| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cell Growth & Differentiation |
Articles |
Department of Cell Biology, University of Cincinnati College of Medicine, Cincinnati, Ohio 45267-0521
Abstract
Prostate cells are dependent on androgen for proliferation, but during tumor progression prostate cancer cells achieve independence from the androgen requirement. We report that androgen withdrawal fails to inhibit cell cycle progression or influence the expression of cyclin-dependent kinase (CDK)/cyclins in androgen-independent prostate cancer cells, indicating that these cells signal for cell cycle progression in the absence of androgen. However, phosphorylation of the retinoblastoma tumor suppressor protein (RB) is still required for G1-S progression in androgen-independent cells, since the expression of constitutively active RB (PSM-RB) or p16ink4a caused cell cycle arrest and mimicked the effects of androgen withdrawal on downstream targets in androgen-dependent LNCaP cells. Since Ras is known to mediate mitogenic signaling to RB, we hypothesized that active V12Ras would induce androgen-independent cell cycle progression in LNCaP cells. Although V12Ras was able to stimulate ERK phosphorylation and induce cyclin D1 expression in the absence of androgen, it was not sufficient to promote androgen-independent cell cycle progression. Similarly, ectopic expression of CDK4/cyclin D1, which stimulated RB phosphorylation in the presence of androgen, was incapable of inactivating RB or driving cell cycle progression in the absence of androgen. We show that androgen regulates both CDK4/cyclin D1 and CDK2 complexes to inactivate RB and initiate cell cycle progression. Together, these data show that androgen independence is achieved via deregulation of the androgen to RB signal, and that this signal can only be partially initiated by the Ras pathway in androgen-dependent cells.
Introduction
Tumorigenic growth is characterized by proliferation occurring under inappropriate environmental conditions (1, 2) . This phenomenon is clearly observed during the progression of prostate cancer. Initially, prostatic epithelia or early-stage prostatic cancer cells are androgen-dependent and thus require androgen to proliferate (3–5) . In the most commonly used therapy for prostate cancer, androgens are removed either by chemical or surgical means, leading to the cessation of proliferation (6) . Despite this initial remission, the tumor cells rapidly become androgen-independent, rendering hormonal treatment ineffective (7, 8) . Strikingly, there is no efficient treatment for androgen-independent tumors. Furthermore, little is known about how these recurrent tumor cells are able to proliferate in the absence of androgen.
Androgen is a known growth factor for prostatic epithelial cells (9) . However, the mechanism through which androgen signals to promote proliferation is poorly understood (10) . In general, mitogenic factors act through the Ras/ERK3 pathway to activate a host of CDK/cyclin complexes, which are required for cell cycle progression and cellular proliferation (1, 2, 11, 12) . Early in the G1 phase of the cell cycle, mitogenic signals activate CDK4/cyclin D complexes, which are required to initiate phosphorylation of RB (2, 13) . Reagents that block RB phosphorylation (e.g., the CDK4 inhibitor p16ink4a) or constitutively active RB proteins (PSM-RB) arrest cells in G1 (2, 14, 15) . Thus, inactivation of RB is typically required for entry into S-phase, although specific cell types acquire resistance to the actions of these reagents, either through the attenuation of downstream signaling or through the expression of viral oncoproteins of DNA tumor viruses (16, 17) . We and others have shown recently that RB inhibits cell cycle progression by attenuating cyclin A expression, thus diminishing CDK2 activity (16, 18–20) . Because both CDK2/cyclin E and CDK2/cyclin A activities are required for entry into S-phase, inhibition of cyclin A expression is a critical component of the antiproliferative action of RB (18, 21, 22) .
The LNCaP (androgen-dependent) tumor cell line is a model system for the study of androgen-dependent proliferation (23, 24) . Androgen deprivation causes a documented reduction in cyclin D protein levels and CDK4/cyclin D kinase activity. As a result, RB remains underphosphorylated/active, and cyclin A protein levels (as well as CDK2 activity) are attenuated (25) . Thus, these cells arrest in G1 when androgen is removed. Conversion to androgen independence in LNCaP cells can be achieved by ectopic expression of viral oncoproteins E1A and T-Ag, which functionally disrupt RB (25) . Together, these results suggest that RB plays an important role in maintaining androgen dependence in prostatic tumor cells. Furthermore, this suggests that androgen independence occurs through signals that disrupt RB function, either through upstream or downstream signaling.
In an effort to understand the molecular changes that underlie androgen independence, we evaluated the ability of androgen-independent cells to transmit signals to RB and for RB to signal to its downstream effectors. We show that androgen-independent cells remained responsive to RB activation and that RB targets were invoked in an appropriate fashion. By contrast, RB inactivation in these cells was deregulated and occurred in the absence of androgen. Thus, signals upstream of RB are misregulated in androgen-independent prostate cells. Because Ras is known to link growth factor signaling to the cyclin D/RB network, we investigated the function of activated Ras in androgen-dependent cells. Surprisingly, although Ras did induce cyclin D1 in the absence of androgen, Ras had an overall negative effect on cell cycle progression that was p53 dependent. Even when the growth-inhibitory activity of Ras was disrupted with HPV-E6 or when CDK4/cyclin D1 was ectopically expressed, androgen-independent cell cycle progression failed to occur. These data indicate that androgen must regulate more than CDK4/cyclin D to initiate cell cycle progression. In fact, we demonstrate that androgen also regulates CDK2/cyclin activity, and the collective activities of CDK4/cyclin D and CDK2/cyclin complexes are required to inactivate RB and initiate cell cycle progression in the absence of androgen. These data demonstrate that maintenance of the androgen to RB pathway requires factors distinct from Ras and involves the combined regulation of CDK4/cyclin D and CDK2/cyclin complexes.
Results
Androgen Signaling to RB Is Deregulated in Androgen-independent
Cells.
The LNCaP cell line is androgen dependent and fails to incorporate
BrdUrd when cultured in CDT serum that lacks androgen (Fig. 1A
). We have shown previously that cell cycle progression in
LNCaP cells is restored by supplementing CDT with dihydrotestosterone
(25)
. In contrast, the androgen-independent cell lines
JCA-1 and TSUPr-1 continue to incorporate BrdUrd when cultured in CDT
(Fig. 1A
). Consistent with the BrdUrd incorporation results,
whereas androgen-independent cells proliferate in the absence of
androgen, androgen-dependent LNCaP fail to proliferate, as determined
by growth curves (not shown). When deprived of androgen, LNCaP cells
exhibited decreased levels of cyclin D1, and as a consequence RB was
dephosphorylated (Fig. 1
B, compare Lanes 5 and
6). Cyclin E expression remained constant in the presence or
absence of androgen, whereas cyclin A protein levels were diminished in
the absence of androgen (Fig. 1
B, Lanes 5 and 6).
These data are consistent with a model wherein androgen acts through
cyclin D1 to stimulate RB phosphorylation, which then inactivates RB
and derepresses cyclin A expression. Thus, we hypothesize that
androgen-independent cells have either lost this signal from androgen
to RB or from RB to its known downstream effectors. To distinguish
between these possibilities, JCA-1 and TSUPr-1 cells were propagated in
either the presence or the absence of androgen. Regardless of androgen
status, these cells maintained high levels of cyclin D1 protein and
demonstrated RB hyperphosphorylation, indicating that CDK4 activity was
also maintained (Fig. 1
B, compare Lanes 1 and
2, Lanes 3 and 4). Cyclin A protein levels
remained unchanged, consistent with the phosphorylation status of RB.
Cyclin E protein levels also remained constant (Fig. 1
B,
compare Lanes 1 and 2, Lanes 3 and 4).
These data indicate that androgen-independent cells have lost the
ability to regulate RB in response to androgen withdrawal. A possible
explanation for this finding would be that the androgen receptor in
these cell lines is constitutively active and thus functions in the
absence of androgen. However, immunoblot analysis of the androgen
receptor showed that JCA-1 and TSU-PR1 cells do not express detectable
receptor (Fig. 1
C, Lanes 1–4), whereas in LNCaP cells the
receptor was readily detectable (Fig. 1
C, Lanes 5 and
6). Thus, these androgen-independent cell lines have
subverted the requirement for the androgen receptor to promote cell
cycle progression.
|
). By contrast, cells transfected with p16ink4a or PSM-RB
showed significant reduction in BrdUrd incorporation in comparison to
vector-transfected cells or nontransfected cells from the same
coverslip. Together with our previous observation that PC3 and TSU-PR1
cells also arrest in response to active RB (17)
, these
results indicate that signaling of RB to its downstream targets is
intact in androgen-independent tumors.
|
B, Lanes 1–6). By
contrast, when the androgen-independent prostatic tumor cells were
either transfected with p16ink4a or PSM-RB, cyclin A protein levels
were reduced in comparison with the same cells transfected with
parental vector (Fig. 2
B, compare Lanes 1 and
3 with Lane 2 or Lanes 4 and
6 with Lane 5). In addition, the endogenous RB
maintained high levels of phosphorylation when transfected with PSM-RB
(Fig. 2
B, Lanes 3 and 6), whereas after
transfection with p16ink4a, only the underphosphorylated form of RB was
predominant in JCA-1 and TSUPr-1 (Fig. 2
B, Lanes 1 and
4). This indicates that as reported previously
(16)
, PSM-RB does not feed back to modify the
phosphorylation of endogenous RB, whereas p16ink4a blocks the CDK4
kinase activity that is required to initiate RB phosphorylation. Taken
together, these data show that the signaling pathway from androgen to
RB is lost in androgen-independent cells, but these cells have not lost
the ability to arrest in response to RB. Moreover, active RB or
p16ink4a transfected into androgen-independent cells mimics the effect
of androgen withdrawal in androgen-dependent cells (i.e.,
down-regulation of cyclin A but not cyclin E). Thus,
androgen-independent cells have achieved independence not by subverting
RB activity directly or by downstream disruption of RB signaling but
rather by dysregulation of upstream signaling pathways.
Ras Induction of Cyclin D Is Insufficient to Promote
Androgen-independent Proliferation.
These data suggest that androgen acts through cyclin D1 to stimulate RB
phosphorylation/inactivation, thus leading to induced cyclin A
expression and cell cycle progression. Consistent with this idea, a
kinetic study of androgen stimulation in LNCaP cells revealed that
increased cyclin D1 expression correlates with increased RB
phosphorylation that preceded the accumulation of cyclin A protein
(Fig. 3
A, Lanes 1–3). It is known that growth factors can act
through the Ras signaling pathway to activate cyclin D1 expression
(1, 2, 11, 12, 26)
. We therefore surmised that androgen
acts through Ras to stimulate cyclin D1 expression and inactivate RB.
To test whether activation of Ras is sufficient to promote
proliferation in LNCaP cells, we used activated V12Ras. To confirm that
the V12Ras (activated Ras) expression plasmid expressed active
protein when transfected into LNCaP cells in androgen-deprived media,
cells were transfected in CDT with either V12Ras expression plasmid or
parental vector. Cells were harvested 48 h after transfection, and
immunoblot analyses were performed. As shown in Fig. 3
B, Ras
protein was overexpressed when the expression plasmid was transfected
into LNCaP cells and could be readily detected through its HA epitope
(Fig. 3
B, compare Lanes 1 and 2). To
test V12Ras activity, ERK phosphorylation was also monitored, because
ERK is a major downstream target of Ras (27–29)
. No
significant difference in ERK protein expression was observed between
the two transfections, although ERK1 and ERK2 phosphorylation was
dramatically induced when LNCaP cells expressed V12Ras (Fig. 3
B, compare Lanes 1 and 2).
Furthermore, V12Ras led to an increase in the expression of cyclin D1
in CDT cultured LNCaP cells (Fig. 3
D, compare Lanes
1–4). Thus, these data confirmed that V12Ras is expressed and
active when transfected into the LNCaP cells in the absence of androgen
and acts to promote the accumulation of cyclin D1.
|
C, cells
transfected with vector alone failed to incorporate BrdUrd in the
absence of androgen. Strikingly, LNCaP cells transfected with activated
Ras also failed to demonstrate androgen-independent cell cycle
progression. By contrast, E1A rescued cell cycle progression of LNCaP
cells to approximately the same level as vector-transfected cells in
the presence of androgen. Using a similar protocol, V12Ras promoted
cell cycle progression in serum-starved Rat-1 cells (Fig. 3E
). Thus, V12Ras signaling is not sufficient to promote
cell cycle progression in LNCaP cells. Consistent with this finding,
although we observed an increase in cyclin D1 expression, we did not
observe an increase in cyclin A expression (Fig. 3D
), nor
did we observe RB phosphorylation (not shown). Cyclin E levels were
unaffected by the expression of V12Ras (Fig. 3D
).
V12Ras Inhibits Cell Cycle Progression in LNCaP Cells via p53.
Although the expression of cyclin D1 was stimulated when V12Ras was
transfected into LNCaP cells cultured in CDT, neither increased RB
phosphorylation nor was cyclin A expression detected. These data
suggested that V12Ras was either incapable of acting to activate the
cyclin D1-associated kinase activity, or alternatively that V12Ras was
stimulating an antiproliferative pathway. Because it has been shown
recently that V12Ras can inhibit proliferation of primary fibroblasts
(30)
, we determined whether V12Ras exerted a
growth-inhibitory effect in LNCaP cells. To test this, LNCaP cells
cultured in FBS were cotransfected with a GFP expression plasmid and
either vector control or V12Ras expression plasmids (Fig. 4B
). Cells were labeled with BrdUrd, and the incorporation of
BrdUrd in GFP-positive cells was determined. Although the vector had no
influence on BrdUrd incorporation, transfection of V12Ras significantly
inhibited BrdUrd incorporation, thus demonstrating that V12Ras inhibits
cell cycle progression in LNCaP cells.
|
A, compare Lanes 1 and
2 or Lanes 3 and 4). To test this
directly, LNCaP cells cultured in FBS were cotransfected with GFP
expression plasmid and either V12Ras or V12Ras and HPV-E6 expression
plasmids. HPV-E6 is known to promote p53 degradation and thereby
overcome the antiproliferative activity associated with p53 (Fig. 4B
). Although the GFP-positive cells cotransfected with
V12Ras expression plasmid arrested, coexpression of HPV-E6 partially
reversed this effect of V12Ras (Fig. 4B
).
Cyclin D Expression Is Not Sufficient for Androgen Independence.
Because the growth-inhibitory effects of V12Ras can be overcome by
HPV-E6, we reasoned that this combination should be sufficient to
render LNCaP cells androgen independent. LNCaP cells cultured in CDT
were cotransfected with GFP expression plasmid and either Vector, E1A,
or V12Ras + HPV-E6 expression plasmids. Consistent with our earlier
observations, expression of E1A promoted BrdUrd incorporation in CDT
(Fig. 4C
). Surprisingly, cotransfection of V12Ras with
HPV-E6 was not sufficient to drive androgen-independent cell cycle
progression (Fig. 4C
). However, E1A rendered LNCaP cells
androgen independent in the presence of V12Ras (Fig. 4D
),
further substantiating the idea that RB inactivation is required for
androgen-independent growth, and that V12Ras is incapable of fulfilling
this requirement in the absence of androgen (Fig. 4D
).
Collectively, these results suggest that additional mechanisms besides
the accumulation of D-type cyclins are at play in androgen-dependent
proliferation. To test this hypothesis, LNCaP cells were cotransfected
with CDK4/cyclin D1 and monitored for CDK4-associated kinase activity,
RB phosphorylation, and BrdUrd incorporation (Fig. 5
). Initially, CDK4 complexes were immunoprecipitated from transfected
cells and tested in in vitro kinase assays. As shown in Fig. 5
A, endogenous CDK4 activity was high in FBS (Fig. 5
,
Lane 2), and this activity was further enhanced by
coexpression of CDK4/cyclin D1 (Fig. 5
, Lane 3). These
observations were supported by in vivo RB kinase assays,
wherein cells were transfected with the LP fragment of RB (Fig. 5B
; Refs. 25, 31, 32
). As has been reported
previously, endogenous CDK4 activity was insufficient to drive LP
phosphorylation (Fig. 5
B, Lane 1; Ref. 25
),
whereas cotransfection of CDK4/cyclin D1 effectively drove LP
phosphorylation in the presence of FBS (Fig. 5
B, Lane 2).
Therefore, ectopic CDK4/cyclin D1 was shown to harbor RB kinase
activity by both in vivo and in vitro kinase
assays. LNCaP cells cultured in CDT also demonstrated enhanced in
vitro RB kinase activity upon transfection with CDK4 and cyclin D1
(Fig. 5
A, compare Lanes 4 and 5).
However, transfection of CDK4/cyclin D1 was insufficient to effectively
drive in vivo phosphorylation of LP (Fig. 5
B, middle
panel, Lanes 3 and 4). Therefore, ectopic CDK4/cyclin
D1 kinase activity is not sufficient to drive RB inactivation.
Consistent with this observation, ectopic CDK4/cyclin D1 expression did
not restore cyclin A expression in CDT (Fig. 5
B, bottom
panel). Moreover, ectopic CDK4/cyclin D1 was insufficient to drive
cell cycle progression in the absence of androgen (Fig. 5C
).
|
A, no changes in p16ink4a or p27Kip1 were observed in the
presence or absence of androgen; accordingly, no changes were observed
upon cotransfection of CDK4/cyclin D1 (compare Lanes 1–4).
As we have reported previously (25)
, cells cultured in the
absence of androgen exhibit low levels of p21Cip1, as compared with
cells cultured in the presence of androgen (compare Lanes 1
and 2 with Lanes 3 and 4). No changes
in p21Cip1 levels were observed upon cotransfection with CDK4/cyclin
D1. Collectively, these data reaffirm that androgen deprivation does
not result in increased expression of these CDK inhibitory proteins,
and in fact lack of androgen actually causes a decrease in p21Cip1.
Moreover, restoration of CDK4 activity does not affect the CDK
inhibitor profile.
|
B, LNCaP cells cultured in the presence of androgen
harbored significant CDK2-associated kinase activity, as determined by
in vitro kinase assays against an RB substrate, and ectopic
expression of CDK4/cyclin D1 did not enhance the endogenous CDK2
activity (top panel, Lanes 1–3). Consistent with previous
reports (25)
, CDK2-associated kinase activity is
attenuated in the absence of androgen, and CDK2 activity was not
restored by the ectopic expression of CDK4/cyclin D1 (Fig. 6
B,
top panel, Lanes 4 and 5). A trivial explanation for
this would be that lack of cyclin A is responsible for the attenuation
in CDK2 kinase activity. Therefore, we analyzed the kinase activity
specifically associated with cyclin E (Fig. 6
B, bottom
panel). To do this, cyclin E complexes were immunoprecipitated and
used in in vitro kinase assays against the RB substrate.
Cyclin E kinase activity was not influenced by ectopic expression of
CDK4/cyclin D1, and cyclin E-associated activity was attenuated in the
absence of androgen (Fig. 6
B, bottom panel, compare
Lanes 2 and 3 and Lanes 4 and
5).
To investigate the activity of CDK2 for in vivo RB
phosphorylation, ectopically expressed cyclin E or cyclin A was
monitored for its ability to phosphorylate the LP fragment of RB. Both
cyclin E and cyclin A overproduction promoted RB phosphorylation in the
presence of androgen; however, in the absence of androgen, RB
phosphorylation was inhibited (Fig. 6C
). Because CDK4/cyclin
D activity is required to prime the phosphorylation of RB for
subsequent hyperphosphorylation by CDK2/cyclin complexes, we
reasoned that the inhibition of cyclin D expression and CDK4 activity
may underlie the inhibition of RB phosphorylation when cyclin E and
cyclin A are overexpressed. To test this idea, CDK4, cyclin D1, cyclin
E, and cyclin A were coexpressed with the LP fragment, and the status
of LP phosphorylation was monitored in the presence or absence of
androgen. As shown in Fig. 6
D, the LP fragment of RB was
efficiently phosphorylated by this combination of CDK/cyclins in
both the presence and absence of androgen (compare Lanes 2
and 4), thereby indicating that these conditions overcome
both the CDK4 and CDK2 inhibition observed in androgen-deprived LNCaP
cells. Because RB is a critical determinant of androgen-dependent cell
cycle progression, we determined whether the cells cotransfected with
CDK4, cyclin D1, cyclin E, and cyclin A were able to progress through
the cell cycle in the absence of androgen. As shown in Fig. 6
E, vector-transduced cells failed to incorporated BrdUrd,
whereas cells transfected with E1A expression plasmid incorporated
BrdUrd in the absence of androgen. Ectopic coexpression of CDK4,
cyclin D1, cyclin E, and cyclin A also promoted cell cycle progression
in the absence of androgen. Together, these findings indicate that the
control of RB phosphorylation occurs at the level of both CDK4/cyclin D
and CDK2/cyclin kinase activities, and combined deregulation of both
components leads to androgen-independent cell cycle progression (Fig. 6E
).
Discussion
Here we show that androgen acts through both CDK4/cyclin D and CDK2/cyclin activities to promote RB inactivation and cell cycle progression. We also show that this process is deregulated in androgen-independent tumor cells, whereas signaling capabilities from RB to its downstream effectors remain intact. Thus, this report delineates mechanisms used by androgen to initiate cell cycle progression and underscores the impact of deregulated RB inactivation in prostatic tumor development.
Initially, we demonstrated that in androgen-dependent prostate
cancer cells (LNCaP), androgen withdrawal led to a cessation of cell
cycle progression that correlated with a reduction in cyclin D1
expression, RB phosphorylation, and cyclin A expression. In contrast,
in androgen independent cell lines (JCA-1 and TSU-PR1), androgen
withdrawal had no influence on the cyclin D/RB pathway (Fig. 1)
. In
fact, these androgen-independent cell lines no longer required
expression of the androgen receptor to simulate cyclin D1 expression.
Thus, mutations in these cells must have rendered them capable of
progressing through the cell cycle independent of androgen action. To
determine whether these lesions act upstream or downstream of RB, we
expressed either p16ink4a or PSM-RB in the androgen-independent cell
lines (Fig. 2)
. These reagents led to the cessation of cell cycle
progression and reduced the levels of cyclin A protein, similar to that
observed in LNCaP cells upon androgen withdrawal. These findings
indicate that the lesions that lead to cell cycle dysregulation in
androgen-independent prostate cancer act upstream of RB. It is known
that Ras activity stimulates cyclin D1 expression, thus activating
CDK4/cyclin D1 complexes and initiating RB phosphorylation/inactivation
(1, 2, 11, 12, 26, 34) . We hypothesized that androgen acts
through Ras signaling to stimulate RB inactivation and cell cycle
progression in androgen-dependent cells. Surprisingly, although V12Ras
stimulated the expression of cyclin D1, it did not promote RB
phosphorylation or androgen-independent proliferation, whereas E1A
rendered LNCaP cells androgen independent (Fig. 3)
. Interestingly,
V12Ras exerted a negative effect on LNCaP proliferation that was p53
dependent. Although HPV-E6 overcame the growth-inhibitory action of
V12Ras in LNCaP cells, it was incapable of cooperating with Ras to
induce androgen independence (Fig. 4)
. Similarly, although ectopic
expression of cyclin D1/CDK4 promoted RB phosphorylation in the
presence of androgen, it was incapable of inactivating RB or driving
cell cycle progression in the absence of androgen (Fig. 5)
. Thus,
cyclin D expression as stimulated by V12Ras or via its ectopic
expression is not sufficient for androgen-independent proliferation,
indicating that androgen acts through an additional signaling pathway.
We show that CDK2 and cyclin Eassociated activity is also
regulated by androgen, and that combined action of CDK4/cyclin D1 and
CDK2/cyclin complexes are sufficient to inactivate RB and drive
androgen-independent cell cycle progression (Fig. 6)
. Thus, the data
presented in this report demonstrate that androgen signals to both CDK4
and CDK2 to inactivate RB and suggest that this requirement is
subverted in androgen-independent prostate cancers.
Ras in Androgen-dependent Proliferation.
Prostatic epithelia and early-stage prostate cancer cells require
androgen for proliferation, as in the absence of androgen, prostate
tumors cease proliferation and undergo a degree of apoptosis
(24)
. At present, the relative contributions of
proliferation and apoptosis to the therapeutic treatment of prostate
cancer remain controversial (7, 8, 24)
. However, it is
clear that if the prostatic cancer cells fail to proliferate, the
cancer will be controlled. How androgen acts as a prostatic growth
factor is largely not understood. Androgen, like many other growth
factors, stimulates proliferation through its cognate receptor. The
androgen receptor acts as a DNA-binding transcriptional modulator,
which presumably changes the pattern of gene expression to stimulate
proliferation (24)
. Although a number of
androgen-responsive genes have been cloned (e.g.,
prostate-specific antigen), the role of these proximal targets in
prostate proliferation is not well defined (24)
. However,
it is apparent that these early events are integrated to stimulate the
activation of cyclin D1 expression and associated kinase activity,
which subsequently initiates phosphorylation/inactivation of RB
(25)
. The importance of these events in cellular
proliferation is underscored by the action of p16ink4a or PSM-RB, as
both of these reagents are effective at inhibiting the proliferation of
all prostate cancer cell lines tested (16, 17)
.
Furthermore, because androgen withdrawal attenuates RB phosphorylation
and cyclin A expression in androgen-dependent cells, p16ink4a or PSM-
RB simulated these same responses in androgen-independent cells.
Therefore, we sought to delineate the signaling pathway between androgen and RB. Ras is known to mediate signals that stimulate both the expression of cyclin D1 and the phosphorylation of RB. In LNCaP cells, we show that V12Ras or V12Ras + HPV-E6 stimulates cyclin D1 expression in the absence of androgen. However, restored cyclin D1 did not lead to RB inactivation or cell cycle progression in the absence of androgen. This contrasts with what is observed in immortalized fibroblastic cells, wherein V12Ras stimulates the expression of cyclin D1 and drives RB phosphorylation and cell cycle progression under conditions of serum starvation or anchorage deprivation (35, 36) .
The role of Ras in the androgen dependence of prostate cancer is
polemical. Activating Ras mutations do occur in prostate cancer,
although the overall frequency of these mutations is quite low
(37–41)
. In contrast, increased loss of RB correlates
with the development of androgen-independent cancer after androgen
blockade therapy, and mutation of RB is observed at relatively high
frequency in prostate cancers (42–44)
. In culture, loss
of RB signaling through the use of viral oncoproteins such as E1A or
T-Ag is sufficient to enable LNCaP cells to progress through the cell
cycle in the absence of androgen (25)
. Although others
have tested the influence of Ras on androgen dependence in LNCaP
cells, the results are controversial (45, 46)
. Our results
clearly show that in bulk-transfected LNCaP cells (which have not been
selected), V12Ras does not induce androgen-independent proliferation,
although Ras activity was apparent (Fig. 3)
. Importantly, under
virtually identical conditions, V12Ras induced serum-independent cell
cycle progression in Rat-1 cells (not shown). Thus, Ras is specifically
defective in promoting androgen-independent growth. Furthermore, we
actually found that V12Ras induces cell cycle inhibition in LNCaP
cells, which is mediated by a p53-dependent pathway. Importantly, even
when we disrupt this antiproliferative signaling from Ras using the
HPV-E6 protein, a condition which transforms primary fibroblasts, we
were unable to achieve androgen-independent cell cycle progression.
Together, these data may explain why Ras mutations are rare events in
prostate tumorigenesis (37)
. Complementary to these
findings, Ravi et al. (46)
used a Raf/estrogen
receptor-inducible system to demonstrate that overt activation of Raf
induces cell cycle arrest in LNCaP cells. By contrast, Voeller et
al. (45)
showed that cadmium-inducible Ras in LNCaP
cells leads to a modest proliferative advantage in the absence of
androgen. However, the extent of proliferation in this system was quite
reduced in comparison with androgen treatment and was clone dependent.
Because V12Ras induces cell cycle arrest in cells containing functional
p53 signaling pathways, the data of Voeller et al.
(45)
could reflect the loss of p53 during the selection of
clones. However, our results with V12Ras and E6 suggest that additional
events must be involved.
Androgen Regulation of CDK4 and CDK2 Activities.
The results presented here suggest that in prostatic epithelia,
androgen-mediated proliferation likely involves two sets of signals;
one that may be mediated by Ras, which stimulates the expression and
activity of cyclin D1, and another signal that is required for complete
RB inactivation and the expression of cyclin A. Although CDK4/cyclin D1
in vitro kinase activity was restored in the absence of
androgen by transfection of CDK4 and cyclin D1 expression plasmids,
ectopic CDK4 demonstrated little RB kinase activity in vivo
and did not relieve RB-mediated repression of cyclin A (Fig. 5)
.
Moreover, ectopic CDK4/cyclin D1 did not promote androgen-independent
proliferation. These data are also consistent with the failure of
V12Ras and E6 to promote cell cycle progression while restoring cyclin
D1 expression. We show that androgen withdrawal also inhibits CDK2 and
cyclin E-associated kinase activity. The ectopic overexpression of
cyclin E or cyclin A drove RB phosphorylation in the presence of
androgen but failed to do so in the absence of androgen. These findings
agree with the hypothesis that CDK4/cyclin D-mediated phosphorylation
was also important for priming RB for subsequent inactivation of RB by
CDK2-mediated phosphorylation. The combined importance of these two
signaling pathways is exemplified by the observation that ectopic
expression of CDK4, cyclin D1, cyclin E, and cyclin A was sufficient to
drive RB phosphorylation in the absence of androgen and render LNCaP
cells androgen independent for cell cycle progression (Fig. 6)
.
Together, these data indicate that androgen acts through both
CDK4/cyclin D and CDK2/cyclin complexes to promote cell cycle
progression.
How androgen regulates the activity of CDK/cyclin complexes is still a subject of study. We demonstrate that cyclin D1 protein is diminished in the absence of androgen. Because cyclin D1 is regulated both at the level of transcription and protein stability (1, 2) , it is probable that androgen regulates cyclin D1 at one of these levels. We have observed that androgen stimulates cyclin D1 promoter activity in reporter assays (data not shown). Similarly, Ras stimulates cyclin D1 promoter activity, suggesting that the androgen receptor may signal through Ras to stimulate cyclin D1 transcription and result in protein accumulation (11, 34, 35) .
Although ectopic CDK4/cyclin D1 expression is capable of phosphorylating RB in vitro, this complex lacks the ability to fully phosphorylate and inactivate RB in vivo. This finding suggested that the CDK2/cyclin complexes that mediate RB hyperphosphorylation (47, 48) must be attenuated by androgen withdrawal independently of CDK4/cyclin D1 activity. Consistent with this idea, we find that CDK2 and cyclin E-associated activity (but not cyclin E expression) are down-regulated via androgen withdrawal, even when CDK4/cyclin D1 activity is restored. CDK2/cyclin complexes are regulated by a myriad of mechanisms. Because cyclin A expression is inhibited, this would partially explain the decrease in CDK2 activity but fails to explain the specific deficiency in cyclin E-associated kinase activity. It is unlikely that accumulation of the CDK2-inhibitory molecules (i.e., p21Cip1 and p27Kip1) or CDK-inhibitor shuffling (as mediated by p16ink4a induction) accounts for the attenuation of cyclin E-associated kinase activity, because none of these inhibitory molecules were induced by androgen withdrawal. Thus, androgen regulates cyclin E-associated kinase activity by as of yet undefined mechanisms.
Acquisition of Androgen Independence.
The progression of prostatic cancer from androgen dependence to
androgen independence represents a catastrophic consequence for
patients suffering from the disease (24)
. At present,
endocrine therapy represents the only viable treatment for those with
early-stage prostate cancer, and no effective therapy exists for those
individuals with androgen-independent prostate cancer (7, 8)
. As with the mechanism through which androgen exerts its
mitogenic effect, little is known regarding how prostate cells can
bypass this requirement (24)
. In most established
androgen-independent tumor cell lines, androgen receptor expression is
actually lost (24)
. For TSU-Pr1, this has been attributed
to methylation of the androgen receptor promoter (49)
.
JCA-1 cells express prostate-specific antigen (a target of the androgen
receptor; Ref. 50
); therefore, we suspected that JCA-1
cells may express the androgen receptor. However, we failed to detect
the expression of AR protein in either cell line. Because no androgen
receptor protein is detectable, both cell lines must have activated
mechanisms for the continued proliferation in its absence.
In cancer, a variety of mechanisms are used to lead to deregulate proliferation. These mechanisms can be loosely divided into those that act upstream of RB (e.g., deregulated mitogenic signaling or cyclin D1 amplification) and those that act downstream of RB (e.g., viral oncoproteins of DNA tumor viruses or deregulated cdk2 activity). Our finding that the TSU-PR1 and JCA-1 cell lines are arrested by PSM-RB and p16ink4a indicates that deregulation of the cell cycle acts upstream of RB. Although we focused on oncogenic Ras, we also tested the action of activated ß-catenin and c-myc. Neither of these proteins stimulated androgen-independent cell cycle progression in LNCaP cells (data not shown), although ß-catenin and c-myc are known to stimulate cyclin D1 expression (51) . Furthermore, ectopic expression of CDK4/cyclin D1 also failed to inactivate RB or stimulate cell cycle progression in CDT. In fact, the only proteins capable of converting LNCaP cells to androgen independence were proteins that lead to RB inactivation in the absence of androgen.
In summary, we show that the signal from androgen to RB is misregulated in androgen-independent tumor cells, whereas the ability of RB to signal to its downstream effectors is precisely maintained. These data suggest that reactivation of RB in androgen-independent tumors should be a goal of anticancer therapeutics. In focusing on the signal from androgen to RB, we also demonstrate that this signal has at least two components: one, which may be Ras mediated, leading to the stimulation of cyclin D expression and CDK4 activity; and the other, which is Ras independent and completes RB inactivation through the action of CDK2/cyclin complexes. These findings underscore the importance of RB as an integrating target for cell cycle control and tumor progression in prostate cancer. Clearly, the pathways regulating RB are tightly controlled, and the deregulated phosphorylation of RB that occurs in androgen-independent tumors requires the dysregulation of multiple signaling pathways.
Materials and Methods
Cell Culture.
The human prostatic adenocarcinoma cell line, LNCaP, was obtained
from the American Type Culture Collection. The TSUPr-1
androgen-independent cell line was kindly provided by Dr. John Isaacs
(Johns Hopkins), and the JCA-1 cell lines was obtained from Dr. J. W.
Chiao (New York Medical College). Cells were passaged fewer than 10
times in the experiments described. For regular passage, cells were
grown in Improved MEM (Biofluids) containing 10% heat-inactivated FBS
(Hyclone) supplemented with 100 units/ml penicillin-streptomycin and 2
mM L-glutamine at 37°C in a humidified
atmosphere of 5% CO2. For growth in an
androgen-depleted media, cells were propagated in Improved MEM
containing 10% CDT FBS (Hyclone). Rat-1 fibroblastic cells were kindly
provided by Dr. D. Green (La Jolla Institute for Allergy and
Immunology, La Jolla, CA). Cells were maintained in DMEM supplemented
with 10% FBS, 100 units/ml penicillin-streptomycin, and 2
mM L-glutamine at 37°C in a humidified
atmosphere of 5% CO2. Transfected JCA-1 and
TSUPr-1 cells were selected 24 h after transfection with 2 µg/ml
puromycin (Sigma) for 48 h of selection.
Plasmids.
The pH2B-GFP plasmid, encoding histone H2B fused to the GFP, was
obtained from Dr. Geoff Wahl (Salk Institute). The pPSM-RB and p16ink4a
plasmids have been described previously (15)
. Dr. Gilbert
Morris (Tulane University) supplied the E1A expression plasmid, Drs.
Kenji Fukasawa and Tony Capobianco (University of Cincinnati) provided
the V12Ras plasmid, and the RSV-cyclin D1 construct was kindly provided
by Dr. C. Sherr (St. Jude Children’s Research Hospital, Memphis, TN).
The puromycin resistance plasmid, pBABE-PURO, was kindly provided by
Dr. H. Land (Imperial Cancer Research Fund, London, United Kingdom).
Dr. Jean Wang (University of California, San Diego, CA) provided the
WTLP expression plasmid, and the CDK4 expression plasmid was obtained
from Dr. Webster K. Cavenee (Ludwig Institute, La Jolla, CA). Dr. Karen
Vousden (National Cancer Institute, Frederick, MD) kindly provided the
HPV-E6 expression plasmid.
Transfections/BrdUrd Incorporation.
Approximately 2.5 x 105 LNCaP cells
were seeded onto poly-L-lysine-coated coverslips in each
well of a six-well dish. After 48 h, wells were washed one time
with serum-free Improved MEM. Lipofectin substrate was then applied and
allowed to incubate for 6 h, after which the serum-containing
medium indicated was added back. To monitor protein expression, cells
were harvested 48 h after transfection. To monitor S-phase
progression, cotransfection with pH2B-GFP (0.12 µg) and a secondary
effector expression plasmid (5 µg) were carried out. Cells were
labeled 48 h after transfection with Cell Proliferation Labeling
Reagent (Amersham Pharmacia Biotech), according to the manufacturer’s
recommended protocol. Pulse labeling was continued for 14 h, at
which time cells were fixed and processed for immunofluorescence. Rat-1
cells were seeded at a density of 2 x 104
cells onto coverslips in each well of a six-well dish. Twenty-four h
later, the cells were transfected with 2 µg of total DNA (as
indicated) with FuGENE 6 transfection reagent, according to the
manufacturer’s protocol (Roche Diagnostics). JCA-1 and TSUPr-1 cells
were seeded on coverslips at a density of 1 x
105 cells/well of a six-well dish. Twenty-four h
later, the cells were transfected with 4 µg of total plasmid DNA (as
indicated) using the BBS/calcium phosphate method described previously
(52)
. Forty-eight h after transfection, Cell Proliferation
Labeling Reagent (Amersham) was added according to the manufacturer’s
protocol. Sixteen h later, cells were fixed with formaldehyde and
processed for indirect immunofluorescence to detect BrdUrd
incorporation as described previously (53)
. For each
experiment, at least 150 transfected (GFP-positive) and untransfected
(GFP-negative) cells were counted. Data shown reflect the average of at
least 2–3 independent experiments. Images were captured using a Nikon
Axiophot at x40 and a SpotCam digital camera.
Quiescence.
Twelve h after transfection, Rat-1 cells were washed and placed in DMEM
supplemented with 0.1% FBS, 100 units/ml penicillin-streptomycin, and
2 mM L-glutamine for 48 h prior to BrdUrd
labeling.
Immunoblotting.
Immunoblotting was carried out by use of standard procedures as
described previously (25)
. Blots were probed for the
following proteins with polyclonal antibodies: RB (851; gift of Dr.
Jean Wang, University of California at San Diego, San Diego, CA);
Cyclin D (sc-182 and sc-753; Santa Cruz Biotechnology); cyclin A
(sc-751; Santa Cruz Biotechnology); cyclin E (sc-198; Santa Cruz
Biotechnology); Ras (sc-519; Santa Cruz Biotechnology); ERK (sc-93-G;
Santa Cruz Biotechnology); phospho-ERK#2 (9101S; New England BioLabs);
androgen receptor (sc-816; Santa Cruz Biotechnology); p21(sc-757; Santa
Cruz Biotechnology) and with monoclonal antibodies: phosphoERK#1
(sc-7383; Santa Cruz Biotechnology); HA-tag (MMS-101R; Covance); p53
(OP-43; Calbiochem). Goat antimouse horseradish peroxidase, goat
antirabbit horseradish peroxidase, protein A-horseradish peroxidase
(Bio-Rad), or antigoat (sc-2020; Santa Cruz Biotechnology) were used
for antibody visualization via enhanced chemiluminescence (Amersham
Pharmacia Biotech).
Immunoprecipitations and CDK Kinase Assays.
For in vitro kinase assays, 2–4 x
106 cells from conditions indicated were
harvested by trypsinization and washed in PBS. To analyze CDK2
activity, cells were lysed and clarified as described previously
(25)
. After quantitation, 160 µg of lysate were
incubated for 2 h (4°C with rotation) with either anti-CDK2
antiserum (Santa Cruz Biotechnology), anti-cyclin E antiserum (Santa
Cruz Biotechnology), or anti-E1A antiserum (Santa Cruz Biotechnology)
as a negative control. Immunoprecipitations and kinase reactions were
performed as described previously (25)
. CDK4 kinase assays
were performed using the technique of LaBaer et al.
(54)
.
Time Course.
LNCaP cells were seeded in medium containing 10% CDT at a density of
1.1 x 106 cells/6-cm dish. After 96 h,
medium was replaced by 10% FBS. Cells were harvested at the times
indicated in the experiment.
Acknowledgments
We thank Dr. Webster K. Cavenee, Dr. Kenji Fukasawa, Dr. Zvjezdana Sever-Chroneos, and Dr. Larry Sherman for critical reading of the manuscript. We also thank Drs. Webster K. Cavenee, Kenji Fukasawa, Peter J. Stambrook, Karen Vousden, and Geoffrey Wahl for a generous supply of reagents.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported in part by Grant
R01-CA82525-01 from the National Cancer Institute-NIH and a grant from
the Sidney Kimmel Cancer Foundation (to E. S. K.). K. E. K. is
supported by Postdoctoral Fellowship CA82034 from the National Cancer
Institute. 
2 To whom requests for reprints should be
addressed, at Department of Cell Biology, Neurobiology, and Anatomy,
University of Cincinnati College of Medicine, P. O. Box 670521, 3125
Eden Avenue, Cincinnati, OH 45267-0521. Phone: (513) 558-6849; Fax:
(513) 558-4454; E-mail: Erik.Knudsen{at}UC.edu
3 The abbreviations used are: ERK, extracellular
signal-regulated kinase; CDK, cyclin-dependent kinase; RB,
retinoblastoma tumor suppressor protein; BrdUrd,
bromodeoxyuridine; CDT, charcoal dextran treated; GFP, green
fluorescent protein; HPV, human papillomavirus; LP, large pocket; FBS,
fetal bovine serum.
Received for publication 1/ 4/00. Revision received 5/11/00. Accepted for publication 5/11/00.
References
This article has been cited by other articles:
 |
|
 |
|
  D. K. Bowles, K. K. Maddali, V. C. Dhulipala, and D. H. Korzick PKC{delta} mediates anti-proliferative, pro-apoptic effects of testosterone on coronary smooth muscle Am J Physiol Cell Physiol, August 1, 2007; 293(2): C805 - C813. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Sharma, C. E.S. Comstock, E. S. Knudsen, K. H. Cao, J. K. Hess-Wilson, L. M. Morey, J. Barrera, and K. E. Knudsen Retinoblastoma Tumor Suppressor Status Is a Critical Determinant of Therapeutic Response in Prostate Cancer Cells Cancer Res., July 1, 2007; 67(13): 6192 - 6203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xu, S.-Y. Chen, K. N. Ross, and S. P. Balk Androgens Induce Prostate Cancer Cell Proliferation through Mammalian Target of Rapamycin Activation and Post-transcriptional Increases in Cyclin D Proteins. Cancer Res., August 1, 2006; 66(15): 7783 - 7792. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. N. Papandreou and C. J. Logothetis Bortezomib as a Potential Treatment for Prostate Cancer Cancer Res., August 1, 2004; 64(15): 5036 - 5043. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Zhang, M. Li, H. Wang, S. Agrawal, and R. Zhang Antisense therapy targeting MDM2 oncogene in prostate cancer: Effects on proliferation, apoptosis, multiple gene expression, and chemotherapy PNAS, September 30, 2003; 100(20): 11636 - 11641. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. E. Petre-Draviam, S. L. Cook, C. J. Burd, T. W. Marshall, Y. B. Wetherill, and K. E. Knudsen Specificity of Cyclin D1 for Androgen Receptor Regulation Cancer Res., August 15, 2003; 63(16): 4903 - 4913. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Solomon, Y. Wang, S. R. Fox, T. C. Lambeck, S. Giesting, Z. Lan, A. M. Senderowicz, and E. S. Knudsen Cyclin D1 Splice Variants: DIFFERENTIAL EFFECTS ON LOCALIZATION, RB PHOSPHORYLATION, AND CELLULAR TRANSFORMATION J. Biol. Chem., August 8, 2003; 278(32): 30339 - 30347. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G.-L. Jiang, C. R. White, H. Y. Stevens, and J. A. Frangos Temporal gradients in shear stimulate osteoblastic proliferation via ERK1/2 and retinoblastoma protein Am J Physiol Endocrinol Metab, August 1, 2002; 283(2): E383 - E389. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. B. Wetherill, C. E. Petre, K. R. Monk, A. Puga, and K. E. Knudsen The Xenoestrogen Bisphenol A Induces Inappropriate Androgen Receptor Activation and Mitogenesis in Prostatic Adenocarcinoma Cells Mol. Cancer Ther., May 1, 2002; 1(7): 515 - 524. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Wu, T. L. Foreman, C. W. Gregory, M. A. McJilton, G. G. Wescott, O. H. Ford, R. F. Alvey, J. L. Mohler, and D. M. Terrian Protein Kinase C{epsilon} Has the Potential to Advance the Recurrence of Human Prostate Cancer Cancer Res., April 1, 2002; 62(8): 2423 - 2429. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. E. Petre, Y. B. Wetherill, M. Danielsen, and K. E. Knudsen Cyclin D1: Mechanism and Consequence of Androgen Receptor Co-repressor Activity J. Biol. Chem., January 11, 2002; 277(3): 2207 - 2215. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cell Growth & Differentiation |
14 h. Subsequent to the pulse, cells were fixed and
processed to monitor BrdUrd incorporation via direct
immunofluorescence. Data shown are the average of at least three
independent experiments in which at least 150 cells/experiment were
analyzed; bars, SD. Representative
immunofluorescence from the experiment described are shown in the
inset. B, JCA-1, TSUPr-1, and LNCaP cells
were propagated for 96 h in the presence of either 10% CDT
(Lanes 2, 4, and 6) or 10% FBS
(Lanes 1, 3, and 5), harvested, and
clarified, and protein lysates were prepared. Equal protein was
resolved by SDS-PAGE, and the endogenous cyclin A, cyclin E, cyclin D1,
and RB proteins were detected by immunoblotting with the respective
antibodies. pRB, underphosphorylated RB;
ppRB, hyperphosphorylated RB. C, lysates
from JCA-1, TSUPr-1, and LNCaP cells as described in
B were resolved by SDS-PAGE, and the androgen receptor
(AR) was detected by immunoblotting.
