CCAAT boxes, but not GATA or CACCC binding sites, are required for interaction between the gamma-globin promoter and the LCR/5'HS2 and that regulatory elements in addition to the core enhancer may be required for the enhancer to act from a distance.
人类β-珠蛋白结构域基因的正常表达依赖于位于β-珠蛋白结构域内的至少三种调控元件:基因座控制区(LCR)、珠蛋白增强子元件(3'β和3'Agamma)以及各个珠蛋白基因启动子和上游区域。据推测,调控是通过与这些元件结合的因子之间的物理相互作用发生的,而这些元件彼此之间距离相当远。为了从远距离鉴定启动子-增强子相互作用所需的元件,我们研究了与5'HS2增强子相连的野生型、截短型和突变型γ-珠蛋白启动子的表达。我们发现,在K562细胞中,5'HS2使野生型和截短型(-135 --> +25)γ启动子的活性增加约20倍,并且串联CCAAT框的截短或定点诱变消除了5'HS2的增强作用。γ-珠蛋白基因启动子GATA-1结合位点的突变既没有降低启动子强度,也没有降低5'HS2对活性的增强作用。为了确定携带与胎儿血红蛋白遗传性持续存在(HPFH)相关突变的γ-珠蛋白基因启动子的增强表达是否是由于与增强子的更大相互作用,我们将这些HPFHγ-珠蛋白基因启动子与5'HS2相连,并证明在MEL细胞中其表达比相应的野生型启动子加增强子高两到三倍。将Agamma-珠蛋白基因3'增强子添加到含有γ-珠蛋白基因启动子和5'HS2的质粒中并没有进一步增强启动子强度。此外,我们已经证明,先前鉴定的核心5'HS2增强子(46 bp串联AP-1/NF-E2位点)仅当位于γ-珠蛋白-荧光素酶报告基因的5'端而非3'端时才增加表达,并表明其增强作用不是通过DNA环化。我们的结果表明,γ-珠蛋白启动子与LCR/5'HS2之间的相互作用需要CCAAT框,而不是GATA或CACCC结合位点,并且增强子远距离发挥作用可能需要除核心增强子之外的调控元件。