|
|
|
|||||||
|
|||||||||
(Received for publication, May 11, 1995) From the
During differentiation of C2C12 myoblasts in vitro,
expression of
Collagens constitute a complex family of extracellular proteins
that are major determinants of the mechanical properties of tissues
(van der Rest and Garrone, 1991). The expression of each collagen type
is specifically controlled in different tissues and is differentially
activated by various stimuli. Although several conditions influencing
expression of different collagen types have been described, the
molecular mechanisms involved have only rarely been determined, the
most notable example concerning the stimulation of collagen I chains by
transforming growth factor- Type VI collagen is composed of three genetically distinct
polypeptide chains, The promoter region of the three genes
coding for type VI collagen chains has been recently cloned from
several species (Koller et al., 1991; Koller and Trueb, 1992;
Bonaldo et al., 1993; Saitta and Chu, 1994). The
characterization of the chicken
Figure 5:
Summary of the structural analyses of the
Figure 3:
DNase I footprinting analysis of the
region extending from -82 to +41 nucleotides from the
transcription start site. The end-labeled fragments were reacted with
80 µg of nuclear extract prepared from differentiating C2C12 cells.
The areas protected from DNase I digestion are marked by brackets. Positions of protected segments were determined by
comparison with a G + A sequencing reaction (G/A) and are
indicated relative to the most upstream transcription start
site.
Figure 1:
Expression of
Figure 2:
Expression of
Figure 4:
Determination of contact sites of nuclear
factors within the -82 to +41 region by methylation
interference assay. The non-coding strand of the fragment was
end-labeled with
Figure 6:
The GA box motif is necessary for
induction of the
Figure 7:
Functional properties of artificial
promoters carrying multiple copies of the individual GA box-containing
elements. C2C12 myoblasts were transfected with the indicated
constructs, and CAT activity from proliferating and differentiating
cells and induction was determined as described in the legend of Fig. 2. Constructs (pAsCAT, pBsCAT, pCsCAT, and pB*sCAT) contain
three copies of element A, B, C, and mutated B, respectively, in the
sense orientation fused with the region +8 to -41 of the
natural
Figure 8:
Electrophoretic mobility shift assays
analyzing the binding properties of GA boxes containing
oligonucleotides A, B, and C (see Fig. 5). 20,000 cpm of each
probe were incubated with 4 µg of nuclear extract from
differentiating cells under the indicated conditions, which included
the presence of 10 mM EDTA or an antibody against
transcription factor Sp1 (
Figure 9:
Competition gel mobility shift assays
identifying common associations of sequences A, B, and C (see Fig. 5) with nuclear factors from C2C12 myoblasts. 20,000 cpm of
each probe were incubated with 4 µg of nuclear extract from
differentiating cells. S is an oligonucleotide containing the
recognition sequence for the transcription factor Sp1 (Briggs et
al., 1986) and C* and B* represent
oligonucleotides C and B with a mutated GA box (GGGGAGGG substituted
with GGACATGG). The criteria followed to identify the bands are
summarized in Table 1.
Competition
between some of the nuclear factors for binding to the GA box motifs is
evident in Fig. 8and Fig. 9. Thus, the intensity of band
A7 increased after inactivation of heat-sensitive factors (Fig. 8, compare lanes2 and 5) or in
the presence of anti-Sp1 antibodies (Fig. 8, lane4). The same band was more prominent when access of
proteins to probe A was inhibited by cold oligonucleotides (the most
clear examples are lanes9, 23, and 26 of Fig. 9). Similarly, band A4b became very strong when the
formation of all the other complexes was blocked by cold
oligonucleotide B (Fig. 9, lane7). An
interesting competition concerned associations A1-A6; when
formation of complexes A1 and A2 was abolished by oligonucleotide S,
which contains the recognition sequence of Sp1, bands A5 and A6 (and
also A7) appeared very prominent and bands A3 and A4 remarkably fainter (Fig. 9, compare lane21 with lanes26 and 27). A phenomenon similar to that
described for A was also detected with probe B; the presence of S
lowered intensity of B3 and B4 and enhanced B5 and B6 (data not shown).
A 1000-fold molar excess of cold C* reproduced partially the effect of
S (Fig. 9, lane23). This result was not
surprising, in view of the fact that labeled C* gave rise to a single,
very faint band corresponding to C1 and should therefore slightly
inhibit complex A1 at high concentrations. It must also be noted that
additional bands became apparent at low concentrations of inhibitor
oligonucleotides (Fig. 9, lanes2, 5, 6, 12). This may indicate the existence of potential
recognition sites for other transcription factors in the
oligonucleotides used. The molecular weight of proteins binding to
the GA box-containing elements were investigated by UV cross-linking
and southwestern blotting assays. With the first technique, probe A
produced seven bands of about 150, 105, 76, 54, 50, 46, and 31 kDa (Fig. 10, lane1). All the complexes were
specifically abolished by an excess of the same oligonucleotide (Fig. 10, lane2). Probes B and C gave rise to
six (105, 76, 54, 50, 46, and 31 kDa) and five (105, 76, 54, 46, and 31
kDa) specific bands, respectively (Fig. 10, lanes3-6). Southwestern blotting assays with probe A
revealed five proteins of 105, 65, 52, 43, and 34 kDa (Fig. 10, lane7). Competition experiments confirmed that
binding of the probe to these proteins was specific (Fig. 10, lane8), whereas Western blotting on the same filters
allowed the identification of the 105-kDa species as Sp1 (Fig. 10, lanes9 and 10).
Figure 10:
Biochemical characterization of proteins
binding to GA box-containing elements. Lanes1-6, UV cross-linking assays in solution.
Radiolabeled double-stranded oligonucleotide A, B, and C (Fig. 5) (100,000 cpm) were incubated with 10 µg of nuclear
extract from differentiating myoblasts and without or with the
indicated inhibitor oligonucleotide (400-fold molar excess). After
treatment with UV light, the samples were separated in a 10%
SDS-polyacrylamide gel, and proteins bound to the probe were identified
by autoradiography. Lanes7-10, Southwestern (S.-W.) and Western (W.) blotting assays. Nuclear
proteins (20 µg) from differentiating cells were separated in a 12%
SDS-polyacrylamide gel and transferred to nitrocellulose membrane. The
filters were then hybridized with radiolabeled double-stranded
oligonucleotide A in the absence (lane7) or in the
presence (lane8) of an excess (400
Given
the activation of the
Figure 11:
Electrophoretic mobility shift assays
comparing GA box binding activity in proliferating and differentiating
myoblasts. 20,000 cpm of radiolabeled oligonucleotides A and B (defined
in Fig. 5) were reacted with 6 µg of nuclear extract
prepared from proliferating (P) or differentiating (D) C2C12 cells under the conditions indicated and resolved in
a 6% polyacrylamide gel. Symbols on the left identify
the different bands as referred in Table 1. The gels were overrun
to allow better separation of the bands.
This study has identified a region of the promoter of the
Electrophoretic mobility shift assays indicate that the
three GA box-containing sequences associate with a common set of five
nuclear factors, including Sp1. Three additional protein complexes are
bound by B and A but not by C element. Finally, one complex is
recognized only by fragment A. Therefore, although the GA box is
essential for binding of all the factors, sequences flanking this motif
also contribute to the binding specificity of the three elements. The
relative affinity of the nuclear proteins and the competition between
the factors for binding to overlapping sequences could determine the
overall properties of each GA box-containing element. The complexity of
nuclear factors recognizing the homopyrimidine/homopurine-rich region
of the The molecular mechanism of transcriptional activation of
the Myoblast
differentiation is just one of several situations characterized by
increased expression of collagen VI. Others include differentiation of
adipocytes (Dani et al., 1989) and chondroblasts (Quarto et al., 1993) and confluence of fibroblasts (Hatamochi et
al., 1989). All these conditions are characterized by a decrease
of cell proliferation, which may be a major determinant of the effect,
as previously suggested (Ibrahimi et al., 1993). Thus, it is
not unlikely that collagen VI induction in these other cells is also
mediated by factors binding to GA boxes, whose activity may be
modulated by biochemical pathways depending on the cell cycle.
Volume 270,
Number 33,
Issue of August 18, pp. 19583-19590, 1995
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
1(VI) Collagen Gene during Myoblast
Differentiation Is Mediated by Multiple GA Boxes (*)
1(VI) collagen mRNA was transiently stimulated
severalfold. Promoter assays on cells transfected with chloramphenicol
acetyltransferase (CAT) chimeric constructs have identified a region of
the 1(VI) collagen promoter that increases CAT activity about
8-fold during differentiation. The region, which overlaps with
transcription initiation sites, was shown to contain three protected
segments (A, B, and C) in DNase I footprinting assays. The contact
points between nuclear factors and the protected segments were
determined by methylation interference assay and included the sequence
GGGAGGG (GA box) in all segments. Experiments in which CAT constructs
were cotransfected with double-stranded oligonucleotides containing the
GA box suggested that this motif was necessary for induction.
Transfections with deletion constructs of the natural promoter and with
minipromoters made of three copies of A, B, or C showed that the
elements have inducing activity and that elements C and, to a lower
extent, B are stimulatory for basal transcription, whereas the
contribution of A in this process is limited. Electrophoretic mobility
shift assays with nuclear extracts from C2C12 cells indicated that the
three GA box-containing elements bound several transcription factors,
including Sp1. Comparison of the properties of the bands shifted under
different experimental conditions (presence of 10 mM EDTA,
heating of the nuclear extracts, addition of different concentrations
of competitor oligonucleotides) established that A, B, and C probes
form nine, eight and five main retarded complexes, respectively, and
indicated that nuclear factors binding to C and B are subsets of
proteins binding to A. UV cross-linking assays identified several
peptides (seven with probe A, six with B, and five with C) in the range
of 150-32 kDa. Comparison of the gel retardation pattern obtained
with nuclear extracts from proliferating and differentiating cells
revealed a particular increased intensity of two retarded bands. The
data establish that multiple GA boxes mediate induction of the
1(VI) collagen promoter during myoblast differentiation and
suggest the attractive hypothesis that the effect may be related to
variations of expression of transcription factors binding to these
motifs.
(Rossi et al., 1988;
Ritzenthaler et al., 1993; Inagaki et al., 1994). 1, 2, and 3, all of which contain
several domains related to the von Willebrand type A repeats (for a
recent review on this collagen type, see Colombatti et
al.(1993)). The protein has adhesive properties, and several
observations suggest that it plays an essential role in regulating the
structural organization of the extracellular matrix through specific
interactions with a number of other components. Collagen VI is
particularly abundant in the pericellular space where it forms
microfibrillar aggregates. Expression of the protein during development
is both stage- and tissue-specific. (
)Several studies have
featured a distinct program of type VI collagen regulation. As in other
collagens, the synthesis of all three chains in fibroblasts is
inhibited by viral transformation, but treatment of the cells with
phorbol esters does not change mRNA levels for collagen VI, although it
causes a 3-5-fold reduction for collagen I and III (Schreier et al., 1988). Hyperglycemia, on the contrary, has been found
to increase expression of the three collagen VI chains (Muona et
al., 1993). The effect of transforming growth factor- and
-interferon is restricted to the
3 chain, the former
stimulating and the latter inhibiting its expression (Heckmann et
al., 1989; Heckmann et al., 1992). A unique feature of
the regulation of collagen VI expression is the considerable induction
of protein and mRNA levels observed during differentiation of
mesodermal cells like adipocytes (Dani et al., 1989),
chondrocytes (Quarto et al., 1993), and myoblasts (Ibrahimi et al., 1993) and by confluence in fibroblasts (Hatamochi et al., 1989).1(VI) collagen promoter has
revealed the presence of two Sp1 and one AP1 binding sites close to the
transcription start sites, which are necessary for full promoter
activity (Willimann and Trueb, 1994). Additional stimulatory elements
are certainly contained in more 5`-end sequences (Koller and Trueb,
1992; Bonaldo et al., 1993); these elements, however, have not
been characterized yet. The availability of the promoter region of
different -chains allows the investigation of the molecular
mechanisms of transcriptional regulation possibly involved in the
different conditions affecting type VI collagen expression. In this
work, we have studied the transcriptional regulation of the 1(VI)
collagen promoter during differentiation of a myoblasts cell line in vitro and have begun to define the molecular details of its
activation.
Cell Cultures and Transfections
C2C12
myoblasts (Yaffe and Saxel, 1977) were grown in Dulbecco's
modified Eagle's medium supplemented with 10% fetal calf serum
(proliferation medium) in an atmosphere of 5% CO
. When
required, cells were induced to differentiate by changing the medium to
Dulbecco's modified Eagle's medium containing 2% horse
serum and 0.5 µg/ml bovine insulin (differentiation medium). For
transfection, cells were plated at a density of 100-200
10
/Petri dish (10-cm diameter). The following day,
DNA-Ca phosphate precipitates (Wigler et
al., 1978) were added directly to the culture medium, and cells
incubated for 16 h. The medium was then removed, the layer was treated
for 2 min with 15% glycerol in 25 mM Hepes, 140 mM NaCl, 0.75 mM Na
HPO
, pH 7.05,
and, after washing with phosphate-buffered saline, fresh proliferation
medium was added and incubation continued for 24 h. At this stage,
one-half of the plates were scraped and processed (proliferating
cells), whereas in the other half the medium was changed to
differentiation medium, incubation was prolonged for two additional
days, and cells were harvested (differentiating cells). Extracts were
prepared by resuspending the collected cells in 0.1 ml of 0.1 M potassium phosphate buffer, pH 7.8, 2 mM dithiothreitol,
and repeated freeze-thawing. CAT ()and luciferase assays
were performed as described (Seed and Sheen, 1988; Brasier et
al., 1989). The amount of DNA used for transfection of each Petri
dish was 10 µg of CAT-promoter plasmid and 1 µg of Rous sarcoma
virus luciferase (De Wet et al., 1987) as internal standard
for differences in transfection efficiencies.RNA Analysis
Total RNA was purified from
10-cm Petri dishes with RNAfast reagent (Molecular Systems) following
the procedure recommended by the manufacturer and analyzed by Northern
hybridization using standard protocols (Sambrook et al.,
1989). Probes used were clone SMP1, encoding the amino-terminal end of
the 1(VI) mRNA (Bonaldo et al., 1993), a rat cDNA clone
for cardiac troponin T kindly provided by Dr. S. Schiaffino (University
of Padova, Italy), and a cDNA clone for human
glyceraldehyde-3-phosphate dehydrogenase (Tso et al., 1985).Plasmid Constructions
Plasmid p1341CAT
was constructed by ligating a 1382-base pair fragment spanning
positions -1341 to +41 of the murine 1(VI) promoter
(Bonaldo et al., 1993) into the promoterless vector pBL6CAT
(Boshart et al., 1992). The two plasmids p215CAT and p82CAT
were constructed by polymerase chain reaction amplification from
plasmid p1341CAT and cloning into pBL6CAT. p215
(-72,
-1)CAT was obtained by ligation of two polymerase chain reaction
fragments into HindIII- and PstI-cut pBL6CAT. One
fragment, which extended from -215 to -73, contained a HindIII site at the 5`-end, whereas the other fragment
(+1 to +41) contained a PstI site at the 3`-end.
Deletions of p82CAT from the 5`- and 3`-ends were developed by Bal31 digestion following established protocols (Sambrook et al., 1989). All plasmids were purified by CsCl gradient
centrifugation and sequenced to verify their correct orientation and
sequence identity. Artificial promoter constructs containing
repetitions of elements A, B, and C identified in Fig. 5were
derived by ligating synthetic double-stranded oligonucleotides,
including the indicated sequences and appropriate sticky ends, and
cloning the ligation products upstream of the promoter segment of
p8CAT. Clones carrying three copies of the elements were selected and
confirmed by sequencing.
1(VI) collagen promoter. Regions protected in the DNase I
footprinting experiments reported in Fig. 3are indicated by doublelinesabove and below the
sequence. Singlelines mark the DNase I footprinting
of an AP1 binding site characterized in unpublished experiments. Dots identify contact sites with transcription factors mapped
by methylation interference assays. The most upstream transcription
initiation site is labeled by an arrow, and transcribed
sequences are lowercaseletters. Squarebrackets delimit the sequence of double-stranded
oligonucleotides A, B, and C used in gel shift experiments reported in
the following figures.
DNA Binding Assays
Nuclear extracts were
prepared as described (Shapiro et al., 1988). C2C12 cells were
plated at a density of 11 10
/cm and
either harvested the following day (proliferating cells extract) or
collected after 1 day in differentiation medium (differentiating cells
extract).DNase I Footprinting
Appropriate DNA
fragments were labeled at one end with [P]dNTPs
and Klenow enzyme and gel purified as described (Ausubel et
al., 1993). 20
10
cpm probe were incubated in
20 mM Hepes, pH 7.9, 50 mM KCl, 0.1 mM EDTA,
0.5 mM dithiothreitol, 10% glycerol with or without nuclear
extract (usually 80 µg) in a total volume of 30 µl for 45 min
at 4 °C. 2 µl of 125 mM MgCl, 25 mM CaCl were then added, and the samples were maintained
at room temperature for 1 min. DNase I (Sigma) was added (100-20
ng/reaction in samples containing the nuclear extract and 5-0.4
ng/reaction in samples without nuclear extract), samples were incubated
for 1 min at room temperature, and the reaction was stopped by addition
of 200 µl of 0.2 M NaCl, 1% SDS, 25 mM EDTA and
100 µg/ml herring sperm DNA. After phenol/chloroform extraction and
ethanol precipitation, the samples were resuspended in loading buffer
(80% formamide, 10 mM EDTA, 0.1% xylene cyanol, 0.1%
bromphenol blue), denatured at 95 °C for 3 min, and resolved in a 8
or 10% sequencing gel. The gel was dried and exposed to x-ray film at
-80 °C with intensifying screen.Electrophoretic Mobility Shift
Assay
Synthesized double-stranded oligonucleotides were
radiolabeled with T4 polynucleotide kinase (Sambrook et al.,
1989). 2-5 µg of nuclear extract were incubated for 10 min at
room temperature in 30 µl of binding buffer (20 mM Hepes,
pH 7.9, 50 mM KCl, 5 mM MgCl, 0.5 mM dithiothreitol, 50 µM ZnSO, 5% glycerol)
containing 0.5 µg of poly(dI-dC). Competitor oligonucleotides and
the probe (20-30,000 cpm) were added, and incubation was
prolonged for an additional 20 min at room temperature. Samples were
electrophoresed in 6% non-denaturing polyacrylamide gels (20:1
acrylamide to bisacrylamide) in 40 mM Tris, 190 mM glycine for 3 h at 150 V. The gels were fixed in 10% acetic acid,
dried under vacuum, and exposed to x-ray films. In supershift
experiments, the samples also contained 0.3-4 µg of either
affinity-purified anti-Sp1 antibody (Santa Cruz Biotechnology, Inc.) or
the same amount of preimmune IgG.Methylation Interference Assay
The
procedure followed was exactly that detailed by Ausubel et
al.(1993).UV Cross-linking
100,000 cpm of labeled
double-stranded oligonucleotide were incubated with 10 µg of
nuclear extract in 30 µl under the same conditions used for
electrophoretic mobility shift assays. Samples were then exposed to UV
light (250 nm) for 50 min in a Stratalinker apparatus (Stratagene), 30
µl of 2 final sample buffer (50 mM Tris-HCl, pH
6.8, 2% SDS, 0.1% bromphenol blue, 10% glycerol) were added, and the
molecules were separated by polyacrylamide gel electrophoresis in the
presence of SDS. DNA-protein complexes were revealed by exposure to
x-ray film.
Southwestern Blotting
20 µg of C2C12
nuclear extract were resolved by polyacrylamide gel electrophoresis in
the presence of SDS and electroblotted into nitrocellulose filters. The
proteins bound to the filter were renatured, and filters were processed
as described (Jackson, 1993).
C2C12 myoblasts can be induced to
differentiate when cultured at high density in medium deprived of fetal
calf serum (Yaffe and Saxel, 1977). The appearance of differentiation
markers in our cultures was very rapid, as demonstrated by the increase
of the mRNA for cardiac troponin T (Fig. 1), a gene that is
transiently expressed also at early stages of skeletal muscle
differentiation (Toyota and Shimada, 1981). Myotubes could be noted at
day 2 after stimulation of differentiation and constituted the main
part of the culture at day 4 (data not shown). During differentiation,
the mRNA encoding the 1(VI) Collagen mRNA Expression in
Differentiating Myoblasts1(VI) chain increased very rapidly in the
first day, peaking at about 24 h, and then declined in the following
days to a level that remained constant for at least 1 week (Fig. 1). The maximal increase of mRNA relative to the level
detected at the time of stimulation of differentiation (time 0 in Fig. 1) was usually 5-15-fold, depending on the state of
confluence of the cells. Very similar results have been described for
the 2(VI) collagen chain during differentiation of the same cells
(Ibrahimi et al., 1993).
1(VI) collagen mRNA
by differentiating myoblasts. A series of Petri dishes containing C2C12
cells were plated at a density of 150,000/dish (10 cm) in proliferation
medium. Cells from groups of dishes were harvested for total RNA
purification with the following schedule: 1 day after plating, which
corresponds to the day before switching to differentiation medium
(-1d); the day of application of differentiating conditions (time 0);
at various times (hours or days) after induction of differentiation. 15
µg of RNA were run on a 1% agarose gel and analyzed by the Northern
blotting procedure using cardiac troponin T, 1(VI), and
glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA probes. The
histogram is a quantitative evaluation of 1(VI) mRNA levels and
was obtained by densitometry of specific bands and normalization to the
glyceraldehyde-3-phosphate dehydrogenase
signal.
Identification of the Promoter Region Activated
during Differentiation
The increased production of mRNA in
differentiating myoblasts could be the effect of either transcriptional
or post-transcriptional regulation. To investigate the contribution of
transcriptional mechanism(s), the promoter region of the mouse
1(VI) collagen gene was isolated (Bonaldo et al., 1993),
and several CAT chimeric constructs were derived. The plasmids were
transiently transfected into C2C12 cells, and their promoter activity
was measured in parallel in exponentially growing cells and in cells
switched to differentiating medium (see ``Materials and
Methods''). Initial experiments were performed with two plasmids
containing fragments extending from +41 to -1341 and from
+41 to -215 base pairs from the most upstream transcription
initiation site. The results indicated that CAT expression in
proliferating myoblasts was about double for the longer construct and
that differentiation stimulated promoter activity about 15- and 8-fold
for the long and the short construct, respectively (data not shown). We
therefore concentrated our study on the region -215 to +41,
from which several 5`- and 3`-deletions were obtained and tested for
promoter activity (Fig. 2). The data showed that the region
extending from -83 to -215, although required for high
levels of expression in proliferating cells, was not important for
induction (compare CAT activities of p215CAT and p82CAT). On the
contrary, the sequence from -1 to -82 produced low basal
expression but was essential for induction (compare the results
obtained with plasmid p215CAT with that of p82CAT and
p215(-72,-1)CAT). Stepwise removal of 5`-sequences
from the region -82 to +41 gradually lowered basal CAT
activity and induction properties of the constructs, indicating the
presence of multiple regulatory tracts. Reduction of the 5`-end to
-42 base pairs caused a drop of induction to 5-fold, and an
additional cut to -8 brought to the complete loss of stimulation.
The results indicated that the region -24 to -9 contributed
most to induction and that the stretch -82 to -25 had low
induction activity but remarkably increased basal expression. This
interpretation was confirmed by assays with 3`-deletions of the
promoter region, in which fragments extending from -16 to
-82 and from -55 to -82 gave an identical low level
of induction (Fig. 2). Induction was specific for the 1(VI)
promoter constructs, since CAT activity expressed by the SV40 early
promoter (pACAT
) and the tropoelastin promoter
(pTE911CAT) did not change during myoblast differentiation (Fig. 2).
1(VI) promoter CAT
constructs in C2C12 myoblasts. Promoter fragments extended from the
indicated base at the 5`-end to base +41 at the 3`-end from the
most upstream transcription initiation site (Bonaldo et al.,
1993). Deletions are indicated by followed by parentheses
comprising the position of the bases delimiting the deleted fragment.
pBL6CAT is the vector into which the DNA fragments were cloned. Plasmid
pACAT
contains the SV40 early promoter without
the enhancer (Laimins et al., 1984). pTE911CAT is a plasmid
enclosing a portion of the human tropoelastin promoter (Marigo et
al., 1993). The data were derived from several experiments in
which four Petri dishes were transfected with the indicated plasmids.
After incubation for 1 day in proliferation medium, cells were
harvested from two dishes, and CAT activity was measured (leftpanel). Differentiation medium was added to the other two
dishes, and cells were grown for 30-36 h. After this time, CAT
activity was tested, and induction was determined (rightpanel) as the ratio of CAT activity obtained from
differentiating and proliferating cells.
Structural Analysis of the Inducible
Region
DNase I footprinting with a DNA fragment extending
from +41 to -82 identified two protected sequences in the
upper strand: one from -75 to -59 and the second, which was
very long, from bases -45 to +4. Three footprints were
detected in the lower strand and comprised bases +8 to -19,
-28 to -45, and -60 to -70 (Fig. 3). The
interaction of the DNA fragment with transcription factors was also
analyzed by methylation interference assay (Fig. 4), and the
results are summarized, together with the DNase I footprinting data, in Fig. 5. The contact points (indicated by dots in Fig. 5) were concentrated in three regions and were Gs in the
sequence GGGGAGGG or GGGAGGG, thus identifying the binding string as a
GA box.
P and 100,000 cpm reacted with dimethyl
sulfate, incubated with 20 µg of nuclear extract from
differentiating C2C12 cells, and resolved by polyacrylamide gel
electrophoresis. The major retarded band (R) and the free
probe (F) were purified, fragmented by piperidine treatment
and analyzed in an 8% sequencing gel (rightpanel). G/A, Maxam and Gilbert G + A sequencing
reaction.
Functional Analysis of GA Box-containing
Elements
To establish the functional importance of this
repeated motif on induction of promoter activity, myoblasts were
cotransfected with construct p215CAT and a molar excess of
double-stranded oligonucleotides derived from the sequence of the GA
box-containing elements marked by brackets in Fig. 5.
Induction of CAT expression during myoblast differentiation was
completely inhibited by the three oligonucleotides A, B, and C, which
contain the wild type sequence, whereas it was not affected by C*, in
which the GA box of C was mutated (Fig. 6). On the contrary,
none of these oligonucleotides had any major effect on the CAT activity
of proliferating cells. These data suggest that the GA box motif is
necessary for induction of the 1(VI) collagen promoter during
myogenesis. To evaluate the inductive potency of the individual A, B,
and C elements (Fig. 5), artificial promoters were created in
which trimers of each element were fused upstream of the region
-8 to +41, which contains the transcription initiation sites
(Bonaldo et al., 1993). Induction was observed for all three
elements, with A having the highest and C the lowest activity (Fig. 7). Three copies of A in the antisense orientation were
also stimulatory (plasmid pAsCAT in Fig. 7). On the contrary,
mutation of the GA box abolished induction (plasmid pB*sCAT in Fig. 7). The experiments described in Fig. 7gave also
important information on the function of the three promoter elements in
basal transcription. In proliferating myoblasts, three copies of C
produced a very high level of CAT expression, which was 30-40
times that of the A trimer and 5-10 times that of the B trimer
and of the natural promoter.
1(VI) collagen promoter during myogenesis. C2C12
myoblasts were transfected with the construct p215CAT (defined in Fig. 2) in the absence or presence of the double-stranded
oligonucleotides A, B, C (defined in Fig. 5), and C*, in which
the GA box of C was mutated (GGGGAGGG to GGACATGG). The molar ratio of
the oligonucleotides to the p215CAT construct was 800-fold. Dishes of
proliferating and differentiating cells were processed for CAT
activity, and induction was determined as described in the legend of Fig. 2. CAT activity is expressed as cpm
butyryl-[
C]chloramphenicol formed/h/10
light units.
1(VI) collagen promoter. In B*, the GA box (GGGGAGGG) was
mutated to GGACATGG. pAsCAT differs from pAsCAT for the antisense
orientation of the A trimer. p215CAT is defined in Fig. 2.
Characterization of the Nuclear Factors Binding to GA
Box
Nuclear factors binding to GA boxes were examined by
mobility shift assays. Incubation of double-stranded oligonucleotides
A, B, or C (defined in Fig. 5) with nuclear extracts from C2C12
cells produced several retarded bands (Fig. 8, lanes2, 7, and 12). When the GA box of B or
C was mutated (GGGGAGGG to GGACATGG), the bands disappeared or were
dramatically decreased (data not shown), indicating that all the
factors binding to B and C required at least part of the GA box
sequence. No mutation of the A oligonucleotide was attempted, since it
is entirely formed by repeated GGGAGGG motifs, and changes of the GA
box would have altered the entire sequence. To better identify each
single band and define the possible correspondence of the bands
obtained with the three different probes, the conditions of the assays
were varied: divalent cations were chelated by EDTA (Fig. 8, lanes3, 8, and 13) or an antibody
against Sp1 was added (Fig. 8, lanes4, 9, and 14) or the nuclear extract was heated for 5
min at 95 °C just before the assay (Fig. 8, lanes5, 10, and 15) or cold competitor
oligonucleotides were included in the mixture (Fig. 9).
Representative results are reported in Fig. 8and Fig. 9,
and the data from several experiments are summarized in Table 1.
The use of the different conditions allowed the clear identification of
nine, eight, and five retarded bands with oligonucleotide A, B, and C,
respectively. The analysis suggested that complexes observed with B and
C were due to subsets of proteins binding to A, and, likewise, C bound
a group of factors recognizing B. We have labeled the bands in Fig. 8and 9 and in Table 1on the basis of the probe used
(A, B, or C) and with progressive numbers that are equal for bands that
we propose are retarded by the same protein(s) on the basis of their
properties reported in Table 1. With all probes, band 1 (Fig. 8, lanes2, 7, and 12)
was composed of two complexes (bands 1a and 1b), one of which was
supershifted by an antibody against Sp1 (Fig. 8, lanes3, 8, and 13). The presence of two
bands is perceptible with probe C (Fig. 8, lane12) and is also deduced from the fact that increasing the
concentration of anti-Sp1 antibody used in Fig. 8did not alter
the amount of the supershifted complex (data not shown). Band A4 (Fig. 8, lane2) also included two overlapping
associations, bands A4a and A4b, of which only the former corresponded
to a complex with similar properties in assays with probes B and C
(bands B4 and C4, respectively). This conclusion was derived from
several observations. After heating the nuclear extract, complex A4 did
not disappear completely (Fig. 8, lane5),
identifying a heat-resistant band unique for probe A (A4b in Table 1). This band was EDTA sensitive, since addition of this
reagent after heating completely suppressed complex A4 (data not
shown). On the contrary, treatment with EDTA without heating did not
abolish band A4 (Fig. 8, lane3), thus
distinguishing an EDTA-resistant, heat-sensitive association (band A4a
in Table 1). Finally, cold oligonucleotide A inhibited all the
complexes generated with probe B (Fig. 9, lanes15-17), whereas a band with the mobility of the A4
complex persisted when probe A was incubated with an excess of cold B (Fig. 9, lanes5-7), suggesting that A
binds all the factors recognized by B and, in addition, a protein(s)
that contributes to the formation of band A4. The disappearance of
bands in the presence of EDTA was reversible since the complexes
reformed by readdition of zinc ions (data not shown).
Sp1) or heating of the nuclear extract
for 5 min at 95 °C just before the assay. The bands that could be
identified on the basis of the data reported in this figure and in Fig. 9are indicated. The criteria followed to identify the bands
are summarized in Table 1.
) of
unlabeled oligonucleotide. The strips of lane7 and 8 were probed with either polyclonal antibodies against Sp1 (lane9) or preimmune IgG (lane10). Arrows on the left of the panels
indicate the migration of radiolabeled complexes. Numbers on
the right mark the mobility of proteins of known molecular
mass (given in kDa).
1(VI) collagen gene during differentiation,
it was of interest to test if there was any obvious difference in the
binding activity of nuclear extracts to the GA box-containing elements
in proliferating and differentiating myoblasts. Electrophoretic
mobility shift assays using probes A, B, and C revealed enhanced
intensity of a few bands in differentiating cells. The most significant
variation concerned complexes 1 and 4: A1 and A4 (Fig. 11, lanes1 and 2), B1 and B4 (Fig. 11, lanes5 and 6), and C1 and C4 (data not
shown). Addition of anti-Sp1 antibodies established that the increase
of bands 1 was mainly due to the b component (Fig. 11, lanes3, 4, 7, and 8 and data not
shown). Densitometric quantitation indicated a relative increment of
3-4-fold for bands A1b, B4, and C4 and of about 2-3-fold
for complexes A4, B1b, and C1b.
1 chain of type VI collagen that plays a major role in activation
of the gene during myoblast differentiation. The region extends from
+8 to -75 base pairs from the transcription initiation site
and is homopyrimidine/homopurine rich. Functional and structural
characterization has restricted the activity to three elements
(identified as A, B, and C) whose common feature is the presence of the
sequence GGGAGGG (GA box). A key experiment in defining the function of
this sequence is that reported in Fig. 6, which implies that GA
boxes are essential for induction of promoter activity during
differentiation; mutation of the GA box in cotransfected
oligonucleotides encompassing the sequence protected in DNase I
footprinting assays abolishes inhibition of CAT activity expressed by
promoter constructs. The same mutation also prevents binding of the
nuclear factors to the oligonucleotides. Promoter assays with 5`- and
3`-deletions of the region suggest that the three elements are not
functionally equivalent; A has the strongest inducing activity, whereas
C is important for high basal expression. Additional information on the
functional properties of the three GA box-containing elements have been
obtained from analysis of CAT expression from constructs carrying
artificial promoters formed by repeated A, B, and C sequences. These
experiments confirm the high inducing capacity of A and show that B and
C are also inducers; in addition, they prove that C is the most
effective in enhancing basal transcription. In fact, CAT activity
expressed by the minipromoter with three copies of C is 30-40-
and 5-10-fold higher than that produced by similar constructs
containing an equal number of copies of A and B, respectively. It is
important to note that a single A copy is very active in inducing
transcription (Fig. 2, plasmid p24CAT) and that additional
copies of either A (Fig. 7) or B plus C (Fig. 2) further
enhance induction only to a limited extent. Owing to these data, it can
be proposed that, in the context of the natural promoter, element A is
mainly responsible for induction, whereas B and C are important for
basal expression. In the light of this hypothesis, it is surprising to
find that B and C oligonucleotides do not decrease CAT expression in
proliferating myoblasts when cotransfected with p215CAT (Fig. 6). This apparent contradiction can be reconciled by
assuming that basal expression and induction depend on distinct
regulatory events and that the latter process is more easily altered by
inhibiting binding of nuclear factors to the GA box-containing
elements.1(VI) collagen promoter is apparent also from the partial
biochemical characterization. Seven, six, and five polypeptides in the
range 150-31 kDa are found to associate with elements A, B, and
C, respectively, by UV irradiation. Again, some are common to A, B, and
C, one is common to A and B, and one is unique to A. Due to the
complexity of the patterns, a correspondence between the gel-shifted
bands and the UV-cross-linked peptides could not be deduced. The
formation of five, four, and three complexes in band shift assays with
probes A, B, and C, respectively, required the presence of zinc ions,
suggesting that an important group of transcription factors binding to
the 1(VI) collagen promoter is represented by proteins with
zinc-finger domains. Accordingly, our data show that one of the
EDTA-sensitive proteins is Sp1. As pointed out above, the functional
data have established that A, B, and C have inductive activity and that
C is particularly important for basal expression. A simple correlation
of these functional data with the pattern of bands detected in
electrophoretic mobility shift assays would predict that induction is
mediated by the factors that bind all three elements (complexes
1-4) and that the function of the additional factors, which
recognize A and B (bands 5-7), may rather depress basal
expression.1(VI) collagen gene during myoblasts differentiation remains
to be established and will require full characterization of the GA
box-binding factors. The finding of a consistent severalfold increase
of intensity of some bands, especially 1b (A1b, B1b, C1b) and 4 (A4,
B4, C4) in gel retardation assays stimulates the hypothesis that
induction is due to increased synthesis or binding activity of these
complexes. Gel mobility shift experiments show that all the nuclear
factors require the GA box for significant binding and also indicate
that the nuclear factors compete for binding to the DNA. It is
attractive to speculate that induction during differentiation is not
only due to higher availability of key factors but is also the
consequence of the fact that the increased factors displace from the
promoter proteins, which are inhibitory for transcription.
))
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
ABSTRACT
This article has been cited by other articles:
|
|
 |
|
  C. Fabbro, P. de Gemmis, P. Braghetta, A. Colombatti, D. Volpin, P. Bonaldo, and G. M. Bressan Analysis of Regulatory Regions of Emilin1 Gene and Their Combinatorial Contribution to Tissue-specific Transcription J. Biol. Chem., April 22, 2005; 280(16): 15749 - 15760. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. Lamande, K. A. Shields, A. J. Kornberg, L. K. Shield, and J. F. Bateman Bethlem Myopathy and Engineered Collagen VI Triple Helical Deletions Prevent Intracellular Multimer Assembly and Protein Secretion J. Biol. Chem., July 30, 1999; 274(31): 21817 - 21822. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Fabbro, P. Braghetta, D. Girotto, S. Piccolo, D. Volpin, and G. M. Bressan Cell Type-specific Transcription of the alpha 1(VI) Collagen Gene. ROLE OF THE AP1 BINDING SITE AND OF THE CORE PROMOTER J. Biol. Chem., January 15, 1999; 274(3): 1759 - 1766. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Girotto, C. Fabbro, P. Braghetta, P. Vitale, D. Volpin, and G. M. Bressan Analysis of Transcription of the Col6a1 Gene in a Specific Set of Tissues Suggests a New Variant of Enhancer Region J. Biol. Chem., June 2, 2000; 275(23): 17381 - 17390. [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 |