1. No category

From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
1992 79: 765-772
Regulation of globin gene expression in human K562 cells by
recombinant activin A
NL Jr Frigon, L Shao, AL Young, L Maderazo and J Yu
Updated information and services can be found at:
http://www.bloodjournal.org/content/79/3/765.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
Regulation of Globin Gene Expression in Human K562 Cells by
Recombinant Activin A
By Normand L. Frigon, Jr, Li-en Shao, Arlene L. Young, Lina Maderazo, and John Yu
Recent studies indicate that a purified protein, activin A,
belongs t o the transforming growth factor p(TGF-p)superfamily. Similar t o TGF-p, activin A can have different biologic
activities, depending on the target tissues. We used recombinant activin A to demonstrate a possible regulatory role of
this protein in modulating human erythroid differentiation in
the human erythroid cell line, K562. Using genomic probes
containing the second exon of a,p, y, and e globins, relative
abundance of various types of globin transcripts in untreated
and activin-treated K562 cells was examined with S, nuclease
analysis. Despite considerable homology amongst various
globin sequences, these globin probes were highly specific
for their unique mRNA species in the analyses. It was shown
that the abundance of specific globin probe fragments for y
and e globins (209 nucleotides) as well as a (180 nucleotides),
which were protected from S, digestion, increased many fold
in K562 cells treated with activin A. In contrast, there were no
specific transcripts of p globin detected in either the control
or activin-treated cells. The increases in the level of fetal and
embryonic p-like and Q globin transcripts also confirmed
earlier studies of Northern and slot-blot analyses using
globin cDNA as probes. In addition, nuclear run-off transcription assay using isolated nuclei indicated that most of the
increase in the globin transcripts after activin treatment
could be attributed t o the stimulation of transcription rate for
globin genes. Transient transfection assays also provide
evidence that activin A significantly stimulated transcriptional activity of an e. globin promoter in K562, but not in the
nonerythroid Chinese hamster ovary cells. Therefore, it was
concluded that activin A exerts its effects on globin gene
expression at the level of transcription in erythroid cells.
o 1992by The American Society of Hematology.
A
pg/mL of streptomycin and 15% fetal calf serum, as previously
described.” Cell cultures at density of 1 to 2 x 10S/mLwere
incubated with recombinant human activin (generous gift from Dr
R. Schwall, Genentech, Inc, South San Francisco, CA) stock
solution added to the specified final concentration. The recombinant activin A (ie, PAPAin this study) was purified from spent
culture medium of mammalian kidney cells after transfection with
a plasmid containing the entire coding region of the PA subunit
under the control of a cytomegaloviruspromoter.
RNApreparation. Total RNA was prepared from 1 to 3 X lo’
K562 cells by the guanidinium isothiocyanate method.”.I4 After
being washed three times with phosphate-buffered saline, the cells
were lysed in 4 mol/L guanidinium isothiocyanate, 20 mmol/L
sodium acetate, pH 5.2 containing 0.1 mmol/L dithiothreitol and
0.5% N-lauryl sarcosine (Sarkosyl; Sigma, St Louis, MO); they
were then sheared through a 22-gauge needle. Samples were spun
in a microfuge to pellet debris, the supernatant was recovered, and
cesium chloride was added to a concentration of 0.4 g/mL. This
mixture was then layered over a cushion of 5.7 mol/L CsCl and
centrifuged at 35,000 rpm in a Beckman SW 50.1 Ti rotor
(Beckman, Palo Alto, CA) for 16 to 20 hours. After centrifugation,
the RNA pellet was dissolved in H,O; NaOAc was added, and the
sample was extracted with phenolichloroform. RNA was then
precipitated with ethanol and dissolved in H,O.
Northern and slot-blot analyses. For Northern analyses, 12.5 pg
of total RNA was loaded on 1% agarose gel containing 1.1 mol/L
formaldehyde in 20 mmol/L 3-(N-morpholino) propanesulfonic
acid, 5 mmol/L NaOAc, 1 mmol/L EDTA, pH 8.0 buffer. After
CTIVIN IS a newly purified protein dimer consisting of
two P-subunits.’ There are two distinct types of
P-subunits, PA and PB, with 116 and 115 amino acid
residues, re~pectively.’.~The complete amino acid sequences of the precursors of these subunits (PA and P,)
have been deduced from cDNA s e q ~ e n c e s .The
~ . ~ mature
14,000-dalton subunits have nine cysteines and each subunit is presumably extensively folded. Activin is an evolutionarily conserved protein. The mature PAsubunits are identical in porcine, bovine, human, and murine species and the
mature PB subunits differ by only a single residue among
these species. Furthermore, PA and PB subunits have 70%
homology with one another. The mature, most abundant
form is activin A (PAPA);the other two are activin B (P,PB)
and activin AB (PAPB).
In early studies, it was found that the addition of activin
A to K562 cells caused increasing numbers of these cells to
produce hemoglobin, whether measured by benzidine stainIn addition, Eto
ing6 or by immunohistochemical ~taining.’.~
et alyisolated a protein factor from conditioned medium of
THP-1 cells, which induced mouse Friend cells to differentiate, and which was later found to be identical with activin
A, because it shared an N-terminal sequence with porcine
activin A. Furthermore, activin A was shown to potentiate
the proliferation and differentiation of colony formation for
both erythroid colony-forming units (CFU-E) and erythroid
burst-forming units (BFU-E) but not CFU-granulocyte/
macrophage colonies in human marrow cultures.631031’
Even
though the mechanism for these activin-induced effects on
erythroid cells was unclear, these data indicate a possible
role of activin A in the regulation of erythroid differentiation. In present studies, the effect of activin A on the levels
of various globin transcripts in K562 cells was examined
using recombinant protein. In addition, the mode of action
for activin A in affecting globin transcripts was analyzed
with nuclear runoff transcription and transient transfection
assays.
MATERIALS AND METHODS
Cell cultures. Stock cultures of the K562 cell line were grown in
RPMI 1640 medium supplementedwith 50 IUlmL of penicillin, 50
Blood, Vol79, No 3 (February I), 1992: pp 765-772
From the Department of Molecular and Experimental Medicine,
The Scripps Research Institute, La Jolla, CA.
Submitted June 28,1991; accepted September 23, 1991.
Supported by Grant DK40218 from the National Institutes of
Health. This is publication no. 7046-MEMfrom The Scripps Research
Institute.
Address reprint requests to John Yu, MD, PhD, Department of
Molecular and Experimental Medicine, BCR 7, The Scripps Research
Institute, 10666 N Torrey Pines Rd, La Jolla, CA 92037.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1992 by The American Society of Hematology.
0006-4971/ 9 2 /7903-0009$3.OOIO
765
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
766
electrophoresis at 175 V for 2.5 hours, samples were transferred to
nitrocellulose.15The filters were then baked at 80°C for 2 hours
under vacuum. Prehybridization of the filters took place at 42°C
overnight in 56% formamide, 5X Denhardt’s solution, 5X 1,4piperazine-diethanesulfonicacid (PIPES), 0.2% sodium dodecyl
sulfate (SDS), 0.2% NaPPi, and 100 pg/mL salmon sperm DNA.
The filters were then hybridized overnight at 42°C with 2 to 5 x lo6
cpm/mL of nick-translated probe in the same prehybridization
solution except that 1X Denhardt’s solution was used. After
hybridization, the filters were soaked at room temperature for 15
minutes in 2X SSC (1X SSC is 0.15 mol/L sodium chloride, 0.025
mol/L sodium citrate) with 0.1% SDS and then washed in
subsequently decreasing concentrations of SSC as follows: 2X SSC
at 42°C for 1 hour, 1X SSC at 42°C for 1 hour, 1X SSC at 58°C for
45 minutes, 0.2X SSC at 58°C for 45 minutes, 0.2X SSC at 68°C for
45 minutes, and 0.1X SSC at 68°C for 45 minutes. Afterward, the
filters were rinsed at room temperature in 1X SSC without SDS
and autoradiographed overnight at -70°C.
For slot-blot analyses, RNA samples equivalent to 10 pg were
loaded into the wells and transferred to nitrocellulose membrane
in a slot-blot apparatus. The wells were then rinsed with 20X SSC
and the filters were baked at 80°C for 2 hours under vacuum.
Prehybridization, hybridization, and washing were similar to those
described above. Afterward, the filters were autoradiographed
overnight at -70°C.
The probes used for these RNA analyses were as follows: a,the
Mbo I1 fragment of plasmid JWlOl carrying human a globin
cDNAI6;p, the Hha I fragment of plasmid JW102 carrying human p
globin cDNAk6;y, the Hha I fragment of plasmid JW151 carrying
human y globin cDNA16;5, the “clone 2” (Pst I fragment) carrying
human 5 globin cDNA, E plasmid pe 0.7 carrying the 0.7-kb BamHI
restriction fragment from the human E globin gene; and the Pst I
fragment of actin cDNA. The a, p, y, and 5 globin cDNAs were
gifts from Dr B. Forget at Yale (New Haven, CT) and the E plasmid
from Dr D. Tuan at MIT (Cambridge, MA). Probes used for the
slot blot analysis also included a 1.0-kb fragment containing 700 bp
human glyceraldehyde-3 phosphate dehydrogenase (G3PD)
cDNA.” After enzyme digestion of these recombinant plasmids,
the restriction fragments were purified by agarose gel electrophoresis and labeled by the nick-translation reaction.
S, nuclease analyses. The genomic DNA probes for the SI
nuclease analyses were as follows: (1) a,the 1.0-kbPst I-Hind I11 5’
fragment of human a2globin gene; (2) p, the 1.9 kb BamHI 5’ p
fragment; (3) y, the 1.7-kb EcoRI-BamHI 5’ A, fragment; and (4)
c, the 0.67-kb BamHI 5’ E fragment. These DNA clones,” obtained
from Dr T. J. Ley (Washington University Medical Center, St
Louis, MO), were grown in Escherichia coli JM 101 and the plasmid
DNA isolated by standard procedures. The globin genomic fragments were prepared by cutting with appropriate restriction
endonucleases, dephosphorylating with calf intestinal phophatase
(Stratagene, La Jolla, CA), and separating by agarose gel electrophoresis. Then the fragments of interest were purified using NA-45
paper (Schleicher & Schuell, Keene, NH), and eluted by the
manufacturer’s protocol. Afterward, the fragments were precipitated and resuspended at a concentration of approximately 100
ng/mL in 1 mmol/L EDTA, 10 mmol/L Tris-CI, pH 7.6. Approximately 500 ng of each fragment was labeled with T4 polynucleotide
kinase (New England Biolabs, Beverly, MA) and 0.5 mCi Y-’~PATP (New England Nuclear, Wilmington, D E 6,000 Cilmmol) to
a specific activity of greater than 10’ cpm/pg. After chromatography on Sephadex G-50 (Sigma), the probes were counted by liquid
scintillation.
From 0.5 to 1 x 105cpm of probe was combined with 1 Fg of total
RNA and ethanol precipitated. The pellet was dried at room
temperature and resuspended in 15 pL of hybridization buffer
FRIGON ET AL
(80% formamide, 1 mmol/L EDTA, 0.4 mol/L NaCI, and 40
mmol/L PIPES pH 6.4).19 Samples were then boiled for 10 minutes
and transferred immediately to a 52°C water bath to hybridize
overnight. All assays were performed in probe excess. Afterward,
samples were digested with 150 U SI nuclease (Bethesda Research
Lab, Gaithersburg, MD) for 30 minutes, and ethanol precipitated
with carrier E coli tRNA. Then the air-dried samples were taken up
in formamide loading buffer and quantitatively loaded onto denaturing 8% sequencing gels. After autoradiography, the samples
were analyzed by cutting out the relevant portion of the gels and
counting in a liquid scintillation counter.
Nuclear runoff transcription. K562 cells were washed three
times in buffer C (1 mmol/L Tris-HCI, pH 7.6 25 mmol/L KCI, 0.9
mmol/L MgCl,, 0.9 mmol/L CaCI,, 0.14 mmol/L spermidine), once
in buffer C plus 1mmol/L phenylmethylsulfonyl fluoride, and lysed
in a Dounce homogenizer (Baxter Sci. Products, McGaw Park, IL).
Nuclei were pelleted, washed twice in buffer C, and resuspended at
40 A,/mL. Afterward, these isolated nuclei were used in the
following in vitro transcription assay. The standard labeling assay
described by Hofer and Darnell’” used 125 pCi of [c~-’~P]UTP
(3,000 Ci/mmol; New England Nuclear) per lo8 nuclei. Incorporation at 37°C was in linear range up to 20 minutes, and all
preparations were labeled for 10 minutes. Purified genomic globin
probes were bound to nitrocellulose by the method of Kafatos et
a12’after denaturation in 0.2 N NaOH at 100°C for 5 minutes and
neutralization with 2 mmol/L HEPES. Approximately 2.5 pg of
DNA was bound per dot. Hybridizations were performed as
described above. After hybridization, the samples were washed
with 2X SSC and then treated with 2.5 pg/mL of RNase A and 5
pg/mL RNase TI at 37°C for 60 minutes. The RNase-resistant
hybrids were quantitated by scintillation counting.
Transfection and chloramphenicol acetyltransferase (CAT) assay.
Approximately 2 x lo7rapidly growing K562 cells were washed and
resuspended in 0.5 mL HEPES-buffered saline containing the
peGLCAT and the pRSVLUC plasmids (gifts from Dr A. Schechter
at NIH, Bethesda, MD, and Dr A. McLachlan at Scripps Clinic).
After incubation at room temperature for 10 minutes, the mixture
was electroporated with 1 pulse from a Bio-Rad (Richmond, CA)
“Gene Pulser” set to deliver 220 V for 18 to 19 milliseconds. In
other experiments, the monolayer Chinese hamster ovary cells
were transfected with standard calcium phosphate precipitation
methods.22The cells were then allowed to stand at room temperature for another 10 minutes before being divided and cultured in
the presence and absence of activin A for 3 days. CAT assays were
performed by incubating cellular extracts with 14C-chloramphenicol
(Amersham, Arlington Heights, IL) and acetyl CoA in a manner
similar to that previously de~cribed.~’
Within each cell type, CAT
activity for each plasmid was normalized to luciferase activity of the
internal control plasmid pRSVLUC. To assay for the CAT activity,
cell extracts obtained from transfected cells were first heated at
65°C for 10 minutes, a step important for CAT enzyme stability in
K562 cell extracts. Then approximately 20 pL of extract was mixed
with 66 pL of 0.25 mol/L Tris-HCI (pH 7.8), 20 pL of 4 mmol/L
acetyl CoA, water, and 0.5 pCi of 14C-chloramphenicol(50 mCi/
mmol; New England Nuclear) and the reaction was performed at
37°C for 30 minutes. Under these conditions, the activity of the
CAT enzyme was linear for greater than 3 hours. The reaction was
then stopped and the chloramphenicol was extracted with 1 mL
ethyl acetate, and spotted on a silica gel thin-layer plate to separate
the chloramphenicol from its acetylated derivatives. Migration was
in chloroform-methanol (95:5) for 20 minutes. After autoradiography overnight, the spots in thin layer chromatography plates were
cut out and counted to quantitate the amount of chloramphenicol
converted into its 1- and 3-acetyl form. CAT activity was expressed
as the percentage of conversion of chloramphenicol into its 1- and
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
NOVEL GLOBIN REGULATOR, ACTlVlN A
767
Globin
Actin
+-
+-
RESULTS
Northem and slot-blot anahses for globin transcripts. To
examine the effect of activin A on the expression of globin
genes, RNAs prepared from control and activin-treated
K562 cells were subjected to Northern analyses using
various specific human globin cDNA probes. Figure 1
shows hybridization experiments in which equal quantities
of RNAs were used for hybridization with a globin and
actin probes. Hybridization of RNAs with a globin cDNA
indicates a significant increase of globin transcripts (600 to
700 nucleotides in size) in the activin-treated K562 cells.
Similarly, the use of y and E globin cDNA probes also
demonstrate a similar enhancing effect of activin A on the
levels of mRNA contents for these @-likeglobin genes
(picture not shown). In other experiments, slot-blot analysis
was performed to compare the expression of various globin
transcripts, using RNAs corresponding to 1 x lo6cells per
slot and the same hybridization and washing conditions (Fig
2). With the use of cDNA probes for various globin
transcripts, these slot-blot experiments confirmed that there
was a significant increase in transcripts for a,y, 5. and E
globin in the K562 cells after incubation with activin A; but
no expression of p globin mRNA was detected in either the
control or activin-treated cells. Hybridization with the actin
probe (as an internal control) showed that there were equal
amounts of RNA in the control and activin-treated samples
(Fig 2). When K562 cells were incubated with hemin, the
increases in a,y, 5, and E globin mRNAs were also noted
(hybridization blot not shown) and the results correspond
to the data previously rep~rted.'~
These studies indicate
that activin causes a great increase in the contents of globin
mRNA.
Accumulationof variousglobin transcripts. To quantitate
the effect of activin A on the expression of various globin
genes in K562 cells, we then measured the relative abundance of a, p, y, and E globin mRNAs in control and
activin-induced samples by S, nuclease analyses using
highly specific genomic probes. As shown in Figs 3 and 4,
the activation of various types of globin genes in K562 cells
after incubation with different concentrations of activin A
was examined. The organization of the probes used is
shown in Figs 3B and 4B. Exon 2 of the a,and a2human
globin genes has the same nucleotide sequencez and has a
Hind111 site 180 nucleotides downstream from the 5' end of
this exon. Therefore, a probe fragment of 180 nucleotides
Fig 1. Northern blot analysis of RNA isolatedfrom K562 cells. K562
cells were incubatedwith (+) or without (-) 25 ng/mL of activin A for
4 days. Approximately 12.5 pg of RNA was fractionated on a denaturing agarose gel, transferred to nitrocellulose, and probed with =Plabeled cDNA probe for a globin and actin.
3-acetyl forms. In our experimental conditions, the measurement
of this CAT activity for a given cell extract was highly reproducible
(less than 10%variation). In addition, the concentrations and the
molar ratio of various plasmids used were also carefully examined,
so that no interference occurred between the expression of one
reporter with that of the other.
cy
Control
Activin
P
Y
E
actin
W
(
I
,
Fig 2. Slot-blot hybridizationof K562 RNA with globin cDNA probes in control and activin-inducedK562 cells. Aliquots of RNA corresponding
to 1 x 1CP cells of the control and activin-treated samples were loaded onto nitrocellulose membrane. These nitrocellulosestrips were then
hybridizedwith the following cDNA probes: a globin, p globin, y globin, 3 globin, e globin, and actin, respectively. The filters were washed and
autoradiographedas described.
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
768
FRIGON ET AL
.a
A
i-L
B
/"'
, 180
W
aprob.
C
I
2
3
I
4
O
5
6
O
7
8
Fig 3. S, nuclease analyses of a globin mRNA using specific a
globin genomic probe. (A) K562 cells were incubated with increasing
concentrations of activin A, as specified in sample nos. 3 through 8 in
the figure for 5 days. Then total RNA was isolated and 1 c g was
hybridized with the specific a globin genomic probe as described in
(6) and in Materials and Methods. Control sample no. 2 did not
contain any RNA. After hybridization and S, nuclease treatment,
protected probe fragments were electrophoresed on 8% denaturing
sequencing gels, and autoradiography was performed at -70°C. The
probe fragment (180 nucleotides) protected by globin mRNA is
shown. Sample no. 1 in the gel represents the undigested a globin
probe used for this study, while no. 9 indicates the sizes of the
pBR322I Mspl digests. Standards consist of fragments with the
following sizes: 622,527,404,309,242 + 238,217,201, 190,180,160,
147,123, and 110 nucleotides. (6)The organization of the probe used
is shown, with exons as black boxes, introns as open boxes, and 5'
and 3' flanking sequences as a thin line. The position of a Hindlll site in
exon 2 is indicated. (C) The autoradiography of a slot-blot analysis for
sample nos. 2 through 8 using a 1.0-kb probe containing the human
G3PD cDNA as described in Materials and Methods.
on the noncoding strand would be expected to be protected
from SInuclease digestion by correctly spliced exons 2 from
both the aIand a2globin genes in these experiments (Fig
3A). On the other hand, the exon 2 sequences of the E, y,
and P globin genes are 77% to 87% conserved, and each
contains a BamHI site 209 nucleotides from the 5' end.
Correctly processed mRNAs of the P-like globin gene
family would then protect probe fragments of 209 nucleotides from SI digestion of the noncoding strand of the
probes (Fig 4A). The hybridization protocol used in these
experiments utilized a 80% formamide solution and highstringency conditions. In Figs 3 and 4, total RNA from
nontreated and activin-induced K562 cells was hybridized
to globin probes. After SI treatment and electrophoresis of
the protected probe fragments, the gels were autoradiographed. Then, accumulation of globin mRNAs in K562
cells was examined by comparison with the relative abundance of the specific globin fragments in the autoradiographs. It was shown that the abundance of specific globin
fragments for y and E globins (209 nucleotides) as well as a
globin (180 nucleotides) increased in K562 cells treated
with increasing amounts of activin (Figs 3 and 4). Slot-blot
hybridization with the probe for G3PD confirmed that
there were equal amounts of RNA in the control and
activin-treated samples (eg, Fig 3C) in these experiments.
The results from four similar SI nuclease digestion experiments are summarized in Fig 5. It was shown that the levels
of a, y, and E globin transcripts were increased in a
dose-responsive manner. Maximum induction of the increase of those globin transcripts was observed with activin
A at approximately 100 ng/mL in the incubation medium
(Fig 5). In addition, there were no specific transcripts of P
globin detected in either the control or activin-treated cells
(Fig 4). The absence of cross-hybridization between the P
globin probe and other P-like globin mRNAs (ie, y and E
globin transcripts) in this analysis shows the specificity of
each probe for its unique mRNA species, despite considerable homology among P, y, and E exon 2 sequences as
discussed above. Furthermore, in the S, analyses of a globin
mRNAs there was one additional band present above the
expected 180-nucleotide band (Fig 3A). This band was
reproducibly detected and its intensity was always proportional to that of the 180 nucleotide band. However, it is of
uncertain significance and may represent an artifact of SI
nuclease digestion." Thus, these data indicate that fetal (y)
and embryonic (E), as well as a,globins are increased by
incubation of cells with activin A, while no expression of P
globins is demonstrated in K562 cells.
Transcriptional activation of globin genes. Previously, a
chemical inducer, hemin, was shown to cause accumulation
of globin mRNA in K562 via a posttranscriptional regulat i ~ n .To
? ~study the mechanism of the effect of activin A on
the contents of globin gene transcripts, the following two
experiments were performed. Using the nuclear runoff
transcription assay, RNA labeled in isolated nuclei was
hybridized to filter-bound DNA fragments containing various globin probes. It was found that the a and y globin
genes were transcribed at a very low rate in uninduced K562
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
NOVEL GLOBIN REGULATOR, ACTlVlN A
769
Fig 4. S,nuclease analyses of various p-like globin mRNA using specific @like genomic probes. (A)
K562 cells were incubated for 5 days with activin A at
the concentrations specified in the figure. Then total
RNA was isolated and analyzed for the abundance of
specific mRNA with genomic probes for E, y, and p
globins. The S,nuclease analyses were performed in
a manner similar t o that in Fig 3. Sample no. 1 in the
gel represents the undigested probes used in each
experiment; sample no. 2 did not include any RNA in
the S,analyses, samples no. 3 through 8 contained 1
p g of total RNA that was isolated from K562 cells
after incubation with activin A; and sample no. 9 in
the gel indicates positions of various size standards,
the pBR322/Mspl digests. The sizes of these standards are identical t o those in Fig 3. Equal amounts of
RNA present in various samples were confirmed with
slot-blot analyses as shown in Fig 3 (picture not
included). ( 8 )The organization of the probes used in
the experiment is shown, with the symbols for exons,
introns, and flanking sequences similar t o those in Fig
38. The position of a BamHT site in exon 2 is
indicated.
0.I
1
I
I 2 3
ndml
12 3 4 5 6 7 8
1
2
3
4
5
6
7
8
9
lo
Y
c
B
5
'
I
10
Actlvln Concantratlon (
4 5 6 7 8 9
)
Fig 5. Effects of increasing concentrations of activin A on globin
gene expression in K562 cells detected by S,nuclease analyses. K562
cells were incubated with various concentrations of activin A and
total RNAs were analyzed with S, nuclease analyses with globin
probes specific for U. y, and e, as shown in Figs 3 and 4. After
autoradiography, the samples were analyzed by cutting out the
relevant portion of the gels and counting in a liquid scintillation
counter. (-A-), a Globin transcript; (---W---), y globin; and (--0--),
globin RNA. Error bars indicate the standard error of four independent
experiments.
4
3
' 209 *
B
'
a,
-
like gene
'I, or B probe
cells, as previously reported." However, the rate of transcription of these two globin genes showed significant increase
(4.9- and 4.0-fold, respectively) after treatment of cells with
activin A. Furthermore, activin A stimulated the rate of
globin gene transcription and this effect of transcriptional
stimulation was comparable with the activin-induced increase in accumulation of globin mRNA as measured with
S, nuclease protection assay in the same cell preparations
(Fig 6).
The effect of activin A on globin gene expression was also
confirmed in a transient transfection assay. In this assay, the
transcriptional activities of the human E globin gene promoter were compared in the cells after incubation with and
without activin A. An E globin-CAT construct containing
approximately 1.46 kb of 5' flanking region relative to the
cap site of the human e globin gene26was transfected into
K562 and Chinese hamster ovary cells, along with a control
plasmid, pRSVLUC." These cells were separated into two
portions and incubated in the presence or absence of
activin for 3 days. The cell lysates were then assayed for
CAT activities after normalization to equal amounts of
luciferase activities which were derived from the pRSVLUC. As presented in Fig 7, the level of CAT expression
from the plasmid containing the E globin promoter sequences in the transfected K562 cells was increased significantly after the treatment with activin A. In contrast, this
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
FRIGON ET AL
770
A
C
B
Nuclear Transcription mRNA Accumulation RNA Content
-+
Activin
+
- +
El
4.9 x
a
4.0 x
(average
4.0 x
Y
G3PD
I
3.5 x
Same
of two experiments)
transcriptional activity in a similarily transfected Chinese
hamster ovary cell line was negligible and not affected by
incubation of cells with activin A. Previous studies had
shown that this nonerythroid cell line possessed receptors
specific for activin A molecule (Matthews LS,unpublished
observation).
DISCUSSION
Activin A is an evolutionarily conserved protein.’ The
P-subunits of activin A have significant homology to transforming growth factor+ (TGF-P) in primary amino acid
sequences and distribution of cysteine residues.%When the
cysteine residues of the P subunit of activin A and TGF-P
molecule are aligned with each other, they have 33 amino
acid residues in identical positions and have 10 more amino
- +
1
2
Fig 6. Comparison between effects of activin A on
globin gene transcription and on mRNA accumulation. (A) Nuclei were isolated from K562 cells, which
were incubated with (+) or without (-1 activin A at 25
ng/mL for 4 days. These nuclei were then incubated
in the presence of labeled uridine triphosphate and
RNA was purified and annealed t o plasmid DNA
spotted onto Immobilon-P (Millipore, Bedford, MA),
as described in Materials and Methods for nuclear
runoff transcription. The same amount of labeled
transcripts for uninduced and induced cells was used.
(E) RNA was purified from the same cell preparations
as in (A). S, nuclease protection analyses were then
performed using specific probes for a and y globins.
(C) As a control, G3PD probe was used t o ensure
same amount of RNA present in the samples using
slot-blot analyses, as described in Fig 3.
I !
Y
a
Increase:
- +
acids with minor conservative changes?” thus giving rise to
an overall sequence homology of about 30% to 40%.
TGF-f3 has a variety of different biologic activities,2y.w
depending on its target tissues. Some members of the
TGF-p superfamily are closely related to TGF-Pl?’.33The
other members, including inhibin and activin, are more
distantly related to TGF-PI. This family of TGF-P now
includes Mullerian duct inhibiting substance,” the decapentaplegic gene complex of Drosophila,3’ and the Vg-1 gene in
Xenopus,”6all of which display differentiation functions.
Recently, activin A has been implicated as an important
mesoderm-inducing factor in embryogenesis of mice” and
plays some role during development of Xenopus.x It is
conceivable that during early development, activin A is an
important regulator. Given the evolutionary conservation
- +
3
4
5
6
Fig 7. CAT activity of E globin gene promoter in erythroid and nonerythroid cells. Extractsfrom either K562 (lanes 1and 2) or Chinese hamster
ovary (lanes 3 and 4) were obtained from cells cotransfected with psGLCAT and pRSVLUC plasmids and then cultured in the absence (-) and in
the presence (+) of 25 ng/pL activin A. They were incubated with “C-chloramphenicol and acetyl CoA for 30 minutes as described in Materials
and Methods. The acetylated products (A and E) were separated from unreacted chloramphenicol (C) by thin-layer chromatography, dried, and
autoradiographed overnight at room temperature. In lanes 5 and 6, cells were either mock-transfected or transfected with pSVOCAT. In each CAT
assay, cellular extracts containing equal amounts of luciferase activities from the pRSVLUC plasmid were applied.
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
NOVEL GLOBIN REGULATOR, ACTlVlN A
771
of activin A among different species and the homology
between this molecule and other members of the TGF-P
superfamily, it is likely that, similar to TGF-P, activin A may
have biologic activities on different tissues, playing many
roles as hormonal, paracrine, and autocrine regulators.'
The chemical and biologic properties of the activin family
have recently been reviewed.' In these and the accompanying ~tudies'~
we investigated the effects of activin A on
human erythroid cells.
It has been suggested that activin A provides a complex
humoral regulatory control over erythropoiesis! Activin A
is known to potentiate the number of erythroid colonies in
human bone marrow cultures.6~'0.11
The potentiation effect
of activin A could be attributed to an increase in the
proliferative state of BFU-E and CFU-E in the bone
marrow cultures." Previously, it was shown that the potentiating effect of activin A on cellular proliferation was mainly
observed in erythroid lineage.10~''.40
It is also possible that
activin A might drive the differentiation of progeny of these
progenitors into hemoglobinized cells, thus rendering the
colonies recognizable as erythroid colonies. We have shown
in present studies that activin A can induce the accumulation of globin transcripts and stimulate the rate of their
transcriptions in K562 cells.
It was found that activin A could increase the expression
of globin transcripts for a, y, and E globins, while no
expression of P globin message was detected. The hybridization protocol used an 80% formamide solution and highly
stringent conditions. The absence of cross-hybridization
between the P globin probe and other P-like globin mRNAs
in this analysis demonstrates the specificity of each probe
for its unique mRNA species. This was observed despite
considerable homology among P, y, and E exon 2 sequences.
In addition, the lack of expression of the P globin gene
seems to be a special property of the K562 cells, because
they were shown to contain intact copies of the adult P
globin
Previous studies concerning the expression
of various globin genes favor an interpretation that K562
cells contain positive trans-acting factors that interact with
sequence elements in and around the embryonic and fetal
globin genes but lack a comparable trans-acting factor
required for P globin gene e ~ p r e s s i o n . ~ ~
The present studies also indicate that most of the
increase in the globin transcripts after activin A treatment
could be attributed to the stimulation of transcription rate
for these globin genes. This observation is in agreement
with the finding that activin A stimulated transcriptional
activity of a E globin 5'-flanking sequence in the transfection
assay. (The characterization of activin-responsive elements
in the globin promoter sequences is under active study and
will be presented elsewhere.) This is similar to the observations that hemin seems to extend its effect on globin
accumulation in K562 cells at the level of transcriptional
r e g ~ l a t i o n However,
.~~
hemin does not induce stimulation
of transcription from the same E globin promoter in a
similar transfection assay.26It should also be noted that
hemin did not cause commitment of cells to "terminal
differentiation" (defined as limited capacity for further
proliferation"), but activin A did induce these cells to
become differentiated and associated with limited proliferation.6 The latter is similar to thymidine and butyrate,
which also cause commitment of K562 cells.45The accumulation of hemoglobin in the activin-treated K562 cells was
shown to represent an approximately %fold increase in
hemoglobin content! We had estimated that activin A is
about twice as effective on a single-cell basis as are chemical
inducers in inducing hemoglobin production in responsive
K562 cells.6 Therefore, activin A is capable of exerting a
potent differentiation-inducing effect on K562 cells. K562
cells are known to possess approximately 600 high-affinity
'ZSI-activinreceptors per
In the accompanying report,3yit was shown that 'ZSI-activinA could be affinitylabeled to the surface binding proteins (45 to 54 Kd) in both
purified erythroid progenitors and K562 cells. This affinity
labeling of surface binding protein could not be displaced
with incubation of excess amounts of erythropoietin. Therefore, the effects of activin A on globin gene expression in
K562 cells are most likely mediated through its specific
surface receptors, and are independent from the action of
erythropoietin.
ACKNOWLEDGMENT
We thank Judith Preston for secretarial support in the preparation of this manuscript.
REFERENCES
1. Vale W, Hsueh A, Rivier C, Yu J: The inhibidactivin family
of hormones and growth factor. Handbook Exp Pharmacol95:211,
1990
2. Vale W, RiYier J, Vaughan J, McClintock R, Corrigan A,
Woo W, Karr D, Spiess J: Purification and characterization of an
FSH releasing protein from porcine ovarian follicular fluid. Nature
321:776,1986
3. Ling N, Ying FY, Ueno N, Shimasaki F, Esch F, Hotta F,
Guittemin R Pituitary FSH is released by a heterodimer of the
beta subunits from the two forms of inhibin. Nature 321:779, 1986
4. Mayo KE, Cerelli GM, Spiess J, Rivier J, Rosenfeld MG,
Evans RM, Vale W Inhibin A-subunit cDNAs from porcine ovary
and human placenta. Proc Natl Acad Sci USA 83:5849,1986
5. Mason AJ, Niall HD, Seeburg PH: Structure of two human
ovarian inhibins. Biochem Biophys Res Commun 135:957,1986
6. Yu J, Shao L, Lemas V, Yu AL, Vaughan J, Rivier J, Vale W:
Importance of FSH-releasing protein and inhibin in erythrodifferentiation. Nature 330:765, 1987
7. Yu J, Maderazo L, Shao L, Frigon NL, Vaughn DJ, Vale W,
Yu A: Specific roles of actividinhibin in human erythropoiesis in
vitro. Ann NY Acad Sci 628:199,1991
8. Frigon NL Jr, Lemas V, Shao L-E, Maderazo L, Yu J:
Induction of erythroid differentiation in a human cell line by a
TGF-P reiated molecule. J Genet Mol Biol 1991 (in press)
9. Eto Y, Tsuji T, Takezawa M, Takano S, Yokogawa Y, Shibai
H: Purification and characterization of erythroid differentiation
factor (EDF) isolated from human leukemia cell line THP-1.
Biochem Biophys Res Commun 1421095,1987
10. Yu J, Shao L-E, Vaughan J, Vale W, Yu A L Characterization of the potentiation effect of activin on human erythroid colony
formation in vitro. Blood 73:952,1989
11. Broxmeyer JE, Lu L, Cooper S, Schwa11 RH, Mason AJ,
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
772
Nikolics K Selective and indirect modulation of human multipotential and erythroid hematopoietic progenitor cell proliferation by
recombinant human activin and inhibin. Proc Natl Acad Sci USA
85:9052, 1988
12. Singh MK, Yu J: Accumulation of a heat shock-like protein
during differentiation of human erythroid cell line K562. Nature
309:631,1984
13. Chirgwin JJ, Przybyla AE, MacDonald RJ, Rutter WJ:
Isolation of biologically active ribonucleic acid from sources
enriched in ribonuclease. Biochemistry 185294, 1979
14. Glisia V, Crkvenjakov R, Byus C: Ribonucleic acid isolated
by cesium chloride centrifugation. Biochemistry 13:2633,1973
15. Towbin H, Staehelin T, Gordon J: Electrophoretic transfer
of proteins from polyacrylamide gels to nitrocellulose sheets:
Procedure and some applications. Proc Natl Acad Sci USA
76:4350,1979
16. Wilson JT, Wilson LB, deRiel JK, Villa-Komaroff L, Efstratiadis A, Forget BG, Weissman SM: Insertion of synthetic copies of
human globin genes into bacterial plasmids. Nucleic Acids Res
5:563,1978
17. Yun Tso J, Xiao-Hung S, Teh-hui Kao K, Reece S, Wu R:
Isolation and characterization of rat and human glyceraldehyde-3phosphate dehydrogenase cDNAs: Genomic complexityand molecular evolution of the gene. Nucleic Acids Res 13:2485, 1985
18. Ley TJ, Maloney KA, Gordon JI, Schwartz A L Globin gene
expression in erythroid human fetal liver cells. J Clin Invest
83:1032,1989
19. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning, vol
I. Cold Spring Harbor, NY, Cold Spring Harbor Laboratory, 1989,
p 7.62
20. Hofer E, Darnell JE Jr: The primary transcription unit of the
mouse B-major globin gene. Cell 23585, 1981
21. Kafatos KC, Jones CW, Efstratiadis A Determination of
nucleic acid sequence homologies and relative concentrations by a
dot hybridization procedure. Nucleic Acids Res 7:1541,1979
22. Potter H, Weir L, Leder P: Enhancer-dependent expression
of human K immunoglobulin genes introduced into mouse pre-B
lymphocytes by electroporation. Proc Natl Acad Sci USA 81:7161,
1984
23. Gorman CM, Moffat LF, Howard BH: Recombinant genomes which express chloramphenicol acetyltransferase in mammalian cells. Mol Cell Biol2:1044, 1982
24. Charnay P, Maniatis T: Transcriptional regulation of globin
gene expression in the human erythroid cell line K562. Science
220:1281,1983
25. Michaelson AM, Orkin SH: Boundaries of gene conversion
within the duplicated human a-globin genes. J Biol Chem 258:
15245,1983
26. Fordis CM, Nelson N, McCormick M, Padmanabhan R,
Howard B, Schechter AN: The 5’-flanking sequences of human
globin genes contribute to tissue specific expression. Biochem
Biophys Res Commun 134:128,1986
27. De Wet JR, Wood KV, DeLuca M, Helinski DR, Subramani
S: Firefly luciferase gene: Structure and expression in mammalian
cells. Mol Cell Biol7:725,1987
28. Mason AJ,Hayflick JS, Ling N, Esch F, Ueno N, Ying S,
Guillemin R, Niall H, Seeburg PH: Complementary DNA sequences of ovarian follicular fluid inhibin show precursor structure
and homology with transforming growth factor-beta. Nature 318:
659,1985
FRIGON ET AL
29. Sporn MB, Roberts AB, Wakefiled LM, Assoian R K
Transforming growth factor-beta biological function and chemical
structure. Science 233:532, 1986
30. Sporn MB, Roberts AB, Wakefield LM, Crombrugghe B:
Some recent advances in the chemistry and biology of transforming
growth factor-beta. J Cell Biology 105:1039,1987
31. Cheifetz S, Weatherbee JA, Tsang ML, Anderson JK, Mole
JE, Lucas R, Massague J: The transforming growth factor-beta
system, a complex pattern of cross-reactive ligands and receptors.
Cell 48:409,1987
32. Dijke PT, Hansen P, Iwata KK, Pieler C, Foulkes JG:
Identification of another member of the transforming growth
factor type beta gene family. Proc Natl Acad Sci USA 85:4715,1988
33. Metcalf D, Begley CG, Johnson GR, Nicola NA, Vadas MA,
Lopez AF: Biologic properties in vitro of a recombinant human
granulocyte-macrophage colony-stimulating factor. Blood 67:37,
1986
34. Hutson JM, Fallat ME, Kamagata S, Donahoe PK, Budzik
GP: Phosphorylation events during Mullerian duct regression.
Science 223586,1984
35. Padgett RW, St Johnston RD, Gelbart WM: A transcript
from a Drosophila pattern gene predicts a protein homologous to
the transforming growth factor-beta family. Nature 325:81,1987
36. Tannahill D, Weeks DL, Melton DA: A localized Xenopus
maternal mRNA encodes a TGF-beta-like factor. J Cell Biochem
12A:182,1988
37. Jones CM, Hogan BLM: Isolation and initial characterization of Vgr-2: A new member of the TGF-beta gene family. J Cell
Biochem 14E:59,1990
38. Smith JC, Price BMJ, Van Nimmen K, Huylebroeck D:
Identification of a potent Xenopus mesoderm-inducing factor as a
homologue of activin A. Nature 345:729,1990
39. Shao L-e, Frigon NL Jr, Young AL, Yu AL, Mathews LS,
Vaughan J, Vale W, Yu J: Effect on activin A on globin gene
expression in purified human erythroid progenitors. Blood 79:773,
1992
40. Yu J, Shao L, Vaughan J, Vale W, Yu AL: Characterization
of the potentiation effect of activin on human erythroid colony
formation in vitro. Blood 72:106a, 1988 (abstr)
41. Donovan-Peluso M, Young K, Dobkin C, Bank A: Erythroleukemia (K562) cells contain a functional p-globin gene. Mol Cell
Biol4:2553,1984
42. Fordis CM, Anagnou NP, Dean A, Nienhuis AW, Schechter
AN: A B-globin gene, inactive in the K562 leukemic cell, functions
normally in a heterologous expression system. Proc Natl Acad Sci
USA 81:4485,1984
43. Fordis CM, Nelson N, Dean A, Schechter AN, Anagnou NP,
Nienhuis AW, McCormick M, Padmanabhan R, Howard BH:
Trans-acting factors are responsible for the lack of p-globin gene
expression in K562 cells, in Stamatoyannopoulos G, Nienhuis AW
(eds): Experimental Approaches for the Study of Hemoglobin
Switching. New York, NY, Liss, 1985, p 281
44. Dean A, Erard F, Schneider AB, Schechter AN: Induction
of hemoglobin accumulation in human K562 cells by hemin is
reversible. Science 212:459, 1981
45. Rowley PT, Ohlsson-Wilhelm BM, Farley BA: K562 human
erythroleukemia cells demonstrate commitment. Blood 65:862,
1985
46. Campen CA, Vale W: Characterization of activin A binding
sites on the human leukemia cell line K562. Biochem Biophys Res
Commun 1572344,1988