The function of the zeta-globin promoter was studied using a series of
zeta-globin promoter deletion constructs to drive luciferase expressi
on in transiently transfected human erythroleukemia cells, The promote
rs were used without enhancers, or with enhancers beta-globin locus co
ntrol region and the alpha-globin HS-40 enhancer. When transfected int
o K562 cells, which express zeta-globin, comparable amounts of activit
y were obtained from the -557 and -417 zeta-luciferase constructs and
the alpha-luciferase constructs when no enhancers or the alpha-globin
locus enhancers were used. When the constructs were transfected into O
CIM1 cells, which do not express zeta-globin, the zeta-globin promoter
s were at best 20% as active as the alpha-globin promoters. When seque
nces from -417 to -207 5' to the zeta-globin mRNA cap site were delete
d, up to 95% of the zeta-globin promoter activity was lost in K562 cel
ls. Reinsertion of these sequences into zeta-luciterase constructs mis
sing the -417 to -207 region showed that the sequences lack classical
enhancer activity, Point mutation of a GATA-1 site at -230 reduced pro
moter activity by 37%. Point mutation of a CCACC site at -240 had no e
ffect. Electrophoretic mobility shift assays indicated that the -230 G
ATA-1 site has a relatively low affinity for GATA-1. These experiments
show the presence of a strong positive-acting element, located betwee
n -417 and -207 bp 5' to the zeta-globin mRNA cap site, is necessary f
or high-level promoter activity in K562 cells. This element requires G
ATA-1 and additional unknown factors for maximal activity, (C) 1995 by
The American Society of Hematology.