(Received for publication, July 23, 1996, and in revised form, November 4, 1996)
From the Department of Experimental Medicine,
University of L'Aquila, 67100 L'Aquila, Italy, the ¶ Department
of Experimental Medicine and Pathology, University "La Sapienza,"
00161 Rome, Italy, the ¶¶ Neurological Mediterranean
Institute, Neuromed, Pozzilli, Italy, the
Department of
Pediatrics and Microbiology, School of Medicine, Children's Hospital
of Los Angeles, Los Angeles, California 90027, ** Dibit, San Raffaele
Scientific Institute, 20132 Milano, Italy, and the
§§ Institute of Histology, School of Medicine,
University of Siena, 53100 Siena, Italy
Transforming growth factor type (TGF
) is a
pleiotropic factor that regulates different cellular activities
including cell growth, differentiation, and extracellular matrix
deposition. All the known effects of TGF
appear to be mediated by
its interaction with cell surface receptors that possess a
serine/threonine kinase activity. However, the intracellular signals
that follow receptor activation and lead to the different cellular
responses to TGF
are still largely unknown. On the basis of the
different sensitivity to the protein kinase inhibitor 2-aminopurine and
the phosphatase inhibitor okadaic acid, we identified two distinct
pathways through which TGF
activates a genomic response.
Consistently, 2-aminopurine prevented and okadaic acid potentiated the
induction of JE by TGF
. The induction of PAI-1 and junB
was instead potentiated by 2-aminopurine, after a transient inhibition
and was unaffected by okadaic acid. The superinducing effect of
2-aminopurine required the presence of a functional RB protein since it
was abolished in SV40 large T antigen-transfected cells, absent in the
BT549 and Saos-2 RB-defective cell lines, and restored in BT549 and Saos-2 cells after reintroduction of pRB. The effects of 2-aminopurine on the TGF
inducible junB expression occur in all the
cell lines examined suggesting that junB, and possibly
other genes, can be regulated by TGF
through a distinct
pRB-dependent pathway.
The three mammalian isoforms of the transforming growth factor
type (TGF
1, TGF
2, and TGF
3),1
belong to a superfamily of related polypeptides involved in the control
of a large number of biological activities, including cell growth,
differentiation, and development (1). Originally identified as a factor
able to induce growth of normal rat kidney fibroblast in soft agar,
TGF
was subsequently shown to be a potent growth inhibitor for most
epithelial cells (1).
Most of the reported actions of TGF were shown to be dependent on
its binding to at least two specific membrane-bound proteins, each
belonging to a recently discovered family of serine/threonine kinase
receptors (2, 3), that are active as hetero-oligomeric complexes (4,
5). Although a few candidate transducing molecules have been identified
(6, 7), the biochemical pathways acting downstream of these receptors
are still largely uncharacterized.
With few exceptions (8), in epithelial cells TGF-mediated growth
inhibition correlates with the G1 inhibition of the
phosphorylation of the retinoblastoma gene product, pRB (1, 9). Several events contribute to preventing pRB phosphorylation in TGF
-treated cells: suppression of CDK4 synthesis (10), down-regulation of cyclins
and cdks expression (11), inhibition of cycE-cdk2 complexes by
p27Kip1 binding (12); induction of p21CIP1, and
p15Ink4B with consequent inhibition of cdk4 and cdk6 kinases
(13, 14).
Most of the other cellular responses to TGF, including control of
extracellular matrix protein deposition, wound healing, and immune
suppression are proposed to be mediated by controlling the expression
of specific genes (1). In many cases this is regulated at the
transcriptional level through the binding of specific transcription
factors to stimulatory sequences, as in the case of PAI-1 (15), JE
(16), p21CIP1 (17), p15Ink4B (18), and
2(I) collagen
(15, 19) gene regulation. The expression of other genes appears to be
regulated through TGF
inhibitory sequences (20-22). In the case of
c-myc the same promoter region, the TGF
control element,
is required for down-regulation of c-myc by both TGF
and
the retinoblastoma protein (21, 22). Furthermore, pRB is required for
TGF
-dependent c-myc down-regulation and
growth inhibition, in skin keratinocytes (21) and for down-regulation of N-myc in embryonic lung organ cultures (23). In contrast, in Mv1Lu cells, pRB appears not to be required for PAI-1,
junB, and fibronectin induction by TGF
(24, 25).
2-Aminopurine (2-AP) is a serine/threonine protein kinase inhibitor initially described for its ability to inhibit the double-stranded RNA-dependent protein kinase (26). It was also shown to inhibit expression of interferon-induced genes, to block serum and platelet-derived growth factor induction of c-fos and c-myc (27, 28) and to modulate the expression of a group of genes specifically expressed in growth-arrested cells (29). Okadaic acid is a fatty acid that has tumor promoter activity on mouse skin (30); however, it does not activate or bind protein kinase C. It is, on the contrary, a potent inhibitor of protein phosphatases 1 and 2A (30). Although their broad and still not completely characterized activity may limit the interpretation of the results obtained by their means, 2-AP and okadaic acid have successfully been employed for testing the involvement of protein kinases and phosphatases in several signal transduction pathways (26-30).
To better understand the role of post-translational modifications, such
as phosphorylation and dephosphorylation of pre-existing proteins, on
the signal transduction pathway(s) activated by TGF we have analyzed
the effect of 2-AP and okadaic acid on the activation of gene
expression and the inhibition of cell proliferation induced by TGF
.
In particular we report here on the regulation of three genes involved
in different biological effects initiated by TGF
: (i) JE, a monocyte
chemoattractant (16); (ii) PAI-1, whose induction by TGF
is an
important step in the control of extracellular matrix deposition (1,
15); and (iii) junB, an early responsive gene involved in
the control of growth and differentiation which is regulated by TGF
in most cell types (1). The effects of 2-AP on TGF
-induced pRB
dephosphorylation and growth inhibition were also investigated.
According to the sensitivity to 2-AP and okadaic acid we identified two
different pathways through which TGF
stimulates gene
expression. 2-AP, which is able to prevent TGF
-inducible expression of several genes, potentiates the induction of PAI-1 and junB by TGF
, after a transient inhibition. In
addition we show that potentiation of TGF
-inducible junB
expression by 2-AP requires a functional RB protein. Therefore 2-AP
unravels a role for pRB in a distinct pathway leading to the
up-regulation of junB and possibly other genes by
TGF
.
All the different cell types were
cultured in Dulbecco's modified Eagle's medium supplemented with 10%
fetal calf serum, 2 mM glutamine, 10 units/ml penicillin,
and 10 units/ml streptomycin, with the only exception of Saos-2
requiring RPMI 1640 medium. pPVU-01.5.3, Saos-2 #84, Saos-2 #1, and
BTB5V4-RB were cultured in the presence of Geneticin or Geneticin and
hygromicin as specified elsewhere (9, 31, 32). For stimulation studies,
unless differently specified, cells were grown to subconfluency and
then transferred in serum-free medium containing 10 µg/ml bovine
serum albumin and incubated with TGF1 (Sigma; 1-5 ng/ml) and/or
2-AP (10 mM), cycloheximide (10 µg/ml), phorbol ester
(TPA, 100 ng/ml), and okadaic acid (10 ng/ml).
Cells were
washed twice with ice-cold phosphate-buffered saline, lysed in
guanidinium thiocyanate buffer, and total RNA was isolated by CsCl
gradient centrifugation. 20 µg of total RNAs were denatured with
formamide and formaldehyde, fractionated by denaturing agarose gel
electrophoresis, and transferred to nylon Gene Screen Plus
hybridization membranes (DuPont) by overnight blotting. Filters were
hybridized overnight with 2 × 106 cpm of
32P-labeled DNA probes/ml. DNA probes were labeled by
random priming to an efficiency of 0.5-1 × 109
cpm/µg. Filters were washed to a final concentration of 0.1 × SSC, 0.1% SDS and autoradiographed at 70 °C with intensifying screens.
Cells were washed twice in ice-cold phosphate-buffered saline, scraped off plates into hypotonic lysis buffer (20 mM Tris-HCl, pH 7.4, 25 mM NaCl, 1 mM sodium orthovanadate, 10 mM sodium orthophosphate, 0.25 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, 10 µg/ml leupeptin, and 1% aprotinin) and then flash frozen in liquid nitrogen. After three cycles of freeze-thaw, the lysates were passed several times through a 25-gauge needle. Lysates were cleared by centrifugation at 15,000 × g for 30 min and protein concentrations were determined using Bio-Rad protein assay reagent. Equal amounts of protein (usually 30-60 µg) were separated by SDS-polyacrylamide gel electrophoresis (8.5%), electrophoretically transferred onto nitrocellulose (Schleicher & Schuell), and probed with mouse monoclonal anti-RB (IF8, Santa Cruz Biotechnology, CA). Immunoreactive bands were visualized by enhanced chemoluminescence (ECL, Amersham Corp.).
Cell Cycle AnalysisCell cycle analysis was performed as described previously (33). Briefly, 1 × 106 cells for each sample were fixed in 70% cold ethanol for 30 min at 4 °C and, after washes in cold phosphate-buffered saline, treated with RNase (0.5 mg/ml) and stained with 40 µg/ml propidium iodide. Cells were then kept in the dark at 4 °C for 30 min and immediately analyzed by flow cytometry in a linear scale using a FACscan cytometer (Becton Dickinson, Mountain View, CA). Cell debris and doublets were excluded from the analysis by appropriate gating using physical parameters. Fluorescence data were analyzed by the Consort 30 software.
The induction of PAI-1 and
junB mRNAs is a relatively early response of different
cell types to TGF treatment, while induction of JE mRNA is
considered a late effect. Despite the differences in their kinetic, the
accumulation of JE, PAI-1, and junB mRNAs upon treatment
with TGF
is known to be regulated at the transcriptional level and
does not require ongoing protein synthesis (1, 16). To verify the
possible role of serine/threonine protein kinases on TGF
-specific
pathway(s), mink lung epithelial Mv1Lu cells were treated with 2-AP
both in the absence and presence of TGF
for 1-6 h. Treatment of
cycling Mv1Lu cells with TGF
resulted in a biphasic up-regulation of
the JE mRNA with an initial peak after 1 h, a decline to the
control level around 3 h, followed by a second gradual increase
beginning at 6 h (Fig. 1A) and reaching the plateau level by 12-24 h (not shown). Addition of 2-AP diminished both the basal and the TGF
-stimulated levels of JE mRNA
irrespective of the length of the treatment (Fig. 1A). As
expected PAI-1 and junB expression was up-regulated within
1 h of treatment with TGF
and gradually decreased between 3 and
6 h (Fig. 1A). Addition of 2-AP showed an inhibitory
effect on the TGF
-mediated induction of PAI-1 and junB
mRNAs after 1 h of treatment. However, when applied for either
3 or 6 h, 2-AP appeared to specifically superinduce PAI-1 and
junB expression activated by TGF
, without a significant effect on their basal level (Fig. 1A). Neither the
inhibitory nor the stimulatory effect of 2-AP were inhibited by
contemporary addition of cycloheximide (not shown).
To better characterize the described effect of 2-AP on PAI-1 and
junB expression induced by TGF, Mv1Lu cells were exposed to 2-AP for time intervals of different length, before they were stimulated with TGF
for 1 h. Addition of 2-AP within 5 min
prior to the stimulation with TGF
consistently reduced the
up-regulation of JE, PAI-1, and junB mRNAs (Fig.
1B). Inhibition of JE induction could also be observed in
cells that were exposed to 2-AP for 2 or 5 h (Fig. 1B).
On the contrary, up-regulation of PAI-1 and junB mRNA by
TGF
was higher in cells subjected to such a long-term pre-exposure
to 2-AP compared to control cells (Fig. 1B).
To investigate on the specificity of action of 2-AP, we also studied
its effect on the genomic response initiated by phorbol ester (TPA), a
known activator of protein kinase C. Treatment of Mv1Lu cells with TPA
for 6 h also increased the level of JE, PAI-1, and junB
mRNAs (Fig. 1C). Contemporary addition of 2-AP slightly
reduced JE up-regulation in cells treated with TPA (Fig. 1C). On the other hand, 2-AP did not modify the response of
PAI-1 and only modestly affected the response of junB to
TPA, suggesting that the late superinducing effect on PAI-1 and
junB expression is specific to the regulation of these genes
by TGF.
Given the opposite sensitivity to 2-AP of PAI-1 and junB
compared to JE expression induced by TGF, we tested the effect of okadaic acid, an inhibitor of phosphatase 1 and 2A. Treatment of cells
with okadaic acid did not affect expression of the basal levels of
PAI-1 mRNA in Mv1Lu cells, while it slightly increased the level of
junB and JE after 6 h (Fig. 2). However,
while okadaic acid did not interfere with the induction of PAI-1 and
junB expression by either TGF
(Fig. 2) or TGF
plus
2-AP (not shown), it caused a small but reproducible increase of the
level of JE mRNA induced by TGF
(Fig. 2). In cells stimulated
with TGF
and okadaic acid, 2-AP induced a reduction of the JE
mRNA (not shown), although less severe than that induced in cells
treated with TGF
alone, meaning that also the superinduction due to
okadaic acid of the TGF
-stimulated expression of JE is negatively
affected by 2-AP.
Effect of 2-AP on the TGF
The retinoblastoma susceptibility gene
product, pRB, is considered one of the most important targets for
TGF action in several cell lines, including Mv1Lu. In this cell type
the growth inhibitory effect of TGF
was shown to be correlated to
the inhibition of pRB phosphorylation (9). To ascertain if treatment
with kinase inhibitor 2-AP could also affect the TGF
-activated
pathways leading to pRB dephosphorylation and/or to the cell cycle
arrest, actively growing Mv1Lu cells were treated with TGF
, 2-AP, or
a combination of the two substances for 1-24 h. The relative protein
extracts were analyzed by Western blot for the detection of pRB
isoforms. In the same experiment duplicates of the different cell
culture samples were stained with propidium iodide and subjected to
cell cycle FACS analysis. One representative experiment is shown in Fig. 3. As previously reported (9), in actively
proliferating Mv1Lu cells the majority of pRB appeared in its
hyperphosphorylated form (Fig. 3). Hypophosphorylated pRB started to
accumulate within 3-6 h of treatment with TGF
and after 24 h
virtually all pRB was in its hypophosphorylated form in TGF
-treated
cells. Although Mv1Lu cells are very sensitive to contact inhibition, a
condition in which accumulation of hypophosphorylated pRB is also
observed, this did not limit our analysis since pRB was still fully
phosphorylated in control cells which were maintained in culture for
24 h without any treatment (Fig. 3, last lane). The
addition of 2-AP to TGF
did not prevent the accumulation of the
hypophosphorylated isoform of pRB induced by TGF
, independent of the
length of the exposure to both agents (Fig. 3). 2-AP itself was able to
induce accumulation of the hypophosphorylated pRB isoform (Fig. 3),
with a maximum effect after 24 h. However, after such a long
treatment a considerable amount of pRB was still hyperphosphorylated in
2-AP-treated cells, but not in cells treated with both TGF
and 2-AP.
The cell cycle analysis of the duplicate samples revealed that
treatment with TGF
induced accumulation of the cells in the
G1 and a strong reduction of cells in the S and
G2/M phases of the cell cycle. This effect was maximum
after 24 h and clearly distinct from the spontaneous increase in
the number of G1 cells due to the contact inhibition
observed in control cells maintained in culture for 24 (Fig. 3,
last lane) (9). In contrast, 2-AP induced an increase in the
number of cells with a G2/M DNA content detected as early as 1-3 h after the beginning of the treatment and maximum after 6 h (Fig. 3). This effect was also accompanied by the reduction in the
number G1 cells, whereas S phase cell number was only
modestly affected by 2-AP, under these conditions. In the presence of
both TGF
and 2-AP the majority of the cells progressively
accumulated in a G2/M DNA content state, similar to
2-AP-treated cells. However, after 24 h we observed a strong
reduction in the number of cells in the S phase, which suggests that
even under these conditions Mv1Lu cells are sensitive to TGF
growth
inhibitory signals.
Superinduction of the TGF
To study whether pRB could be involved
in the specific response generated by 2-AP on TGF-induced gene
expression, we investigated the effects of 2-AP on the induction of JE,
PAI-1, and junB mRNAs by TGF
in the pPVU-01.5.3, a
Mv1Lu clone transfected with the SV40 large T antigen (9). In these
cells, large T antigen binding to pRB has been related to the loss of
response to the growth inhibitory effect of TGF
(9), although they
retain the capability to up-regulate junB and extracellular
matrix proteins upon TGF
treatment (24). In keeping with these
observations we found that TGF
induced PAI-1, junB, and
JE expression in pPVU-01.5.3 clone (Fig. 4) with a
similar kinetic when compared to the parental Mv1Lu cell line (Fig.
1A). Contemporary addition of 2-AP was able to inhibit the
induction of JE throughout the 1-6 h treatment and the induction of
PAI-1 and junB mRNAs after 1 h of treatment (Fig.
4), as previously observed for Mv1Lu cells. However, we failed to
detect the expected superinducing effect of 2-AP, which instead
inhibited the TGF
-inducible expression of both PAI-1 and
junB in the pPVU-01.5.3 clone even after 3 and 6 h of
treatment (Fig. 4). These results indicate that, while the early
inhibitory effect of 2-AP on TGF
-induced expression of JE, PAI-1,
and junB is unaffected by the presence of the large T
antigen, the late superinducing effect of 2-AP is negatively regulated
by the large T antigen possibly through its binding to pRB. Since large
T antigen binds and possibly inactivates several cellular proteins in
addition to pRB, we sought to investigate the effect of 2-AP on
TGF
-induced gene expression on different cell lines expressing a
functional pRB (A549, PMC42, and HaCat) compared to cell lines
harboring inactive RB alleles (BT549 and Saos-2) (Ref. 34 and
references therein). The above mentioned cell lines were treated with
TGF
, 2-AP, or a combination of the two for 1-6 h and
junB expression was investigated by Northern blot and
quantified by densitometric scanning of the x-ray films. To ensure a
better comparison of the rough data, the absolute numeric values were
converted to junB percent fold induction, with 100%
arbitrarily assigned to the highest value of junB induction
after TGF
treatment. On the contrary we could not evaluate the
effect of 2-AP on TGF
-inducible expression of the PAI-1 mRNA
since several cell lines did not up-regulate PAI-1 expression in
response to TGF
. In the RB-positive A549, PMC42 (Fig.
5A), and HaCat cells (not shown) 2-AP
prevented the induction of junB mRNA occurring after
1 h of treatment with TGF
, but it superinduced the same
mRNA after 3 and 6 h. On the contrary 2-AP inhibited the
TGF
-regulated junB expression even after 3 and 6 h
in RB-defective Saos-2 and BT549 cell lines (Fig. 5B).
To confirm the involvement of pRB in the superinduction of
junB mRNA upon treatment with both TGF and 2-AP, we
also analyzed the response of three different RB-reconstituted clones
(Fig. 6). The BTB5V4-RB is a BT549 stable transfectant
in which the full-length human RB cDNA is located under a
tetracycline-controlled promoter (31). As previously observed (31),
tetracycline-treated cells expressed a discrete amount of pRB, which
could be increased after 48 h of culture in the absence of
tetracycline (Fig. 6A). Saos-2 cells express a truncated and
nonfunctional p95RB (Fig. 6B). Saos-2 number 1 and
Saos-2 number 84 are two RB-reconstituted clones (32) expressing low
levels of the exogenous p105RB (Fig. 6B). All these
cell lines responded to TGF
with an early up-regulation of
junB mRNA (Fig. 6C). Addition of 2-AP to the treatment prevented junB induction by TGF
after 1 h,
whereas it clearly showed a superinducing effect after 3 h of
treatment in RB-reconstituted BTB5V4-RB cells which was further
enhanced under conditions in which pRB expression was further induced
by the removal of tetracycline from the culture medium (Fig.
6C; compare with parental BT549 in Fig. 5B). A
similar response to 2-AP was found in the pRB-reconstituted Saos-2 #1
and Saos-2 #84 clones (Fig. 6C; compare with parental Saos-2
in Fig. 5B), confirming that the introduction of a
functional RB gene is sufficient to reconstitute the
TGF
-dependent superinduction of junB mRNA
in response to 2-AP. Taken together these observations clearly indicate that the late superinducing effect of 2-AP on TGF
inducible
junB expression is dependent on the presence of functional
pRB while its early inhibitory effect is pRB independent.
The G2/M Arrest Induced by 2-AP Requires a Functional pRB
The temporal association between the RB-dependent
superinducing effect of 2-AP on the expression of PAI-1 and
junB induced by TGF and its ability to block Mv1Lu cells
in the G2/M phase of the cell cycle prompted us to verify
whether also the latter effect was dependent on the presence of a
functional RB protein. In order to test this possibility, parental
Saos-2 cells and the RB-reconstituted Saos-2 #1 and the Saos-2 #84
clones were treated with 10 nM 2-AP for 6 h and
subjected to cell cycle FACS analysis. The average values of
G2/M and of the sum of G1 and S cells obtained from three different experiments are summarized in Table
I. Treatment with 2-AP failed to block pRB-defective
Saos-2 cells in the G2/M phase of the cell cycle. Rather it
induced a slight decrease in the number of cells with a
G2/M DNA content compared to untreated control cells.
However, the ability of 2-AP to induce a G2/M cell cycle
arrest was restored in the RB-reconstituted Saos-2 #1 (Table I) and
Saos-2 #84 (not shown) clones. These results suggest that a functional
RB protein is required for the induction of the G2/M arrest
as well as for the potentiation of TGF
inducible junB expression by 2-AP.
|
In this report we describe the effect of
2-aminopurine and okadaic acid, an inhibitor of serine-threonine
protein kinases and an inhibitor of protein phosphatases, respectively,
on TGF transduction pathway(s) leading to the regulation of gene
expression. The use of these two drugs allowed us to identify two
different pathways involved in TGF
signal transduction and gene
regulation. Specifically, we found that induction of JE expression by
TGF
requires the action of a 2-AP-sensitive protein kinase. Similar results were also obtained for TGF
-induced expression of
c-jun, vimentin, thymosin-
4,2 and
RYR3 (35). The same kinase may also be required for the basal
expression of JE and for its up-regulation by protein kinase C
activation. Indeed 2-AP is also able to inhibit basal level as well as
TPA-induced JE expression. On the other hand we found that okadaic acid
slightly stimulates both basal and TGF
-induced JE expression,
confirming that JE induction may be mediated by increasing the level of
phosphorylation of an unknown protein in the cells. Okadaic acid was
also shown to stimulate basal and TGF
-induced expression of
urokinase receptor and collagenase mRNAs in A549 cells (36, 37) and
2(I) collagen mRNA in CF-37 fibroblasts (38), suggesting that a
similar mechanism can be involved in the regulation of several genes by
TGF
.
Our data indicate that the regulation of other genes by TGF can
utilize a different signaling pathway. PAI-1 and junB
induction by TGF
, in fact, is independent from the action of okadaic
acid-sensitive phosphatases. More interestingly, although the early
induction of these genes by TGF
still rely on the activity of a
2-AP-sensitive kinase, as suggested by the early inhibitory effect of
2-AP, an alternative TGF
-dependent pathway is activated
after prolonged exposure to this agent and leads to PAI-1 and
junB superinduction. Under this condition the pathway
leading to the induction of the JE gene is still repressed, suggesting
that 2-AP inhibitory potential is still intact. In addition, the
delayed action of 2-AP is specific to the TGF
-induced pathway since
it does not significantly affect PAI-1 and junB basal or
TPA-stimulated levels.
Together with the observation that an active kinase domain in both
TGF receptor type I and -II are required for signal transduction (39-41), our results also suggest that the inhibitory action of 2-AP
on the TGF
-induced expression of JE is likely to involve cellular
kinases acting downstream from the TGF
receptors complex. A complete
inhibition of the receptor kinase activity should interfere with any
downstream event following TGF
treatment. On the contrary TGF
is
still able to activate an alternative pathway leading to induction of
PAI-1 and junB genes even after 2-5 h of exposure to 2-AP.
The recent description of a cytoplasmic 78-kDa protein kinase (42)
activated within minutes by TGF
in PC3 cells and the isolation of
TAK1 (43), a member of the MAPKKK family, linked to TGF
signal
transduction in both Mv1Lu and MC3T3-E1 cells, identifies these
proteins as candidate targets for the early inhibitory action of 2-AP
on TGF
signal transduction pathway(s).
TGF generate antiproliferative
signals in several cell types, possibly through its ability to prevent
pRB phosphorylation in the G1 phase of the cell cycle (9).
This event may contribute to the control of S phase-specific genes by
TGF
. Indeed pRB is required for the TGF
-dependent
down-regulation of c-myc in skin keratinocytes (21, 22) and
of N-myc in lung organ cultures (23). In Mv1Lu cells the
pathway activated by TGF
and leading to the inhibition of pRB
phosphorylation is inhibited by the serine/threonine kinase inhibitors
H7, H8, and H9, which can also prevent the induction of PAI-1,
junB, and fibronectin expression by TGF
(44). On the
contrary we found that 2-AP, which by itself caused accumulation of
hypophosphorylated pRB isoforms, did not interfere with the induction
of pRB dephosphorylation in response to TGF
, since TGF
-treated
cells showed substantially the same content of hypophosphorylated pRB
as cells treated with both TGF
and 2-AP. The failure of 2-AP to
inhibit the TGF
growth inhibitory pathway is also indicated by the
fact that TGF
can consistently reduce the number of Mv1Lu cells
entering the S phase of the cell cycle even in the presence of 2-AP.
Such observations suggested that pRB could play a role in the
superinduction of PAI-1 and junB in cells treated with TGF
and 2-AP. Indeed this is demonstrated by three different lines
of evidence. 2-AP could no longer superinduce PAI-1 and junB
expression regulated by TGF
in the pPVU-01.5.3 cell line, in which
pRB is inactivated by the presence of the SV40 large T antigen. In
addition, although 2-AP potentiated the TGF
-inducible junB expression in the RB positive A549, HaCat, and PMC42
cell lines, it failed to do so in the BT549 and Saos-2 tumor-derived cell lines, which were previously shown to lack a functional pRB. In
both these cell lines the superinducing effect of 2-AP was restored
after transfection of the human RB cDNA. Taken together our data
demonstrate that TGF
can regulate the expression of junB,
and possibly other genes, through a distinct pathway. A similar
conclusion is also supported by the evidence that TGF
-inducible junB expression, but not c-jun expression, is
selectively inhibited by the E1A adenoviral oncogene (45).
Despite its function as a transcriptional repressor, the necessity of a
functional pRB for the 2-AP dependent superinduction of junB
by TGF suggests that pRB can positively regulate transcription. This
conclusion is in agreement with previous reports which demonstrated that pRB stimulates the SP1/SP3-mediated transcription by binding and
inactivating a SP1-inhibitory protein (46, 47). More recently pRB was
shown to potentiate transcriptional activation induced by
glucocorticoid receptor in complex with the hBrm protein (48).
It has been shown that pRB is not necessary for TGF- inducible
expression of several genes, including junB (24, 25). This
observation is not in conflict with our data which demonstrate that,
under particular circumstances, uncovered by 2-AP, pRB can potentiate
TGF
-inducible junB expression. In contrast to other serine/threonine kinase inhibitors (44), 2-AP did not prevent accumulation of hypophosphorylated pRB induced by TGF
, implying that
a potentially active pRB is still able to contribute to TGF
signal
transduction in 2-AP-treated cells and could be responsible for
junB superinduction. However, contact inhibited Mv1Lu cells, which are synchronized in the G1 phase of the cell cycle
and accumulated hypophosphorylated pRB, do not respond to TGF
with a
stronger induction of junB mRNA compared to actively
proliferating cells (not shown and Ref. 25). Therefore pRB
dephosphorylation per se is not sufficient to mimic 2-AP
effect on TGF
induced expression of junB. Cell cycle
analysis of Mv1Lu cells treated with TGF
and 2-AP revealed that,
coincident with the superinducing effect on the TGF
inducible
junB expression, 2-AP caused accumulation of the cells in
the G2/M phase of the cell cycle. The G2/M
arrest induced by 2-AP is not unique to Mv1Lu cells since it was also observed on other cell lines (49). Metheny et al. (50) have shown that release of CV-1P cells from nocodazole block in the presence
of 2-AP prevented cells from progressing through methaphase and
anaphase concomitant with pRB being in the hypophosphorylated form. Our
results also indicate that RB-negative Saos-2 cells, in which 2-AP
fails to potentiate the TGF
-inducible junB expression, are insensitive to the effect of 2-AP on the cell cycle. However, 2-AP
was consistently able to increase the G2/M cell population in Saos-2 #1 and Saos-2 #84 pRB-transfected clones, providing evidence
for a role of pRB in 2-AP induced G2/M arrest.
It is accepted that the TGF growth inhibitory activity takes place
within mid-late G1 in order to prevent
G1-arrested cells from re-entering the cell cycle (9). On
the contrary it is still debated how TGF
inhibits proliferation of
asynchronously growing cells. It was shown that the growth inhibition
of Mv1Lu cells by TGF
does not occur within the ongoing cycle, but
only after actively proliferating cells have undergone division in the
presence of TGF
(51), indicating that important events are likely to
occur in the G2 and/or M phases of the cell cycle. Taken
together these observations raise the possibility that as cells enter a
restricted point of the cell cycle, in a particular condition in which
pRB is mostly in its hypophosphorylated form, as in the case of the
2-AP block, they acquire an increased responsiveness to TGF
in terms
of induction of specific genes like junB, and perhaps other
growth regulatory genes, which in turn may contribute to the regulation
of cell cycle progression. Further work is required to ascertain
whether the potentiation of junB induction by TGF
is
selectively occurring in the G2/M cell population which
accumulates upon treatment with 2-AP.
In conclusion, we have provided evidence that TGF regulates gene
expression through at least two different signaling pathways, one of
which is dependent on a functional pRB and possibly restricted to the
G2/M phase of the cell cycle.
We thank Dr. J. Lukas, Dr. M. Strauss, and Dr. T. Gjetting for providing BTB5V4-RB cells, Dr. Y-K. Fung for Saos-2 number 1 and Saos-2 number 84 clones, and Dr. J. Massague and Dr. M. Lahio for pPVU-O1.5.3 cells. We also thank Dr. B. Cardinali, Dr. C. Thiele, Dr. J. Letterio, Dr. G. Piaggio, Dr. M. Crescenzi, Dr. M. Maroder, and Dr. F. Navid for comments on the manuscript, and Dr. A. Del Nero for artwork.