(Received for publication, April 18, 1997)
From the Baker Medical Research Institute, Commercial Road, Prahran, Victoria 3181, Australia
Most cancer cells have increased levels of
telomerase activity implicated in cell immortalization. Activation of
telomerase, a ribonucleoprotein complex, catalyzes the elongation of
the ends of mammalian chromosomal DNA (telomeres), the length of which regulates cell proliferation. Currently, how telomerase is regulated in
cancer is not yet established. The present study shows that telomerase
activity is regulated by protein phosphorylation in human breast cancer
cells. Incubation of cell nuclear telomerase extracts with protein
phosphatase 2A (PP2A) abolished the telomerase activity; in contrast
cytoplasmic telomerase activity was unaffected, and protein
phosphatases 1 and 2B were ineffective. Inhibition of telomerase
activity by PP2A was both concentration- and time-dependent and was prevented by the protein phosphatase inhibitor okadaic acid. In
addition, nuclear telomerase inhibited by PP2A was reactivated by
endogenous protein kinase(s) in the presence of ATP, but not in the
presence of ATPS. Furthermore, telomerase activity in cultured human
breast cancer PMC42 cells was stimulated by okadaic acid, consistent
with a role for PP2A in the regulation of telomerase activity in intact
cells. These findings suggest that protein phosphorylation reversibly
regulates the function of telomerase and that PP2A is a telomerase
inhibitory factor in the nucleus of human breast cancer cells.
One of the critical events in the oncogenic progression of normal
human cells is the escape from the limitations on proliferation imposed
by cellular senescence. One of the primary mechanisms underlying
cellular senescence in normal diploid cells involves controlled
telomere shortening in cell division (for reviews, see Refs. 1-4).
Telomeres are defined as the ends of all eukaryotic chromosomes
comprising an array of tandem repeats of the hexanucleotide 5-TTAGGG-3
and its binding proteins. These nucleoprotein structures function to protect chromosomes against exonucleases and ligases, to
prevent the activation of DNA-damage checkpoints, and to counter the
loss of terminal DNA segments that occurs when linear DNA is
replicated. Mutation of telomere repeat, however, blocks chromosome separation suggesting a regulatory role for telomeres in anaphase mitosis (5). Since conventional DNA polymerases cannot replicate the
very 5
end of linear DNA, telomeres shorten as a function of each cell
division in normal human somatic cells, contributing to the mechanisms
underlying cell senescence (6-9). However, several specialized
proteins are involved in the regulation of telomere metabolism,
including the yeast double strand telomere-binding protein Rap1 (10)
and single strand telomere-binding proteins Rlf6p (11), Cdc13 (12), and
Est1 (13), the human duplex telomere-binding protein TRF (14),
telomerase, and telomerase-associated protein TP1 (15).
Telomerase, a primary determinant of telomere length, is a ribonucleoprotein complex specific for telomere DNA synthesis (1-4). It is composed of two protein subunits of 95 kDa and 80 kDa in the ciliate Tetrahymena thermophila (16) and of 120 kDa and 43 kDa in the ciliate Euplotes aediculatus (17). While the 95-kDa subunit binds substrate oligonucleotides, the 80-kDa subunit binds to a telomerase RNA component of 159 nucleotides (16, 18-21). Thus, telomerase uses a portion of this internal RNA moiety as a template to replicate telomeres. Consistent with the shortening of telomeres with each normal cell division, telomerase activity is below detectable levels in normal somatic cells. In most primary human malignancies, however, telomerase is activated by an as yet unknown mechanism. This suggests not only that telomerase activity may be a marker for malignancy but also crucial for unlimited cell division in tumor proliferation through de novo synthesis of telomeres (2, 3, 8).
Recent studies indicate that telomerase activity increases as normal human T cells enter the cell cycle (22, 23), and increased telomerase activity in human promyelocytic leukemic HL60 cells is repressed by cellular differentiation (24-27). In addition, telomerase activity in human breast cancer cells is regulated in a cell cycle-dependent manner, with increased telomerase activity in the S phase and decreased telomerase activity in the G2/M phase (28). These findings suggest that telomerase activity in cell growth and development is reversibly regulated by a molecular switch. Since vertebrate telomerase is a ribonucleoprotein complex with at least two protein components containing multiple potential phosphorylation sites (16), we tested the hypothesis that protein phosphorylation might play a role in telomerase function. The present study shows that nuclear telomerase activity in breast cancer PMC42 cells is specifically abolished by protein phosphatase 2A (PP2A).1 The effect of PP2A is in turn blocked by okadaic acid and reversed by protein phosphorylation through endogenous protein kinase(s), indicating a reversible regulation of telomerase activity by protein phosphorylation and dephosphorylation.
Human breast cancer PMC42 cells have
been described earlier (29). The cells were maintained in the medium of
RPMI 1640 supplemented with fetal calf serum (10%), insulin (0.6 µg/ml), hydrocortisone (1 µg/ml), and galactose (0.1%) at an
enclosed humidified atmosphere of 95% O2 and 5%
CO2 at 37 °C. Cell nuclei were isolated by
centrifugation (3000 × g) after homogenization of the
cells with a Teflon/glass homogenizer (10 ml), and the cytosolic and
membrane fractions were separated further by ultracentrifugation
(100,000 × g). In some experiments, cells at 80%
confluence were treated with okadaic acid (1 µM, from
Calbiochem) as indicated. PP2A (>95% pure) was obtained from Biomol
Research Laboratories (Plymouth Meeting, PA) or partially purified from
rat liver as described previously (30). PP1 and PP2B were from UBI
(Lake Placid, NY). Taq DNA polymerase, dNTP, and T4 gene 32 protein were from Boehringer Mannheim. [-32P]ATP was
from Amersham. Gel electrophoresis reagents and equipment were from
Bio-Rad.
Telomerase activity in the
membrane, cytosolic, and nuclear fractions or extracted from whole
cells with buffer containing 2.5% CHAPS was determined in duplicate by
a slightly modified TRAP assay as described previously (31). Briefly, 1 µg of protein extracts were incubated at 30 °C for 20 min in a
50-µl reaction buffer containing 10 mM Tris, pH 8.3, 0.1 µg of telomerase substrate (5-AATCCGTCGAGCAGAGTT-3
), 50 µM each deoxynucleoside triphosphate, 0.5 µCi of
[
-32P]dATP, 1.5 mM MgCl2, 1 mM EGTA, 68 mM KCl, 0.05% Tween 20, 1 µg of
T4 gene 32 protein. To each reaction was then added 2 units of
Taq DNA polymerase and 0.1 µg of downstream primer CX
(5
-CCCTTACCCTTACCCTTACCCTAA-3
), followed by a 31-cycle polymerase
chain reaction (94 °C for 30 s, 50 °C for 30 s, and
72 °C for 45 s). Twenty µl of the polymerase chain reaction
was then analyzed for de novo synthesized
32P-telomeres by electrophoresis on a 10% non-denaturing
polyacrylamide gel followed by overnight exposure for autoradiography.
For relative quantification, telomerase activity was measured by
scintillation counting of 32P-telomeres resolved on
non-denaturing polyacrylamide gel electrophoresis and sliced from the
gels with autoradiographs as templates. For experimental controls of
telomerase activity assay, treatment of telomerase extracts with 2 µg
of RNase A or heating at 90 °C for 10 min was routinely
included.
Inappropriate telomerase activity has been found to be associated
with a variety of human cancers (8). The mechanisms of regulation of
telomerase activity have not been established. The primary aim of these
studies was to test whether telomerase activity might be regulated by
protein phosphorylation. To this end, we have determined effects of the
three major cellular protein phosphatases (1, 2A, and 2B) on the
activity of telomerase extracted from human breast cancer PMC42 cells.
Subcellular fractionation of the cultured cells was performed for
nuclear, membrane particulate, and cytoplasmic fractions. Telomerase
activity was then examined after incubation of each fraction with
exogenous protein phosphatases 1, 2A, and 2B under dephosphorylation
conditions (30). Interestingly, PP2A markedly inhibited telomerase
activity in nuclear extracts (Fig. 1A), an
effect which was concentration-dependent (ED50
10 ± 2 units, Fig. 1B) and rapid (significant
inhibition at 10 s, complete inhibition by 2 min; Fig.
1C). PP2A inhibited telomerase activity in the membrane
particulate fraction to a lesser extent than in nuclear and appeared
inactive in cytosol (Fig. 1C). Substantial levels of
telomerase activity were associated with the nucleus (45%) and
membrane particulate fractions (38%), compared with 17% in the
cytosol (Fig. 1C) of the cultured PMC42 cells. Telomerase activity in different subcellular locations may thus ultimately be
achieved by a balance of complex control mechanisms, including the
inhibitory effect by PP2A. The inhibition of nuclear telomerase activity by PP2A is specific, since neither protein phosphatase 1 nor
2B affected telomerase activity (Fig.
1B).
Inhibition of telomerase activity by PP2A may reflect direct dephosphorylation of telomerase protein components or indirect protein-protein interactions. To verify that PP2A-induced inhibition of telomerase activity involves protein dephosphorylation, nuclear telomerase extracts were coincubated with PP2A and various concentrations of okadaic acid, which binds to the catalytic motif of PP2A (32). The inhibition of telomerase activity by PP2A was prevented by okadaic acid in a concentration-dependent fashion, with an ED50 of 0.2 nM (Fig. 2A), suggesting that the catalytic activity of PP2A is required for its inhibition of telomerase activity. This is also supported by the observation that the effect of PP2A on telomerase activity was avoided after the PP2A was absorbed in advance with a specific antibody against PP2A catalytic subunit (data not shown).
The hypothesis that protein dephosphorylation is involved in PP2A
inhibition of telomerase activity is further supported by the
correlation between protein phosphorylation and reactivation of
telomerase activity. In these protein phosphorylation experiments, nuclear telomerase extracts were first incubated with PP2A for 3 min;
after addition of the protein phosphatase inhibitor sodium fluoride,
protein phosphorylation was allowed to proceed by further incubation in
the presence of ATP, followed by telomerase activity analysis. In the
presence of ATP, but not ATPS, telomerase inhibited by PP2A was
reactivated back to levels equivalent to those prior to PP2A
dephosphorylation (Fig. 2B), strongly suggesting that reactivation of telomerase is through an ATP
hydrolysis-dependent mechanism and that protein
phosphorylation by an as yet unidentified protein kinase(s) is likely
to mediate this process after dephosphorylation by PP2A. In support of
this concept of protein phosphorylation, telomerase reactivation was
inhibited by the protein kinase inhibitor H-7 (Fig. 2B).
These data unequivocally suggest that protein phosphorylation and
dephosphorylation are involved as a molecular switch in up- and
down-regulating telomerase activity in the nucleus of human breast
cancer cells.
To determine if the PP2A-mediated dephosphorylation event occurs in
intact cells, we treated the human breast cancer PMC42 cells with
okadaic acid for different periods of time and found that telomerase
activity in these cells was markedly stimulated by okadaic acid (1 µM) (Fig. 3). Okadaic acid stimulation of
telomerase activity was seen after 30 min and peaked at 2-3 h (Fig.
3). This effect on telomerase activity is thus consistent with the
in vitro findings that telomerase is reversibly regulated by
protein phosphorylation and dephosphorylation, through unidentified
protein kinase(s) and the protein phosphatase 2A. In addition, it is
likely that in the control of telomerase activity, at least one of the
telomerase protein components or telomerase regulatory proteins is
phosphorylated and that protein phosphorylation is required for its
activity of telomere synthesis in cancer, suggesting a molecular
mechanism that may explain the reversible regulation of telomerase
activity through cell division cycles (22-28). Furthermore, these
findings suggest a link between telomerase proteins and PP2A and a
potential target of PP2A in controlling telomerase activity in cancer
therapy.
Several lines of evidence from previous studies suggest that PP2A is involved in the oncogenic process. PP2A is present at high levels in the nucleus of non-transformed mammalian fibroblasts (33), and it has been suggested to play a regulatory role during the cell cycle (33, 34). Tyrosine phosphorylation of PP2A catalytic subunit by receptor-linked and non-receptor tyrosine kinases inhibits the activity of PP2A (35, 36). Moreover, PP2A is a direct target of at least three distinct DNA tumor virus families, simian virus 40, polyoma virus (37, 38), and adenovirus (39); in addition, binding of the viral oncogenic protein small-t antigen to PP2A inhibits the activity of PP2A and stimulates cell proliferation (40). It is thus conceivable that inhibition of PP2A by oncogenic signals or viral tumor antigens may be involved in telomerase reactivation in certain cancers and that the pharmacological stimulation of PP2A dephosphorylation of telomerase in cancer may be of potential therapeutic significance.
We thank Bob Whitehead for supplying the breast cancer cell line, Michael Berndt, Alex Bobik, Noel Fidge, Ban-Hock Toh, and Elizabeth Woodcock for their advice, and Zhi-Yong Yang for technical assistance.