| 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 Functional Genetics, Glaxo Wellcome Inc., Research Triangle Park, North Carolina 27709
Abstract
Polo-like kinase 1 (PLK1), which has been shown to have a critical role in mitosis, is one possible target for cancer therapeutic intervention. PLK1, at least in Xenopus, starts the mitotic cascade by phosphorylating and activating cdc25C phosphatase. Also, loss of PLK1 function has been shown to induce mitotic catastrophe in a HeLa cervical carcinoma cell line but not in normal Hs68 fibroblasts. We wanted to understand whether the selective mitotic catastrophe in HeLa cells could be extended to other tumor types, and, if so, whether it could be attributable to a tumor-specific loss of dependence on PLK1 for cdc25C activation. When PLK1 function was blocked through adenovirus delivery of a dominant-negative gene, we observed tumor-selective apoptosis in most tumor cell lines. In some lines, dominant-negative PLK1 induced a mitotic catastrophe similar to that published in HeLa cells (K. E. Mundt et al., Biochem. Biophys Res. Commun., 239: 377–385, 1997). Normal human mammary epithelial cells, although arrested in mitosis, appeared to escape the loss of centrosome maturation and mitotic catastrophe seen in tumor lines. Mitotic phosphorylation of cdc25C and activation of cdk1 was blocked by dominant-negative PLK1 in human mammary epithelial cells as well as in the tumor lines regardless of whether they underwent mitotic catastrophe. These data strongly argue that the mitotic catastrophe is not attributable to a lack of dependence for PLK1 in activating cdc25C.
Introduction
Accurate transmission of genetic information during cell division is dependent on the fidelity of chromosome repair, replication, and segregation. Checkpoint controls monitor these processes for accuracy and completion and, when errors are detected, arrest the cell cycle to allow sufficient time for correction of the errors and completion of these events (1, 2, 3) . Deregulated expression of, or mutations in, cell cycle checkpoint genes can lead to genetic instability and have been associated with the development of cancer (4, 5, 6, 7) .
In mammalian cells, the G2-M checkpoint that responds to unreplicated or damaged DNA can be activated by phosphorylation of Thr-14 and Tyr-15 on cdk12 (also called cdc2; Refs. 8, 9, 10, 11, 12 ). Removal of these phosphorylations by the dual-specificity phosphatase cdc25C is a key event in the prophase-to-metaphase transition of the cell cycle (13) . cdc25C itself undergoes NH2-terminal phosphorylation to become fully active (14, 15, 16, 17) . Early on, cdk1 was identified as a kinase capable of activating cdc25C, which suggested that these two partners could participate in an autoamplification loop (18 , 19) . However, the specific kinase that triggers cdc25C activation was not known (20) . In 1996, Kumagai and Dunphy (21) identified a cdc25 phosphorylating kinase, Plx1, from frog oocytes that turned out to be a homologue of the conserved Drosophila gene called polo (see below). Both biochemical and cell-cycle-timing experiments have been used to make a strong case for Plx1 in activation of cdc25 function (21, 22, 23) .
The polo gene was first identified in mutant screens of Drosophila as causing multiple types of mitotic abnormalities (24) . Homologues have been identified in frogs (Plx1), yeast (Cdc5 in Saccharomyces cerevisiae and Plo1 in Schizosaccharomyces pombe), and mammals (Plk1) that likewise have critical roles in mitosis (25 , 26) . All of the PLK1 homologues share a conserved 30-amino-acid sequence called the polo-box as well as additional conserved sequences in the NH2-terminal kinase domain (25 , 26) . Other polo family members [e.g., proliferation related kinase/fibroblast growth factor inducible kinase and serum inducible kinase (PRK/FNK and SNK)] have been identified through sequence homology, although their functional roles may not be exclusive to mitosis (27, 28, 29, 30, 31) . Interestingly, human PRK has been shown to phosphorylate cdc25C in vitro (30) . Whether PLK family members are the sole activating kinases for cdc25C or can play other roles in activating cdk function is not known at this time (25 , 26 , 32) .
The change in localization of PLK1 during mitosis suggests that PLK1 substrates could be located at multiple sites including centrosomes in prophase, spindle pole bodies in metaphase, and cytokinetic bridges in telophase (33 , 34) . This pattern parallels that seen with a mAb MPM-2, which recognizes mitotic phosphorylations on at least 50 proteins (35, 36, 37, 38) , including cdc25C (21 , 39) . Moreover, the Drosophila polo protein appears to be required for MPM-2 reactivity in vivo (40) . Consistent with its changing localization, PLK1 has been implicated in functions as diverse as centrosome maturation, microtubule dynamics, and cytokinesis (34 , 41, 42, 43) . The substrates identified to date, including tubulins and mitotic kinesin-like proteins as well as anaphase promoting complex subunits, are consistent with the multiple roles of PLK1 in the G2-to-M transition and mitotic exit (34 , 41 , 44 , 45) . PLK1 is also necessary for the destruction of mitotic substrates at the end of mitosis (46 , 47) . In view of its multiple targets and functions, PLK1 appears to be a pleiotropic regulator of mitosis.
Abrogation of PLK1 function through antibody microinjection was shown to block a centrosome maturation step in both normal H68 human fibroblasts and HeLa cervical carcinoma cells (42) , but only the latter entered into mitosis and cell death. Because microsurgical dissection of centrosomes has been shown to block all aspects of mitosis (48) , Lane and Nigg (42) hypothesized that PLK1 may regulate a centrosome maturation checkpoint that is defective in HeLa cells. They further suggested that the HeLa cell defect resulted from a loss of dependence for PLK1 in cdc25C activation. To more closely study the role of PLK1 in tumor-specific apoptosis and checkpoint function, we used adenovirus to deliver and overexpress dnPLK1 in multiple cell types. Our studies demonstrate that inhibition of PLK1 function is proapoptotic in most tumor lines compared with normal cell lines and that this antitumor activity is independent of the role of PLK1 in activating cdc25C phosphatase.
Results
dnPLK1 Is Proapoptotic in Most Tumor Cells Relative to Normal Cell
Counterparts.
Previous studies showed that microinjection of PLK1 antibodies induced
mitotic catastrophe in a HeLa cervical carcinoma cell line, whereas
Hs68 normal human diploid fibroblasts arrested in the
G2 phase of the cell cycle (42)
.
Likewise, transient transfection of a dnPLK1 plasmid has been shown to
elicit mitotic delay and cytokinetic abnormalities in HeLa cells,
although the same comparison to normal cells was not reported
(43)
. The mitotic catastrophe in HeLa cells injected with
PLK1 antibodies was attributed to a defect in a centrosome maturation
checkpoint, perhaps related to a loss of PLK1-dependent cdc25C
activation, although the results were not extrapolated to other tumors,
nor was an assessment of cdc25C function performed (42)
.
To address whether PLK1 inhibition caused tumor selective apoptosis in a range of tumor types, a dnPLK1 mutant (K82M) was cloned behind a cytomegalovirus promoter in an E1-defective adenoviral vector, and the gene product delivered via infection into 10 tumor lines and 3 normal lines. The 10 tumors were selected based on their diverse origins and included breast carcinomas (MDA MB 231, MDA MB 435S, HS578T, MCF-10A, and MCF7), osteosarcomas (SAOS-2 and U-2OS), an ovarian carcinoma (SKOV3), a colon carcinoma (SW480), and a lung carcinoma (NCI H596). They were also selected based on their relative inability to complement E1 function in the adenovirus, thus circumventing the complicating cytopathic effects that result from viral DNA replication seen in HeLa and other tumor cell types (49) . The tumors were compared with two normal cell lines of epithelial origin—HMECs and NHKs—as well as a normal human diploid fibroblast cell line (IMR90). Expression of dominant negative protein was accurately quantitated relative to the endogenous protein (see "Materials and Methods") at the 24-h time point largely to avoid any effects of cell death on the quantitation. A high multiplicity of infection, 500 MOI (equivalent to 500 pfu per cell) was used to ensure a sufficient expression of dominant negative protein to abrogate PLK1 function. The percentage and fold increases in Sub-2N DNA content relative to control adenovirus infection were quantified at several time points postinfection by flow cytometric analysis of propidium iodide-stained cells.
The results in Fig. 1
clearly reveal that expression of the dnPLK1 protein caused significant
cell death by 3 days in 6 of 10 tumor lines, SAOS-2, U-2OS, SK-OV-3,
MCF10A, NCI-H596, and SW480. Cell death was considered significant if
dnPLK induced at least a 20% increase in Sub-2N DNA content and at
least a 2-fold increase in fold apoptosis. In many tumor lines,
significant apoptosis was achieved at high ratios of
dominant-negative:endogenous PLK. Some cell lines
(e.g., SK-OV-3 and NCI H596) showed significant apoptosis at
much lower dominant-negative:endogenous PLK1 expression ratios. Three
tumor lines were classified as "insensitive" to dnPLK1 because two
(MDA 435S and MCF7) were relatively unaffected, and a third (HS578T)
was only slightly affected despite equivalent or greater overexpression
of the dominant-negative protein compared with the "sensitive"
tumors.
|
. Even tumors that grew as slowly as HMECs
(e.g., SAOS-2 and H596) showed more rapid and profound cell
death, which suggested that apoptosis can be independent of the rate of
cell proliferation.
Normal human diploid fibroblast IMR90 cells demonstrated significant
apoptosis in response to dnPLK1 (Fig. 1)
. Because the dnPLK1:endogenous
PLK1 ratio was much higher in IMR90 cells than in most tumor lines, we
suspected that the dn-PLK1-induced apoptotic response might still be
tumor-selective. To examine this question, we quantitated the
dnPLK1:endogenous protein ratio in SAOS-2 and IMR90 cells over a range
of MOI at a time when the apoptotic responses of both cell lines were
similar. The results clearly show that the level of apoptosis is
directly correlated with the expression of dnPLK1 in both of the cell
lines and that significantly more dnPLK1 is required to elicit an
equivalent apoptotic response in IMR90 compared with SAOS-2 cells (Fig. 2)
.
|
dnPLK1 Induces Mitotic Catastrophe in SAOS-2 and U-2OS Tumor Cells
but Spares Normal HMECs.
To determine whether the G2-M arrest and
apoptosis attributable to dnPLK1 expression were linked or separate
events, we used flow cytometry of propidium iodide-stained SAOS-2 and
U-2OS cells to examine how the percentage 4N DNA content and Sub-2N DNA
content changed with respect to time and virus dose. At 24 h,
expression of the dnPLK1 caused dramatic increases in 4N DNA content in
both of the cell types (Fig. 3, A and B)
. The time-dependent loss in 4N cells
corresponded with an increase in cells with Sub-2N DNA content. The
cell-death rate was dependent on the MOI and, thus, the amount of
dominant-negative protein expressed. These results suggested that
ablation of PLK1 function through expression of a dnPLK1 protein could
cause SAOS-2 and U-20S cells to arrest in G2 or
mitosis before progressing to apoptosis.
|
-tubulin to see spindle components
and 
-tubulin to stain centrosomes, and included DAPI to visualize
chromatin condensation. With these markers we could quantify the
percentage mitotic figures (Table 1)
and assess the integrity of the mitotic apparatus (Fig. 4)
. At 24 h postinfection with 200 MOI of virus, 44% of dnPLK1
Ad-infected SAOS-2 cells were in mitosis compared with 4% in control
Ad-infected cells. More strikingly, 63% of the dnPLK mitoses were
abnormal compared with 8% in the control (Table 1)
. Similar results
were seen in U-20S cells in which dnPLK1 induced a >5-fold increase in
abnormal mitoses. Abnormal mitoses in SAOS-2 and U-20S cells showed
evidence of disrupted spindle formation, -tubulin staining that was
abnormally reduced compared with interphase cells, and nonuniform
distribution of condensed chromatin (Fig. 4, A–D)
. These
results suggest that apoptosis in SAOS-2 and U-2OS cells was a
consequence of catastrophic mitoses.
|
|
). dnPLK1 induced only a 3-fold increase in mitotic cells,
which suggested that fewer cells were being delayed (Table 1)
. Fully
formed spindles and a normal distribution of chromosomes characterized
HMEC mitoses whether the cells were in prophase or metaphase (Fig. 4, E and F)
. Perhaps the biggest distinction from
SAOS-2 cells was that the -tubulin immunostaining in the mitotic
cells was unchanged relative to interphase cells (Fig. 4, E and F)
or control Ad-treated HMECs (data not shown). Because
proportionally more mitotic cells were arrested at the prophase stage
of mitosis in HMECs, we wanted to determine whether or not mitotic
catastrophe was specific to the later-stage mitotic arrests prevalent
in SAOS-2 cells. We addressed this question by quantifying the fraction
of monoastral mitoses in each cell type that were abnormal (Table 1)
.
In SAOS-2 cells, dnPLK1 Ad induced a >2-fold increase in monoastral
mitoses, of which 41% had defects in centrosomal -tubulin staining
and spindle formation. In contrast, of the almost 4-fold increase in
monoastral mitoses in HMECs, only 2% were morphologically abnormal.
These results suggest that dnPLK1 induces a defective prophase in
SAOS-2 tumor cells, whereas HMECs have a normal prophase. These results
confirm that normal HMECs show an intrinsic difference in their
response to dnPLK1.
Tumor-specific Apoptosis Is Independent of cdc25C Function.
In Xenopus, xPLK1 is essential for the mitotic
phosphorylation and activation of cdc25C on its
NH2 terminus (4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17
, 21, 22, 23)
. In
turn, activated cdc25C is essential for entry into, and progression
through, mitosis through its activation of cdk1 in frog and mammalian
systems (8, 9, 10, 11, 12)
. Lane and Nigg (42)
hypothesized that microinjection of PLK1 antibodies blocked this
cdc25C-regulated checkpoint pathway in Hs68 cells but not in HeLa
cervical carcinoma cells, which allowed the tumor cells to enter
mitosis prematurely. Because adenovirus can infect 100% of the cells
in a population, we could make the first assessment of the effect of
PLK1 ablation on mitotic phosphorylation of cdc25C in mammalian cells.
To measure the effect of dnPLK1, we infected
G1-synchronized SAOS-2 and U-2OS cells and then
treated with the microtubule inhibitor nocodazole to arrest cells in
mitosis. As expected, nocodazole induced a broad band of lower-mobility
cdc25C protein (Fig. 5A
, open arrow) that represented the mitotically
phosphorylated form because treatment with 
phosphatase increased
its mobility to that seen with interphasic cdc25C (compare Lanes
2 and 4, Fig. 5A
). Mitotically
phosphorylated cdc25C was seen in control virus-infected cells treated
with nocodazole but was completely blocked at 100 MOI and partially
blocked at 30 MOI of dnPLK1 Ad in both cell lines (Fig. 5B)
.
Similar results were seen in normal HMECs (Fig. 5C)
at the
same MOI as SAOS-2 and U-2OS tumor cells, although the degree of
mitotic phosphorylation was less evident attributable in part to a less
efficient nocodazole-induced mitotic arrest. Mitotic phosphorylation of
cdc25C in an insensitive MCF7 tumor (Fig. 5D)
was
also blocked by dnPLK1, although higher MOI (650 and 300) were required
because dnPLK1 was not as effectively overexpressed in the mitotic
population compared with the interphase population of cells (data not
shown). Concordant with the effects on phosphorylation of cdc25C, 650
MOI were sufficient to considerably ablate cdc25C-dependent activation
of cdk1 (see Table 2
) and had no greater effect on apoptosis
than that observed at 500 MOI (Fig. 1
, data not shown).
|
|
. These experiments showed that phosphorylation of the histone H1
substrate in vitro was most inhibited in HMECs (98%),
although it required 48 h in nocodazole to obtain a good mitotic
arrest with nocodazole. cdk1 kinase activity was equally inhibited in
U-20S cells and MCF7 cells at 24 h postinfection (88 and 85%,
respectively), although MCF7 cells were insensitive to the dnPLK1
virus. These data argue that the mitotic catastrophe observed in the
U-20S tumor line is not attributable solely to a lack of dependence for
PLK1 in activating cdc25C, and that loss of cdc25C activation is not
sufficient to induce apoptosis in MCF7 cells. Discussion
Overexpression of PLK1 could have important consequences in tumor
formation. PLK1 overexpression has been shown to be transforming in
NIH3T3 cells (52)
and has been correlated with poor
prognosis in small cell lung (53)
and squamous cell
carcinomas of the head and neck (54)
. Previous work
(42)
suggested PLK1 could have a different function in
HeLa tumor cells versus normal Hs68 fibroblasts.
Microinjection of PLK1 antibodies caused mitotic catastrophe in HeLa
cells, whereas the same treatment of nonimmortalized Hs68 cells
elicited a G2 arrest (42)
. Transient
transfection of a dnPLK1 gene into HeLa cells had somewhat different
results from antibody microinjection studies in that most of the
mitotic arrests were bipolar and cytokinesis appeared disrupted
(43)
. To study PLK1 function in more detail, we used an
adenoviral vector to deliver a dnPLK1 gene to 10 tumor lines and 3
normal cell lines. Here we demonstrated that 6 of 10 tumor lines were
at least 2-fold and up to 11-fold more sensitive to inhibition of PLK1
function than normal epithelial cells (Fig. 1)
. Even the one normal
cell line, IMR90, which demonstrated a significant increase in Sub-2N
DNA content response, required more dnPLK1 expression to elicit the
same apoptotic response as that in the SAOS-2 tumor line (Fig. 2)
. In
the sensitive tumor lines, SAOS-2 and U-20S, we showed that dnPLK1
induced a G2-M cell cycle arrest within 24 h
but that these arrested cells declined coincident with the onset of
apoptosis (Fig. 3)
. These data extend previous observations made with
dnPLK1 in HeLa cells to other tumor lines and further show that these
sensitive tumors show a selective apoptotic response versus
normal epithelial counterparts.
Previous experiments unequivocally demonstrated that microinjected PLK1
antibodies blocked centrosome maturation in both HeLa tumor cells and
normal Hs68 fibroblast cells as assessed by size of the centrosomes and
lack of -tubulin staining (42)
. The effects of
overexpression of dnPLK1 protein on centrosome maturation were not
reported (43)
. To further understand why tumor cells
respond differently to a loss of PLK1 function, we used immunostaining
to examine mitotic structures in two sensitive tumor lines, SAOS-2 and
U-20S, versus unaffected normal human mammary epithelial
cells (Fig. 4)
. To ensure that the phenotypic differences were not
attributable to proliferation rates, we focused on the effects of dnPLK
at 48 h in HMECs versus 24 h in the slow growing
SAOS-2 cell line. It was obvious that dnPLK1 Ad caused a mitotic arrest
in all cell types. SAOS-2 (and U-20S) cells were arrested with bipolar
anaphase-like spindles or as prophase-like mitoses with monoastral
arrays emanating from a single organizing center (Fig. 4, A–D)
. Most of the monoastral and bipolar spindles were
morphologically abnormal (Table 1)
. The weak immunostaining of the
centrosomes using the -tubulin antibody is consistent with the
failure of centrosome maturation noted by others (42)
. The
mitotic prophase cells in HMECs had fully formed, morphologically
normal monoastral spindles (Fig. 4, E and F)
. The
diminished -tubulin staining seen in SAOS-2 cells was clearly not
evident in HMECs, although a 3-fold increase in these prophase-like
cells suggested that they, too, were arrested (Table 1)
. We suggest
that dnPLK1 is arresting these cells by blocking the separation of the
centrosomes. This effect on the HMECs appears to differ from the loss
of centrosome maturation that occurs in Hs68 cells after microinjection
of PLK1 antibodies (42)
. These differences may reflect the
type of functional ablation technique or the type of normal cell
studied (42
, 43)
.
Centrosome maturation is essential for the nucleation of mitotic
microtubules. To confirm that abnormal centrosome maturation in SAOS-2
cells could, in part, account for the ensuing mitotic catastrophe, we
quantitated the normal and abnormal monoastral mitoses in each cell
type. Twenty-four % of all of the abnormal mitoses in control
virus-infected SAOS-2 cells (7.7% of total cells) were of the
monoastral type. The number of abnormal monoastral spindles increased
from 24 to 41% in dnPLK1 Ad-infected SAOS-2 cells. Unlike SAOS-2
cells, the number of abnormal monoastral mitoses increased only from 1
to 2% in HMECs (Table 1)
. These data suggest that the loss of PLK1
function exacerbates existing mitotic defects in these tumor lines. Our
observation that mitotic prophase is normal in HMECs suggests that
HMECs escape mitotic catastrophe because their centrosome maturation is
less affected by a loss of PLK1 function. Because PLK1 is important for
the activation of the anaphase-promoting complex necessary for exit
from mitosis, we also hyopthesize that apoptosis results when the
sensitive tumor lines such as SAOS-2 or U-2OS cannot exit mitosis.
We also wanted to investigate the role of cdc25C in the differential
apoptotic response between normal and tumor cells. cdc25C is required
for the prophase-to-metaphase transition through the removal of
inhibitory Thr-14 and Tyr-15 phosphorylations on cdk1 (9
, 10
, 13)
. Lane and Nigg (42)
suggested that HeLa cells
bypassed a centrosome maturation checkpoint whereas normal cells did
not, because PLK1 was not required for cdc25C activation in the former.
However, an alternative model might be that PLK1 has independent roles
in cdc25C activation and centrosome maturation. Our results clearly
demonstrate not only that PLK1 is necessary for mitotic
hyperphosphorylation of cdc25C in mammalian cells but that dnPLK1
blocked mitotic hyperphosphorylation of cdc25C (Fig. 5)
and activation
of cdk1 (Table 2)
regardless of whether the cell line was sensitive or
insensitive to dnPLK1. These data suggest that the inhibition of cdc25C
inhibition by dnPLK1 is not sufficient to induce apoptosis in MCF7
cells. These data also argue that the mitotic catastrophe observed in
the sensitive U-20S tumor line is not attributable to a lack of
dependence for PLK1 in activating cdc25C. Although they do not exclude
a role for cdc25C in the apoptotic response in sensitive tumor lines,
they do indicate that the loss of cdc25C function itself is not
sufficient.
Our data suggest that centrosome maturation can be a distinguishing
feature in the response to loss of PLK1 function. Clearly, a prophase
arrest prevents the HMECs from entering mitosis and catastrophe,
presumably because bipolar movement of the centrosomes cannot occur
without active cdk1. Improper nucleation of the mitotic spindle by the
centrosomes likely leads to mitotic catastrophe in tumor lines like
SAOS-2. Considering that PLK1 can directly interact with - and
-tubulins (44)
we hypothesize that tubulins may be key
PLK1 substrates in tumor cell lines. We think it is likely that the
loss of substrate adhesion and cell death in IMR90 cells mimics anoikis
and may contribute to cell death in some tumor lines. It would be
interesting to know whether disrupting PLK1 function through other
techniques (e.g., antisense or small molecule inhibitor)
causes the same loss of substrate adhesion because anoikis is being
exploited as an antimetastatic and antiangiogenic strategy for cancer
(55)
.
This intrinsic determinant of sensitivity to the loss of PLK1 function might reflect how PLK1 interacts with, or is activated in, essential mitotic complexes. In view of the fact that PLK1 has multiple roles in mitotic progression, it is perhaps not surprising that there is cell-to-cell variability in the dependence of each function on PLK1. Although we see effects of dnPLK1 on phosphorylation of mitotic substrates such as cdc25C and other MPM-2 substrates (data not shown), the predominant effect of overexpressing a high amount of dnPLK1 protein may be disruption of essential mitotic complexes. In fact, overexpression of wild-type PLK1 was reported to block centrosome maturation and induce aberrant mitotic structures in HeLa cells (43) . Preliminary experiments in SAOS-2 and U-2OS cells using a wild-type PLK1 adenovirus suggest that hyperphosphorylation of cdc25C is revealed at lower MOI whereas higher MOI can block cdc25C phosphorylation and cause many of the mitotic defects seen with the dominant-negative (data not shown). Future studies will be aimed at comparing the consequences of expressing wild-type- versus dnPLK1 in these cells.
Why then do the tumor and normal cell lines used in this study show such a wide range of "sensitivity" to expression of the dominant negative protein when PLK1 should be essential for all cells? Perhaps overexpression of PLK1 protein in the sensitive tumor lines makes these cells more dependent on PLK1 for formation of essential mitotic complexes and consequent mitotic phosphorylations. Normal cells that have very low levels of PLK1 protein could use different mitotic complexes that are not as easily disrupted by dnPLK1 and could use PLK1 only to phosphorylate a few essential mitotic substrates. Perhaps overexpression of other mitotic kinases results in the formation of novel complexes or causes overlapping mitotic phosphorylation of essential substrates that make these cells insensitive to a loss of PLK1 function. Future work on the roles of PLK1 and other mitotic kinases in mitotic progression will help to answer these questions.
Materials and Methods
Cell lines, Maintenance, and Cell Cycle Synchronization.
The U-20S osteosarcoma cell line was obtained from the Lineberger
Comprehensive Cancer Center Tissue Culture Facility (University of
North Carolina at Chapel Hill, NC). All of the other tumor lines were
obtained from the American Type Culture Collection (Rockville, MD).
MCF-10A cells were maintained in DMEM/F12 supplemented with 10
µg/ml bovine insulin, 20 ng/ml epidermal growth factor, 0.5 µg/ml
hydrocortisone, and 5% horse serum. All of the other tumor lines were
maintained in DMEM-H (high glucose) with 10% fetal bovine serum
from HyClone (Logan, UT). HMECs were purchased from Clonetics (San
Diego, CA) and used at or before passage 9. The
G1 cell cycle synchronization protocol consisted
of treatment with 1 mM hydroxyurea for 16 h, release
for 10 h, and retreatment with 1 mM hydroxyurea
overnight. For HMECs, only one treatment with hydroxyurea was
sufficient. The G2 cell synchronization included
G1 synchronization, followed by adenovirus
infection, then the addition of 0.2 µg/ml nocodazole 8 h
postinfection, and finally harvest 24 h postinfection in tumor
lines and 48 h postinfection in HMECs.
dnPLK1 Adenovirus and Infections.
The full-length human PLK1 gene was obtained from Incyte as clone
3180142. PCR mutagenesis was used to change Lys-82 to Met (AAG
to ATG). Sequence was confirmed by double-stranded dideoxy sequencing.
The cytomegalovirus promoter-driven, dnPLK1-expressing, E1(-),
E3(-) Ad5 virus was constructed by transposon 7 transposition
in Escherichia coli, as described previously
(56)
. The virus was titered and purified as described
previously (57)
, except that cesium chloride was removed
by dialysis against 10% glycerol, 10 mM Tris-Cl
(pH 8.0), 10 mM MgCl2, and
150 mM NaCl. The purified virus was stored at
20°C. Infections were done in 0.2 ml of DMEM-H per 2 x 105 cells containing 2% FCS at 37°C
for 1 h with continuous rocking. For lipid-enhanced delivery, 0.1
ml each of adenovirus and lipofectamine diluted in serum-free growth
media (6 x 106 pfu of adenovirus per 1 µg
of lipofectamine) were mixed and incubated for 15 min at room
temperature before addition to cells. Lipid delivery was essential for
HS578T and MDAMB435S and was used to enhance dnPLK1 expression in
MDAMB231 and MCF10A cells. For Western blots and immunofluorescence,
cell lines were plated at 105 cells per
9.6-cm2
well. Twenty-four h after plating, the
cells were infected with dnPLK Ad in DMEM-H plus 2% FCS (for
tumor lines) or serum-free growth media (for HMECs or NHKs) with
rocking for 1 h at 37°C. For kinase assays, cells were plated at
8 x 105 per 100-mm dish and infections were
scaled up by 8-fold.
Western Blotting and Kinase Assays.
For Western blots, cells were lysed in RIPA buffer containing
protease inhibitor cocktail (Boehringer Mannheim), 50 mM
sodium fluoride, 2 mM sodium vanadate, and 10 µg/ml
microcystin LR (Sigma) at 4°C. Protein concentrations were
determined by the bicinchoninic acid assay (Pierce) and equal
concentrations loaded per well. Ten % nonfat dry milk was used for
blocking nitrocellulose and for antibody incubations. Western blotting
antibodies were used at 0.5 µg/ml and included mouse mAb anti-PLK1
(No. 33–170; Zymed) and anti-cdc25C (#sc-327, Santa Cruz
Biotechnology). Quantitation of dnPLK1 overexpression was performed
making serial 1:4 dilutions of dnPLK1-infected lysates and determining
the dilution that was equivalent to control virus infected lysate.
Immunoprecipitations used 100 µg of cell lysate with 10 µg of
cyclin B1 mAb (No. 14541C; PharMingen) and protein G Sepharose
(Pharmacia Biotech). Precipitates were washed twice in RIPA, twice in
RIPA plus 1 M NaCl, then twice in 50 mM HEPES
(pH 7.5), and 150 mM NaCl. Beads were resuspended in
histone H1 assay mix containing 50 mM HEPES (pH
7.5), 150 mM NaCl, 10 mM
MgCl2, 1 mg/ml histone H1, 1 mM DTT,
2 mM EGTA, 50 mM ATP, and 50 µCi/ml
[-32P]ATP and were incubated for 15 min at
30°C. Reactions were stopped with an equal volume of Laemmli sample
buffer, heat-denatured, and electrophoresed on 12% SDS-PAGE mini-gels
(Novex). Kinase activity was quantitated after a 4-h exposure on a
PhosphoImager (Molecular Dynamics).
FACS Analysis.
FACS analysis for cell cycle and apoptosis was performed with a Becton
Dickinson FACS Calibur. After infection, cells were washed in PBS
without calcium and magnesium salts and then fixed in 70% methanol.
Fixed cells were washed once with a solution containing 10
mM HEPES (pH 7.4), 150 mM NaCl, 4% FCS, 0.5%
Tween, and 0.1% sodium azide. Cell pellets were resuspended in PBS
(Ca+2, Mg+2-free) with 0.1% Triton X-100, 100 µg/ml RNase A,
and 40 µg/ml propidium iodide for staining of DNA. Cell cycle
analysis was performed using Cell Quest software.
Immunocytochemistry.
SAOS-2, U-20S, and HMECs were plated at 105 per
well of a 2-well glass chamber slide (Nalge; Nunc International) and,
24 h after plating, were infected with 200 pfu of virus per
cell. Fixation and staining were done as described previously
(58)
using 1 h for each antibody-binding step.
Primary antibodies were rat anti--tubulin (MCP77; Serotec) and mouse
anti--tubulin (GTU-88; Sigma), each at 1:100. Secondary antibodies
were fluorescein-conjugated donkey antirat IgG and rhodamine-conjugated
donkey antimouse IgG (Jackson Immunoresearch) at 1:1000. Vectastain
plus DAPI was added to stain DNA.
Microscopy.
Sections at x200 and cells at x40 were examined on a Leitz DMRX
microscope and captured with Adobe Photoshop using a Hamamatsu chilled
3CCD camera, model C5810.
Acknowledgments
PLK1 clone 3180142 was obtained from Incyte Pharmaceuticals Inc. We thank Jui-Lan Su for antibody optimizations and Cathy Finlay, Gordon McIntyre, Sally Kornbluth, Bob Abraham, and Tony Means for their scientific input.
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 To whom requests for reprints should be
addressed, at Department of Functional Genetics, Glaxo Wellcome Inc., 5
Moore Drive, Research Triangle Park, NC 27709. Phone: (919) 483-3449;
E-mail: jpc30240{at}glaxowellcome.com 
2 The abbreviations used are: cdk,
cyclin-dependent kinase; PLK, polo-like kinase; dnPLK,
dominant-negative PLK; dnPLK1 Ad, adenovirus expressing dnPLK1; MPM-2,
mitotic protein monoclonal 2; FACS, fluorescence-activated cell
sorting; MOI, multiplicity/multiplicities of infection; NHK, normal
human keratinocyte; HMEC, human mammary epithelial cell; DAPI,
4',6-diamidino-2-phenylindole; RIPA, radioimmunoprecipitation assay;
pfu, plaque-forming unit(s); mAb, monoclonal antibody.
Received for publication 6/22/00. Revision received 10/10/00. Accepted for publication 10/25/00.
References
is the major chromosome protein recognized by the mitotic phosphoprotein antibody MPM-2. Proc. Natl. Acad. Sci. USA, 90: 8407-8411, 1993.-, ß-, and -tubulins in a stable complex. Biochem. J., 339: 435-442, 1999.
This article has been cited by other articles:
 |
|
 |
|
  X. Wang, X. Wang, R. K. Varma, L. Beauchamp, S. Magdaleno, and T. J. Sendera Selection of hyperfunctional siRNAs with improved potency and specificity Nucleic Acids Res., October 21, 2009; (2009) gkp864v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. G. Gilmartin, M. R. Bleam, M. C. Richter, S. G. Erskine, R. G. Kruger, L. Madden, D. F. Hassler, G. K. Smith, R. R. Gontarek, M. P. Courtney, et al. Distinct Concentration-Dependent Effects of the Polo-like Kinase 1-Specific Inhibitor GSK461364A, Including Differential Effect on Apoptosis Cancer Res., September 1, 2009; 69(17): 6969 - 6977. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Yim and R. L. Erikson Polo-Like Kinase 1 Depletion Induces DNA Damage in Early S Prior to Caspase Activation Mol. Cell. Biol., May 15, 2009; 29(10): 2609 - 2621. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Sherwood, R. Manbodh, C. Sheppard, and A. D. Chalmers RASSF7 Is a Member of a New Family of RAS Association Domain-containing Proteins and Is Required for Completing Mitosis Mol. Biol. Cell, April 1, 2008; 19(4): 1772 - 1782. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. Leoni, B. Bailey, J. Reifert, H. H. Bendall, R. W. Zeller, J. Corbeil, G. Elliott, and C. C. Niemeyer Bendamustine (Treanda) Displays a Distinct Pattern of Cytotoxicity and Unique Mechanistic Features Compared with Other Alkylating Agents Clin. Cancer Res., January 1, 2008; 14(1): 309 - 317. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Stegmeier, M. E. Sowa, G. Nalepa, S. P. Gygi, J. W. Harper, and S. J. Elledge The tumor suppressor CYLD regulates entry into mitosis PNAS, May 22, 2007; 104(21): 8869 - 8874. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. J. Lansing, R. T. McConnell, D. R. Duckett, G. M. Spehar, V. B. Knick, D. F. Hassler, N. Noro, M. Furuta, K. A. Emmitte, T. M. Gilmer, et al. In vitro biological activity of a novel small-molecule inhibitor of polo-like kinase 1 Mol. Cancer Ther., February 1, 2007; 6(2): 450 - 459. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kappel, Y. Matthess, B. Zimmer, M. Kaufmann, and K. Strebhardt Tumor inhibition by genomically integrated inducible RNAi-cassettes Nucleic Acids Res., September 11, 2006; 34(16): 4527 - 4536. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Spankuch, S. Heim, E. Kurunci-Csacsko, C. Lindenau, J. Yuan, M. Kaufmann, and K. Strebhardt Down-regulation of Polo-like Kinase 1 Elevates Drug Sensitivity of Breast Cancer Cells In vitro and In vivo Cancer Res., June 1, 2006; 66(11): 5836 - 5846. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Malorni and C. Fiorentini Is the Rac GTPase-activating toxin CNF1 a smart hijacker of host cell fate? FASEB J, April 1, 2006; 20(6): 606 - 609. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Y. Chuang, X. Chen, M.-H. Tsai, H. Yan, C.-Y. Li, J. B. Mitchell, H. Nagasawa, P. F. Wilson, Y. Peng, M. M. Fitzek, et al. Abnormal gene expression profiles in unaffected parents of patients with hereditary-type retinoblastoma. Cancer Res., April 1, 2006; 66(7): 3428 - 3433. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Bhonde, M.-L. Hanski, J. Budczies, M. Cao, B. Gillissen, D. Moorthy, F. Simonetta, H. Scherubl, M. Truss, C. Hagemeier, et al. DNA Damage-induced Expression of p53 Suppresses Mitotic Checkpoint Kinase hMps1: THE LACK OF THIS SUPPRESSION IN p53MUT CELLS CONTRIBUTES TO APOPTOSIS J. Biol. Chem., March 31, 2006; 281(13): 8675 - 8685. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Eckerdt, J. Yuan, K. Saxena, B. Martin, S. Kappel, C. Lindenau, A. Kramer, S. Naumann, S. Daum, G. Fischer, et al. Polo-like Kinase 1-mediated Phosphorylation Stabilizes Pin1 by Inhibiting Its Ubiquitination in Human Cells J. Biol. Chem., November 4, 2005; 280(44): 36575 - 36583. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Guan, P. Tapang, J. D. Leverson, D. Albert, V. L. Giranda, and Y. Luo Small Interfering RNA-Mediated Polo-Like Kinase 1 Depletion Preferentially Reduces the Survival of p53-Defective, Oncogenic Transformed Cells and Inhibits Tumor Growth in Animals Cancer Res., April 1, 2005; 65(7): 2698 - 2704. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Mook Jr. The Importance and Complexities of Target Class Selectivity in Drug Discovery Am. Assoc. Cancer Res. Educ. Book, April 1, 2005; 2005(1): 223 - 226. [Full Text] [PDF]
|
|
|
|
|
|
L. Tsvetkov and D. F. Stern Interaction of Chromatin-associated Plk1 and Mcm7 J. Biol. Chem., March 25, 2005; 280(12): 11943 - 11947. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W. Gunawardena, H. Siddiqui, D. A. Solomon, C. N. Mayhew, J. Held, S. P. Angus, and E. S. Knudsen Hierarchical Requirement of SWI/SNF in Retinoblastoma Tumor Suppressor-mediated Repression of Plk1 J. Biol. Chem., July 9, 2004; 279(28): 29278 - 29285. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Gray Jr, D. J. Bearss, H. Han, R. Nagle, M.-S. Tsao, N. Dean, and D. D. Von Hoff Identification of human polo-like kinase 1 as a potential therapeutic target in pancreatic cancer Mol. Cancer Ther., May 1, 2004; 3(5): 641 - 646. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Spankuch-Schmitt, J. Bereiter-Hahn, M. Kaufmann, and K. Strebhardt Effect of RNA Silencing of Polo-Like Kinase-1 (PLK1) on Apoptosis and Spindle Formation in Human Cancer Cells J Natl Cancer Inst, December 18, 2002; 94(24): 1863 - 1877. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-S. Seong, K. Kamijo, J.-S. Lee, E. Fernandez, R. Kuriyama, T. Miki, and K. S. Lee A Spindle Checkpoint Arrest and a Cytokinesis Failure by the Dominant-negative Polo-box Domain of Plk1 in U-2 OS Cells J. Biol. Chem., August 23, 2002; 277(35): 32282 - 32293. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Yuan, A. Kramer, F. Eckerdt, M. Kaufmann, and K. Strebhardt Efficient Internalization of the Polo-Box of Polo-like Kinase 1 Fused to an Antennapedia Peptide Results in Inhibition of Cancer Cell Proliferation Cancer Res., August 1, 2002; 62(15): 4186 - 4190. [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 |
).
