|
|
|
|||||||
|
|||||||||
(Received for publication, September 22, 1994; and in revised form, November
22, 1994) From the
Despite their descent from a common ancestral gene and the
requirement for coordinated, tissue-specific regulation, the
The expression of The differences in regulation of the human However, the particular sequences within
this CpG island that account for the enhancer independence of the human
Figure 1:
An
Fusions between Fusions between
Figure 2:
HeLa cells grown
in Eagle's minimal essential medium with Earle's salts and L-glutamine (MEM), supplemented with 10% fetal calf serum and
1% penicillin/streptomycin, were transfected by the calcium phosphate
procedure (Wigler et al., 1978). The HeLa cells (5
Luciferase-encoding RNAs
were detected by the RNase protection assay of Melton et
al.(1984) as described by Ausubel et al.(1993), using a
176-nucleotide probe generated against the 3` end of the luciferase
region by transcribing pGEMluc (Promega) digested with HpaII with T7 RNA polymerase in the presence of 30 µCi of
[
Figure 3:
S1 nuclease protection assays on RNA from
K562 and HeLa cells transfected with
Figure 5:
Summary of the relative levels of rabbit
globin RNA in transfected K562 and HeLa cells. The constructs used in
the transfection assays are shown in a simplified form and are not
drawn to scale; the conventions in the drawing are the same as in Fig. 1. Multiple determinations of the amount of RNA produced
from individual constructs were normalized to the pBS
As expected from earlier work (Banerji et
al., 1981), this contrasts sharply with the requirement of an
enhancer for expression of the rabbit
Figure 4:
S1 nuclease protection assays on RNA from
K562 and HeLa cells transfected with
Figure 6:
S1 nuclease protection assay on RNA from
transfected HeLa cells. The transfecting DNAs included
Figure 7:
S1 nuclease protection assays on RNA from
transfected K562 cells. The transfecting DNAs include the
pBS
Figure 8:
Test of the ability of
Because measurement
of enzymatic activity was not a reliable indicator of expression, the
production of RNA from these
Figure 9:
RNase protection assays on RNA from K562
cells transfected with
Like its homolog in humans, the Internal
regulatory sequences have been discovered in a growing number of genes,
often within the introns. In some cases, these operate independently of
position or orientation, forming enhancers in introns (e.g. Banerji et al.(1983)). In other cases, the internal
regulatory sequences are active only in their native position, as in
the genes for c-myc (Yang et al., 1986) and ribosomal
protein L32 (Atchison et al., 1989; Chung and Perry, 1989).
The analogous internal sequences in the rabbit Consistent with the absence of a
strong internal enhancer, the intragenic sequences that confer
enhancer-independent expression do not localize to one precise region.
Although a segment extending into exon 2 will produce a significant
amount of RNA in The DNA segments
required for enhancer-independent expression of the rabbit and human
Thus, the CpG island with appropriate
binding sites for transcription factors may constitute an extended
promoter, including both 5`-flanking and internal sequences, that
operates independently of an enhancer. Understanding the mechanisms
that allow this deregulated promoter to be expressed in a wide range of
transfected cell types should provide the basis for determining how,
during normal development, it is turned off in non-erythroid cells and
activated in erythroid cells to produce tight, tissue-specific
regulation. Enhanced expression in erythroid cells is dependent on the
major control region located 40 kilobases 5` to the gene cluster (Higgs et al., 1990). Although the proximal 5`-flanking region of the
human
Volume 270,
Number 8,
Issue of February 24, 1995 pp. 3965-3973
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
-Globin Gene (*)
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
- and
-globin genes in many mammals are regulated in distinctly
different ways. Unlike the -globin gene, the rabbit -globin
gene is transiently expressed at a high level without an added enhancer
in transfected erythroid and non-erythroid cells. By examining a series
of / fusion genes, we show that internal sequences of the
rabbit -globin gene (within the first two exons and introns) are
required along with the 5` flank for this enhancer-independent
expression. Furthermore, deletion of the introns of the -globin
gene, or replacement by introns of the -globin gene, results in
severely decreased expression of the transfecting genes. Hybrid
constructs between segments of the -globin gene and a luciferase
gene confirm that internal -globin sequences are needed for high
level production of RNA in transfected cells. The flanking and internal
sequences implicated in regulation of the rabbit -globin gene
coincide with a prominent CpG-rich island and may comprise an extended
promoter (including both flanking and intragenic sequences) that is
active in transfected cells without an enhancer.
- and -globin genes must be
coordinated and balanced to produce the functional
hemoglobin in erythrocytes, but the
mechanisms leading to this coordination are surprisingly complex. In
particular, the promoters of the - and -globin genes are
regulated differently. The human -globin gene itself, with no
added enhancer, is expressed after transient transfection of
non-erythroid cells (Mellon et al., 1981; Humphries et
al., 1982) and constitutively at a high level after stable
integration in transformed murine erythroleukemia (MEL) cells (Charnay et al., 1984). In contrast, the human and rabbit -globin
genes require the presence of a viral enhancer in cis for
transient expression in non-erythroid cells (Banerji et al.,
1981; Treisman et al., 1983) and are inducible in stably
transformed MEL cells (Chao et al., 1983; Wright et
al., 1983). Although the -globin gene appears to be
deregulated when introduced as a DNA fragment in transfected cells,
both the human - and -globin genes are appropriately
inducible when carried on an entire chromosome in hybrid human 
MEL cells (Deisseroth and Hendrick, 1978; Willing et al.,
1979; Pyati et al., 1980). This indicates that the genes are
regulated at least in part by distal DNA sequences, and, in fact,
linkage to a locus control region (Grosveld et al., 1987) or a
major control region (Higgs et al., 1990) allows the -
and -globin genes to be expressed at a high level in a
position-independent, erythroid-specific manner in transgenic mice. - and -globin
genes correlate closely with the striking differences in their DNA
sequences and genomic context (reviewed in Hardison et
al.(1991) and Hardison and Miller(1993)). Both human and rabbit
-globin genes are largely contained within CpG islands embedded in
a long stretch of G + C-rich DNA that constitutes a very dense
isochore, whereas the -like globin gene cluster is contained
within an A + T-rich isochore characteristic of the bulk of
mammalian genomic DNA (Bernardi et al., 1985). The human
-globin CpG island is not methylated in any tissue or stage of
development examined (Bird et al., 1987), whereas critical
sequences around the -like globin genes (Shen and Maniatis, 1980;
van der Ploeg and Flavell, 1980) are methylated in nonexpressing
tissues. The presence of CpG islands encompassing the -globin gene
may be a general requirement for its enhancer-independent expression.
Indeed, the mouse 1-globin gene is not in a CpG island, and it
requires an enhancer for expression in transfected cells (Whitelaw et al., 1989).-globin gene have not been identified precisely (Charnay et
al., 1984; Whitelaw et al., 1989; Brickner et
al., 1991), nor is it clear whether this effect is derived from
specific activating proteins or is a more general effect of the genomic
DNA context (e.g. being in a CpG island). Further information
can be gleaned from analysis of a similar mammalian -globin gene
that is related to the human gene but differs in some potentially
important internal and flanking sequences. Like the human gene, the
rabbit -globin gene is part of a CpG island (Hardison et
al., 1991), and it is transcribed when transfected into HeLa cells
(Cheng et al., 1988). The present study examining the
expression of various hybrids of the rabbit -globin gene with
either a -globin gene or a luciferase gene suggests a model of an
extended promoter (encompassing both 5`-flanking and internal sequences
of the -globin gene) within the CpG island with multiple, positive
regulatory elements.
Recombinant Plasmids
The structures of plasmids
containing the -globin, -globin, or hybrid genes are
summarized in Fig. 1. All the constructs except CAJO were cloned
into the polylinker region of the plasmid vector Bluescript I
KS(+) from Stratagene, which will be referred to here as pBS. The
rabbit -globin gene is on a SacII to BglII
fragment that extends from -226 to +942 (200 bp (
)past the polyadenylation site) (Cheng et al.,
1988), corresponding to nucleotides 6517 to 7684 in the sequence of the
rabbit -globin gene cluster (GenBank accession number M35026,
Hardison et al.(1991)). A HindIII site was
subsequently inserted at the former BglII site (pBS in Fig. 1). The parental fragment containing the rabbit
-globin gene is a SacI to KpnI fragment that
extends from -1221 to +3325 (pBS4.5 in Fig. 1),
corresponding to nucleotides 29696 to 34241 in the sequence of the
rabbit -like globin gene cluster (EMBL accession number X07786,
Margot et al.(1989)). Shorter fragments were truncated in the
5` flank at the sites for PstI (-100, pBS3.4) or PvuII (-12, pBS3.3) or truncated in the 3` flank at
the HpaI site (+1766, pBS3.0), which is 477
nucleotides past the polyadenylation site. The 196-bp SV40 enhancer
fragment present in pBS.en, CAJO, and pBS.en includes the two
72-base pair repeats (Banerji et al., 1981). This enhancer was
inserted at the BglII site at -424 in the -globin
5` flank (pBS.en, Fig. 1) and in the polylinker of pBS 5`
to the -globin gene (pBS.en, Fig. 1).
-Globin and
-globin gene constructs used in transfection assays. The rabbit
-globin gene is shown with dotted lines for flanking
sequences, dark dotted boxes for exons, and light dotted
boxes for introns. The rabbit -globin gene is shown with black lines for flanking sequences, black boxes for
exons, and white boxes for introns. The restriction
endonuclease cleavage sites used in the construction of each
recombinant are shown. The box labeled enh is the
SV40 enhancer, which includes the two 72-bp
repeats.
-globin gene fragment from NcoI (+35, the ATG
initiation codon) to PvuII (+796, or 84 nucleotides past
the polyadenylation site) was inserted into pBS3.0 at the HpaI site to generate pBS and pBS.in
(opposite orientation, Fig. 1).- and
-globin genes were made at the -globin gene NcoI
site (+35) and the -globin gene PvuII
site(-12) for the exon 1 fusions, at the AccI sites in
exon 2 of the -globin gene (+356) and -globin gene
(+283), at the BalI sites in exon 3 of the -globin
gene (+525) and -globin gene (+1165), and at the EcoRI sites in exon 3 of the -globin gene (+545) and
the -globin gene (+1116). Single sites were used for the
/ and / fusion gene constructs, and combinations of
sites were used for the () and () replacement
constructs (Fig. 1).-globin and
luciferase genes are shown in Fig. 2. The -Luc construct
consists of the rabbit -globin gene from the PstI site at
-1096 to the PstI site at +494 fused in-frame to
the luciferase coding segment (nucleotides 1757 to 45 from plasmid
pGEM-luc from Promega). The fusion is at the 3` end of intron
2 of -globin, maintaining the splice junction. This results in a
hybrid protein encoded by exons 1 and 2 of -globin and the
luciferase cDNA. Nucleotides +544 to +941 of rabbit
-globin, containing the 3` half of exon 3 and the polyadenylation
site, are fused to the 3` end of the luciferase coding region. In
(inverted)-Luc, the 5` rabbit -globin fragment (-1096
to +494) is inserted in the opposite orientation with respect to
-Luc. The construct (
e12)-Luc has a 206-bp deletion in
the 5` -globin fragment from nucleotides +105 to +309
(inclusive). In the p-Luc construct, rabbit -globin
5`-flanking region (nucleotides -241 to +34 relative to the
cap site) was inserted upstream of the luciferase coding region in the
plasmid pGL2Basic (Promega). The -globin start codon was deleted,
such that the luciferase start codon was utilized.
-Luciferase fusion constructs used in
transfection assays. The rabbit -globin gene is shown with black boxes for exons, wide white boxes for introns,
and thin white boxes for flanking sequences. The luciferase
coding region is shown as a dotted box. SV40 untranslated
sequences are shown with diagonal stripes, with the large T
antigen intron indicated by a light stipple
pattern.
Transfections of Cells with DNA
K562 cells grown
in Dulbecco's modified Eagle's medium supplemented with 10%
bovine calf serum, 2% penicillin/streptomycin, and 0.5 µg/ml
amphotericin B, were transfected by electroporation (Neumann et
al., 1982; Potter et al., 1984) at 450 V, 500 µF for
500 ms (Promega electroporator). For transfections with the /
fusion constructs, 1 ml of cell suspension (10
cells/ml)
was mixed with 100 µg of the test DNA, electroporated, and
harvested after 48 h of growth. In transfections for transient
expression of luciferase, 5 10
K562 cells were
mixed with 4 µg of test DNA, 10 µg of pRSVlacZ, and 36 µg
of carrier DNA in a total volume of 700 µl, electroporated, and
allowed to grow as above. To generate pools of cells stably transfected
with the luciferase constructs, 10 cells in a 1-ml
suspension were mixed with 90 µg of linearized test DNA and 10
µg of linearized pM5neo (Laker et al., 1987), which
contains the gene for neomycin resistance driven by the
promoter-enhancer of myeloproliferative virus M5. 24 h after
electroporation, G418 was added to each culture to a final
concentration of 1.2 mg/ml. Cultures were maintained until a pool of
10 G418-resistant cells per transfection culture was
obtained, at which point the cells were harvested.
10
cells/ml, 10 ml per 10-cm Petri dish) were
transfected with a calcium phosphate precipitate containing 50 µg
of test DNA. The media were replaced after 24 h, and the RNA was
harvested after 48 h.RNA Analysis
After transfection, RNA was isolated
from both HeLa and K562 cells by the guanidine thiocyanate-acid phenol
procedure (Chomczynski and Sacchi, 1987). Several different probes were
used in the S1 nuclease protection assays (Favaloro et al.,
1980). To analyze the 3` end of -globin RNA, a 402-bp EcoRI-HindIII fragment from pBS was 3`
end-labeled at the EcoRI site in exon 3 with
[-
P]dATP and Klenow polymerase. Other
probes were uniformly labeled (Bentley, 1984) by copying the inserts in
the single-stranded M13 clones As (NcoI to PstI,
detecting exon 1 and exon 2 of the -globin gene), A2s (PstI to BamHI, detecting exon 1 and exon 2 of the
-globin gene), and Cs (BglII to BglII,
detecting exon 3 of the -globin gene) (Rohrbaugh et al.,
1985; Vandenbergh et al., 1991). For each RNA analysis from a
set of transfection assays, 0.1, 1.0, or 10 ng of rabbit reticulocyte
poly(A)
RNA (Life Technologies, Inc.) was mixed with
100 µg of Escherichia coli tRNA to serve as a positive
control and quantitation standard. Hybridizations were in a solution of
80% formamide, 40 mM Hepes, pH 7.5, 400 mM NaCl, 1
mM EDTA, at temperatures of 51 °C for the probes, 42
°C for the 5` probe, and 37 °C for the 3` probe.
RNA-DNA hybrids were digested with 400 units of S1 nuclease and
resolved on denaturing polyacrylamide gels.-P]UTP. The fragments protected from RNase
digestion were 125 nucleotides long when annealed to -Luc and
(e12)-Luc RNA and 115 nucleotides long when annealed to
p-Luc RNA.Quantification of the Relative Levels of Rabbit Globin
RNA
Transfections followed by S1 analysis of RNA were performed
multiple times for each construct, and the resulting autoradiographs
were quantified by densitometry. The signal for each construct was
normalized to the pBS.en or pBS signal from the same
autoradiograph, and the values from different experiments were
averaged. For transfections into K562 cells, the amount of -globin
RNA from pBS was consistently about 10 times that of -globin
RNA from pBS.en (by reference to the rabbit reticulocyte RNAs used
as controls), so all values could be reported relative to the
-globin RNA from pBS. However, in HeLa cells, the amount of
RNA from pBS.en relative to pBS varied widely in different
experiments, so the values for HeLa cell transfections are reported
separately as relative to -globin RNA from pBS or
-globin RNA from pBS.en (depending on the probe used in the
analysis). The results of the RNase protection assays of
luciferase-containing constructs were quantified on a -scope
(Betagen).Chloramphenicol Acetyltransferase Assay
A series
of restriction fragments from the flanking and internal regions of the
rabbit -globin gene (Yost et al., 1993) were inserted
into the enhancer trap plasmid pCATpromoter (Promega), which contains
the gene for chloramphenicol acetyltransferase (CAT) driven by the SV40
early promoter, but no enhancer. K562 cells were transfected with 100
µg of the pCAT promoter as above, and freeze-thaw extracts of
transfected cells were assayed for CAT activity (Gorman et
al., 1982), which is reported relative to the activity generated
by pSV2CAT. The amount of CAT activity for pSV2CAT in 5 independent
transfections ranged from 30 to 77 pmol of chloramphenicol acetylated
per min per mg of protein in the extract (corresponding to 24 to 64%
acetylation of the chloramphenicol in 60 min by 25 µl of extract).Luciferase Assay
Luciferase assays were performed
on transfected cells as described in the Promega Luciferase Assay
System technical bulletin. The same extracts prepared by this procedure
were used to assay for -galactosidase activity as described in
Sambrook et al.(1989). Luciferase activities in relative light
units per second (RLU
s
) as detected by a
Berthold luminometer were divided by the -galactosidase activities
(expressed as A
) to correct for transfection
efficiency.
Transcription of the Rabbit
The ability of
the rabbit -Globin Gene in
Transfected Cells Does Not Require an Enhancer-globin gene to be transiently expressed was tested in
both erythroid (human K562 cells, which express endogenous -,
-, 
-, and 
-globin genes, but not the -globin gene)
and non-erythroid (HeLa) cells. Cells transfected with plasmids
containing the -globin gene (clone pBS, Fig. 1)
produced about 2 ng of RNA in K562 cells (Fig. 3A, lanes 2, 3, and 13) and 6 ng in HeLa cells (Fig. 3B, lanes 1 and 2). Inclusion
of the SV40 enhancer (pBS.en, Fig. 1) caused only a modest
increase in the amount of RNA (Fig. 3A, lanes
6, 7, and 14; Fig. 3B, lanes
5 and 6). The results from several determinations show
the SV40 enhancer caused at most a 2-fold increase in -globin gene
expression in either K562 cells or HeLa cells (summarized in Fig. 5).
-globin gene constructs.
Autoradiographs of the gels resolving fragments protected from nuclease
S1 digestion are shown. Abbreviated names of the DNA constructs are
given at the top of each lane: M,
mock-transfected cells; pr, input probe; rR, rabbit
reticulocyte poly(A) RNA; ex1, ex2, ex3, protected fragments from exons 1, 2, or 3, respectively. A, RNA from transfected K562 cells was hybridized to a
uniformly labeled probe extending from the NcoI site in exon 1
to the PstI site in intron 2 of the rabbit -globin gene,
and the portions of the probe protected by RNA from digestion by S1
nuclease are shown. The multiple bands probably result from S1 nibbling
into the ends of the duplex. B, RNA from transfected HeLa
cells was hybridized with a 3` end-labeled probe extending from the EcoRI site in exon 3 to an artificial HindIII site
inserted 234 bp 3` to the polyadenylation site of the -globin
gene. Lanes 2-9 in A and 1-6 in B show the results of duplicate transfections (separate plates
of cells transfected with the same DNA).
or
pBS.en signals. The averages ± S. D. (or half the range for n = 2) are reported for n determinations.
Based on Student's t test, the p values for
pBS/.2 and pBS/.3 versus pBS4.5 are
<0.001, except for pBS/.2 versus pBS4.5 in
K562 cells (p < 0.01).
-globin gene in HeLa cells.
The S1 analysis in Fig. 4B (lanes 3, 4, and 8) shows that the rabbit -globin gene
(clones pBS3.0 and pBS4.5, Fig. 1) directs the
synthesis of barely detectable RNA in HeLa cells, but introduction of
the SV40 enhancer (pBS.en, Fig. 1) causes a large increase
in the amount of RNA produced (13- to 25-fold, Fig. 5). Like the
endogenous homolog, the rabbit -globin gene is also not actively
expressed in K562 cells (Fig. 4A, lane 5), but
even addition of the SV40 enhancer (CAJO, Fig. 1) does not
rescue its expression (Fig. 4A, lane 9).
/ hybrid genes and
-globin genes. A, RNA from transfected K562 cells was
hybridized with uniformly labeled probes extending from the BglII site in exon 3 to a BglII site located 350 bp
3` to the polyadenylation site of the -globin gene (lanes
1-9) or from a PstI site located 100 bp 5` to the
cap site to the BamHI site in exon 2 of the -globin gene (lanes 10-18). Lanes 10-15 show the
results of duplicate transfections with the same DNA. Lanes
10-18 are from a longer exposure than lanes
1-9, so that 1.0 ng of rabbit reticulocyte RNA in lane
17 gives a signal comparable to 10 ng of RNA in lane 3.
Bands resulting from cross-hybridization between the endogenous human
-globin RNA and the rabbit -globin probe are labeled (). B, RNA from HeLa cells was hybridized with the uniformly
labeled probe for the 3` end of -globin RNA described for A. M
, size markers of pBR322 digested
with HinfI.
A Promoterless
Addition of the rabbit
-Globin Gene Does Not Serve as an
Enhancer of -Globin Gene Expression-globin gene (minus its promoter) to the -globin gene in
either orientation (clones pBS and pBS.in, Fig. 1) does not cause an increase in the amount of -globin
RNA produced in transfected cells. As shown in Fig. 4B (lanes 9 and 10), no RNA is seen when the
constructs are transfected into HeLa cells. Fig. 4A (lanes 10-13) shows a small amount of -globin
RNA (about 0.1 ng) produced in K562 cells transfected with these
constructs, equal to that obtained from the -globin gene with or
without the SV40 enhancer in this same experiment (pBS3.0,
pBS.en; data not shown). The summary in Fig. 5confirms
this absence of enhancement in multiple determinations.Internal Sequences of the
A series of -Globin Gene Are Required
for Transcription/ fusion genes was
tested in the transient expression assay. The 5` portion of the
-globin gene was joined to the 3` portion of the -globin gene
in either exon 1, exon 2, or exon 3 (clones pBS/.1,
pBS/.2, and pBS/.3, respectively; Fig. 1).
Fusion of the 5` flank of the -globin gene onto the body of the
-globin gene (pBS/.1) does not confer active expression
in either K562 cells (Fig. 4A, lane 6) or HeLa
cells (Fig. 4B, lane 5). However, inclusion of
internal -globin sequences up to exon 2 or exon 3 does generate a
substantial amount of hybrid RNA both in transfected K562 cells (Fig. 4A, lanes 7 and 8) and HeLa
cells (Fig. 4B, lanes 6 and 7),
detected with a 3` end-labeled -globin probe. Analysis of the
hybrid RNA with a uniformly labeled probe from the 5` portion of the
-globin gene confirms that the / genes fused at exons 2
or 3 are transcriptionally active (Fig. 3A, lanes
8, 9, 15, and 16), producing about as
much RNA as the parental -globin gene. The summary of several
experiments (Fig. 5) shows that a construct with both the 5`
flank and internal -globin gene sequences (pBS/.3)
produces about 15 times as much RNA as the -globin gene construct
pBS4.5 in K562 cells and about 10 times as much in HeLa cells,
whereas fusion of the -globin gene 5` flank onto the -globin
gene has little effect.Deletion or Replacement of Internal
The introns of the -Globin Gene
Sequences Reduces RNA Production-globin
gene were removed by replacing the internal sequences with an
-globin cDNA (pBSc, Fig. 1). Transfection of either
K562 or HeLa cells with this construct produced very little
-globin RNA (Fig. 3A, lanes 4 and 5; Fig. 3B, lanes 3 and 4).
This decrease in production of RNA could result from the loss of
positive cis-acting sequences in the -globin gene introns
or from the absence of splicing, or a combination of both. Thus, the
internal regions of the -globin gene were replaced with analogous
regions of the -globin gene, leaving the splice junctions intact
(clones pBS().12, pBS().23, and pBS().13; Fig. 1). Transfection of either HeLa cells (Fig. 6, lanes 9-11) or K562 cells (Fig. 7, lanes
11-13) with these plasmids produced very small amounts of
hybrid RNA. The replacement of the internal -globin gene sequences
with -globin sequences caused about a 5- to 7-fold decrease in the
amount of RNA produced, compared to the parental -globin gene (Fig. 5).
-globin
genes with and without an enhancer, -globin genes with internal
regions replaced by -globin gene sequences (the pBS()
series), a / hybrid gene (pBS/.2) and a
-globin gene with internal sequences replaced with -globin
gene sequences (pBS().23). The RNA was hybridized with the
uniformly labeled probe specific for exons 1 and 2 of the -globin
gene (Fig. 4).
() replacement constructs and / as well as
/ fusion genes. The RNA was hybridized with the uniformly
labeled probe for exons 1 and 2 of the -globin gene (Fig. 4).
Fusion of Internal
The
-Globin Gene Sequences into a
-Globin Gene Does Not Activate Its Transcription-globin gene promoter could not be activated by internal
-globin gene sequences in reciprocal / hybrid gene
constructs (pBS/.2 and pBS/.3; Fig. 1).
Transfections with these plasmids, as well as a plasmid with the second
intron of the -globin gene replaced by that of the -globin
gene (pBS().23; Fig. 1), produced only very small
amounts of RNA in both HeLa (Fig. 6, lanes 12-14)
and K562 cells (Fig. 7, lanes 14-16). Thus, the
positive cis effects of these -globin gene internal
sequences do not act on the -globin gene promoter, whether the
internal sequences are in the normal position (as in the hybrid genes
in pBS/.2 or pBS().23) or inserted at a distant
site, as with the promoterless -globin genes added to plasmids
with the -globin gene (pBS and pBS.in; Fig. 5).Internal and Flanking Sequences of the
A complementary
study showed that several nuclear proteins, including an Sp1-like
protein, will bind specifically in both the 5` flank and internal
regions of the -Globin Gene
Can Increase Expression from the SV40 Promoter-globin gene (Yost et al., 1993). This
suggests that these DNA sequences could perhaps act as enhancers of
promoters containing Sp1-binding sites, despite the fact that the
-globin gene sequences were not effective as enhancers of the
-globin gene promoter. To test this hypothesis, four fragments of
the -globin gene (Fig. 8A) were inserted 3` to the
coding sequences of the chloramphenicol acetyltransferase gene in the
expression vector pCATpromoter, which is driven by the SV40 promoter
(but not the enhancer). After transfection into K562 cells, all four
fragments cause a clear but moderate increase in CAT activity relative
to the parental pCATpromoter (6- to 18-fold, Fig. 8B).
This is a specific effect of adding these fragments since an insertion
of a fragment containing hypersensitive site 3 of the human
-globin locus control region (Philipsen et al., 1990) did
not increase the CAT activity (only 0.4% relative to pSV2CAT, data not
shown); this DNA fragment is not an enhancer of the SV40 promoter (Tuan et al., 1987). These data indicate that several internal and
flanking regions of the rabbit -globin gene are able to increase
expression from the SV40 promoter, but individually they are not as
effective as the SV40 enhancer.
-globin gene
fragments to enhance expression of the CAT gene from the SV40 promoter. A, four fragments of the rabbit -globin gene were placed
3` to the CAT gene driven by an SV40 promoter (pCAT promoter),
transfected into K562 cells and CAT activity was measured. Proteins
implicated in binding to the DNA motifs are indicated. CP1 is
a CCAAT-box binding protein, IRP is the -globin
inverted repeat binding protein (a relative of Sp1), TBP is
the TATA-box binding protein, YY1 is involved in both positive
and negative regulation of various genes, Sp1 is a
transcriptional activator, CACBP refers to any protein binding
to a CACC motif, and CnBP refers to the protein binding to a
string of Cs in the 3`-untranslated region. B, an
autoradiograph of the thin layer chromatogram separating
chloramphenicol (Cm) from its acetylated products (1-Ac-Cm and
3-Ac-Cm) is shown for one set of duplicate transfections with each
construct. CAT activity was calculated as nanomoles of chloramphenicol
acetylated min mg of protein and
is reported relative to the activity of pSV2CAT. The first column of numbers gives the activities (± half the range)
determined for the experiment shown in the autoradiograph. The second column gives the results (±S.D.) for several
independent experiments. All values for the constructs containing
-globin gene fragments are significantly greater than those for
pCATpromoter (p < 0.001 by Student's t test).
Internal Sequences of the Rabbit
Since the -Globin Gene Are
Required for Production of High Levels of RNA from Luciferase Reporter
Gene Fusion Construct-globin gene is not
expressed in K562 cells, interpretation of the results from the
/ fusions is complicated by the possible presence of negative
regulatory elements in the -globin gene sequences. In order to
address this concern, segments of the -globin gene were fused with
a different reporter gene that should not contain any negative
elements. The luciferase reporter gene was chosen because of the
relative ease with which enzyme activity could be measured in
transfected cells. However, transfection of cells with the constructs
shown in Fig. 2showed the fusion of rabbit -globin coding
sequences to the luciferase coding region to generate a hybrid enzyme
caused a decrease in luciferase activity (-Luc versus p-Luc, Table 1), which was contrary to the result
expected based on previous data with the / fusion constructs.
Subsequent observations on the hybrid enzyme produced in stably
transfected K562 cells and in bacteria indicate that the -globin
amino acid sequence has a deleterious effect on the activity of the
hybrid luciferase protein, making enzymatic activity an unreliable and
unpredictable measure of expression for the -luciferase fusions.
This likely explains the increase in luciferase activity seen for the
deletion construct, (e12)-Luc (Table 1), which has less
-globin coding sequence in the hybrid gene.
-luciferase fusion constructs was
measured directly using an RNase protection assay. RNA from pools of
K562 cells stably transfected with -Luc (containing the internal
sequences, Fig. 2) yielded a clear, luciferase-specific,
protected fragment of 125 nucleotides, whereas transfection with a
construct with the -globin sequences in the reverse orientation,
(inverted)-Luc, produced no detectable RNA (Fig. 9). In
contrast, no 115-nucleotide protected fragment was detected above
background in RNA from cells transfected with p-Luc, which does
not contain the internal -globin gene sequences. These data using
the luciferase reporter constructs are congruent with the results from
the / fusion gene experiments; in both cases, the
-globin internal sequences are required for production of high
levels of RNA. The construct using only 5`-flanking sequences as a
promoter is expressed, as shown by the enzymatic activity in Table 1, but from an amount of RNA that is not detectable
relative to that from a construct containing internal sequences
(-Luc, Fig. 9).
-luciferase fusion constructs. RNA from
pools of stably transfected K562 cells was hybridized to a uniformly
labeled RNA probe for the luciferase portion of the hybrid message. An
autoradiogram on the gel resolving the protected fragments is shown.
The positions of the 176-nucleotide probe and the expected protected
fragments (125 nucleotides for -Luc, 115 nucleotides for
p-Luc) are indicated by arrows. The positive control in
the second lane is a clone (D8) of K562 cells stably
transformed with the -Luc construct.
Deletion of a 206-bp Fragment from the
Previous
binding studies with the rabbit -Globin
Structural Sequence That Contains a YY1-binding Site and Two
Sp1-binding Sites Has No Effect on RNA Production-globin gene identified a
YY1-binding site near the 3` end of exon 1 and two tandem Sp1-binding
sites within intron 1 (Yost et al., 1993). To test the
contribution of these sites to the requirement for internal sequences
for high level -globin expression, a 206-bp fragment was deleted
from the construct -Luc by oligonucleotide-directed mutagenesis,
generating the construct (e12)-Luc (Fig. 2). After
stable transfection into K562 cells, the amount of RNA produced is
about the same as that from the parental construct -Luc (Fig. 9), showing that these three sites are not required for
the enhancer-independent expression of the -globin gene.
-globin gene from
rabbits does not require an added enhancer for expression in
transfected erythroid and non-erythroid cells. The role of internal
-globin gene sequences in expression has been controversial.
Studies with human / hybrid genes showed that sequences
internal or 3` to the -globin gene are required for its
constitutive expression in stably transfected MEL cells (Charnay et
al., 1984). However, Whitelaw et al.(1989) argued that
the human -globin gene is expressed without an added SV40 enhancer
only when replicating in HeLa cells, and no erythroid-specific
enhancers were found in or around the gene. Our studies show that
internal regions of the rabbit -globin gene are required for
efficient expression. Inclusion of these internal sequences in
/ hybrid genes allows expression without an enhancer, whereas
their replacement with internal segments from the -globin gene
causes a loss of expression. Furthermore, inclusion of internal
sequences caused a large increase in RNA production from
-luciferase hybrid genes. The regulatory regions implicated for
the rabbit gene are similar to those recently mapped for the human
-globin gene by Brickner et al.(1991), who found that a
DNA segment extending from the 5` flank through exon 1 and intron 1
efficiently drove expression of a CAT reporter gene.-globin gene have
only a modest enhancing effect on the SV40 promoter (Fig. 7),
and this effect may not be sufficient to explain the
enhancer-independent phenotype of the intact gene. The internal
regulatory sequences of the rabbit -globin gene may work best
within their natural context, perhaps constituting part of the
promoter. In fact, the internal regulatory segments of the human
-globin gene are not effective when placed 3` to the gene or in
the distal 5` flank (Brickner et al., 1991). Thus, the
-globin gene may not contain a classic internal enhancer, but
rather the promoter of the gene may be unusually long, extending into
the first two exons and introns./ hybrid genes, inclusion of DNA up to the
5` end of exon 3 will produce more RNA (Fig. 4). Also,
replacement of the region surrounding either intron 1 or intron 2 with
equivalent sequences from the rabbit -globin gene will decrease
the amount of RNA produced. All the internal and flanking regions
tested increased CAT expression from the SV40 promoter to a comparable
extent. Surprisingly, deletion of an internal region containing YY1-
and Sp1-binding sites has no effect on RNA production. It is possible
that multiple, redundant positive elements may be involved in
establishing enhancer-independent expression.-globin genes correspond to the CpG islands encompassing the genes
(Bird et al., 1987; Hardison et al., 1991). The CpG
island in rabbits begins about 500 bp 5` to the -globin cap site
and extends to the third exon, and it contains all the sequences shown
here to be important in transient expression. The CpG islands are not
found around the mammalian -like globin genes or around the mouse
1-globin gene, all of which require enhancers in transient
expression assays. Thus, the CpG island may be critical for generating
a widely expressed, enhancer-independent promoter, either by the
binding of ubiquitous trans-activating proteins (Whitelaw et al., 1989; Yost et al., 1993) or by establishing a
unique chromatin structure that is readily transcribed (Charnay et
al., 1984). Chromatin from CpG islands differs dramatically from
that of bulk chromatin, with much less histone H1 and an elevated level
of acetylation of histone H4 (Tazi and Bird, 1990), which may
facilitate access of transcription factors to the promoter. These two
hypotheses are not mutually exclusive, and the combination of specific
binding by transcriptional activators within a type of chromatin that
is permissive for transcription could account for the ability of the
rabbit and human -globin genes to be expressed after introduction
into a wide variety of cells.-globin gene is sufficient to respond to this enhancer
(Pondel et al., 1992; Ren et al., 1993), ()inclusion of the internal gene region leads to an even
greater level of expression (Ren et al., 1993). In addition,
given the capacity of the -globin gene to express after
transfection into a wide variety of cells, one could propose that a
negative regulator is required to prevent expression in non-erythroid
cells. Further experiments are required to test this possibility.
))
We thank W. Schaffner and E. Schreiber for gifts of
the SV40 enhancer and the CAJO construct and Steve Pullen, Hania
Petrykowska, and Martin Sigg for help with the experiments.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Morel and J. Massoulie Comparative Expression of Homologous Proteins. A NOVEL MODE OF TRANSCRIPTIONAL REGULATION BY THE CODING SEQUENCE FOLDING COMPATIBILITY OF CHIMERAS J. Biol. Chem., March 15, 2000; 275(10): 7304 - 7312. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Grad, J. Le Dai, S. Wu, and K. L. Burnstein Multiple Androgen Response Elements and a Myc Consensus Site in the Androgen Receptor (AR) Coding Region Are Involved in Androgen-Mediated Up-Regulation of AR Messenger RNA Mol. Endocrinol., November 1, 1999; 13(11): 1896 - 1911. [Abstract] [Full Text]
|
|
|
|
|
|
N. Neznanov, A. Umezawa, and R. G. Oshima A Regulatory Element within a Coding Exon Modulates Keratin 18 Gene Expression in Transgenic Mice J. Biol. Chem., October 31, 1997; 272(44): 27549 - 27557. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. H. Surinya, T. C. Cox, and B. K. May Transcriptional Regulation of the Human Erythroid 5-Aminolevulinate Synthase Gene. IDENTIFICATION OF PROMOTER ELEMENTS AND ROLE OF REGULATORY PROTEINS J. Biol. Chem., October 17, 1997; 272(42): 26585 - 26594. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. C. Hardison, J. Oeltjen, and W. Miller Long Human-Mouse Sequence Alignments Reveal Novel Regulatory Elements: A Reason to Sequence the Mouse Genome Genome Res., October 1, 1997; 7(10): 959 - 966. [Full Text] [PDF]
|
|
|
|
|
|
J. D. Jackson, H. Petrykowska, S. Philipsen, W. Miller, and R. Hardison Role of DNA Sequences Outside the Cores of DNase Hypersensitive Sites (HSs) in Functions of the beta-Globin Locus Control Region J. Biol. Chem., May 17, 1996; 271(20): 11871 - 11878. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |